Report No. PAKISTAN Sindh: Public Expenditure Review SOUTH ASIAN REGION JUNE 2017 THE WORLD BANK GROUP Sindh: Public Expenditure Review 2017 STANDARD DISCLAIMER This volume is a product of the staff of the International Bank for Reconstruction and Development/The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. COPYRIGHT STATEMENT The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, Email: pubrights@worldbank.org THE WORLD BANK GROUP 1 Sindh: Public Expenditure Review 2017 Government of Pakistan Fiscal Year: July 1-June 30 Currency Equivalents Currency Unit: Pakistani Rupees (PRs) US$1.00 = PRs 104.85 as of June 30, 2017 Regional Vice President: Annette Dixon Country Director: Patchamuthu Illangovan Senior Director: Carlos Felipe Jaramillo Practice Manager: Manuela Francisco Task Team Leaders: Muhammad Waheed THE WORLD BANK GROUP 2 Sindh: Public Expenditure Review 2017 Table of Contents ACRONYMS AND ABBREVIATIONS ....................................................................................................... 4 ACKNOWLEDGEMENTS ....................................................................................................................... 6 Executive Summary ............................................................................................................................ 7 1. Sindh Province: Unleashing Its Potential................................................................................... 7 2. Maximizing the Resource Envelope .......................................................................................... 7 3. Improving the Efficiency of Public Spending ............................................................................. 9 4. Development Expenditures: Getting Better Value for Money ................................................ 14 5. Recommendations .................................................................................................................. 15 Chapter 1: Macroeconomic and Fiscal Trends .................................................................................... 20 1.1 The Socioeconomic Profile of Sindh ........................................................................................ 20 1.2 Macro-Fiscal Outcomes ........................................................................................................... 23 Chapter 2: Revenues and Expenditures .............................................................................................. 27 2.1 Revenues: The Revenue Challenge and Fiscal Space .............................................................. 27 2.2 Expenditures: The Challenge of Funding Development Priorities .......................................... 35 2.3 Recommendations .................................................................................................................. 54 Chapter 3: Sindh Education Sector ..................................................................................................... 57 3.1 Introduction............................................................................................................................. 57 3.2 Sindh Education Sector Overview ........................................................................................... 57 3.3 Context and Policy Framework ............................................................................................... 58 3.4 Education Sector Profile Analysis ............................................................................................ 59 3.5 Profile and Trend Analysis of Education Sector Expenditure .................................................. 66 3.6 Expenditure Composition Analysis .......................................................................................... 68 3.7 Allocative Efficiency of Education Expenditures ..................................................................... 75 3.8 Sector Effectiveness and Equity Analysis of Expenditure........................................................ 82 3.9 Recommendations .................................................................................................................. 84 Chapter 4: Sindh Health Sector .......................................................................................................... 86 4.1 Introduction............................................................................................................................. 86 4.2 Sindh Health Sector Overview ................................................................................................. 88 4.3 Context and Policy Framework ............................................................................................... 89 4.4 Sindh Health Sector Profile...................................................................................................... 91 4.5 Public Health Expenditure ....................................................................................................... 92 4.6 Allocative Efficiency of Health Expenditures........................................................................... 98 4.7 Recommendations ................................................................................................................ 100 Chapter 5: Protecting the Poor and Vulnerable through the Public Sector ......................................... 103 5.1 Introduction........................................................................................................................... 103 5.2 Social Protection Policy in Sindh ........................................................................................... 104 5.3 Impact of the 18th Constitutional Amendment and the Role of the Province in Social Protection .............................................................................................................................. 105 5.4 Poverty, Inequality and Vulnerability in Sindh Province ....................................................... 105 5.5 Labor-Force Participation in Sindh ........................................................................................ 108 5.6 Social Protection Expenditure in Sindh ................................................................................. 110 5.7 Institutional Arrangements for Social Protection in Sindh .................................................... 113 5.8 Inventory and Analysis of Sindh SPL Programs ..................................................................... 114 5.9 Efficacy of Sindh Social Protection Initiatives........................................................................ 116 5.10 Recommendations ................................................................................................................ 117 Annexes.......................................................................................................................................... 119 THE WORLD BANK GROUP 3 Sindh: Public Expenditure Review 2017 ACRONYMS AND ABBREVIATIONS ADB Asian Development Bank NGOs Non-Governmental Organization ADP Annual Development Plan NMR Neonatal Mortality Rate AG Accountant General NPPI Norway-Pakistan Partnership Initiative AIT Agriculture Income Tax NSER National Socio-Economic Registry ARV Annual Rental Value OOP Out-of-Pocket ASER Annual Status of Education Report OOSC Out-of-School Children ASPIRE Atlas of Social Protection Indicators of OPEC Organization of Petroleum Exporting Countries Resilience & Equity BBSYDP Benazir Bhutto Shaheed Youth OSR Own-Source Revenues Development Program BCG Bacille Calmette Guerin P&DD Planning and Development Department BHU Basic Health Unit PDMA Provincial Disaster Management Authority BISP Benazir Income Support Program PEFA Public Expenditure and Financial Assessment BOR Board of Revenue PER Public Expenditure Review BPS Basic Pay Scale PFM Public Financial Management BSP Budget Strategy Paper PGDP Provincial Gross Domestic Product CCT Conditional Cash Transfers PHC Primary Health Care CDL Cash Development Loans PIFRA Project for Improving Financial Reporting and Auditing CIF Community Investment Fund PIM Public Investment Management CIP Costed Implementation Plan PIMIS Public Investment Management Information System CPI Consumer Price Index PITE Provincial Institute of Teacher Education DE Development Expenditure PMIU Project Management Implementation Units DGM&E Director General Monitoring and PMT Proxy Means Test Evaluation DGSE Directorate General of School Education POL Petroleum, Oil and Lubricants DMIC Development and Maintenance of PPHI President’s Primary Health Care Initiative Infrastructure Cess DOTS Directly Observed Treatment, Short- PPP Public-Private Partnership course DUHS Dow University of Health Science PSLMS Pakistan Social and Living Standards Measurement Survey E&LD Education and Literacy Department PST Problem Solving Teams ECE Early Childhood Education RE Recurrent Expenditure EFF Extended Fund Facility RFA Revolving Fund Account FATA Federally Administrated Tribal Area RSU Reform Support Units FBR Federal Board of Revenue SA Situation Analysis FD Finance Department SAT Standardized Achievement Test /Student Achievement Test FMF Fiscal Management Framework SBA Shaheed Benazir Abad FY Fiscal Year SCARP Salinity Control and Reclamation Program GDP Gross Domestic Product SED School Education Department GER Gross Enrolment Rate SEDD GFMIS Government Financial Management SEF Sindh Education Foundation Information System GoS Government of Sindh SEMIS Sindh Education Management Information System 4 Sindh: Public Expenditure Review 2017 GPF General Provident Fund SERP Sindh Education Reform Program GPs General Practitioners SESP Sindh Education Sector Plan GST General Sales Tax SGPIF Sindh General Provident Investment Fund HD Health Department SGRRP Sindh Growth & Rural Revitalization Program HEC Higher Education Commission SMC School Management Committee HED Higher Education Department SN- Subnational Debt Management Performance DeMPA Assessment HIES Household Income & Expenditure Survey SOE State-Owned Enterprises HIV human immunodeficiency virus SP Social Protection HR Human Resource SPL Social Protection and Labor IBRD International Bank for Reconstruction SRB Sindh Revenue Board and Development ICT Information and Communication SSB School Specific Non-Salary Budget Technology IDA International Development Agency SSSD Sindh Strategy for Sustainable Development IFAD International Fund for Agriculture SSTSA Sindh Sales Tax on Services Act Development IMR Infant Mortality Rate STEDA Sindh Teacher Education Development Authority KP Khyber Pakhtunkhwa STEVTA Sindh Technical Education and Vocational Training Authority LCHS Low cost housing schemes STIs Sexually Transmitted Infection LHW Lady Health Worker STR Student-Teacher Ratio M&E Monitoring and Evaluation STS Sales Tax on Services MCH Maternal and Child Health SUSSESS Sindh Union Council and Community Economic Strengthening Support Program MDA Ministries/Department/Autonomous SWD Social Welfare Department Institutions MDG Millennium Development Goals TB Tuberculosis MHI Micro Health insurance TFR Total Fertility Rate MICS Multi-Indicator Cluster Survey TRF Technical Resource Facility MMR Maternal Mortality Rate TSA Treasury Single Account MTBF Medium-Term Budgetary Framework UCBPRP Union Council-based Poverty Reduction Program NAM New Accounting Model UIPT Urban Immovable Property Tax NCD Non-Communicable Diseases US United States NEC National Economic Council VTPs Vocational training programs NER Net Enrolment Rate W&SD Works & Services Department NFC National Finance Commission WAPDA Water and Power Development Authority 5 Sindh: Public Expenditure Review 2017 ACKNOWLEDGEMENTS This “Sindh: Public Expenditure Review� focuses on provincial finances and their utilization with the objective of identifying possible reforms to expand the resource envelope and ensure better value for money by improving the management and efficiency of public spending. It includes two important components: (i) a detailed analysis of the major revenue challenges and the various expenditures, including development spending and; (ii) and in-depth assessment of how some of the key government priorities are undertaken, such as education, health, and social protection. The review looked at the issues of allocative and operational efficiencies of public finances. In addition, report also observed the element of equity in the expenditures. This report was prepared under the Sindh Fiscal Management-ASA (P159810). The strategic level guidance for this task was provided by Illango Patchamuthu (Country Director, SACPK) and Manuela Francisco (Practice Manager, GFM06). The technical team was led by Muhammad Waheed (Task Team Leader, Senior Economist, GMF06). Several colleagues contributed to the report. These include: Adnan Ashraf Ghumman (Economist, GMF06) Aijaz Ahmad (Senior Public Private Partnerships Specialist, GTPPP) Ali Ansari (Economist, GED06) Amna Sehar (Consultant, GGO18) Carolina Romero Robayo (Research Analyst, GSPGL) David Duarte (Senior Public Private Partnerships Specialist, GTPPP) Hammad Younas (Consultant, GMF06) Irum Touqeer (Public Sector Specialist, GGO18) Kiran Afzal (Senior Private Sector Specialist, GFCSN) Mehwish Ashraf (Economist, GMF06) Minhaj ul Haq (Consultant, GMF06) Noaman Ali (Consultant, GMF06) Rajul Awasthi (Senior Public Sector Specialist, GGOGT) Raul Felix Junquera-Varela (Lead Public Sector Specialist, GGOGT) Saarim Saghir (Consultant, GED06) Shaheen Malik (Consultant, GMF06) Sohail Saeed Abbasi (Senior Social Protection Specialist, GSP06) Tayyeb Masood (Senior Health Specialist, GHN19) Umbreen Arif (Senior Education Specialist, GED06) Waseem Shahid Malik (Consultant, GMF06) Special thanks to the insightful comments and advice provided by Enrique Blanco Armas (Lead Economist, GMF06) and Hanid Mukhtar. Peer review of the document was carried out by Habib Nasser Rab (Program Leader, ECCTR) and Volker Treichel (Principal Country Economist, CEDVP) and their comments have enabled the team to improve the various aspects of the analysis. Team appreciates their time and insights. This report would have been impossible without the administrative assistance of Mohammad Aslam Malik (SASEP). The editorial support was provided by Peter Milne (Consultant, GMF06) during the whole process. Finally, the team is also grateful for the close collaboration, support and encouragement given at every step by counterparts from the Government of Sindh. 6 Sindh: Public Expenditure Review 2017 Executive Summary 1. Sindh Province: Unleashing Its Potential Sindh has the potential to become a high middle-income province, but it lags far behind in terms of economic, social and development indicators. Sindh is the most urbanized of Pakistan’s four provinces, with a large population of about 49 million—roughly one-quarter of the population of Pakistan—and is strategically located on the coast with access to the Arabian Sea. Karachi, the largest city and port in Pakistan, is the main financial and commercial hub of the country, and the main contributor to national GDP. Sindh is also the most industrialized province in Pakistan, resource-rich and endowed with the country’s largest natural gas and coal reserves. In sum, it has the potential to become a high-growth and high-income region. Nonetheless, Sindh has yet to translate this potential into commensurate economic and social development. Instead of becoming an economic powerhouse, Sindh’s rate of growth has decelerated and is no longer the fastest-growing province in Pakistan. While still contributing 28 percent1 to national GDP—higher than its share of the total population, at 24 percent—economic growth in Sindh over the past 7 years has fallen behind both Punjab and Khyber Pakhtunkhwa (KP). Although some social indicators in Sindh have improved over the past decade, they lag behind those of comparable countries and even other provinces in Pakistan. Sindh faces major developmental challenges. Sindh’s weak social indicators are partly the result of the inadequate reach and low quality of public service delivery. This in turn points to inefficiencies in provincial government expenditures on both physical infrastructure and public services. While there are many reasons behind poor public sector performance in the province, two key issues stand out: first, the total resource envelope of the Sindh province needs to be increased to meet the financing needs and, second, the quality of public spending both in terms of allocative and operational efficiency could be significantly improved. This Public Expenditure Review focuses on provincial finances and their utilization with the objective of identifying possible reforms to expand the resource envelope and ensure better value for money by improving the management and efficiency of public spending. It includes two important components: (i) a detailed analysis of the major revenue challenges and the various expenditures, including development spending and; (ii) and in-depth assessment of how some of the key government priorities are undertaken, such as education, health, and social protection. The hope is that by better understanding the constraints, reforms can be designed and implemented to maximize Sindh’s potential and promote a more equitable and productive path.2 2. Expanding the Resource Envelope Significant additional resources are now available from the federal government but these should be complemented with increased revenue from Sindh’s own-source revenues (OSR). The resource envelope from the federal government increased significantly following the 7th NFC award in 2010, adding about one-third to Sindh’s total previous revenue. While this is a welcome and much needed 1 World Bank staff estimates. 2 The public expenditures, which are analyzed in this report, only cover registered outlays in the budget. 7 Sindh: Public Expenditure Review 2017 development, on its own, it may not be enough to increase the quantity and quality of provincial infrastructure and public services required to make Sindh a fast growing province. The Sindh government’s total expenditure averaged 6.7 percent of provincial GDP in the period FY08- 15, which is insufficient given the multiple service-delivery responsibilities and infrastructure gaps at the provincial level. The Ninth NFC Award is not expected to make any large-scale increase in transfers for provinces. As such, the Sindh government needs to focus on maximizing its own-source revenue (both tax and non-tax revenue) collection. Up to now, own-source revenue (OSR) remains very low despite huge potential for generating income from property taxes and general service taxes (GST) on services. Improved collection of OSR will help to generate much needed additional resources to plug social sector and infrastructure gaps. There is considerable potential to expand OSR, starting with the Sales Tax on Services. Now that the Sales Tax on Services (STS) is collected by the province, STS revenues have soared, from just PRs 1.1 billion in the FY11 collection to PRs 64 billion in FY16. This is a very positive step and shows the way forward for other potential sources of OSR. However, more can be collected from STS, as the services brought under the STS umbrella so far are only those that have well-defined bases and are easy to tax and collect. Although the Sindh government has now imposed STS on construction, personal care, and professional services, the wholesale, retail, and transport sectors, remain almost entirely outside the tax net. This is because they comprise a large number of small units that are difficult to identify and where access to accurate sales records is problematic. Expansion of the STS tax base to include more challenging sources of revenue will require innovative thinking and management, which in turn will call for attention to be given to improving the Sindh Revenue Board’s manpower capacity and technology. In addition, there are opportunities for increased revenues from other provincial taxes. First among these is the Urban Immovable Property Tax (UIPT), which has significant untapped revenue potential. This tax is levied on property owners but the tax base is badly eroded by numerous exemptions. For example, all government property is exempt from payment of UIPT, as are the properties of religious, charitable and educational institutions. In addition to this erosion of the tax base, UIPT is marred with a large number of structural and administrative issues. These include the gross under-valuation of the annual rental value (ARV) of properties, on which the UIPT is based, and the way in which owner-occupied properties can pay just 10 percent of the normal tax liability. Both these issues create incentives for rampant tax evasion, including commercial properties pretending to be residential and thus paying just 10 percent of the normal commercial property tax rate. Second, Agriculture Income Tax (AIT) also has huge potential, although up to now revenue collection has remained woefully low. If AIT were to be properly applied on all farmers using its “income mode�, the potential revenue would be about PRs 4 billion, or 12.5 times the revenue collected at present. If the “land mode� were used the revenue potential would be even higher. There is also potential to tap user charges to enhance the resource envelope. The Sindh government has not exploited user charges to enhance the resource envelope. User charges as a share of OSR declined from 14 percent in FY97 to less than 1 percent in FY16, which means that the government is effectively giving an “untargeted subsidy� to services that should normally be charged some tax or a fee. An argument could be made to keep user charges on services low (or even at zero) for those services used by the poor, such as primary and secondary education, primary health care, rural water supply and sanitation. However, economic services (such as irrigation, roads, etc.), and tertiary level social and community services could be more adequately charged. 8 Sindh: Public Expenditure Review 2017 3. Improving the Efficiency of Public Spending Although the Sindh government has increased its allocations for the key priority sectors in recent years, much of this additional funding has gone towards increased pay, allowances, and pensions of government employees. This means that while total allocations may seem to be increasing handsomely, in fact spending on key inputs into the social sectors remains roughly the same and, as a result, there is little, if any, impact on service quality or coverage. In addition, there are instances where the spending is not on the right kinds of human resources. For example, while there is sizeable pool of secondary school teachers, there continues to be significant shortages of subject specialists, particularly for mathematics and science (see Section 3.7). The large and frequent wage increases led to a sharp increase in wage (and other employee-related) expenditure, which can erroneously be interpreted as adequate spending on high-priority sectors such as provincial infrastructure and service delivery. The reality, however, is that not only significant gaps remain, but also that much higher level of budgetary resources be required in the future to narrow these gaps. Fiscal rigidities have increased. These wage increases heighten budgetary rigidities, as both wage bill and pension payment increase sharply. This makes future budgetary adjustments even more difficult. While the government acknowledges the inadequacy of allocations for operational and maintenance (O&M) expenditures, the common solution adopted to remedy this budgetary shortcoming is to allocate, every now and then, a block allocation earmarked for O&M. The inadequate allocation of O&M budgets leads to public infrastructure erosion at a much faster rate than if adequate O&M allocations were made on time. As such, much higher level of fiscal resources (provided through block allocations) are required to restore the quality and efficiency of infrastructure. 3.1 The Allocative Efficiency at the Macro Level Is Good The Sindh “Vision 2030� outlines the priorities of the Sindh government. It is organized according to the thematic areas of poverty, health, education, employment, effective governance, land, water, and infrastructure. This PER analyzes the extent to which the allocation of funds is in accordance with the priorities set out in the “Vision 2030�. With a significant increase in resources, both from federal transfers and OSR, the Sindh government has had extra fiscal space to enhance expenditure. While a significant portion of this increased fiscal space was consumed by government employee-related spending, the Sindh government was nonetheless able to increase spending on development as well. Government development expenditures grew 2.5 times in real terms in the 6 years from FY09 to FY15, while recurrent expenditures increased 1.7 times in real terms in the same period. This meant that development expenditures as a share of total expenditures increased from 13 to 30 percent in this period. Nonetheless, government employee-related salaries and allowances have become the largest component of recurrent expenditures, rising from 33 percent in FY12 to 46 percent in FY15. Similarly, government employees’ pensions increased from 7 percent of recurrent expenditures in FY12 to 10 percent in FY15. These increases, which are of a permanent nature, imply that employee-related expenditures will continue to take a sizeable portion of provincial fiscal space in foreseeable future, limiting the room for new developmental initiatives that the provincial government may wish to undertake. In addition, data reveal that these increases in employee-related expenditures have had little or no impact on the quality or quantity of service delivery, or on developmental outcomes. Despite these increasing budget rigidities, the government was able to increase its operating expenditures. The Sindh government has made progress in its policy goal of increasing spending on operating expenditures, which grew by an annual average of 16 percent from FY09 to FY15, only 9 Sindh: Public Expenditure Review 2017 marginally lower than the annual rate of increase in overall recurrent expenditures, at 18 percent. These expenditures were allocated to provide better communications, transportation, office consumables, and utilities—expenditure that is generally associated with the quality of public services. Spending on repairs and maintenance has increased, but levels are still very low. Repair and maintenance funds were used mainly for roads and buildings, plant and machinery, especially in hospitals, vehicles, and furniture and fixtures. The share of these expenditures in overall recurrent expenditure increased from 1 percent in FY09 to 3 percent in FY15. Despite this increase, repairs and maintenance spending remains far below what is needed and the international benchmarks. This implies that the provincial infrastructure and other assets continue to be eroded at a rate faster than optimal. This adversely affects the operational life of public infrastructure and undermines the efficiency of public services. Expenditure on all the social sectors averaged about one-third of total recurrent expenditures in the 6- year period between FY09 and FY15, but the share is subject to volatility because of the ad-hoc nature of much social protection spending. Recurrent expenditures on education and health increased, while recurrent expenditures on social protection decreased during this period. Over the period, recurrent expenditures on the social sectors increased 3.3 times in nominal terms and 1.9 times in real terms, with the highest increase in health, followed by education. Education now consumes about one-quarter of total recurrent expenditures. In real terms, public spending on education grew at an average rate of 11 percent per annum between FY09 and FY15, slightly faster than overall real recurrent expenditures, at 9.5 percent annually over the same period. This increase in education sector spending has not been uniform over time. After the passage of 18th Constitutional Amendment, the Sindh government started to increase resources for education and health. Expenditure on education was just 17 percent of total recurrent expenditures in FY12, but this share jumped to 25 percent in FY13, and has remained more or less at this level subsequently. The health sector is ranked fifth in terms of average resource allocation by the Sindh government over the 6-year period between FY09 and FY15. Health recurrent expenditures increased from 7 percent of total recurrent expenditures in FY09 to 9 percent in FY15. Much of this spending was allocated to health operating expenses, including spending on medicine, repairs and maintenance, and petroleum, oil and lubricants (POL) for generators. 3.2 There Are Issues of Allocative and Operational Efficiency within the Sectoral Level As mentioned above, in terms of the macro-level allocations for the key social sectors, the Sindh government has increased spending on education and health, but a closer look at disaggregated data reveals issues in regional and needs-based allocations. In addition, further analysis reveals no, or a very weak, relationship between public expenditures and sector outcomes, pointing to low levels of operational efficiency. The following sections provide a brief overview of the operational efficiency issues as they affect the three main social sectors. 3.2.1 Education Spending Inefficiencies There is considerable inefficiency in primary school distribution. Ideally, there should be a strong positive relationship between the number of schools in a district and the number of students, but in Sindh, this is not the case. For instance, Mitthi ranks first in terms of the number of schools in a district (4,008 schools), but ranks 15th in terms of the number of students in the district (with an average of 37 students per school). Larkana, on the other hand, has the highest number of students per school (187). This suggests that schools are not optimally located to cater to demand. Furthermore, an analysis of 10 Sindh: Public Expenditure Review 2017 primary school proximity indicates that 48 percent of boys’ primary schools and 41 percent of girls’ primary schools are located within 1 kilometer of the nearest neighboring primary school of the same gender. Per-student expenditure is skewed towards relatively wealthier urban districts due to more teachers being located in those districts. Analysis shows that annual per-student expenditure is relatively skewed towards few districts in Karachi Division and Hyderabad district, explaining why the two cities have the lowest student-teacher ratios (STRs), of 18 and 19, respectively. In contrast, annual per-student expenditure in relatively poorer districts, such as Ghotki, Badin and Tando Allah Yar, is less than one- quarter of the expenditure in Karachi. Ghotki, with the lowest per-student expenditure in the province at just PRs 14,673 per student per year, also happens to have the highest STR in the province, at 40 students per teacher. This clearly highlights the inter-district misallocation of teaching staff, which has a negative effect on education sector expenditure, as STR at the two extreme of the developmental distribution of districts is definitely sub-optimal. Most primary schools lack basic facilities. Despite a strong focus and commitment over the last many years to “complete� primary schools in the province by providing them with basic facilities, the analysis suggests poor state of basic facilities in many primary schools in Sindh. For example, in the Tharparkar (Mitthi) district only 2 percent of primary schools have all four basic facilities,3 followed by the district of Thatta (3 percent). Urban districts such as Karachi and Hyderabad have the largest share of primary schools with all four basic facilities, at 50 and 47 percent, respectively. Overall, there are 12 districts in Sindh where more than 75 percent of primary schools do not have all four basic facilities. While most teachers have at least a bachelor’s degree, far too few teachers have science backgrounds. With the introduction of merit-based teacher recruitment in Sindh, the average qualification of teachers has improved. Eighty percent of teachers now have a bachelor’s or higher academic degree. However, only 5 percent of public school teachers have sufficient science or mathematics qualification to teach these subjects to primary school students. The education system spends a large share of its budget on the salaries of teachers with higher qualifications. This shortage of qualified teachers to teach the two compulsory and critically important courses implies that the students may not be getting the adequate level and mix of primary school education for these expenditures. There is no, or at best a weak, relationship between learning outcomes and public expenditures at the district level. Analysis of total expenditures by district and the learning outcomes of Grade 5 and 7 students on the Standardized Achievement Test (SAT) indicate that there is no apparent relationship between the two. For instance, the overall score for Grade 7 in Mirpurkhas is 28 percent with annual per- student expenditure of PRs 15,000, and in Sukkur, the score is 21 with annual per-student expenditure of PRs 21,000. Even in Karachi, which has the highest per-student expenditure on education, the average SAT score is only 24. Most importantly, the effectiveness of education expenditure appears to be on a decline. Analysis of the effectiveness of public education expenditure (the relationship between inputs and outputs) achieved by matching data from household surveys at the district level also points towards declining levels of efficiency in public education spending. This analysis indicates that average enrolment in public schools in different districts over time has failed to improve, despite a four to fivefold increase in per- capita expenditures. This provides the clearest evidence that the effective utilization of resources is 3 The four basic facilities required by schools: (i) a fully enclosed boundary wall, (ii) functional toilets, (iii) electricity, and (iv) drinking water. 11 Sindh: Public Expenditure Review 2017 failing to have an impact on educational outcomes and that the education sector is facing huge challenges in improving quality despite higher levels of funding. 3.2.2 Health Spending Inefficiencies The pace and breadth of improvements in health indicators have fallen short of expectations, due to operational inefficiencies and unresponsiveness to public health needs. The challenges facing the health sector in Sindh are already daunting, with its high rate of population growth overshadowing the capacity and ability of public health infrastructure to provide quality services and guaranteed access. Despite being a main target of government policy, and rapid increase in sectoral expenditure, the health sector has not fully succeeded in reducing the urban-rural disparity in access to health services. As such, the huge urban-rural disparity in health indicators in Sindh continues to be a major concern that requires immediate attention in order to improve the health outcomes across the province. Failure to achieve that will imply continue adverse health outcomes, especially for rural households, which comprise more than half the population of the province. A part of the problem is the inability of the health sector to appropriate deploys and retains human resources of acceptable quality and capacity in rural areas, particularly female health staff. There is a chronic shortage of all categories of female staff and specialists in rural areas, while at the same time the overall sectoral employment is excessively bloated with staff belonging to general cadre and support staff at all levels. This imbalanced staff recruitment and deployment is a prime factor for low efficiency of provincial health sector. There are major gaps in primary health care coverage even for the urban poor. Karachi has the highest population growth rate in Pakistan, at 3.2 percent per year. The poor living standards in slums need publically financed support if they are to have access to primary health care. With increasing urbanization (and burgeoning urban slums) and an ever increasing pressure on existing health infrastructure and services, major gaps in the coverage of primary health care services have developed, resulting in poor health indicators and repeated outbreaks of infectious disease, including polio. Weak governance and management is the one underlying issue behind all the ills that plague the health sector. The Health Department in Sindh is faced with an over-centralization of authority, too many vertical programs, and a lack of mechanisms to enhance accountability and transparency. There is also weak oversight of service delivery and no clear M&E framework. Improving sector governance is now all the more important in the post-devolution environment, given that the focus of the mandate has shifted from purely one of service delivery to include the new challenge of the stewardship of the entire health sector. Real health sector spending is low, resulting in high out-of-pocket (OOP) expenditure, which generally is very regressive. Although primary health care provides a cost-efficient means for disease control and management, this has not been the focus of increases made in health sector financing. There are also serious inefficiencies within the public sector, due to the proliferation of parallel vertical programs, redundant posts, an overly ambitious but sub-optimal district health system, and an input-based financial system. Meanwhile, about 66 percent of total health spending in Sindh is financed through out-of-pocket expenditure made by households with the rest mainly accounted for by public sector sources. Based on households’ ability to pay, OOP spending is highly regressively, with the bottom income quintile spending on average 7 percent of monthly household income on health care, compared with the provincial average of 5.2 percent across all quintiles. Medicines account for the largest share of OOP spending, mainly due to their inappropriate prescription by the private sector and their low availability in the public sector outlets. 12 Sindh: Public Expenditure Review 2017 Within the public sector, the non-salary operational budget at the district level is particularly inadequate, ranging between 17 to 25 percent. Meanwhile, allocations for development expenditures at both the provincial and district levels are heavily skewed towards facility construction, which has high operational cost implications for coming years. 3.2.3 Social Protection and Labor Spending Inefficiencies While the government seems aims at protecting the poorest of the poor through social protection interventions, there is a no coherent strategy or a targeting mechanism. Currently, Sindh SPL initiatives are conceived, planned, and executed in a piecemeal fashion, scattered across several line departments and autonomous governing bodies. There is no central entity to plan, consolidate, coordinate, and monitor these SPL interventions. Consequently, the SPL sector is replete with duplication, including between federal and provincial programs, inadequate coverage, and major programmatic inefficiencies. There is little or no meaningful consideration of pre-identified SP risks and vulnerabilities, nor are the SPL programs sustainable at current levels of funding. Therefore, a coherent social protection and labor strategy and implementation program is urgently called for. SPL programs have a limited outreach focussed on a few geographical areas, and cover only a fraction of Sindh’s vulnerable and at-risk population. The number of beneficiairies and services/benefits being offered are too few or too small to have any meaningful impact. In FY15, the most recent year for which data are available, at most only 1.3 million people benefited from SPL intervetions, a meer 2.5 percent of Sindh’s population and only 7.5 percent of Sindh’s poor population. Sindh’s SPL sector is primarily driven by untargeted regressive subsidies, benefiting the non-poor more than the poor due to the former’s greater purchasing power. Most SPL programs either do not target the poor at all, or rely on very loosely defined community-targeting mechanisms. The absence of a central coordinating agency and robust analysis of SP-related risks and vulnerabilities has resulted in a weak understanding of the role of SP interventions. This in turn has led to a plethora of inadequate and inequitable SP programs that have little impact. Besides its low budgetary allocation, the SPL sector has been further weakened by poor utilization due to weak institutional and administrative capacities. The overall budget utilization pattern against yearly allocations for SP shows regular underspending. The lowest utilization was just 29.7 percent in FY13, although there was a subsequent gradual improvement to 51.8 percent in FY14 and 81.7 percent in FY16. There is a lack of a long-term focus on SPL policy. The absence of any long-term SPL vision clearly articulated through SPL policy, together with the absence of a resultant strategy, has led to short- termism with regard to issues such as human capital development that in fact require long-term investment. This in turn has resulted in a piecemeal, ad-hoc, and “needs-based� budgetary allocation process and, despite the benefits being insufficient in the first place, allocations have remained largely under-utilized due to low capacity. 3.3 Budget Credibility and Fiscal Risks The credibility of the budget is low. In addition to the inadequacy of some allocations and the low overall efficiency of public spending, one additional challenge—impacting the transparency of the government’s fiscal operations and the implementation of government policies and programs—is the questionable credibility of the budget. Budget credibility in Sindh is severely undermined by the misuse of contingency budgetary instruments. Pakistan’s budgetary system does not allow any allocation for contingency expenditures. Instead, it allows central and line departments to change the approved budget allocations in response to “emerging situations� provided these changes are to “enhance the efficiency of public spending�. The line departments are allowed to re-appropriate funds from items in their approved 13 Sindh: Public Expenditure Review 2017 budgets for allocation to other items if there is a sound reason to do so. Similarly, if any line department feels that the budget allocated to it is insufficient to meet its ongoing or emerging expenditure needs, it can ask for a supplementary budgetary grant, which can be granted if approved by the Finance Department (and if approved by the Planning and Development Department [P&DD] in the case of the development budget). Excessive use of budget re-appropriations and supplementary grants reduce fiscal discipline, thwart efficiency, cloud the transparency of the budget, and undermine the right of the provincial assembly to approve budgetary allocations. There are significant downstream fiscal risks for the government emanating from large scale increase in employee-related expenditure and servicing liabilities of provincial debt. The significant increase in government employee-related expenditure on salaries, allowances, and pensions, has also increased rigidities and fiscal risk. The rapid increase in pension liabilities of the Sindh government over the 6 years from FY09 to FY15 has meant that spending on pensions almost doubled, both in absolute terms and as a ratio of total recurrent expenditures. If the current rate of increase in pension spending continues, which is likely, it will rise from 22 percent of employee-related spending to about 48 percent in 15 years’ time. Besides pensions, the General Provident Fund4 has become another significant liability of the Sindh government, as it is now the second-largest component of the government’s total debt. The General Provident Fund (GPF) liability that was PRs 12 million in FY72, PRs 836 million in FY87, and PRs 6.5 billion in FY97, and jumped to over PRs 100 billion in FY14. The GPF liability of the Sindh government increased fivefold from FY02 to FY15. Moreover, in FY07, GPF liability was about one-quarter of the total debt of the Sindh government and, in FY15, this liability increased to 35 percent of total debt. If this trend continues, which can be expected, GPF liability will reach PRs 228 billion by 2030. This unfunded nature of the GPF poses another significant risk to the sustainability of public finances in Sindh. 3.4 Development Expenditures: Getting Better Value for Money While development expenditures have increased, so have the inefficiencies in development budget and expenditure. With a significant increase in resources, both from OSR and federal transfers, the Sindh government has used the extra fiscal space to enhance development expenditures. Government development expenditures grew by 2.5 times in real terms in the 6 years from FY09 to FY15. However, systemic and transitional inefficiencies have also increased, as the system’s capacity to adequately plan, program, execute and monitor has not increased in line with the large increase in the development portfolio. Moreover, there is growing political interference in the identification, preparation, approval and execution of development projects, further weakening the Public Investment Management (PIM) System and undermining the efficiency of development expenditure. In particular, planning and execution of development projects have been found wanting, with major time and cost overruns. Weak planning and implementation of development projects have led to a significantly lower spending than envisaged in budget allocations. Higher allocations indicate that the Sindh government wants to focus resources on major infrastructure, with the goal of spurring economic growth and creating jobs. However, these allocations do not guarantee the desired outcomes, especially in the absence of improvements in the management of these investments, such as the efficiency of allocations and operations through better project selection and implementation. As a result, budget utilization has been very weak. Budget execution is also weak and, in some years, only half of the planned expenditures are undertaken. Therefore, a complete assessment of public investment management (PIM) is required as a prerequisite to focusing on getting better value for money from these investments. 4 GPF, or the General Provident Fund account, is a provident fund account that is available for government employees. A government employee can become a member of the GPF by contributing a certain percentage of their salary to the account. 14 Sindh: Public Expenditure Review 2017 4. Recommendations Sindh’s developmental challenges are undoubtedly huge, but so are the opportunities. The benefits of achieving success in using public funds to help turn Sindh into a regional economic powerhouse that can offer employment and support to all its citizens are huge. Sindh is the most urbanized province in Pakistan and is already a hub of economic activities. In order to achieve its full potential, the Sindh government will need to focus on resource mobilization and enhancing the efficiency of expenditures. The following are some recommendations and policy measures for achieving this goal. The chapters discuss these recommendations in more detail. Key Areas to Reform/measure in the short term Reform/measure in the medium to long Expected Results be (within 12 months) term (12 months to 3 years) strengthened Fiscal and Financial Management Revenues 1. Each level of government to take a 1. Bring fiscal transfer mechanism much broader and longer-term view of into line with the needs of both levels revenue needs and potential collections of government. at the 9th NFC Award. Property Tax 2. Expand the own source revenues 1. Improving policy and administration of the provinces by having more realistic valuation rolls, 3. Revamp property tax to raise 2. Rationalization of exemptions and collection tax rates, levying tax on vacant properties, 4. Enhance revenue yields of 3. Indexing the tax base, and reducing agricultural income and equity in the the differential in rates between owner- tax system. occupied and rented properties, and 5. Broaden the base of GST on between residential and commercial services properties. Agriculture income Tax 4. Rationalize exemptions, 5. Indexing the tax base (especially of land tax) 6. Enhancing capacity of the BOR to enable it to better assess tax from the income mode, GST 7. Gradually bring commercial and transportation services into the tax net, 8. Merging the professional tax into the GST on services, together with as many stamp duties as possible. Recurrent 2. Undertaking a complete stock-take 6. It is to ensure that sufficient Expenditures of public physical assets and resources are available for maintaining infrastructure, noting attributes such as existing infrastructure so that it pays age and usage, and updating this off in the longer term. database annually. 3. Drawn up a district-level requirement for repair and maintenance expenditures and link it with between development expenditures and recurrent budgeting to better understand the future spending needs for repairs and maintenance. 4. Prepare an HR assessment of the 7. It is to make public sector public sector to better understand the remuneration system more skills available, the future HR needs, and transparent and improve the 15 Sindh: Public Expenditure Review 2017 the skills required to fill those jobs. credibility of budgetary framework. 5. Undertake an actuarial analysis of the pension system to better understand pension costs in the medium to longer term, and use this analysis in the medium-term budget framework to improve its credibility. 9. Prepare sector strategies and 8. Align budget allocations with budgets according to required outputs medium- to long-term sectoral goals. and outcomes of sectors, 10. Strengthen the coordination mechanism between recurrent and development budgets i.e. allocate a one- line budget to line departments without hard distinctions between two, and allow reallocation between recurrent and development budgets based on the approved sector strategy. 11. Establishment of a Fiscal 9. Assess, monitor, forecast, and Management Framework (FMF) to report the whole range of fiscal undertake an evaluation of fiscal cost of commitments and contingent PPPs. liabilities5 that can be created by 12. Strengthening the SOE monitoring PPPs. mechanism and establishing an up-to-date database. 6. Prepare development framework 10. Translate ambitious objectives establishing broad fiscal parameters, and targets of Vision 2030 into development objectives, and priority actions. programs. Carry out consultations with the legislature and the public dissemination of a draft document for public feedback within a given timeline; 7. Update and approve a development framework from the Cabinet; 8. Develop sector strategies in line with the development framework; and identify priority projects and prepare a consolidated portfolio of pipeline projects. 13. Prepare development budgets using 11. Ensure transparency and the New Accounting Model (NAM) chart completeness of budgets relating to all of accounts for all public investment public investment projects projects. 14. Development budgets should be appropriated across all relevant elements 6 of the chart of accounts . 15. Implement a “Revised accounting procedure for Revolving Fund Accounts (RFA)� for both development partner- and GoS-funded projects. The rules require that project authorities must reconcile and report expenditure to the Accountant General’s (AG) office on a monthly basis, 5 We refer to the explicit contingent liabilities, namely those that are subject to a contractual engagement. 6 Currently budgets for most projects are lumped into only two items: salaries and operating expenses. 16 Sindh: Public Expenditure Review 2017 thus enabling detailed accounting of the 7 RFA funds in the GFMIS. 16. Develop an integrated database for public investment projects to have a consolidated view of the progress of public investment projects within the DG M&E in the P&DD. 17. Prepare an asset safeguard policy and plan to develop a comprehensive non-financial-asset register at the P&DD. 18. Develop an integrated PIM information system (PIMIS) and implemented, with an interfaced created with the PIFRA system for financial data and a PIM reporting framework developed. A PIM monitoring committee should also be established within the P&DD to review development progress and to issue directives for project authorities. 9. Use a criteria-based approach for 12. Ensure systematic decision- project screening and project approvals. making, planned selection and enable 10. P&DD to issue an evaluation policy makers to evaluate the framework that will serve all decision opportunity costs of projects. nodes in the project screening process. 11. Develop protocol for early and extensive engagement of various stakeholders in development budgeting. The stakeholder spectrum will include legislatures, civil society members, media, etc. 12. At the project level, P&DD should review the “Manual for Development Project� in light of emerging requirements for projects. 13. Develop guidelines to comply with PC-1 requirements. 14. Provide legal cover (P&DD 8 notification) so that DG M&E can ensure compliance with PC-III (cash flow projections), PC IV (project completion) and PC-V (monitoring project outcomes). Education 19. Rationalize the recruitment policy 13. An adequate and equitable 20. Ensure outcome-based education distribution of teachers, particularly at expenditures; and the primary level 21. Review school specific non-salary 14. Ensure that funds are available in budgets (SSB) and school management a timely manner and are effectively committee (SMC) interventions utilized. Health 22. There are several positive programs underway, including the introduction of subsidized user charges, and results-based financing in disadvantaged districts, with 7 RFA is a designated bank account (notional) under the Treasury Single Account (TSA) concept generally used for large projects, where single line transfer is made into the RFA and, thus, no break downs of the budget or expenditure are available in the GFMIS. The detailed accounting is undertaken by the project authorities. 8 DG M&E is part of P&DD responsible for project monitoring 17 Sindh: Public Expenditure Review 2017 support from independent monitors. 23. Successful evidenced-based pilots of the LHW program to provide neonatal care, childcare, and nutritional support have been implemented in rural Sindh, but the lessons learned have yet to be mainstreamed. Moreover, alternative outreach strategies are needed for remote areas not covered by LHWs. 24. Pilot programs using GPs in urban areas have been launched by NGOs covering NCDs, TB DOTS, family planning, and STIs and have proved successful. However, once again the lessons learned from these pilots have yet to be scaled up into a systematic strategy. 25. Develop a comprehensive pharma strategy to avoid issue like proliferation of shadow pharmacies, inappropriate prescriptions by medical practitioners, and the relatively low use of recommended and cost-effective generic medication, as well as the supply management issues, such as a high frequency of stock-outs, low quality parameters in purchasing, and a lack of transparent checks in the management of inventories. 26. Regulatory work has been initiated by the Sindh Blood Transfusion Authority in the area of registering blood banks. Similar work also needs to be extended to diagnostic facilities, private health service providers, and shadow pharmacies, and will require the setting-up of a new regulatory authority. Social 27. Existing SP programs/interventions 15. Under an appropriate legal 15. Consolidate and coordinate the Protection need to be reviewed for an increased framework, establish a central SP entity, SP sector under a unified policy focus on the marginalized groups (i.e., such as an SP Authority or Board, framework, duly supported by a rural poor, women) and alignment with entrusted to consolidate, coordinate and responsive institutional and key policy objectives. monitor SP portfolio in Sindh. implementation arrangement. 28. New interventions aimed at reducing 16. Initiate the evidence based SP policy 16. Equitable growth (lessening the inequality and promoting equity, such as formulation process, either through the capability deprivation) should be the CCTs for health and education, and labor- newly formed SP entity, or in the interim, focus of Sindh SP policy reform market programs designed for the most under the P&DD, ensuring adequate agenda. marginalized, should be introduced with a stakeholder consultation, analysis of 17. Rationalization of general particular focus on human capital risks and vulnerabilities, clear subsidies and, possibly provide development. articulation of SP vision; identification of targeted subsidies to the most 29. All cash-based SP interventions key policy objectives (such as equity, vulnerable, while using the freed-up should be brought under a unified resilience, social cohesion, gender equity resources towards core SP programs. targeting and disbursement mechanism. and opportunity) and prioritized 18. Increase coverage to the poorest 30. Review the distributional impact of intervening areas. of the province. untargeted subsidies across economic group quintiles. Develop an action plan to phase out untargeted subsidies based on outcomes of the analytical review. 31. Develop mechanisms for better targeting of subsidies that cannot be 18 Sindh: Public Expenditure Review 2017 phased out. Explore possibilities for targeting through the PMT score database (NSER) 32. Review and analyse the targeting mechanisms of existing SP interventions and introduce unified targeting through the national poverty database (NSER). 33. Develop data-sharing and coordination mechanisms between the provincial and federal governments and, subsequently, with provincial line departments. 19 Sindh: Public Expenditure Review 2017 Chapter 1: Macroeconomic and Fiscal Trends 1.1 The Socioeconomic Profile of Sindh Sindh Province is the most urbanized of the four provinces that comprise Pakistan and is endowed with many of the characteristics of a high-growth region. The province has a large population of about 49 million—roughly one-quarter of the population of Pakistan—and is strategically located on the coast with access to the Arabian Sea. This means that Sindh has the potential to become one of the largest logistical and business hubs in South Asia, with its two major ports of Karachi and Bin Qasim. Karachi, the largest city in Pakistan, is the main financial and commercial hub of the country, and the city therefore makes the largest contribution of all Pakistani cities towards national GDP. While Sindh is the most industrialized province in Pakistan, it is also resource-rich and endowed with Pakistan’s largest natural gas and coal reserves. However, despite these attributes, Sindh has yet Figure 1.1: GDP growth rate (%): Sindh and Pakistan to succeed in converting its potential into 6.0 Sindh Pakistan commensurate economic and social 4.9 5.0 development. In fact, over time Sindh has ceded 4.1 3.9 3.8 3.9 4.0 its status of the fastest-growing province in 4.0 3.6 3.7 Pakistan to Punjab. While still contributing 28 3.1 3.2 9 3.0 2.6 percent to national GDP—higher than its share in the country’s total population, at 24 percent— 2.0 1.6 economic growth in Sindh over the past 7 years has fallen behind that of Punjab and Khyber 1.0 0.4 Pakhtunkhwa (KP). Over the past 10 years, the 0.0 provincial GDP of Sindh has increased at an annual average of 3.7 percent. This partly reflects -1.0 -0.4 FY09 FY10 FY11 FY12 FY13 FY14 FY15 the general slowdown of the national economy (which grew at an average rate of 3.8 percent during 2009-15), and partly the result of natural Source: Data from the Government of Sindh. disasters that the province has had to endure, including devastating floods in 2009-10 and 2010-11 (Figure 1.1). Social indicators in Sindh, as in the rest of Pakistan, have improved over the past decade, but still lag behind those of comparable countries and regions. In many cases, Sindh performs below the national average and the rate of improvement has almost stalled. Being the most urbanized province in Pakistan, Sindh has relatively better access to piped tap water and better access to assisted births for expectant mothers, but other development indicators are not so positive. Urban Sindh is similar to the urban national levels for most indicators but, in contrast, rural Sindh is well below rural national levels across all indicators. 9 World Bank staff estimates. 20 Sindh: Public Expenditure Review 2017 Figure 1.2: Selected social indicators, 2014-15 120 Rural Sindh Urban Sindh Rural national Urban national 94 97 100 74 76 80 69 70 62 60 40 40 33 31 28 24 20 0 Child immunization GER matric (14-15) GER primary (6-10) Flushing toilet Female literacy Literacy (full, records) Source: PSLMS 2014-15 For example, the percentage of births handled by a skilled health worker is highest in Sindh, at 60.5 percent compared with 52.1 percent for Pakistan as a whole. However, the infant mortality rate, at 82 per 1,000 live births, is also the highest in Pakistan (average 74 per 1,000 births), indicating the poor performance of mother and infant health services in the province. The literacy rate (for the population of 10 years old or older) was only 60 percent in 2015, lower than 63 percent for Punjab, but higher than 53 percent for Khyber Pakhtunkhwa and 44 percent for Baluchistan. The gross enrolment rate of Sindh is below the national average and is the lowest among all of Pakistan’s provinces. The ratio of out-of-school children aged 6-15 has also failed to decline from a level of above 35 percent over the past decade. Meanwhile, the contraceptive prevalence rate is only 30 percent compared with 35 percent for Pakistan as a whole (Figure 1.2). In addition, gender and urban-rural disparities continue to be very wide, with huge gaps in social indicators between urban males and rural females, and little improvement seen over the past decade. For example, in FY05, the primary school gross enrolment rate (GER) for urban males was 103 percent, while that for rural females was just 54 percent—a huge gap of 59 percentage points. Similarly, the literacy rate varies from 62 percent for urban males to just 28 percent for rural females. To better understand these wide gaps, we have disaggregated them into their regional (urban-rural) and gender components (Figure 1.3a). This breakdown indicates that regional disparities are more pronounced than gender disparities. For example, from a total gap of 59 percentage points in primary school GERs between urban males and rural females, 40 percentage points are due to regional inequalities, while the remaining 19 percentage points are accounted for by gender disparities. A similar weighting towards regional inequalities holds true for other social indicators. 21 Sindh: Public Expenditure Review 2017 Figure 1.3a: Sindh social disparities, FY05 Figure 1.3b: Sindh social disparities, FY15 Gender Regional Total Disparities Gender Regional Total Disparities 70% 70% 62% 62% 59% 60% 60% 46% 50% 19% 28% 50% 24% 40% 28% 40% 17% 28% 30% 30% 0% 20% 40% 34% 31% 20% 38% 10% 29% 28% 10% 0% -3% GER (Primary) Literacy Immunization 0% -10% GER (Primary) Literacy Immunization Source: World Bank staff calculations based on PSLMS data. The regional and gender disparities have narrowed over time, but not to the desired level. Over the past 25 years, public policy has focused specifically on reducing regional and gender inequalities, with some success. Although there has been significant variation in the progress made in narrowing regional and gender disparities across social indicators, on average regional disparities have narrowed more than gender disparities. By FY15, for example, the overall gap between urban males and rural females dropped from 59 percent in FY05 to 46 percent, i.e., by 14 percentage points. The regional disparity in primary school GER narrowed by 11 percentage points, whereas the gender disparity declined by only 2 percentage points (Figure 1.3b), while there was no change in the gaps in the literacy rate and the immunization rate. However, the components of the gaps have changed: for the literacy rate, the regional gap increased, while gender inequalities declined, whereas for the immunization rate the situation was reversed. Despite the poor performance in social indicators, significant progress has been made in poverty reduction over the past 12 years. Based on the FY14 Household Income and Expenditure Survey (HIES), the poverty line for Pakistan was estimated at PRs 3,030.32 per adult equivalent per month, which translated into a national poverty incidence of 29.5 percent in 2014. In absolute numbers, and based on population estimates of 180 to 200 million in FY14, some 6.8 to 7.6 million households (or 53 to 59 million individuals) were classified as poor. Based on these new poverty metrics, about 34.2 percent of the population in Sindh lives in poverty. Sindh’s poor social indicators are both an outcome of, and contributor to, high poverty levels in the province. The high incidence of poverty and low private-sector participation in service delivery together give rise to poor social indicators, as a result of the inadequate reach and low quality of public service delivery in Sindh. This in turn points to inefficiencies in provincial government expenditure on both physical infrastructure and public services. The poor quality of public expenditure contributes to poor service delivery. There are three constraints to improving public-sector efficiency, loosely categorized as follows: (i) the inadequacy of laws, rules, procedures and processes governing the running of the provincial government; (ii) the weak administrative and technical capacity of public-sector institutions; and (iii) these two weaknesses then combine to manifest themselves in the poor utilization of public funds in terms of value for money. This Public Expenditure Review focuses primarily on the issue of provincial finances and their utilization, while 22 Sindh: Public Expenditure Review 2017 recognizing that the other two constraints have a strong bearing on the management and efficiency of public finances. 1.2 Macro-Fiscal Outcomes While there may now be additional resources, the efficiency of public spending remains suboptimal, and in some cases is actually declining. In addition to the recent natural disasters, Sindh has been prevented from achieving higher development outcomes by relatively low quantity and poor quality public expenditure, improved fiscal space notwithstanding following the 7th National Finance Commission (NFC) Award.10 Increased federal transfers, along with improved own-source revenues, especially from general sales taxes (GST) on services, provided additional resources to the Sindh government to meet the minimal investment needs of the province, and to adequately allocate resources for high-priority sectors and areas. Nonetheless, huge infrastructure gaps continue to exist and social outcomes are far below desirable levels. In addition, on top of this, in some cases the efficiency of public spending has also declined. Sindh’s fiscal management has been prudent, and on average the province has maintained a marginal fiscal surplus. On average, Sindh maintained a positive fiscal balance between FY09 and FY15. Total revenues as a percentage of provincial GDP increased to 7.1 percent in FY15, compared with only 5.6 percent in FY09. With the federal government assigned most of the broad-based taxes, Sindh’s finances are heavily dependent upon revenue transfers from the federal government. Currently, these transfers constitute about 80 percent of overall revenue available to the provincial government. With provincial own-source revenue (OSR), barely enough to meet 20 percent of provincial expenditure (Table 1.1), any shortfall in revenue adds directly to the level of fiscal stress at the provincial level. There is added pressure from the federal government on Sindh to ensure that it posts a surplus. This implies that most provincial-level fiscal management boils down to managing provincial expenditure in an environment in which revenue transfers from the federal government are uncertain. In addition, performance in collecting OSR is not working very well. Under the IMF’s Extended Fund Facility (EFF), the federal government is required to maintain a certain level of consolidated fiscal deficit. In order to achieve this, an agreement between the federal government and the provinces has been reached whereby the provinces should maintain a certain level of overall budget surplus. Table 1.1: Government of Sindh fiscal accounts (% of provincial GDP) PRs million FY09 FY10 FY11 FY12 FY13 FY14 FY15 Total revenues 5.6 5.7 6.2 6.8 6.9 7.2 7.1 Federal transfers 4.8 4.8 5.4 5.4 5.5 5.9 5.7 Revenue assignment 3.3 3.5 4.0 4.3 4.4 4.6 4.7 Straight transfers 1.3 1.1 1.3 0.9 0.9 1.2 0.9 Development grants 0.2 0.2 0.1 0.2 0.2 0.1 0.1 Provincial revenues 0.8 0.9 0.9 1.3 1.4 1.3 1.4 Sales tax on services 0.0 0.0 0.1 0.4 0.6 0.6 0.7 Other tax receipts 0.6 0.6 0.5 0.7 0.6 0.6 0.6 Non-tax 0.2 0.3 0.2 0.2 0.2 0.1 0.1 Total expenditures 6.0 5.8 5.7 6.7 7.0 7.1 7.3 Recurrent expenditures 3.9 3.6 4.1 4.8 5.3 5.1 5.4 General administration 0.7 0.7 0.7 0.8 0.8 0.8 0.8 Law and order 0.7 0.7 0.7 0.8 0.8 0.8 0.9 10 The National Finance Commission (NFC) Award is a series of planned economic programs enacted since 1951. Constituted under Article 160 of the Constitution, the Awards manage financial resources to Pakistan’s four provinces to meet their th expenditure liabilities, while alleviating horizontal fiscal imbalances. In 2010, the 7 NFC Award was enacted, which included a revision of the distribution formula. 23 Sindh: Public Expenditure Review 2017 Social services 1.5 1.5 1.8 2.3 2.2 2.1 2.1 Economic services 0.3 0.4 0.5 0.5 0.5 0.4 0.6 Community services 0.2 0.1 0.2 0.1 0.1 0.2 0.2 Subsidies 0.0 0.0 0.0 0.0 0.0 0.0 0.1 Interest payments 0.3 0.2 0.2 0.2 0.2 0.2 0.2 Others 0.1 0.0 0.0 0.1 0.7 0.6 0.6 Development expenditures 2.2 2.2 1.6 1.8 1.7 1.9 1.9 Social services 0.9 0.8 0.6 0.9 0.9 1.0 1.1 Economic services 0.4 0.5 0.3 0.6 0.5 0.6 0.5 Community services 0.8 0.6 0.5 0.4 0.3 0.3 0.3 Others 0.1 0.2 0.2 0.0 0.0 0.0 0.0 Overall balance (0.4) (0.1) 0.6 0.1 (0.0) 0.2 (0.2) Source: Data from Office of Accountant General, the Government of Sindh. Combined with these limited resources, the Figure 1.6: Health and education public expenditures spending patterns in terms of allocations and the (% of provincial GDP) weak capacity of line departments to spend 5.0 FY01-FY05 FY06-FY10 FY11-FY15 according to budget allocations all contribute 4.5 towards unsatisfactory low outcomes. Moreover, 4.0 the rigidity of current expenditures and the Sindh 3.5 government’s limited efforts to increase province OSR have resulted in low spending on physical 3.0 and social infrastructure. Public spending on key 2.5 services remains low. An inter-provincial 2.0 comparison indicates that Sindh underspends in 1.5 education and health, and its share of these 1.0 critical expenditures is even lower when 0.5 compared with other developing countries (Figure 0.0 1.6). Punjab Sindh KP Balochistan In the past few years, sizeable increases have Source: Data from Accountant General of Pakistan and Author been made in the allocations for some key public calculations. services. However, most of these were prompted by significant and frequent increases in the salaries of government employees. As elaborated in detail in the Section 1.3.1, the 7th NFC Award led to a significant increase in the fiscal resources available to Sindh, whereas the 18th Constitutional Amendment devolved additional functions from the federal to the provincial governments. These two developments led to sizeable increases in the allocations to some key public services, including both health and education. However, the bulk of these increases went towards salaries, allowances, and pensions of government employees. In real terms, the inputs for public services, such as qualified teachers and doctors, hardly changed with the increased funding. As such, large infrastructure and service-delivery gaps have continued and, in some cases, worsened due to the inadequate maintenance of roads, water supplies, irrigation and other infrastructure, leading in turn to an erosion in the productivity of the provincial economy. 1.2.1 Sindh Debt: Overview of Costs and Risks to the Current Portfolio At end-June 2016, Sindh’s total debt was PRs 267 billion, or only 1.3 percent of total public debt held by the Government of Pakistan.11 Given the historic constraints on borrowing, it is not surprising that Sindh’s debt as a share of federal borrowing is small (Table 1.2).12 In terms of PGDP (estimates), Sindh’s 11 At end-June 2016, Pakistan’s public debt stood at PRs 19.97 trillion, or 67.5 percent of GDP. 12 Provinces in Pakistan, in principle, face hard budget constraints. Clause 3, Article 167 of the Constitution prohibits provinces from borrowing without the consent of the federal government if there is any outstanding loan to the federal government. Since 24 Sindh: Public Expenditure Review 2017 debt is also modest, at only 3.2 percent. Financing costs are also meager, amounting to just PRs 4 billion in FY16, and debt service constitutes under 2.0 percent of total provincial revenues (including federal transfers). Table 1.2: Sindh debt profile, June 30, 2016 (FY16) Stock Indicators Flow Indicators PRs PRs Percent Percent billion billion Foreign Debt 51.4 94 Debt Service 12.4 Domestic Debt 16.4 6 Of which: Interest Payments 4.3 Total Debt 267.8 100 Debt Service/Revenue 1.9 Total Debt/PGDP 3.2 Debt Service/Current Expenditure 2.7 Source: Provincial authorities and World Bank staff calculations. Note: PGDP is a World Bank estimate for FY15/16 and not the official government figure. Sindh’s debt is highly concentrated in external debt. Foreign debt accounts for PRs 251 billion, with the World Bank being the principal creditor in FY16, Table 1.3: Sindh outstanding debt portfolio, June holding 58 percent of Sindh’s total foreign debt (Table 30, 2016 (FY16) 1.3), followed by the Asian Development Bank (ADB, Stock Avg. Avg. Percent 37 percent), and the Government of Japan (5 percent). PRs interest remaining (of total) (billio rate (%) maturity Meanwhile, Sindh’s domestic debt is very small, n) (no. of yrs) accounting for just 6 percent of total provincial debt Foreign debt (or PRs 16 billion) at end-June 2016. The domestic IDA* 146.3 1.08 20 58 ADB 92.3 1.30 17 37 debt portfolio of Sindh in FY16 was composed mainly Japan 11.7 1.70 15 5 of Cash Development Loans (CDLs) with long-term IBRD* 0.3 0.75 9 0 maturities, lent by the federal government to finance IFAD 0.2 1.00 13 0 irrigation needs. These loans are now being repaid, OPEC 0.5 2.50 5 0 which is drawing down the share of domestic debt in Fund Total 251.4 1.39 13 100 the total debt portfolio. The most recent domestic Domestic debt loan was received in FY16 as part of the ongoing CDL 14.6 12.62 11 89 Pakistan Water and Power Development Authority (Scarp) CDL 1.8 11.35 8 11 (WAPDA) Salinity Control and Reclamation Program (Norma Total 16.4 12.48 11 100 l) * IDA & IBRD are World Bank lending arms. (SCARP). Source: Provincial authorities and World Bank staff External debt is highly concessional and has long calculations. maturities. Almost 95 percent of the total foreign debt portfolio of the Sindh government is held by the World Bank and the ADB. IDA lending to Sindh amounts to around PRs 146 billion at an average interest rate of 1.08 percent and an average remaining maturity of 20 years. ADB lending to Sindh carries an average interest rate of 1.3 percent and an average remaining maturity of 17 years (Table 1.3). However, foreign currency denominated debt is exposed to exchange rate risk. The main exposure of this exchange rate risk comes from US dollar-denominated loans (93 percent of total external debt), followed by those in Japanese yen (5 percent). A depreciation of the Pakistani rupee would affect both the stock of government debt and debt-servicing flows. For example, a 10-percent depreciation of the Pakistani rupee against both foreign currencies would lead to an increase in debt stock of about PRs 25 billion (or 0.3 percent of PGDP and 3.8 percent of total provincial revenues). all foreign borrowing is on-lent by the federal government to the provincial governments, this means that provinces are constantly indebted to the federal government, requiring explicit consent of the federal government to initiate any other form of borrowing. 25 Sindh: Public Expenditure Review 2017 The Sindh government is fully cognizant of the recent changes to the federal landscape on borrowing. Pakistan’s fiscal architecture underwent a fundamental change with the adoption of the 18th Constitutional Amendment in 2010. The 18th Amendment permits borrowing, both domestic and external, by provinces, subject to limitations imposed by the National Economic Council (NEC).13 On July 1, 2015, the NEC set new limits for sub-national domestic borrowing.14 Sindh Province has identified debt management as one of the three pillars of the Sindh PFM Strategy. In FY13, a Public Expenditure and Financial Assessment (PEFA) were undertaken for Sindh Province. In light of the PEFA findings, provincial authorities were keen to develop an integrated PFM reform strategy and decided to prepare a detailed reform action plan in three areas: (i) financial management; (ii) procurement; and (iii) debt management. In FY14, the World Bank undertook an assessment of Sindh’s debt management functions and practices (using the World Bank’s Subnational Debt Management Performance Assessment tool, SN-DeMPA). Based on this diagnostic, Sindh has developed a reform plan with a matrix of reform activities with a mix of short-, medium- and long-term horizons. 13 th Clause (4) of Article 167, of the Constitution (a new clause inserted by the 18 Constitutional Amendment) reads: “A province may raise domestic or international loan, or give guarantees on the security of the Provincial Consolidated Fund within such limits and subject to such conditions as may be specified by the National Economic Council.� 14 On July, 1, 2015, the NEC took a decision to allow the provinces to borrow in the domestic market. This in FY16 translated into a gross borrowing limit of 0.5 percent of GDP (or PRs 153.4 billion) for the provinces or net domestic borrowing limit of PRs 61.75 billion for Punjab, PRs 20.05 billion for Sindh, PRs 16.88 for KP and PRs 13.91 billion for Baluchistan. 26 Sindh: Public Expenditure Review 2017 Chapter 2: Revenues and Expenditures 2.1 Revenues: The Revenue Challenge and Fiscal Space Fiscal space between provincially collected revenues and provincial expenditures is plugged by an elaborate system of fiscal transfers from the federal government, for which the Constitution has a dedicated institution, the National Finance Commission (NFC). The NFC, which has representation from the federal and all four provincial governments, is tasked with designing a new formula for sharing federal-collected revenue—the NFC Award—every five years. However, in many previous instances, the NFC has failed to deliver its awards on time. Figure 2.1: Sindh government's expenditure and its mode of financing (FY09 to FY16) Federal Transfers Provincial Revenue Fiscal Deficit Total Expenditure 10.0% 7.5% 7.6% 8.0% 6.8% 7.1% 0.8% 6.7% 6.3% 6.2% 6.2% 1.7% 0.5% 0.3% 6.0% 1.5% 1.3% 1.4% 0.8% 1.3% 0.8% 0.9% 4.0% 5.8% 5.7% 6.0% 5.8% 6.5% 5.0% 5.1% 5.4% 2.0% 0.0% -0.4% -0.5% -0.4% -0.1% -0.6% -2.0% FY09 FY10 FY11 FY12 FY13 FY14 FY15 FY16 Source: Data from Accountant General of Pakistan and Author calculations. The Sindh government’s total expenditure averaged 6.8 percent of provincial GDP (PGDP) annually during the period FY09 to FY16 (Figure 2.1), seemingly insufficient given the multiple service-delivery responsibilities at the provincial level. A significant portion of provincial expenditure goes towards the payment of salaries, allowances, and pensions of government employees, leaving only a small share for improving the quality of public services and infrastructure (see Section 2.2.2 on recurrent expenditures). The main reason for this relatively low level of public expenditure is the paucity of fiscal resources. During the 7-year period from FY09 to FY16), Sindh received transfers worth 3.4 percent of PGDP, financing 50 percent of total expenditure. Meanwhile, Sindh mobilized only 0.6 percent of PGDP from its OSR, which funded the remaining 8 percent of provincial spending. Competing demands by the federal government, limited borrowing options, and constitutional provisions, have limited the fiscal space available to the provinces to increase provincial spending. While in some years in the period FY09 to FY16, Sindh did run relatively small fiscal deficits to finance its expenditure, these were more than counterbalanced by fiscal surpluses in other years of the period. Deficits are limited partly because of constraints on provincial borrowing by the Constitution.15 During 5 of the past 6 years,16 Sindh has posted modest fiscal surpluses, further limiting the envelope available for 15 th Prior to the 18 Amendment to the Constitution of Pakistan (Article 167c) allowed a province to borrow only with prior th consent of the federal government (if it owes any debt to the federal government). The 18 Amendment (through insertion of Article 167d) allowed the provinces to borrow as per terms and conditions imposed by the National Economic Council. However, as Article 167c was not removed, the consent of the federal government remains as a prohibiting factor. 16 th These 6 years are covered under the 7 NFC Award. 27 Sindh: Public Expenditure Review 2017 provincial spending. These provincial surpluses have been negotiated by all provinces with the federal government, in light of the large increases in fiscal transfers made to the provinces under the 7 th NFC Award, in order to keep the consolidated (federal and provincial) fiscal deficit low. 2.1.1 The 7th NFC Award The provincial share of federally collected revenue has been increasing over time. Over the past decade, the transfer of resources to the provinces has increased. Through a Presidential Order revision in 2006, the 6th NFC Award was able to give provinces a larger share of the federal divisible pool (Figure 2.2). To carry this momentum forward, a new equilibrium was achieved under the 7th NFC Award, which, in FY11, increased the share going to the provinces from 46.25 percent (under the revised 6th NFC Award) to 56 percent, and subsequently to 57.5 percent for the remaining 4 years of the Award. This caused a sharp increase in federal transfers to Sindh. Between FY09 and FY16, transfers from the divisible pool increased by 20 percent per year. Although energy-related straight transfers (with an average annual growth of only 2 percent) more or less stagnated, overall federal transfers to Sindh increased at a healthy rate of 16 percent annually, implying an almost tripling of these transfers over the period FY09 to FY16 (Figure 2.3). Figure 2.2: Provincial share of divisible pool (percent) Figure 2.3: Federal revenue transfers to Sindh (PRs billion) 60.0 57.5 57.5 57.5 57.5 Divisible Pool Straight Total 450 56.0 400 55.0 350 7th NFC 300 50.0 250 45.0 6th NFC (Revised) 7th NFC 200 46.3 Award 45.0 150 43.8 40.0 41.5 42.5 100 50 35.0 0 FY07 FY08 FY09 FY10 FY11 FY12 FY13 FY14 FY15 FY09 FY10 FY11 FY12 FY13 FY14 FY15 Source: Data from the Government of Sindh. The annual increase of 20 percent in divisible pool transfers was the cumulative effect of various changes made by the 7th NFC in the revenue-sharing arrangement. As indicated in Figure 2.2, the most important change was the reduction in the federal government share in the vertical distribution of divisible pool revenue. Under the 6th NFC Award, the federal government received 53.7 percent of divisible pool revenue, while the 7th NFC Award reduced this to just over 44 percent. 2.1.2 Provincial Fiscal Space The impact of the 7th NFC Award on the revenue transfers to Sindh can be better understood if the individual impacts of each of the three major changes brought about by 7 th NFC Award are explicitly quantified. 17 In addition to this, a comparison is made between the revenue transfers under the 7th NFC 17 For horizontal distribution, all previous NFCs had relied on population shares as the sole criterion. The 7th NFC made a bold deviation from this past practice to include other criteria in the revenue-sharing formula. Population shares continued to command the largest weight (82 percent), but inverse population density (with a weight of 2.7 percent), and a backwardness index (10.3 percent) were added to the formula. Sindh’s demand that revenue collection be included in the formula was also 28 Sindh: Public Expenditure Review 2017 Award and the amount that could have been transferred if the 6th NFC Award had continued to be valid for the duration of the 7th NFC Award period (FY11 to FY15). The comparison shows that during this 5- year period, Sindh received an additional sum of PRs 342 billion in federal revenue transfers, solely due to the changes effected in the revenue-sharing formula by the 7th NFC Award. As expected, the bulk of this increase (24 percent out of the 33 percent increase) was due to a reduction in the federal government’s share, while the reduction in collection charges contributed an additional 4 percent to this increase. Despite all the positive proclamations, changes in the formula (for the horizontal distribution of revenue) added only 3 percent in revenue for Sindh. Another 2 percent of revenue was contributed by the interplay of the above-mentioned three changes (Figure 2.4). Figure 2.4: Composition on revenue increase for Sindh under the 7th NFC Award 30% 25% 24% 20% 15% 10% 4% 5% 3% 2% 0% Federal Share Formula Collection Charges Combinition Source: Data from the Government of Sindh. Although the increase in federal-revenue transfers significantly expanded the province’s fiscal space, there were some unwarranted consequences. First, with 80 percent of provincial expenditure financed from revenue collected at the federal level, the accountability of provincial expenditure is brought into question. Not only does the province have a rather weak record in public financial management,18 but stakeholder accountability is compromised, as the provincial government does not have to raise taxes to finance its own expenditure and therefore may not feel answerable to its own citizens regarding the best possible use of these resources. More importantly, with the large increase in revenue transfers, provincial political will to raise more revenue was further diluted. 2.1.3 Sindh Province’s Own-Source Revenue (OSR) The main contribution to Sindh government’s taxes came from GST on services, followed by taxes on the development and maintenance of infrastructure cess (DMIC). These two sources of OSR contributed more than three-quarters (78 percent) of total provincial tax revenues in FY15 (60 percent in FY09). The remaining one-quarter (22 percent) was collected through six sources, namely stamp duty, motor vehicle tax, provincial excise duty, capital value tax, urban immovable property tax, and property transfer tax (Figure 2.5). agreed to (giving it a weight of 5 percent). However, the combined impact of all these changes on horizontal distribution was that the revenue share of Baluchistan increased, while that of the other three provinces declined. Nonetheless, due to the large reduction in vertical share of federal government, all provinces saw a substantial increase in overall revenue transfers. Finally, the collection charge on divisible pool taxes was reduced from 5 to 1 percent, implying a marginal expansion in the divisible pool. 18 World Bank (2013), Public Financial Management and Accountability Assessment – Pakistan Sindh Province. 29 Sindh: Public Expenditure Review 2017 Sindh has simplified its tax structure and yet it retains a number of taxes with low revenue yields and significant economic costs. Some of the taxes that Sindh levies yield negligible amounts of revenue, but they still impose significant non-monetary costs on taxpayers. Early in the past decade, the provincial tax regime was greatly simplified by reducing the number of provincial taxes from 23 to only 10.19 However, further simplification may be required, as some of the taxes currently levied still have low revenue yields and continue to impose significant economic costs on taxpayers, thereby stifling economic activity. Some of these taxes could be merged with other taxes (e.g., lumping profession and calling tax into sales tax on services), others could be modernized, and still others could be devolved to local (district) governments. In terms of revenue yield, the Sales Tax on Services (STS) provides almost half of the revenue collected from provincial OSR (Figure 2.5). This is followed by the DMIC, which resembles a surcharge on all (non- liquid) imports passing through Sindh. The DMIC provides another one-quarter of provincial OSR. All other taxes combine to provide the remaining one-quarter of OSR and indicate the relatively low yields of most provincial taxes. Figure 2.5: Revenues from different sources (% province OSR) FY15 FY14 FY13 FY12 FY11 FY10 FY09 GST of Services (Provincial) S.D&M of I Structure Stamp Duties Motor Vehicles Provincial Excise Capital Value Tax Urban Immovable Property Tax Transfer of Property Tax (Registration) Others 0 10 20 30 40 50 60 Source: Data from the Government of Sindh. Despite the unfavorable assignment of revenue authority, provinces must do a greater effort to mobilize revenue. This, along with lack of tax policy and extremely weak tax administration, defines a revenue system that is neither prepared, motivated, nor even used for mobilizing additional revenue. As such, Sindh’s OSR has always remained at less than 2 percent of provincial GDP, falling far short of the province’s needs. On one hand, this perpetuates and strengthens the province’s dependence of fiscal transfers from the federal government, and on the other it undermines the Sindh government’s ability to undertake initiatives to accelerate growth and development in the province. 19 Sales tax on services (STS) and capital value tax (CVT) were added to the list of provincial taxes in 2010 as the federal government devolved these taxes to provinces. 30 Sindh: Public Expenditure Review 2017 The remainder of the revenue section in this chapter describes some of the major provincial taxes, identifying their limitations and indicating possible ways to make them more efficient and productive. 2.1.4 Sales Tax on Services (STS) Under the Constitution, the Sales Tax on Services (STS) is a provincial tax. However, as the Sales Tax on Goods (STG) is collected by the Federal Board of Revenue (FBR), it was considered appropriate that STS should also be collected by the FBR, with the collected revenue then returned to the provinces.20 By FY10, the FBR was taxing 10 services under the STS regime. However, in 2010, during the 7th NFC Award discussions, provinces (especially Sindh) asked for, and received, the right to collect STS themselves. As a result, in 2011, Sindh promulgated the Sindh Sales Tax on Services Act (SSTSA), and established the Sindh Revenue Board (SRB) to collect the tax. Compared with the FY10 STS-related revenue transfers of PRs 1.1 billion, the FY11 collection by the SRB was 200 percent higher. This was mainly because, in promulgating SSTSA, the Sindh government brought some additional services under the STS umbrella.21 This base-broadening of STS continued in subsequent years and, as a result, Sindh’s STS revenue has increased exponentially, reaching PRs 64 billion in FY16.22 Despite this phenomenal progress in STS revenue collection, significant challenges remain. The services brought under STS to date are those that have relatively well-defined bases and are easy to tax. Although the Sindh government has imposed STS on some construction, personal care, and professional services, the wholesale, retail, and transport sectors remain almost entirely outside the tax net.23 These sectors are composed of a large number of small units that are difficult to identify and where access to accurate sales records is problematic. Adding income from services24 given in the estimated provincial accounts, the present effective tax rate of STS is still very low (1.7 percent in FY15). Expansion of the STS base to more challenging bases (e.g., commerce and transport) would require some innovative thinking and management, which in turn would require considerable improvement in the SRB’s manpower capacity and technology. 2.1.5 Urban Immovable Property Tax (UIPT) This is the oldest form of provincial taxation, providing revenue for provincial governments and tehsil municipal administrations (TMAs).25 This tax is levied on property owners, but is limited to buildings and land located within the limits of urban areas.26 The tax base is the assessed (imputed) rent in the case of (owned) rented buildings. The tax base is badly eroded by numerous exemptions. For example, all government property is exempt from payment of UIPT, and so are the properties of religious, charitable and educational institutions. Similarly, cantonment properties are also exempt from the tax. Other exemptions include: (i) all properties commanding annual rent of less than a specified value; (ii) single 20 With federal government retaining 2 percent of revenue as collection charges. 21 Some of these additional services were taxed by the federal government under the Federal Excise Duty (FED) regime, but vacated these bases for the provincial governments to tax in FY10. 22 This increase was despite a reduction in standard STS rate from 15 to 14 percent in FY15. The rate was further reduced to 13 percent in FY17. 23 There is a positive list of services sub-sectors that are taxed under SST regime. There is a need to issue a negative list of sub- sectors, which are exempted instead of positive list. 24 These however exclude value-added in gas distribution and ownership of dwellings. Taxation of former is a federal subject, while the latter is taxed by the provincial government under Urban Immovable Property Tax (UIPT). 25 TMAs are district governments and district councils created by the Local Government Ordinances (LGO) of 2001. 26 The Local Government Ordinances (LGO) of 2001 abolished the distinction between urban and rural municipalities, making it possible for the local governments to levy the tax on all properties in the province. However, no local government decided to tax rural properties. With the lapsing of 2001 LGO, the former distinction of rural and urban properties was reinstated with tax levied only on urban properties. 31 Sindh: Public Expenditure Review 2017 homes owned by a widow, an orphan, or a disabled person with an assessed annual rental value of below PRs 5,000; (iii) all new residential buildings for a period of 3 years; and (iv) single-owner occupied houses with assessed rental value of less than a specified limit. In addition to this erosion of the tax base, the tax is marred with a large number of structural and administrative problems. First, there is gross under-valuation of annual rental value (ARV) of properties. The provincial excise and taxation department conducts a periodic survey of properties, last conducted in 2001, and the results are used to prepare valuation tables for properties in different localities (zones) of an urban area. Anecdotal evidence suggests that assessed ARV of a property is 5 to 10 percent of actual market rent of the same or similar property. Second, owner-occupied properties are assessed in a way that the imputed ARV is about 10 percent of the ARV of a comparable rented unit. This means that owner-occupiers pay 10 percent of the normal tax liability. As a large proportion of properties are owner- occupied, this concession to owner-occupied properties erodes the tax base substantially. In addition, this concession opens up the possibility for tax evasion by declaring rented properties to be owner- occupied. The same tax-evasion opportunity exists for commercial properties, given the 10-times-higher tax rate on commercial properties compared with the tax rate on residential properties. The tax base is too low. There are problems of under-assessment and an incomplete valuation roll. There are five urban areas in Sindh, namely Karachi, Hyderabad, Mirpurkhas, Nawabshah and Sukkur, but only three levy and collect property tax, namely Karachi (90 percent), Hyderabad (5 percent) and Sukkur (5 percent). The number of tax-base residential properties in Karachi of about 700,000 appears incompatible with a population of 22 million. The numbers for Hyderabad and Sukkur also appear to be significantly underestimated. These structural issues with UIPT, coupled with weak tax administration, result in extremely poor collection of UIPT. Figure 2.6 provides an international comparison of revenue collection from property tax. Sindh, similar to other provinces in Pakistan, compares unfavorably with other countries in terms of collection rates.27 Figure 2.6: Cross-country comparison of property tax collection 5 4 3 Percent of GDP 2 Sindh, Punjab, KP & 1 Balochistan 0 6.0 7.0 8.0 9.0 10.0 11.0 -1 . Log(Per-Capita GDP) Source: World Bank staff calculations based on IMF and provincial data 27 In 2014-14, UIPT collection from Karachi was less than PRs 2.5 billion. In the same years, property tax collection in Indian cities was many time larger (e.g., Mumbai collected an equivalent of PRs 45 billion, Delhi, PRs 110 billion and Bangalore PRs 15 billion). 32 Sindh: Public Expenditure Review 2017 2.1.6 Agriculture Income Tax (AIT) Under the Constitution, all personal and corporate incomes can be taxed by the federal government, with one notable exception, namely income from agriculture, which can only be taxed by provincial governments. In Sindh, the Agriculture Income Tax (AIT) is collected by the Board of Revenue (BOR) by levying a presumptive tax on agricultural land. The Sindh Agricultural Income Tax Act of 2000 provides two alternative calculations of the tax. The first is based on land area (area sown), while the second is based on agricultural income. However, hardly any tax is collected on the basis of income. AIT is levied at the rates shown in Box 2.1. Box 2.1: Tax rates of the Agricultural Income Tax (AIT) Annual rates of tax on land tax Irrigated land PRs 200/- per acre Unirrigated land PRs 100/ per acre Matured orchards, banana and betal leaf: a) Irrigated PRs 700/- per acre b) Unirrigated PRs 350/- per acre Exemptions 1. No land tax is levied on holdings up to 4-0 acres irrigated and 8-0 acres unirrigated. This exemption does not apply to matured orchards, banana and betal leaf. 2. No land tax is levied in the Thar Desert and the Kohistan areas. Rates of tax on total agricultural income (1) Where the total income does not exceed PRs 100,000 5% of the total income. (2) Where the total income exceeds PRs 100,000 but does not PRs 5,000plus 7.5% of the amount exceed PRs, 200,000 exceeding PRs 100,000 (3) Where the total income exceeds PRs 200,000 but does not exceed PRs 12,500 plus 10% of amount exceeding PRs 300,000 PRs 200,000 PRs 22,500/- plus 15% of the amount (4) Where the total income exceeds PRs 300,000 exceeding PRs 300,000 Revenue collection has remained woefully low, with the highest tax collected in FY01, when the BOR collected PRs 547 million (Figure 2.7). Since then, collection has been on a declining trend until FY12. The inadequacy of collection can be gauged from the “effective tax rate�, which provides the ratio of AIT collection to value-added in cropped agriculture (in Sindh). This has never exceeded 0.5 percent and is currently (FY15) below 0.1 percent. Figure 2.7: Trend in AIT collection (PRs million and percent) 600 Agriculture Income Tax Effective Tax Rate--RHS 0.5% 0.5% 500 0.4% 0.4% 400 0.3% 300 0.3% 0.2% 200 0.2% 0.1% 100 0.1% 0 0.0% Source: Data from the Government of Sindh. 33 Sindh: Public Expenditure Review 2017 One reason given for the poor revenue collection of AIT is the low revenue potential of the tax. However, using data for cropped acreage under every crop from the Sindh Agricultural Census 2010, per- acre yield of every crop from agriculture statistics, and farm-gate prices and production cost of crops from various research studies, the tax revenue from agricultural income is estimated to be PRs 35 billion in FY16. Table 2.1 estimates the potential of AIT if it were applied on all farmers in the income mode. However, if all farms were taxed under the land-tax mode, the revenue potential drops to PRs 4 billion. This is one-ninth of the revenue potential under the income-tax mode, but still 12.5 times the revenue collected at the moment. Table 2.1: Estimates of agriculture income tax revenue potential Estimated income per farm (PRs p.a.) Potential tax revenue (PRs billion) Farm size in acres Owners Owner- Tenants Owners Owner- Tenants Total Tenants Tenants Under 1.0 15,280 7,930 2,508 0.0 0.0 0.0 0.0 1.0 to under 2.5 49,447 39,094 29,953 0.0 0.0 0.0 0.0 2.5 to under 5.0 107,961 81,268 64,091 0.1 0.0 0.0 0.1 5.0 to under 7.5 158,977 134,037 101,386 0.8 0.0 0.0 0.9 7.5 to under 12.5 268,750 203,945 153,146 2.2 0.2 0.1 2.5 12.5 to under 25.0 461,591 335,548 282,073 5.8 0.2 0.2 6.2 25.0 to under 50.0 752,393 597,150 395,158 9.7 0.4 0.1 10.2 50.0 to under 100.0 1,377,261 821,004 797,256 9.6 0.4 0.1 10.1 100.0 to under 150.0 2,090,973 1,487,860 780,871 2.0 0.3 0.0 2.3 150.0 and above 3,439,837 2,617,138 1,207,992 2.3 0.4 0.1 2.8 Total 32.4 1.9 0.7 35.0 Source: WB staff estimates. 2.1.7 User Charges Sindh government has made little effort to use user charges to enhance the resource envelope. A large share of provincial services is financed from Figure 2.8: Use charges (% of recurrent expenditure) general purpose revenue and reliance on user Community Services Social Services charges has declined sharply over time. User Economic Services 30.0% charges as a share of OSR declined from 14 28 percent in FY97 to less than 1 percent in FY16. 25.0% This lack of exploitation of user charges to finance public services implies giving an “untargeted 20.0% subsidy� to these services. User charges currently 15.0% finance only 4 percent of Sindh government’s recurrent expenditures on community services 10.0% (13 percent in FY97), 1 percent of recurrent expenditures on economic services (down from 5.0% 28 percent in FY97), and 0.4 percent of recurrent 0.0% expenditures in the social sectors (Figure 2.8). An argument could be made on equity grounds to keep user charges on services low (or even at Source: Data from the Government of Sindh. zero) on those services used by the poor and vulnerable segments of the population, including primary and secondary education, primary health care, rural water supply and sanitation. However, there is 28 A part of this decline, however, is due to administrative reforms undertaken by the province in irrigation and health sectors. In irrigation sector, the government established autonomous Farmers’ Associations (FOs) and Area Water Boards (AWBs) to collect irrigation user-charges from the farmers in order to manage and maintain non-trunk water courses. Similarly, larger hospitals in the province had been made administratively and financially autonomous. These hospitals now retain the user-charges that they collect to part finance their operations. 34 Sindh: Public Expenditure Review 2017 hardly any rationale for keeping user charges low for economic services (such as irrigation, roads, etc.), and tertiary level social and community services.29 2.2 Expenditures: The Challenge of Funding Development Priorities 2.2.1 Allocation of Fiscal Resources: Recurrent vs Development The long-term development plan of the Government of Sindh, “Sindh Vision 2030�, envisages working towards the creation of an educated, healthy, productive and prosperous Sindh. The Sindh Vision 2030 was structured to be aligned with the pillars of the national poverty reduction strategy, the Millennium Development Goals, and the Global Competitiveness Index by devising an integrated framework. The document highlights the overall vision as: “A united, just, resilient, productive, innovative, industrialized, and prosperous Sindh with a disciplined, caring society comprising of healthy, happy and educated people and built upon the enduring pillars of self-reliance, respect, tolerance, equity and integrity�. The Sindh Vision 2030 document is organized according to the thematic areas of poverty, health, education, employment, effective governance, land, water, and infrastructure. This PER analyzes the extent to which the allocation of funds is actually in accordance with the priorities set out in the Vision 2030. With a significant increase in resources, both from OSR and federal transfers, the Sindh government has been able to use this extra fiscal space to enhance investment expenditure. While recurrent expenditures grew significantly due to increased employee-related spending and consumed a significant portion of the available fiscal space, the Sindh government was able to increase spending on development as well. Government development expenditures grew 3.3 times in nominal terms (2.5 times in real terms) in the 6 years from FY09 to FY15, while recurrent expenditures increased 2.3 times in nominal terms (1.7 times in real terms). Development expenditures as a share of total expenditure increased from 13 to 23 percent (Figure 2.9). Figure 2.9: Recurrent and development expenditures of Sindh Structure of public expenditure (share) Structure of public expenditure (% PGDP) Recurrent Development Recurrent Development 100% 7% 13% 17% 90% 25% 24% 23% 32% 26% 6% 80% 70% 5% 60% 4% 50% 7% 87% 83% 3% 6% 6% 6% 6% 6% 40% 75% 76% 77% 68% 74% 30% 2% 4% 20% 3% 1% 2% 2% 2% 2% 10% 1% 1% 0% 0% FY09 FY10 FY11 FY12 FY13 FY14 FY15 FY09 FY10 FY11 FY12 FY13 FY14 FY15 Source: Data from Accountant General of Pakistan and Author calculations. Around half of the recurrent expenditures are consumed by employee-related expenses (salaries, allowances, and pensions). Following the 7th NFC Award, provinces received larger shares from the 29 Especially designed and well-targeted cash transfer programs can mitigate the ill-effects of such a policy on poor. 35 Sindh: Public Expenditure Review 2017 federal collection of taxes, which the Sindh government largely spent on increasing compensation for government employees, or in the development sector. Since FY13, government employee-related benefits (salaries and allowances) have become the largest component of recurrent expenditures, rising from 33 percent in FY12 to 46 percent in FY15. Similarly, government employees’ retirement benefits (pensions) increased from 7 percent of recurrent expenditures in FY12 to 10 percent in FY15. These increases in employee-related expenses may pose fiscal risks for the future, as committed expenditures increase fiscal rigidity and reduce fiscal space. In addition, as discussed in other chapters, increases in government employee-related expenditures have no impact on the quality and quantity of service delivery, or sector outcomes. 2.2.2 Trends and Composition of Recurrent Expenditures Recurrent expenditures, which constitute a large portion of total public expenditure, have increased threefold since FY09. On average, recurrent expenditures constitute about three-quarters of all Sindh government outlays. Partly due to the additional fiscal space provided by the 7th NFC Award and partly because of large and frequent increases made in Sindh government’s employee-related spending, recurrent expenditures increased at a rapid rate (at an average of 19.4 percent per year) between FY09 and FY15. In real terms, recurrent expenditures on average grew by 10 percent over the period. This caused recurrent expenditures to increase from about 4 percent of PGDP in FY09 to 6 percent in FY15 (Figure 2.10a/b). Despite this large increase, the share of recurrent expenditures in overall public spending has been declining (falling from 86 percent in FY09 to 74 percent in FY15).30 This implies an even faster increase in development spending, as an increasing portion of additional fiscal resources available to the province under the 7th NFC Award and from its OSR was channeled towards development spending. Figure 2.10a: Recurrent expenditures (% of total Figure 2.10b: Recurrent expenditures (PRs billion) expenditure and GDP) Curr Exp Tot Exp Curr Exp / GDP Nominal Real 100% 8% 500 450 7% 80% 400 6% 350 60% 5% 300 4% 250 40% 200 3% 150 2% 100 20% 1% 50 0 0% 0% FY09 FY10 FY11 FY12 FY13 FY14 FY15 FY09 FY10 FY11 FY12 FY13 FY14 FY15 Source: Data from Accountant General of Pakistan and Author calculations According to the functional classification, only five sectors take almost all (96 percent) of the recurrent budget. These are the sectors that the government has identified as priorities, namely: general public services (38 percent); education (22 percent); economic services (16 percent); law and order (13 percent); and health (7 percent). 30 The sharp increase in the share or recurrent expenditures (in total expenditure) in FY11 were due to a roughly 50 percent increase in employees’ salaries. 36 Sindh: Public Expenditure Review 2017 There has been an apparent shift in spending from general public services towards economic and social services. In the 6-year period from FY09 to FY15, the share of general public services, which includes general administration and other expenditure items, declined sharply from 46 percent of overall recurrent expenditures in FY09 to just 29 percent in FY15. The bulk of this decline instead went towards economic services (e.g., irrigation, roads, etc.), the share of which increased by 13 percentage points, while the share of health services increased by 2 percentage points, and that of education and other services increased by 1 percentage point each (Figure 2.11). The share of recurrent expenditures on law and order remained unchanged. As mentioned earlier, to better understand these changes, they should be seen in the context of the rolling back of devolution. General public services include grants and transfers to district governments. Hence, the 17-percentage-point decline in this share was mainly due to the decline in grants as a result of the rolling back of devolution. Similarly, the increase in the shares of other sectors is partly the result of this roll-back.31 In short, the changes in sectoral composition of recurrent expenditures may not be a sound indicator of shifts in government priorities. Figure 2.11: Sectoral composition of recurrent expenditures 100% Other Services 90% 80% Education Services 70% Health Services Percent of total 60% 50% Economic Services 40% Law & Order 30% 20% General Public Services 10% 0% FY09 FY10 FY11 FY12 FY13 FY14 FY15 Source: Data from Accountant General of Pakistan and Author calculations. On average, education consumes about one-quarter of total recurrent expenditures. In real terms, public spending on education grew on average by 11 percent annually in the 6 years between FY09 and FY15. In the same period, overall real recurrent expenditures grew by about 9.5 percent annually. In the first 2 years of the 7th NFC Award, expenditure on education did not increase significantly and therefore decreased as a percentage of total recurrent expenditures. However, after the passage of 18th Constitutional Amendment, education and health fell under provincial responsibility, and the Sindh government therefore started to increase resources to both sectors. Expenditure on education was just 17 percent of total recurrent expenditures in FY12, but this jumped to 25 percent the following year, and remained almost at the same level thereafter. For a more detailed discussion on public expenditure on education and its impact on outputs and outcomes, see Chapter 3 on education. The crux of the analysis in Chapter 3 highlights the paucity of resources, together with the very poor effectiveness and efficiency of public expenditure. 31 This is because the salaries of employees of various sectors that were previously paid from district budgets (but that expenditure was classified in the provincial budget under grants, i.e., general public services), are now being paid from the sector allocations in the provincial budget. In other words, the larger the employment in the sector and the greater the roll-back in devolution, so the greater will be the increase in sector expenditure, although little or no real change may have occurred on the ground or in the Sindh government’s priorities. 37 Sindh: Public Expenditure Review 2017 The health sector is ranked fifth in terms of average resource allocation by the Sindh government over the 6-year period between FY09 and FY15, but it is the fastest growing sector. Health recurrent expenditures increased from 7 percent of total recurrent expenditures in FY09 to 9 percent in FY15. Most of these expenditures were allocated to health operating expenses, including spending on drugs and medicines, repairs and maintenance, and petroleum, oil and lubricants (POL) for generators. A more detailed discussion on public health expenditures can be found in Chapter 4 on health. During the 6 years between FY09 and FY15, expenditure on the social sectors averaged about one-third of total recurrent expenditures, but the share remained volatile. Expenditures on education and health increased, while expenditures on social protection decreased during this period. Over this 6-year period, expenditures on the social sector increased 3.3 times in nominal terms and 1.9 times in real terms. The highest increase was seen in the health sector, followed by the education sector. With such a large portion of recurrent expenditures going towards employees’ salaries and pensions, this will tend to increase the rigidity of the Figure 2.12: Composition of recurrent expenditures, recurrent budget. Over 88 percent of the FY09-FY15 recurrent budget is spent on three items, with Operating salaries and pensions alone taking almost half of Expenses, Interest , 11% the recurrent budget, another 28 percent going Transfers & 4% in grants, transfers and subsidies, while Subsidies, 28% operating expenses (including utility charges) Asset Other, 12% Creation & account for about 11 percent of recurrent Acquisition, spending (Figure 2.12). The remaining 12 percent Pay & 6% Pension, of recurrent expenditures are spent on interest 49% Repairs & payments (4 percent), and asset acquisitions Maintenanc (i.e., furniture, vehicles, etc., 6 percent), while a e, 2% grossly inadequate 2 percent is spent on repairs and maintenance. The high proportion of salaries and pensions expenditure indicates strong Source: Data from Accountant General of Pakistan and Author calculations. rigidities in the recurrent budget. Low spending on asset acquisitions, and repairs and maintenance, signifies the tight controls that the Finance Department has applied on non-salary spending. As a result, a significant portion of asset acquisitions and repairs that are needed by the line departments is met through the development budget (Figure 2.13). 38 Sindh: Public Expenditure Review 2017 Figure 2.13: Trends in various economic components of recurrent expenditures Pay and Pension Operating expenses Repairs And Maintenance Interest Payment Physical Assets Transfers and Subsidies Other 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% FY09 FY10 FY11 FY12 FY13 FY14 FY15 Source: Data from Accountant General of Pakistan and Author calculations. Not only are recurrent budget rigidities high, they appear to have grown sharply over the 6 years to FY15. Government employee-related spending is not only the largest component in recurrent spending, but it has been growing at a disturbingly high average annual rate of 23 percent. Although interest payments constitute only about 4 percent of recurrent expenditures, these have grown more, at an average annual rate of 49 percent over the 6 years to FY15. As such, the share of “rigid expenditures� (i.e., the sum of salaries, allowances, pensions and interest payments) in total recurrent spending increased by 5 percentage points (from 53 to 58 percent) between FY09 and FY15 (Figure 2.12). The increase in budgetary rigidities is, at least in part, self-inflicted. While the rapid increase in interest payments is real, the increase in government employee-related spending is partly due to large increases in salaries and pensions that the Sindh government has effected and partly the outcome of the sharp increase in the number of retiring employees. It is also interesting to observe that the gap between public sector and private sector employees’ wages is widening (Figure 2.14). This situation implies that the Sindh government lacks the fiscal space to allocate funds to much needed priority areas. While the number of government employees and their salaries are increasing, expenditure on operations, and repairs and maintenance has failed to keep pace with spending on salaries, eroding the quality of service delivery. 39 Sindh: Public Expenditure Review 2017 Figure 2.14: Monthly wages of paid employees, age 15-64, CPI adjusted Public Sector Private Sector 20000 18000 16000 14000 12000 10000 8000 6000 4000 2000 0 1999 2005 2008 2010 2012 2013 2014 Source: Labor Force Surveys, Pakistan Bureau of Statistics. On the positive side, despite these increasing rigidities, the government has been able to make some positive changes in budget allocations. As a policy, the Sindh government has increased its spending on operating expenditures. These expenditures grew by an average annual 16 percent from FY09 to FY15, marginally lower than the rate of increase in overall recurrent expenditures, at 18 percent. These expenditures are allocated to provide better communications, transportation, office consumables, and utilities—expenditure that is generally associated with the quality of public services. The significant increase in the allocations to these operational items, while less than overall recurrent expenditures growth, is mainly due higher spending on medicine and lump-sum provisions kept in the budget for specific purposes of non-salary expenditures in health, education, and the maintenance of law and order. The Sindh government plans to continue this trend of higher allocations for these items. Spending on repairs and maintenance is increasingly aligned with the government’s objective to better maintain public infrastructure, but levels are still very low. The funds were allocated for repairs and maintenance of roads and buildings, plant and machinery, especially in hospitals, vehicles, and furniture and fixtures. These expenditures increased from 1 percent of current expenditures in FY09 to 3 percent in FY15. Despite the increase, repairs and maintenance spending remains far below what is needed and the international benchmarks, adversely affecting the overall life of public infrastructure and undermining the efficiency of public services. 2.2.2.1 Budget Credibility In addition to the inadequacy of some allocations and the low overall efficiency of public spending, one additional problem—impacting the transparency of the government’s fiscal operations and the implementation of government policies and programs—is the questionable credibility of the budget. Budget credibility in Sindh is severely undermined by the misuse of contingency budgetary instruments. Pakistan’s budgetary system does not allow any allocation for contingency expenditure. Instead, it allows line and central departments to effect changes in the approved budget in cases where these are necessitated by “emerging situations� and where they can “enhance the efficiency of public spending�. The line departments are allowed to re-appropriate funds from items in their approved budgets for allocation to other items if there is a sound reason to do so. Similarly, if any line department does not consider that the budget allocation is adequate, it can request for a supplementary budgetary grant, which can be granted if approved by the Finance Department (and if approved by the Planning and Development Department [P&DD] in the case of the development budget). 40 Sindh: Public Expenditure Review 2017 Excessive use of budget re-appropriations and supplementary grants have the potential to greatly reduce fiscal discipline, thwart efficiency, cloud the transparency of the budget, and undermine the right of the provincial assembly to approve budgetary allocations. Re-appropriations are generally affected by drawing and disbursing officers (DDOs) at the spending unit level, with DDOs able to change the composition but not the size of the budget. However, supplementary grants impact both the composition and the size of the budget. Grants range from 10 percent to almost 100 percent of the budget (Table 2.2) and, as such, can severely undermine budget credibility. A Sindh Public Financial Management and Accountability Assessment undertaken in 2013 give poor scores (D/C) to the Sindh government on budget credibility. Since this 2013 assessment, matters have improved somewhat, but budget credibility remains a major concern. Table 2.2: Post-approval changes in the budget (PRs billion and percent) PRs billion FY09 FY10 FY11 FY12 FY13 FY14 FY15 Original budget 51.5 59.0 84.2 85.1 194.4 223.5 255.6 Plus supplementary grants 5.2 14.4 10.9 81.3 46.0 44.1 33.1 Plus net re-appropriations 0.1 0.7 0.1 0.5 -0.2 -0.1 0.1 Final budget 56.8 74.2 95.2 167.0 240.2 267.5 288.9 Releases 49.0 68.0 82.4 126.9 197.3 215.3 245.0 Expenditure 42.8 56.2 73.0 140.0 197.8 219.8 236.0 Memo items: Final budget as a % of original budget 110 126 113 196 124 120 113 Supplementary as a % of original budget 10 24 13 96 24 20 13 Releases as a % of final budget 86 92 87 76 82 81 85 Expenditure as a % of releases 87 83 89 110 100 102 96 Source: Data from Accountant General of Pakistan, Author’s calculations Despite budget-enhancing provisions, overall expenditure generally tends to fall short of the budget. Control measures have been effective in maintaining the budget, but perhaps at the Figure 2.15: Budget and utilization (PRs billion and expense of efficient public spending. While the percent) Utilization (Right Axis) Original Budget re-appropriation provision and especially Expenditure supplementary grants create a situation in which 500 120% expenditure could significantly exceed the 450 400 100% approved budget, various control measures 350 adopted by the Finance Department (FD), 80% Rs Billion 300 together with the weak absorptive capacity of line 250 60% departments, have avoided substantial breaches 110% 200 100% 95% 85% 91% in the overall budget. In general, expenditure falls 150 77% 40% short of the approved budget (Figure 2.15). The 100 54% 20% main tool adopted by the FD to avoid any breach 50 of fiscal discipline is the control of budgetary 0 0% releases. Budgetary funds, also for recurrent expenditures, are released in multiple tranches and on the basis of line departments meeting Source: Data from FD & Accountant General of Pakistan, Authors certain criteria and the availability of funds. As can calculations. be seen in Table 2.2, the release of funds falls short of the revised budget (original budget plus approved supplementary grants) by as much as one-quarter. Nevertheless, this still does not show the overall impact of budgetary “control� measures. Generally, a large portion of budgetary funds is released to the line departments late in the fiscal year, making its full use by the line departments almost impossible. Although this helps to keep overall expenditure within budget limits, it may also undermine the objectives for which these budgetary allocations are made. 41 Sindh: Public Expenditure Review 2017 2.2.2.2 Fiscal Risk Much of the recurrent budget is already committed to government employee-related spending and the servicing of provincial debt, while on top of these commitments there are significant potential downstream fiscal risks for the government. Following the 7th NFC Award, there has been a significant increase in government employees-related expenditure on salaries, allowances, and pensions, and this has also increased rigidities and fiscal risk. An important issue in this regard is the rapid increase in pension liabilities of the Sindh government. Over the 6 years from FY09 to FY15, spending on pensions has almost doubled, both in absolute terms and as a ratio of total recurrent expenditures (Figure 2.16). Figure 2.16: Changing structure of recurrent expenditures (percent) 90 Employees related exp Employees retirement benefits Grants, subsidies and write off loans 80 70 33 18 14 14 60 38 37 50 25 8 10 10 7 40 7 6 6 30 43 46 47 46 20 31 31 33 10 0 . FY09 FY10 FY11 FY12 FY13 FY14 FY15 Source: Data from Accountant General of Pakistan and author's calculations Looking forward, pension liabilities are expected to remain on a rapidly rising trend, limiting fiscal space to increase allocations for other needs. The age profile of government employees in Sindh is such that pension liabilities will increase significantly over the coming 15 years. Currently, there are more than 140,000 pensioners of the Sindh government and more than 480,000 government employees with a median age of 45 years. This implies that if the government retirement age remains at 60 years, more than 240,000 additional government employees will join the pool of pensioners over the next 15 years. Moreover, a significant number of government employees retire before the age of 60. Assuming that half of the current beneficiaries survive, the total number of beneficiaries will be about 310,000 after 15 years. Pension spending that used to be about PRs 42 billion could increase to PRs 93 billion (in constant Pakistani rupees of FY15). If the present rate of increase in pension spending remains at the current level, it will rise from 22 percent of government employee-related expenditure currently to about 48 percent in 15 years’ time. 42 Sindh: Public Expenditure Review 2017 Besides pensions, the General Provident Fund32 Figure 2.17: GPF liabilities of GoS (PRs billion) has become another significant liability of the 120.00 Sindh government, as it is now the second- 105.52 largest component of the provincial 98.18 100.00 87.69 government’s total debt. The General Provident Fund (GPF) liability that was PRs 12 million in 80.00 FY72, PRs 836 million in FY87, and PRs 6.5 billion in FY97, jumped to over PRs 100 billion in FY14 60.00 (Figure 2.17). The GPF liability of the Sindh government increased fivefold from FY02 to FY15. 40.00 28.46 Moreover, in FY07, GPF liability was about one- 17.45 20.00 quarter of the total debt of the Sindh government 6.50 and, in FY15, this liability increased to 35 percent 0.01 0.07 0.84 3.00 0.00 of total debt. If this trend continues, which is FY72 FY77 FY87 FY92 FY97 FY02 FY07 FY13 FY14 FY15 highly likely, GPF liability will reach PRs 228 billion by 2030.33 Source: Finance Department, GoS This unfunded nature of the GPF poses another significant risk to the sustainability of public finances in Sindh. The Sindh government has long been aware of the fiscal risk emanating from the GPF liability and, in order to cover the risk, established the Sindh General Provident Investment Fund (SGPIF) in 2007. The Sindh government has been building up the assets of the SGPIF to meet future liabilities than it could incur from the GPF. The SGPIF was established with seed money of PRs 2 billion in the FY08. The main objective of the SGPIF was to use the returns earned to pay off, including profits, which is generally fixed at a rate determined by the government on an annual basis. The closing balance of the SGPIF was PRs 28 billion on June 30, 2013, and PRs 37.6 billion on June 30, 2015, or 43 percent of the GPF liabilities at that time. There is a need to accelerate the funding of this vehicle. Private-public partnerships, although a definite opportunity, may also pose some actual and contingent fiscal risks. The Sindh government is relying increasingly on public-private partnership (PPP) schemes to develop infrastructure, with the goal of increasing the efficiency of investments and asset management. 32 GPF, or the General Provident Fund account, is a provident fund account that is available for government employees. A government employee can become a member of the GPF by contributing a certain percentage of their salary to the account. 33 The rapid increase in GPF liability is attributable to a higher-than-market rate of interest applied to the amount contributed by employees as a mandatory deduction from their salaries, despite the fact that no asset is actually created. The government- determined interest rate paid on the GP Fund amounted was 9 percent in the 1970s, 20 percent in the mid-1980s to mid-1990s, 21 percent in second half of 1990s, and 13 percent in the first half of the 2000s. 43 Sindh: Public Expenditure Review 2017 However, PPPs can also create challenges for provincial finances, given that PPP projects entail long-term fiscal implications and can be contingent on risks. These could impose fiscal risks for provincial financial management in the form of direct liabilities (commitments not dependent on the occurrence of an uncertain future event) and contingent liabilities (commitments whose occurrence, timing and magnitude depend on uncertain future events outside the government’s control). Through PPP projects, the Sindh government is acquiring a series of fiscal commitments that may constitute a significant fraction of the provincial budget in future. Six PPP Figure 2.18: Trend in interest payments, standardized projects with a total cost of PRs 45 billion have base, FY09 to FY15 already been signed by the Sindh government, Interest Payments Total Current Expenditures 900 adding to its liabilities. Analysis shows that these direct and contingent liabilities from PPP projects 800 will add to gross debt, equivalent to an additional 700 1 to 2 percent of provincial GDP. 600 Any deterioration in overall macroeconomic 500 stability could pose fiscal challenges to the Sindh 400 government through higher interest rates or a 300 depreciating currency. Either of these external 200 events could significantly increase debt servicing 100 in domestic currency terms. Increased interest 0 payments on foreign loans would reduce fiscal FY09 FY10 FY11 FY12 FY13 FY14 FY15 space available to the Sindh government and development expenditure would be badly Source: Data from Accountant General of Pakistan and Author affected (Figure 2.18). calculations 2.2.3 Development Expenditures: Poor Planning and Implementation Sindh’s current Vision 2030 was derived from Pakistan’s Vision 2025 document at the federal level, published in 2014. The Chief Minister of the province has officially approved and presented the Sindh Vision 2030 to the province, as its own version of the national vision for development. However, Sindh’s Vision 2030 has neither been debated nor ratified by the assembly. The Vision 2030 specifically includes the following six themes: (i) self-sufficiency in energy; (ii) quality education, health and safe drinking water; (iii) conservation of irrigation water, agriculture development and food security; (iv) an effective communications system; (v) industrial development through both the public and private sectors; and (vi) women’s empowerment and equal opportunities. The Sindh government does not have any all-encompassing provincial development strategy. There is, however, a Sindh Strategy for Sustainable Development (SSSD), which was prepared in 2007. Currently, the Sindh government is planning to develop a “Provincial Growth Strategy� with the help of the World Bank. Sector plans have been developed for education, health, agriculture, irrigation, energy, communications and mass transit.34 For education and health, plans were prepared in 2014 and 2012, respectively, but have not been updated since then, and the link between the sector plans and the budget preparation process remains weak. No district plans exist from a policy/financing perspective. District administrations are considered to be the executive arm of the provincial government and, as a result, no separate planning exercises are conducted. For the purpose of regional planning and infrastructure planning, the Government of Sindh passed legislation in February 2014 requiring every district to have a Master Development Plan. 34 Budget Strategy Paper2016-19, Government of Sindh. 44 Sindh: Public Expenditure Review 2017 The Budget Strategy Paper (BPS) presents projections for aggregate development spending, but the linkages with other budget documents are unclear. The BSP does not break down the forecasts of total development spending, either with respect to ministry/department/autonomous institution (MDA), or in terms of functional classifications as defined in the annual budget document. Therefore, it is not possible to compare the BSP forecasts of development spending with the numbers published in the annual budget appropriations, the PC-1, or other project documents. 2.2.3.1 Investment priorities A brief overview of sector plans presented in the BSP is based on the Sindh Vision 2030. However, the linkage between these two documents is not clearly established in the BSP. The priority sectors identified for the next 3 years are: (i) energy; (ii) Figure 2.19: Development expenditures communications; (iii) agriculture and irrigation; (PRs billion) and (iv) health and (v) education. However, it is 180 Nominal Real unclear how project costs in the sector plans are 160 incorporated into the expenditure forecast. 140 The Medium-Term Budgetary Framework (MTBF) 120 only covers recurring budget estimates with 100 expenditure ceilings that are usually incremental, but with no formal negotiation process occurring 80 between individual departments and the Finance 60 Department. The MTBF has been implemented in 40 eight (out of 33) departments since its inception in 20 2009.35 Core budget committees have been 0 established in each department in which the FY09 FY10 FY11 FY12 FY13 FY14 FY15 MTBF has been implemented. These are the primary decision-making groups within each Source: Data from Accountant General of Pakistan and Author department for all budget-related matters (target- calculations. setting, output indicators and strategic direction). Figure 2.20: Development budget execution (percent) Except for education and health, no other sector 0.90 plans or costed strategies have been developed. 0.80 0.77 0.71 0.71 In recent years, the composition of public 0.70 0.68 expenditure has tilted towards development 0.59 0.60 expenditures. Development expenditures, which 0.52 were only PRs 23 billion in FY09, increased to 0.50 PRs 179 billion in FY15, an average increase of 40 0.40 0.30 percent per year (Figure 2.19). In real terms, 0.30 development expenditures grew on average by 23 0.20 percent per year during this period. This is a significant increase compared with real recurrent 0.10 expenditures, which grew on average by about 9.5 0.00 percent per year (Figure 2.20). FY09 FY10 FY11 FY12 FY13 FY14 FY15 Nevertheless, weak planning and Source: Data from Accountant General of Pakistan & Author implementation have led to significantly lower calculations. spending than envisaged in the budget (Figure 2.21a/b). Higher allocations in the budget indicate the Sindh government’s desire to focus resources on major infrastructure, with the goal of spurring growth 35 Education and literacy, health, irrigation, agriculture, livestock and fisheries, social welfare, special education, and energy. 45 Sindh: Public Expenditure Review 2017 and creating jobs. However, these allocations do not guarantee the desired outcomes, especially in the absence of improvements in the soft aspects of investment, such as the efficiency of allocations and operations through better project selection and implementation. As a result, budget utilization is very weak. Budget execution is also weak and, in some years, only half of the planned expenditures are undertaken (Figure 2.20). There is a major difference between budgeted allocations and actual development spending. Economic affairs and general public services have seen significant increases in their respective shares over the past few years. The challenging law-and-order situation has also resulted in more allocations and spending on law and order. It is also clear from Figure 2.21a/b that spending patterns reflect ad-hoc decision-making, with actual expenditures varying considerably from year to year. Figure 2.21a/b also illustrates a significant difference between budgeted and actual expenditures. Figure 2.21a: Share in development budget Figure 2.21b: Share in development expenditures Others Others Social Protection Social Protection Housing and Community Amenities Housing and Community Amenities Education Education Health Health Economic Affairs Economic Affairs General Public Service General Public Service 100% 100% 0% 0% FY09 FY10 FY11 FY12 FY13 FY14 FY15 FY09 FY10 FY11 FY12 FY13 FY14 FY15 Source: Data from Finance Department & Accountant General of Pakistan and the author’s calculations. 2.2.3.2 Performance of Sindh Annual Development Plan (ADP) Portfolio The Sindh government has earmarked PRs 196.2 billion for development in FY17, a massive increase of 38 percent on the previous year (Figure 2.22). The Annual Development Plan (ADP) for FY17 places a major emphasis on projects that are aligned to the Sindh government’s development priorities. The major thrust of the ADP is towards irrigation, the improvement of infrastructure, and investments in the energy sector, with special emphasis on Thar coal infrastructure development. The education, health, nutrition, and youth affairs sectors were also prioritized. In addition, focus has been given to making agriculture and industry more economically viable. Moreover, special schemes have received an allocation of PRs 29 billion, while PRs 24.3 billion has been allocated to the Local Government Department (Figure 2.23). 46 Sindh: Public Expenditure Review 2017 Figure 2.22: Annual growth in ADP collection, Figure: 2.23: Sector wise ADP allocation in FY17 FY11-FY17 50% 45% 35 44% 29.0 30 26.0 38% 24.3 40% 25 21.7 20 17.2 PRs billion 16.015.0 30% 15 9.2 10 7.5 20% 5.5 5.4 4.0 3.9 3.8 3.2 3.1 5 1.3 10% 0 3% Irrigation Local Govt. Dept Industries & Inv Agriculture Matching Allocat Planning & Devel Sports, Culture Education & Lite W&S and Home Dep Others Special Schemes Energy Health Revenue and Finance SGA&CD Transport & Mass Social Protectio 3% 0% 0% FY11 FY12 FY13 FY14 FY15 FY16 FY17 -10% -20% -15% Source: Planning and Development Department, GoS. The size of the development portfolio has increased due to the inclusion of many new projects. The total number of projects has now reached to 2,686, including 934 new schemes added in the ADP FY17, compared with 2,352 projects including new 592 schemes in the previous year (Figure 2.24a).36 At the same time, the share of new projects in the ADP allocation has also increased, jumping from 25 percent to 41 percent in the past 2 years, while allocations to ongoing projects have significantly declined from 75 percent to 59 percent in the total allocations for FY17 (Figure 2.24b). This decline in allocations could result in time and cost overruns. Figure 2.24a: ADP composition - New vs ongoing Figure 2.24b: ADP allocation - New vs ongoing projects projects (% of total projects, RHS) (% of total allocation, RHS) New On-going Total Projects New On-going 100% 3,500 0.8 76% 75% 70% 90% 0.7 55% 3,000 61% 63% 80% 60% 58% 49% 0.6 70% 2,500 60% 74% 72% 71% 57% 0.5 42% 65% 2,000 40% 39% 37% 50% 0.4 75% 30% 40% 1,500 25% 0.3 24% 30% 51% 1,000 0.2 20% 45% 43% 35% 500 0.1 10% 26% 28% 29% 25% 00 0% 0 0 FY10 FY11 FY12 FY13 FY14 FY15 FY16 FY17 Source: Planning and Development Department, GoS. In contrasts to the principles of good budgeting, the Sindh ADP contains a significant number of unapproved projects. This practice creates a concern that some of the projects included in the portfolio may not have been adequately prepared. Many new projects enter the portfolio without prior approval with, for example, almost 34 percent of projects in the ADP FY17 being unapproved (Figure 2.25a). 36 A separate sector is called the “Thar-Coal Infrastructure Development�, including ongoing schemes in the Thar Desert with an allocation of PRs 11.53 billion. 47 Sindh: Public Expenditure Review 2017 Moreover, these unapproved projects command a significant share of the development budget, a share that has been on an upward trajectory in recent years. For instance, these projects have been allocated about 35 percent of the development budget in FY17, a significant increase on the 24 percent allocated in FY14 (Figure 2.25b). The large number of unapproved projects indicates inadequate capacity to prepare projects, likely leading to low implementation readiness and a limited pipeline of worthwhile projects. The result is likely to be delays and wastage. Figure 2.25a: ADP composition - Approved vs Figure 2.25b: ADP allocation - Approved vs unapproved unapproved projects (% of total projects) projects (% of total allocation) Approved Un-approved Approved Un-approved 100% 90% 83% 90% 15% 80% 76% 25% 24% 29% 25% 73% 73% 70% 80% 34% 70% 67% 67% 65% 43% 50% 70% 60% 60% 50% 50% 35% 40% 33% 33% 40% 85% 30% 75% 76% 71% 75% 27% 27% 66% 30% 24% 30% 57% 17% 45% 20% 20% 10% 10% 0% 0% FY10 FY11 FY12 FY13 FY14 FY15 FY16 FY17 FY10 FY11 FY12 FY13 FY14 FY15 FY16 FY17 Source: Planning and Development Department, GoS. Projects are selected for the ADP without having to go through an appraisal process, which is instead undertaken after a project has been selected (Box 2.2). New projects within the current year’s (FY17) ADP do not have an underlying feasibility study (PC-II), or a project planning document (PC-1) at the time the ADP is approved.37 Instead, the appraisal is done after selection. The reason for this is that the estimated costs of a project vis-à-vis the annual allocations in the ADP are unreliable and change significantly over time. It also leads to delays in starting the project, as the best part of the first year of execution is consumed in preparing the project documents (PC-II and PC-I). However, anecdotal evidence from other provinces suggests that projects that enter the ADP unapproved are prepared and implemented to lower standards than projects that undergo proper preparation before entering the ADP. Box 2.2: Project appraisal It is not possible to trace individual projects to the provincial strategy, because there is no single document that can be viewed as a long-term provincial strategy. While the Sindh government has prepared the Sindh Vision 2030, no medium- to long-term plan has been developed to translate this strategic vision into a definite public investment management plan. Thus, as a broad outline of strategic direction, the Sindh Vision 2030 may be a good reference point. The Sindh Vision was developed as part of the National Vision 2025 exercise in 2014. The Sindh government, with assistance from the World Bank, is currently developing a provincial growth strategy that would serve as a foundation for a public investment management program/development action plan. For those sectors where sector plans have been developed, the annual schemes are generally in line with sector policies. However, sector plans do not identify the actual projects and the funding requirements are provided at the policy level (aggregated). Sector plans have now been prepared for education, health, energy, agriculture, irrigation and drainage, communications and mass transit. The project appraisal process is centralized at the Planning and Development Department (P&DD) and 37 The provincial ADP is the approved (by the legislature) in the development budget for the year. The ADP contains a list of development schemes, along with the estimated total costs and allocations for the year. 48 Sindh: Public Expenditure Review 2017 undergoes various stages of screening before final approval is granted. Each project passes through multiple reviews at different forums depending upon their size and complexity. Before a project can be submitted for approval, it has to be cleared by a technical group comprising officials from P&DD and the relevant line department. For large and more complex projects, such as those in the energy or communications (multi-billion- dollar mass transit projects) sectors, a feasibility study or survey is mandatory before a PC-1 can be prepared. Large projects with high visibility tend to have greater political backing and more support during the project appraisal and approval process. There are no formal guidelines/checklist for appraisal, but a PC-1 requires that project objectives should be clearly traceable to the sector plan with proper justification being provided. However, during the project appraisal process very few queries are raised on the justification and its relevance to the sector plan. Although the approved unit cost rates are available for infrastructure projects, these are not updated in a timely manner, often resulting in significant under-budgeting of the project cost. There is also a lack of national guidelines for planning capital projects, which is an essential requirement. Compliance with PC-1 requirements pertaining to financial and economic analysis is weak. Due to the lack of capacity within line departments, compliance is limited to stating economic assumptions and future cash flows. Development projects are either identified by the line department (bottom up), or instructions can come directly from chief ministers (top down). For line departments, the P&DD annually issues guidelines for ADP preparation, which set out the criteria for project selection. While these guidelines are generally adhered to by the line departments, they are limited in their timeframe (to 1 year) and focus on annual budget allocation instructions (between new and old projects), rather than on establishing preliminary screening criteria for project selection. After the approval of the relevant minister, the project proposal/concept is submitted to the P&DD for review and finalization for inclusion in the ADP.38 There are significant time and cost overruns in the development portfolio. The ADP portfolio is highly fragmented, with a large number of small Figure 2.26: Operational efficiency (no. of projects, projects. The number of new projects has RHS) increased over the past 2 years, after a trend Total Projects Time Overrun (%) towards consolidation in FY15. By adding new Cost Overrun (%) 70% 3500 projects while also continuing with all existing 65% projects, this tends to dilute scarce resources for 60% 3000 55% the existing projects in the portfolio. 50% 53% 50% 49%2500 Consequently, around 50 percent of projects take 45% 40% 41% 42% 2000 longer than envisaged at the planning stage. In addition, cost overruns have occurred in about 10 30% 1500 percent of projects that should be completed 22% 20% 18% 1000 according to the ADP plan in FY17 (Figure 2.26). 16% 17% 12% 13% These delays and cost overruns together result in 10% 9% 10%500 lower economic returns to investment and the 0% 0 deferral of projected benefits into the future, in turn eroding the efficiency of the overall Source: Planning and Development Department, GoS. investment portfolio. 38 Those projects selected on the basis of instructions from the chief minister automatically qualify for the ADP. They do not go through the preliminary screening process for selection, but instead have very strong political ownership. A block allocation in the ADP is made for such projects and, similar to others, the detailed concept and project planning documents are prepared after the ADP approval. These projects not only receive preference on allocation and releases, but their execution is also monitored closely. There are instances where these projects are not part of the original ADP and have been conceived and approved during the financial year. In such cases, they are classified as a non-ADP project. 49 Sindh: Public Expenditure Review 2017 Part of the problem is due to low capacity in the line departments. The departmental staff responsible for preparing projects often does not have the required skills or the information needed to prepare high- quality projects. This results in continuous revisions in the timeframe and the costs of the projects. As a result, it is difficult to manage the overall ADP portfolio, resulting in significant throw forward of the portfolio. In addition, line departments tend to under-cost projects so they can be included in the ADP. Project authorities do not have reliable information for committing expenditure to capital projects and, in most cases, the commitments are much more than the funds available. Once the ADP is approved, the line departments have no discretion regarding prioritization. The release of funds for development expenditure is at the discretion of the P&DD, which releases funds for projects based on numerous variables. In the current year, FY17, the Sindh government has issued a release strategy splitting funds between ongoing and new projects, where Figure 2.27: Release pattern across fiscal years (% of ongoing projects are given priority. The strategy the original budget) also takes into account the size and stage of completion of capital projects. 120 Ongoing New Total In the current year, releases for development 100 projects are once again suffering from an inconsistent release pattern. Annual releases are 80 skewed towards the final quarter of the fiscal year, as evident in the current year, FY17. In this 60 fiscal year, after 6 months only 30 percent of the total allocations had been released. In addition, 40 the release of funding for new development 20 projects suffers from the consistently late approval of the PC-1 documents, which is one of 0 the key factors in the delayed start of projects. FY14 FY15 FY16 FY17 The process for PC-1 preparation takes around 3-6 Source: Finance Department, GoS. months after the start of the fiscal year. All of these issues result in significant delays in the actual execution of projects. For instance, during FY13, about 58 percent of projects were present Figure 2.28: Projects with no expenditures (% of total in the portfolio for 1 year but no spending projects) occurred, while 17 percent of projects were 1yr > 1yrs 70% present for more than 1 year, but no spending occurred. This situation has deteriorated 60% 58% subsequently, as about 25 percent of projects in 48% FY17 have remained inactive for more than 1 year 50% 41% 41% after their inclusion in the ADP (Figure 2.28). 40% 34% 37% 33% Interestingly, some of these projects have been in 29% 31% 30% the portfolio for more than 5 years, implying that 30% 26% 25% 19% 21% the portfolio would benefit from a major 20% 17% 14% “cleaning out�, with the removal of all those projects that are no longer economically viable. 10% Over the years, the proportion of relatively large 0% projects has remained high. About 65 percent of FY10 FY11 FY12 FY13 FY14 FY15 FY16 FY17 the projects in the FY17 ADP have a total Source: Planning and Development Department, GoS. estimated cost of more than PRs 50 million, of which 44 percent of total projects have an estimated cost of more that PRs 100 million, while only 8 50 Sindh: Public Expenditure Review 2017 percent of all projects in the FY17 ADP have a cost Figure 2.29: Portfolio distribution in terms of cost, FY10 below PRs 20 million (Figure 2.29). There may be a to FY17 (% of total projects) need to consolidate small projects, as smaller Less than Rs. 10 million Rs. 10-20 million projects still require significant administrative Rs. 20-50 million Rs. 50-100 million resources. This implies higher implementation Rs. 100-500 million Over Rs. 500 million costs and greater resources at the P&DD for 35% monitoring and implementing projects, resulting 30% in lower economic or social returns. 25% Meanwhile, the average allocation is increasing 20% at a far slower pace than the average project 15% size. ADP allocations averaged around 20 percent of project costs (Figure 2.30a). 10% 5% 0% FY10 FY11 FY12 FY13 FY14 FY15 FY16 FY17 Source: Planning and Development Department, GoS. Figure 2.30a: Ratio of average allocation to average Figure 2.30b: Distribution of portfolio allocation, FY10 to cost, FY10 to FY17 FY17 (% of total projects) 25.0 Less than Rs. 10 million Rs. 10-20 million 23.8 22.8 22.9 Rs. 20-50 million Rs. 50-100 million 21.7 21.5 20.0 Rs. 100-500 million Over Rs. 500 million 18.6 70% 17.8 15.9 60% 15.0 50% 40% 10.0 30% 5.0 20% 10% 0.0 0% FY2010 FY2011 FY2012 FY2013 FY2014 FY2015 FY2016 FY2017 FY2010 FY2011 FY2012 FY2013 FY2014 FY2015 FY2016 FY2017 Source: Planning and Development Department, GoS. Despite a higher average cost than average allocation ratio, projects are likely to be completed within 5 years. For instance, almost 84 percent of projects in the ADP portfolio (59 percent of them are ongoing) are expected to be completed within 5 years. Nevertheless, the current level of allocations for some projects is clearly insufficient to ensure timely completion. Around 154 projects (6 percent of all projects) in the FY17 ADP portfolio will require between 11 and 100 years completing at the current levels of allocation (Figure 2.31). Some of these projects may not be receiving their required allocations due to changes in priorities. Therefore, a complete review and overhaul of the portfolio should be undertaken. 51 Sindh: Public Expenditure Review 2017 Figure 2.31: Project completion time with current level of allocation 70% less than 1 yr 1-5 yrs 6-10 yrs 11-20 yrs More than 21 yrs 59% 60% 50% 42% 40% 38% 40% 35% 30% 28% 30% 24% 20% 10% 0% FY2010 FY2011 FY2012 FY2013 FY2014 FY2015 FY2016 FY2017 Source: P&DD, GoS. Source: Planning and Development Department, GoS. As a consequence of implementation delays and Figure 2.32: Throw forward by sector, FY17 the dilution of overall development spending, throw forward of the development portfolio is 18% 17% significant. The inclusion of new projects every 16% 14% 13% year, coupled with the slow implementation of 11% 12% 12% the existing portfolio, adds to substantial throw 10% 10% forward of commitments—funds needed to 8% 7% 6% complete existing development projects —which 6% 4% 4% 4% 4% 3% 3% 3% has increased by 13 percent compared with FY16. 2% 2% 2% 1% 1% During FY17, around 83 percent of projects have 0% throw-forward of over PRs 20 million. Much of the throw forward is found in the Irrigation Department, followed by the Works and Services Department, the Local Government Department and the Education Department (Figure 2.32). The Sindh government needs to focus on efficient Source: Planning and Development Department, GoS. completion, as well as rationalizing the size of the portfolio by eliminating non-performing and unapproved projects. District analysis shows that per-capita ADP allocations vary significantly across districts. Per capita average allocation over the past 3 years shows that Tharparkar and Sujawal districts are the top recipients, with PRs 16,184 and PRs 14,796 per capita allocations, respectively, followed by Thatta, SBA, Khairpur, and Jamshoro. The districts receiving the lowest per capita allocations include Ghotki, N. Feroze, Shikarpur, Sanghar, and Umerkot. Moreover, a review of the past 3 years’ average per capita allocations shows that the top 25 percent of provincial population districts received around 40 percent of total ADP allocations, whereas the bottom 25 percent received only 11 percent of total ADP funds (Figure 2.33). It is important to review these allocations to ensure that the disparities correspond to explicit policy objectives. 52 Sindh: Public Expenditure Review 2017 Figure 2.33: Per capita ADP allocation, average FY15 to FY17 (PRs billion) Top 25% of population on average recieves more than 40% of ADP allocation 18.0 16.0 14.0 12.0 10.0 8.0 Bottom 25% population on average recieves 6.0 11% of ADP allocation 4.0 2.0 0.0 Source: Planning and Development Department, GoS. Note: This analysis excludes the allocation for province-based projects amounted PRs 48 billion, PRs 45.6 billion and PRs 68 billion for FY15, FY16 and FY17, respectively. 2.2.3.3 Management of project implementation The P&DD carries out centralized monitoring through its M&E section, which regularly submits its reports to the secretary of the P&D. The M&E section has a network district monitoring officer, who visits each project once a month within his/her respective district. These reports include issues regarding the project’s financial and physical progress, quality, and reasons for any delays. The major focus of M&E is on infrastructure projects and very few reports are issued in respect of social sector reform programs. Similarly, there is no consolidated view of physical or financial progress of capital investment projects. In addition, it is not evident just how these reports are used or discussed in monitoring committee meetings for effective monitoring and mid-course correction. For all mega infrastructure projects, Project Management Implementation Units (PMIU) are established with a project director directly appointed by the Sindh government. The project director is selected by the Chief Minister or the minister of the relevant line department from an existing pool of government officers, who may or may not have the relevant qualifications or experience for project management. The turnover of project directors is low and generally the senior staff does not change during the life of the project. For technical expertise, consultants are hired on short-term contracts. Project authorities find it difficult to appoint consultants on a market-based salary due to existing government hiring policies, which offers remuneration significantly lower than that of the market. As a result, consultant turnover is high, disrupting efficient work flows vis-à-vis project management. Owing to weak contract management capacity, delays occur due to frequent disagreements between contractors, consultants and the PMIU. For mega infrastructure projects, there is a three-way contract agreement between the government (PMIU), the contractor and the consultant (engineer/supervision). The roles and responsibilities shared between the PMIU and consultants/firms are demarcated into the contract, but there is no standard practise of carefully vetting or seeking legal opinions before awarding a contract. This has resulted into many contracts being open to interpretation and, as a result, at risk of being abused. The payment process is smooth and efficient, as the PMIU is provided with a dedicated special bank account, a Revolving Fund Account (RFA). An RFA is part of the Treasury Single Account (TSA), where funds are earmarked for the project in the form of a release. This arrangement allows project authorities 53 Sindh: Public Expenditure Review 2017 to make direct payments from the RFA to the contractor.39 The contractor’s bills are verified by an independent consultant for work done, which is subsequently approved by the project director prior to payment being made.40 The Sindh government has issued clear guidelines for payment processing and the maintenance of underlying records, which include a “measurement book�—a core document that forms the basis of a contractor invoice and payment system. No project planning is carried out and throughout the project life the PC-1 remains the only document used as a reference for planning. The PC-1 contains a high-level project implementation schedule, but without any details or step-by-step implementation process. Project authorities do not use any project planning tools to identify critical activities and flag potential bottlenecks (network analysis). Most importantly, for small infrastructure projects executed by the Works and Services Department (W&SD)41 most of the project management is carried out via informal communication and documentation. 2.3 Recommendations Revenues Review the fiscal transfer mechanism to bring it more into line with the needs of both levels of government and make it more “tax friendly�. This requires each level of government to take a much broader and longer-term view of revenue needs and potential collections at the 9th NFC Award. Revamp property tax to raise collection by improving policy and administration. This requires having more realistic valuation rolls, rationalization of exemptions and tax rates, levying tax on vacant properties, indexing the tax base, and reducing the differential in rates between owner-occupied and rented properties, and between residential and commercial properties. Enhancing revenue yields of agricultural income to enhance equity in the tax system. This can be done by rationalizing exemptions, indexing the tax base (especially of land tax), and enhancing capacity of the BOR to enable it to better assess tax from the income mode. Broaden the base of GST on services. It is important to gradually bring commercial and transportation services into the tax net, merging the professional tax into the GST on services, together with as many stamp duties as possible. Recurrent Expenditures There is need to ensure that sufficient resources are available for maintaining existing infrastructure so that it pays off in the longer term. This requires undertaking a complete stock-take of public physical assets, noting attributes such as age and usage, and updating this database annually. In addition, a district-level requirement for repair and maintenance expenditures should be drawn up, together with a link between development expenditures and recurrent budgeting to better understand the future spending needs for repairs and maintenance. A complete review of public-sector remuneration and the pension system is required. It is proposed that the Sindh government undertake a medium-term assessment of the skills required to provide its services and to publish this for complete transparency. This should then be followed by the preparation 39 In Pakistan, the accounting and payments are centralized with Controller General of Accounts (CGA) who works through network of district account officer to record and process payments. All bills are sent to DAO for payments but in case of RFA, the project authorities can make payments directly to the contractor. 40 The project director has financial management team for accounting reporting and processing of the payments. 41 Each project under the W&SD is headed by an executive engineer (XEN). An XEN can have more than on project. 54 Sindh: Public Expenditure Review 2017 of an HR assessment of the public sector to better understand the skills available, the future HR needs, and the skills required to fill those jobs. The Sindh government should also consider undertaking an actuarial analysis of the pension system to better understand pension costs in the medium to longer term, and use this analysis in the medium-term budget framework to improve its credibility. A better forecast of the resource envelope should be considered to improve budgetary accuracy, together with a complete evaluation of the release mechanism. This requires capacity building at the Finance and P&D Departments, strengthening the medium-term budget framework by including the development budget in the medium-term budget strategy paper and making it a part of the MTBF. This would help the government to link future operating costs with current investment decisions and establish fiscal discipline over the medium term. The government should also consider introducing a rules-based release mechanism in case of fiscal constraints, while also enforcing a ceiling for supplementary grants above which ex-ante parliamentary approval is mandatory. There is a need to align budget allocations with medium- to long-term sectoral goals. At present, Sindh prepares its budget on an incremental basis that does not allow a complete re-evaluation of future requirements. There is a need to prepare sector strategies and budget these according to required outputs and outcomes, together with a need to strengthen the coordination mechanism between recurrent and development budgets. Ideally, this would allocate a one-line budget to line departments without hard distinctions between two, and allow reallocation between recurrent and development budgets based on the approved sector strategy. Undertake an evaluation of the fiscal costs of different fiscal arrangements, such as SOEs and PPPs. This requires the establishment of a Fiscal Management Framework (FMF) for PPPs. The objective of the FMF is to assess, monitor, forecast, and report the whole range of fiscal commitments and contingent liabilities42 that can be created by PPPs, as well as strengthening the SOE monitoring mechanism and establishing an up-to-date database. Development Expenditures The Sindh government has developed a Vision 2030 that should be followed by the development strategy to translate aspirations into actions. This will require the preparation of a development framework establishing broad fiscal parameters, development objectives, and priority programs. There is also a need to carry out consultations with the legislature and the public dissemination of a draft document for public feedback within a given timeline; update and approve a development framework from the Cabinet; develop sector strategies in line with the development framework; and identify priority projects and prepare a consolidated portfolio of pipeline projects. Budgets relating to all public investment projects should be entered into the government's financial management information system (GFMIS). There is need to prepare development budgets using the New Accounting Model (NAM) chart of accounts for all public investment projects to ensure transparency and completeness. Development budgets should be appropriated across all relevant elements of the chart of accounts. Currently budgets for most projects are lumped into only two items: salaries and operating expenses. A “Revised accounting procedure for Revolving Fund Accounts (RFA)� should be implemented for both development partner- and GoS-funded projects. The rules require that project 42 We refer to the explicit contingent liabilities, namely those that are subject to a contractual engagement. 55 Sindh: Public Expenditure Review 2017 authorities must reconcile and report expenditure to the Accountant General’s (AG) office on a monthly basis, thus enabling detailed accounting of the RFA funds in the GFMIS.43 An integrated database for public investment projects should be developed. There is a need to have a consolidated view of the progress of public investment projects within the DG M&E in the P&DD. Currently, this view is restricted to individual progress reports. Similarly, there is a need to prepare an asset safeguard policy and plan to develop a comprehensive non-financial-asset register at the P&DD. An integrated PIM information system (PIMIS) should also be developed and implemented, with an interfaced created with the PIFRA system for financial data and a PIM reporting framework developed. A PIM monitoring committee should also be established within the P&DD to review development progress and to issue directives for project authorities. Use a criteria-based approach for project screening and project approvals. International experience suggests that political considerations trump objective evaluations when it comes to making choices for public spending.44 There is a growing realization that the issue of unsystematic decision-making and ad- hoc selection can only be addressed if appropriate information/analysis is presented to decision-makers, enabling them to evaluate the opportunity costs of their decisions.45 Therefore, there is a strong case for the P&DD to establish project appraisal criteria. Such a framework would provide basic guidelines to differentiate projects with respect to the government’s development priority ladder. The criteria should be simple and easy to use. The P&DD should issue an evaluation framework that will serve all decision nodes in the project screening process (development strategy  approved pipeline projects  sector plans proposed project) and should develop a protocol for early and extensive engagement of various stakeholders in development budgeting. The stakeholder spectrum would likely include legislatures, civil society members, media, etc. At the project level, the P&DD should review the “Manual for Development Project� in light of emerging requirements for projects. The two chapters in the manual that will specifically require updating are Chapter 4: Project Preparation and Chapter 5: Project Appraisal.46 (Project preparation covers defining project objectives, scope, stakeholders, justification, geographical information, resource requirements and outcomes.) The government should develop guidelines to comply with PC-1 requirements and also provide legal cover (P&DD notification) so that the DG M&E47 can ensure compliance with PC-III (cash flow projections), PC-IV (project completion), and PC-V (monitoring project outcomes). 43 RFA is a designated bank account (notional) under the Treasury Single Account (TSA) concept generally used for large projects, where single line transfer is made into the RFA and, thus, no break downs of the budget or expenditure are available in the GFMIS. The detailed accounting is undertaken by the project authorities. 44 Public Project Selection and Evaluation, Alex Mosesova and Sudhakar Kota. 45 An Alternative Approach to Project Selection: The Infrastructure Prioritization Framework Darwin Marcelo, Cledan Mandri- Perrott, Schuyler House, Jordan Z. Schwartz World Bank PPP Group April 14, 2016. 46 Currently, PC-1 form has 14 different formats as per specific requirements of each sector. “Manual for Development Project 2010�- Ministry of Planning Development and Reforms is also adopted and used by the provincial government. 47 DG M&E is part of the P&DD’s responsibility for project monitoring 56 Sindh: Public Expenditure Review 2017 Chapter 3: Are we getting the value for money? Sindh Education Sector 3.1 Introduction This chapter presents a Public Expenditure Review (PER) of the education sector in Sindh Province. The main objective of the chapter is to analyze the performance and adequacy of public expenditure in meeting the education policy objectives of Sindh Province. The chapter provides three analyses in order to achieve this goal: (i) a profile and trend analysis of service delivery statistics to present a profile of education sector expenditure and also comment on their adequacy; (ii) a profile and trend analysis of education sector expenditure in relation to the economic, functional, and objective classifications of that expenditure and its allocative efficiency; and (iii) an analysis of the effectiveness and equity of education sector expenditure by linking the expenditure to educational outcomes. This third analysis also highlights regional (urban/rural) and gender-based inequalities of expenditure, together with the absorptive capacity of schools to cater for the enrolment of additional children into the education system. Finally, conclusions and recommendations are provided at the end of the chapter for improving the effectiveness of fund utilization in the education sector. The analyses utilize secondary data from the following sources: the Project for Improving Financial Reporting and Auditing (PIFRA), the Sindh Education Management Information System (SEMIS), the Pakistan Social and Living Standards Measurement (PSLM) survey, and Student Achievement Test (SAT). The first two sources provide data on education sector inputs, while the second two sources provide information on educational outcomes. 3.2 Sindh Education Sector Overview The education sector of Sindh comprises of three major departments responsible for six sub-sectors as identified in Table 3.1 below.48 Mainly due to its relative importance in the Sindh government’s development vision, the scope of this PER includes expenditure from the pre-primary through to higher secondary education levels.49 Table 3.1: Public education sector overview 50 Department/ Responsible agency Education sub-sectors Grade levels/years of education School Education Department Katchi/ Early childhood education (ECE) Katchi and pre-primary (SED) Primary education Grades 1-5 Middle/ Elementary education Grades 6-8 Secondary education Grades 9-10 College Education Department Higher secondary education Grades 11-12 Sindh Technical Education and Technical/ Vocational education Vocational Training Authority 48 Sindh Education Sector Plan 2014-2018. 49 Throughout this chapter the “education sector� refers to primary, middle and secondary education (Grades K -10) unless stated otherwise. 50 Pakistan’s education system is divided into the following levels: primary school, or Grades 1-5 and middle school, or Grades 6- 8. Primary and middle school combined are often referred to as elementary school (i.e., Grades 1-8). Secondary school is divided into two levels, namely high school, or Grades 9-10, and higher secondary school, or Grades 11-12. Grades 13+ are classified as tertiary education. 57 Sindh: Public Expenditure Review 2017 (STEVTA) Higher Education Department Tertiary education 13-years of education and (HED) and Higher Education above Commission (HEC) Source: Staff exposition The public sector remains the primary service provider of education in Sindh, catering to 60 percent of enrolled children (ages 5-16) in the province. However, the private provision of education services is sizable and growing, although data are lacking.51 Currently, there are more than 45,000 public schools in Sindh with a total enrolment of about 4 million students and a teaching force of about 150,000 spread across the province. Overall, the education system is highly centralized, with administrative powers concentrated at the provincial level within the Education and Literacy Department. The Education and Literacy Department (E&LD) consists of two main branches: a secretariat and the allied institutions. The latter include various directorates, including the Directorate General of School Education (DGSE), which has a wide presence at the provincial and district levels, and is responsible for the implementation and management of education service delivery. Other institutions that play a critical role in the implementation of basic education reforms include the Reform Support Units (RSU), the Sindh Education Foundation (SEF), which is the conduit for public-private partnerships (PPPs) in education, the Sindh Textbook Board, the Curriculum Wing, Provincial Institute of Teacher Education (PITE), Sindh Teacher Education Development Authority (STEDA) and the Directorate General of Monitoring and Evaluation (DGM&E). 3.3 Context and Policy Framework The Sindh government, in accordance with the National Education Policy 2009, passed the Sindh Right of Children to Free and Compulsory Education Act in 2013, as envisaged in Article 25A of the Constitution. This 2013 law on the right to free education stipulates that it will be provided to children from the age of 3 onwards until their completion of secondary education. Education is a priority that is also reflected in the province’s Budget Strategy Paper (FY17 to FY19). The Budget Strategy Paper (BSP) of the Sindh Finance Department (GoS) declares that education is a priority and aims to address three challenging areas highlighted in the Sindh Education Sector Plan 2014-2018 (SESP), namely: (i) access; (ii) quality; and (iii) governance with a focus on eliminating social exclusion. The Sindh government has implemented two phases of the Sindh Education Reform Program (SERP), SERP I and II, beginning in 2009. The objectives of SERP are to increase school participation, retention, and completion rates by improving sector governance and accountability, strengthening administrative systems, and providing direct incentives to students to enroll in education. To support these objectives, the program has introduced various reforms, including merit- and needs-based teacher recruitment; teacher education and development reforms; increasing access to schooling through public-private partnerships; establishing an assessment regime to conduct Grade 5 and 8 assessments; providing school specific non-salary budgets (SSB) directly to schools; and school consolidation. The SESP serves as a guiding document to achieve education sector goals. In order to implement the SERP reform agenda, the RSUs together with development partners and the E&LD have developed a wide ranging SESP 2014-2018 that represents the input of 12 technical groups. Education has a critical role to play in achieving the provincial objectives of reducing poverty and regional inequality, and the 51 The share of enrolled children is derived from PSLM FY15 household survey. Data on private school enrolment are not available. The service delivery statistics on public schools are made available annually through the Annual School Census (ASC). 58 Sindh: Public Expenditure Review 2017 SESP serves as a guiding document to help achieve the education goals in the sector. The priorities highlighted in the SESP are:52 (i) Increasing equitable access to early childhood education (ECE), primary, middle/elementary and secondary education; (ii) Improving the curriculum and learning outcomes; (iii) Improving teacher quality; (iv) Strengthening governance and service delivery; (v) Improving resource allocation; (vi) Addressing adult literacy and non-formal basic education; and (vii) Addressing cross-cutting areas, such as ICT, education in emergencies, gender equity, social cohesion, and PPPs for education. 3.4 Education Sector Profile Analysis The School Education Department (SED) provides education services to more than 4 million students through its 46,000 schools and 144,170 teachers (SEMIS, FY15). The Sindh E&LD’s responsibilities include catering for school education, technical education, incentive programs, and development schemes to promote education in Sindh. 3.4.1 Sindh Enrolment and Learning Outcome Trends Enrolment rates have registered minimal growth over the past decade. Figure 3.1 highlights school enrolment rate trends over the past 10 years utilizing data from the nationally and provincially representative Pakistan Social and Living Standard Measurement (PSLM) survey. Compared with FY05, there was a modest annual increase of 0.5 of a percentage point across almost all enrolment rates. Overall, enrolment rates increased until FY09. However, after FY11, they fell marginally until FY13. The reason for this more recent decline in enrolment as cited in the PSLM report was the devastating impact of the floods that occurred in 2010 and in subsequent years. However, over the 2 years from FY13 to FY15, marginal improvements in primary education net enrolment rates (NER) were evident for both boys and girls (2- to 3-percentage-point increases), indicating that NERs are back on a positive track. The middle-school level NER for both genders stagnated over the 2 years from FY13 to FY15, while the gender gap in enrolment narrowed at the secondary level. Interestingly, the secondary (matriculate) level NER for girls showed an improvement and was 3 percentage points higher than for boys, revealing a reduction in the gender gap at this level. 52 Sindh Education Sector Plan 2014-18, E&LD, Government of Sindh (Page 16). 59 Sindh: Public Expenditure Review 2017 Figure 3.1: Net enrolment trends by gender and level, 2005 to 2015 GER Boys GER Girls Primary Middle Secondary Primary Middle Secondary 100 100 80 80 60 60 40 40 20 20 0 0 04-05 06-07 08-09 10-11 12-13 13-14 14-15 04-05 06-07 08-09 10-11 12-13 13-14 14-15 NER Boys NER Girls Primary Middle Secondary Primary Middle Secondary 100 100 80 80 60 60 40 40 20 20 0 0 04-05 06-07 08-09 10-11 12-13 13-14 14-15 04-05 06-07 08-09 10-11 12-13 13-14 14-15 Source: PSLM Survey Reports. There was no reduction in the percentage of Figure 3.2: School participation trends by schooling out-of-school children in the period since FY09 status to FY15. The proportion of out-of-school children Public Private OOSC (OOSC) remained stagnant from FY11 to FY15. 46 This is shown in Figure 3.2, with the overall 43 41 39 trends of school participation in Sindh for 35 36 36 36 children between the ages of 5 and 16. However, the share of private school participation 26 22 increased by 7 percentage points, while public 19 21 school participation decreased by the same proportion from FY09 to FY15. At the district level, school participation rates for children between the ages of 5 and 16 vary, 08-09 10-11 12-13 14-15 and were low for most districts in the province. Source: PSLMS Household data. Only three districts—Karachi, Dadu and Naushahro Feroz—had more than 70 percent of children enrolled in school. Fourteen districts of the 60 Sindh: Public Expenditure Review 2017 total 24 districts53 in the province have less than half the school age population enrolled in school, and two of these districts (Tando Mohammad Khan and Kashmore) have overall participation rates of below 40 percent. Figure 3.3: School participation rates by district, FY15 85% Dadu 73% 71% Hyderabad 63% 62% Larkana 62% 60% Shaheed Benazir Abad 60% 59% Matiari 52% 49% Shahdadkot 49% 49% Tharparkar 48% 47% Umer Kot 46% 46% Shikarpur 46% 44% Jacobabad 43% 43% Sujawal 41% 38% Tando Mohammad Khan 35% Source: PSLMS data FY15 There is a large differential in gender participation rates for children between age 5 and 16 across Sindh. As reflected in Figure 3.4, across all districts (with the exception of Karachi) female participation rates are lower than male participation rates. The districts of Ghotki, Tharparkar, and Sanghar have the greatest gender differential in participation rates, at more than 25 percentage points. Conversely, Karachi and Dadu districts have the closest gender parity.54 Figure 3.4: Female and male participation rates, by district Female Male 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% . Source: PSLMS data FY15 53 In fact, there are 29 districts in Sindh, as there are several individual districts in the urban areas of Karachi, Hyderabad and Larkana. However, in this research the districts in urban areas are combined, hence the appearance of 24 districts instead of 29. 54 Source: PSLM data FY15. 61 Sindh: Public Expenditure Review 2017 There are a high percentage of out-of-school children, particularly in the early years and at the secondary level of schooling, with a persistent gender gap (across all age levels) in rural areas of Sindh. In Pakistan in general, but in Sindh in particular, there is a growing concern about out-of-school- children (OOSC) and early dropouts. Figure 3.5 shows school participation rates in both public and private schools by age, urban/rural location, and gender from household survey data. Most notably, urban girls are on a par with urban boys and the gender gap in school enrolment is eliminated at all ages in the urban areas of Sindh. However, in rural areas the story is very different, with school participation much lower than in urban areas, and the gender gap far more apparent. Rural girls are less likely than rural boys to be enrolled in school are and the gender gap increases with age, particularly when girls reach the age of 11 and beyond.55 Figure 3.5: School participation rates by age, gender and urban/rural Rural Boys Rural Girls Urban Boys Urban Girls 100 80 60 40 20 0 4 5 6 7 8 9 10 11 12 13 14 15 16 Age in years Source: PSLMS data FY15 Late entry into the schooling system is a major challenge in expanding early childhood education (ECE) coverage in Sindh. Although boys fare better than girls at all ages in terms of enrolment outcomes, it is clear that early childhood enrolment (ages 4 and 5 years) is much lower than desired for both boys and girls. Children in urban areas are more likely than rural children to enroll in school at earlier grades. However, parents in both rural and urban areas appear to have no apparent bias against girls when it comes to early childhood enrolment, suggesting that age and urban/rural location affect girls’ enrolment only at a later age. Student learning outcomes are poor particularly in public schools. However, there has been some improvement in learning outcomes over the 5 years to 2015. Figure 3.6 shows Grade 3 student learning outcomes for public and private school students for Urdu/Sindhi and arithmetic in rural Sindh, as reported in the Annual Status of Education Report (ASER). It is evident that private school students on 55 Statistics derived from PSLM FY15. 62 Sindh: Public Expenditure Review 2017 average perform better than public school students in both language and arithmetic. For public school students, there was significant improvement in Urdu/Sindhi learning outcomes between 2011 and 2015, and a 9-percentage-point increase in Grade 3 arithmetic learning outcomes during the same period. Despite this positive trend, only 32 percent of Grade 3 students can perform subtraction—one of the expected Grade 3 student learning outcomes. Figure 3.6: Grade 3 student learning outcomes in Urdu/Sindhi and arithmetic Grade 3 Student learning outcomes (Urdu/Sindhi) Grade 3 Student learning outcomes (Arithmetic) Public Grade 3 Students: can read a sentence in Urdu Public Grade 3 Students: Can atleast do subtraction Private Grade 3 Students: can read a sentence in Urdu Private Grade 3 Students: Can atleast do subtraction 70 70 60 60 Percentage of students Percentage of students 50 50 40 40 30 30 20 20 10 10 0 0 2011 2012 2013 2014 2015 2011 2012 2013 2014 2015 Source: Data from Annual Status of Education Report. 3.4.2 Public School Education Trends The number of public schools has declined over the past five years, although the Sindh government has taken some steps to address the issue of closed schools. The service delivery profile of the education sector in Sindh between FY11 and FY15 is shown in Table 3.2. As expected, public education service delivery is dominated by primary schools. The number of primary schools declined continuously over the period of this review (from 48,914 to 46,039). This decline, however, was mainly due the government’s efforts to consolidate primary schools, close non-functioning schools, and eliminate ghost schools. Similarly, middle schools also declined (from 2,505 to 2,316). However, high and higher- secondary schools increased from 1,887 to 1,999. The Sindh government has taken steps to address the issue of closed schools: in 2012, the government identified those schools that had been closed due to a lack of teachers but that still had potential for enrollment. These schools were then re-staffed with teachers and subsequently reopened. Student enrolment in public primary schools decreased by about 0.5 million over the 5 years to 2015. Of particular concern is the fact that enrolment in public primary schools declined by 15 percent between 2010 and 2015, although there was a marginal improvement in enrolment at the elementary (primary and middle) level, with an 8-percentage-point increase in enrolment over the same period. At the secondary level, there was a sustained increase in enrolment, with a 31 percent increase in enrolment following FY11. This trend suggests that even if the transition rate from primary to middle levels is low, a large portion of students who do manage to progress to middle school continue their education to higher (secondary) grades. There is a huge gender gap in the public education system and also a significant number of students (both boys and girls) who do not transition to the secondary level. The enrolment of boys in Sindh far outnumbers girls across all levels of the public education system (Figure 3.7). Gender parity is relatively higher at the middle-secondary level: data show that for every 100 boys, 64 girls are enrolled at the 63 Sindh: Public Expenditure Review 2017 primary level, 74 girls are at middle-secondary level, and 54 girls are attending higher secondary school. Annual School Census data from FY15 also show that enrolled girls fare better than boys in the transition from the primary to the middle-secondary level, but face higher drop-out rates than boys at the post- secondary level. Figure 3.7: School enrolment pyramid (number of students by level and gender) Hi. Secondary 119,643 221,010 Boys Middle - Secondary 449,572 608,383 Girls Primary 1,030,313 1,615,555 Source: Data from SEMIS FY15, and staff exposition Table 3.2: Education profile by year and level of public sector schools EMIS Survey Year Description (Unit) 2010-11 2011-12 2012-13 2013-14 2014-15 Trend Total School (numbers) 48,914 47,557 47,394 46,724 46,039 Primary 44,522 43,089 42,900 42,342 41,724 Middle/ Elementary 2,505 2,554 2,429 2,336 2,316 Secondary/ Hi Secondary 1,887 1,914 2,065 2,046 1,999 Total enrollment (numbers) 4,215,076 4,222,160 4,249,033 4,085,415 4,044,476 In primary school 3,108,274 2,980,446 2,963,622 2,802,824 2,645,868 In middle/ elementary schools 233,279 280,128 263,910 246,893 252,824 In secondary/ Hi. Secondary schools 873,523 961,586 1,021,501 1,035,698 1,145,784 Total teachers (numbers) 144,610 146,103 142,639 145,438 144,170 In primary school 102,061 99,254 96,401 92,521 87,085 In middle/ elementary schools 9,959 12,063 10,755 11,980 12,278 In secondary/ Hi. Secondary schools 32,590 34,786 35,483 40,937 44,807 Overall student per teacher (ratio) 29.1 28.9 29.8 28.1 28.1 In primary school 30.5 30.0 30.7 30.3 30.4 In middle/ elementary schools 23.4 23.2 24.5 20.6 20.6 In secondary/ Hi. Secondary schools 26.8 27.6 28.8 25.3 25.6 Source: Data from SEMIS FY15, and staff exposition Four out of every 10 public primary school children were found to be absent in Sindh, and student absenteeism in girls’ primary schools is higher than in boys’ primary schools. Even those children who 64 Sindh: Public Expenditure Review 2017 are enrolled in public schools do not necessarily regularly attend school, as illustrated in Figure 3.8. An analysis of the Sindh Schools Monitoring System data collected by the DGM&E56 finds that for primary schools that were visited by monitors over the past 5 months, student presence ranged from 59 to 68 percent. Student presence was lowest for middle schools (ranging between 51 and 57 percent), followed by high schools (52 to 58 percent). Across all levels of schooling (except primary and middle schools), student absenteeism was lowest in the month of January, which could be attributed to the fact that the winter break was extended to the second week of January 2017 due to a cold wave in Sindh. For primary schools, student absenteeism in girls’ schools is consistently higher than in boys’ schools, ranging between 3 and 8 percentage points during the FY11 to FY15 period. Figure 3.8: Public school student attendance, by level and gender Primary Elementary Boys Primary Schools Girls Primary Schools 66% Middle Secondary 66% 64% 64% 62% 62% 60% 60% 58% 58% 56% 56% 54% 54% 52% 52% 50% 50% Oct Nov Dec Jan Feb Oct Nov Dec Jan Feb 2016 2017 2016 2017 Source: DG Monitoring and Evaluation, Oct 2016 to Feb 2017. Student-teacher ratios have remained constant at the primary level and declined at higher levels, as both the numbers of teachers and students at the primary level have decreased. The number of teachers at the primary level has decreased substantially over the 4 years to 2014-15, falling by 15 percent. Given that primary enrolment also decreased over the same period, student-teacher ratios (STR) were relatively stable at the primary level. At the middle to higher secondary levels, the number of teachers actually increased substantially, by 34 percent and, as a result, the STR in these school levels declined. Teacher recruitment has been unable to keep pace with the demand for primary school teachers. It is estimated that about 12,000 teachers retired between FY11 and FY16 (Table 3.3). This reduction in the teaching force was only partially compensated for by the recruitment of 9,444 teachers during the period from FY15 to FY16. During the same period, the biometric verification system resulted in 6,000 education sector employees being removed from the payroll.57 In addition, another explanation for the decline in the number of primary school teachers is the change of levels of some schools: with the introduction of the school consolidation policy some primary schools were designated “campus� schools and are now included with high schools. The result is that 1,144 primary schools have now been consolidated with nearby high schools. As a result, it seems that with the decreasing supply of teachers and the lack of teacher recruitment, there is now a growing need to recruit additional teachers in the province. 56 DGME data are collected by monitors who visit schools unannounced on a monthly basis in 15 districts (out of 24) in the province. 57 The biometric verification was initiated by the DL&E to develop an HR database of all employees in the system. The biometric data have become a platform for developing a HRMIS and for developing the Sindh Schools Monitoring system. Source: RSU. 65 Sindh: Public Expenditure Review 2017 Table 3.3: Numbers of teachers by fiscal year Description (Unit) 2010-11 2011-12 2012-13 2013-14 2014-15 2015-16 Teachers in primary school (A+B) 102,061 99,254 96,401 92,521 87,085 92,942 A- Government teachers in 96,433 94,015 91,552 88,437 83,141 89,060 primary schools B- Non-government teachers in 5,628 5,239 4,849 4,084 3,944 3,882 primary schools Annual retirement (assumed) 2,000 2,000 2,000 2,000 2,000 2,000 New PST inducted in Round-III N/A N/A N/A N/A 1,003 8,441 Source: RSU staff calculations. Teacher absenteeism is a chronic problem across Sindh, with teacher absenteeism in girls’ primary schools consistently higher than in boys’ primary schools. Many public school teachers are not regularly present at their schools in Sindh, as preliminary data collected by the DGME show that teacher presence ranged at around 70 to 74 percent during the 5-month period of the monitoring (Figure 3.9). Stated differently, on average, 3 out of 10 teachers were found to be absent during the monitors’ visits. Non-teaching staff are even more likely to be absent than teaching staff, with attendance rates ranging between 62 and 66 percent.58 When comparing teacher presence by school gender, it is apparent that girls’ primary school teachers are consistently more likely to be absent than boys’ primary school teachers. In fact, the difference in teacher presence between the two types of school ranged from 7 to 8 percent in the first 2 months of 2017 (Figure 3.9). Figure 3.9: Teaching and non-teaching staff presence Teacher Presence Non-teacher Presence Boys Primary School Girls Primary School School Visited School Visited 80.0% 28,000 85% 25000 75.0% 24,000 20,000 80% 80% 78% 79% 20000 70.0% 77% 16,000 75% 65.0% 73% 15000 12,000 71% 70% 70% 60.0% 69% 69% 8,000 65% 66% 10000 55.0% 4,000 50.0% - 60% 5000 OCT NOV DEC JAN FEB OCT NOV DEC JAN FEB 2016 2017 2016 2017 Source: DG Monitoring and Evaluation, October 2016 to February 2017. 3.5 Profile and Trend Analysis of Education Sector Expenditure This section presents the public financial landscape of the SEDD. Sindh’s Provincial Public Financial Management Reform Strategy FY15 to FY20 paper is a policy document that identifies priority areas for accelerated development. Its primary aim is to enhance efficiency and effectiveness of expenditure to 58 Non-teachers include watchmen, sweepers, gardeners and attendants. 66 Sindh: Public Expenditure Review 2017 improve the quality of services. The E&LD is one of the pilot departments that follow the Medium-Term Budgetary Framework (MTBF). The education sector is a priority sector for the Sindh government, as demonstrated through increased education expenditure in the province. Overall, fiscal space has improved substantially with the increase in transfers from the federal government following the 7th NFC Award. Following the General Public Services’ budget, the education sector has received the second-highest budgetary allocation in the province. Between FY09 and FY15, nominal education spending grew on average by about 20 percent annually, while in real terms education expenditure increased by 10 percent annually during the same period (Figure 3.10). As a percentage of provincial GDP, education expenditures are growing steadily, and amounted to 1.6 percent of provincial GDP in FY15. Figure 3.10: Annual expenditure on education and share of provincial GD P Eduaction Expenditutes: Nominal and Real Education as % of PGDP Nominal Real 140000 1.8 1.6 120000 1.4 100000 1.2 PRs. million 80000 1 60000 0.8 0.6 40000 0.4 20000 0.2 0 0 FY09 FY10 FY11 FY12 FY13 FY14 FY15 FY09 FY10 FY11 FY12 FY13 FY14 FY15 Source: Data from Accountant General of Pakistan. Over the 2 years from FY13 to FY15, education Figure 3.11: Share of education and health spending sector expenditure accounted for about one- Others Health Education Affairs and Services quarter of total provincial expenditure. The 100% education sector budget and expenditure 90% incurred over a 7-year period as a share of 80% overall provincial government expenditure is 70% 66% 64% shown in Figure 3.11. Over the 3 fiscal years, 60% 70% 78% 77% 77% 70% FY13, FY14 and FY15, the education budget 50% accounted for about 25 percent of total 40% provincial expenditure. The share has increased 10% 30% 9% substantially following SERP-II, with a substantial 7% 8% 20% 5% 5% 7% increase in education spending from FY12 25% 26% 10% 23% 17% 18% 16% 22% onwards. The Sindh Education Sector Reform 0% Program is also integrated into the Sindh FY09 FY10 FY11 FY12 FY13 FY14 FY15 government’s Medium-Term Development Framework (MTDF) to ensure the financial Source: Data from Accountant General of Pakistan. sustainability of resource allocations to the education sector.59 59 Source: PIRFA. 67 Sindh: Public Expenditure Review 2017 3.6 Expenditure Composition Analysis Recurrent expenditures continue to dominate Figure 3.12: Actual education expenditures by spending overall expenditure in the education sector. classification Education service delivery involves significant Development Recurrent human resources and, as a result, recurrent 100% 7% 6% 6% 6% expenditures vastly exceed capital expenditures. 14% 12% 15% Until FY05, development expenditures were very 80% low and remained below 3 percent of total education spending. However, between FY09 and 60% FY15, development expenditures started to 93% 94% 94% 94% increase their share and reached a peak in FY12 40% 86% 88% 85% (when they accounted for 15 percent of total education spending). Since FY13, development 20% expenditures have remained at around 6 percent of total education expenditure (Figure 3.12).60 0% FY09 FY10 FY11 FY12 FY13 FY14 FY15 Personnel costs continue to account for the largest share of total education expenditure, Source: Data from Accountant General of Pakistan. while repairs and maintenance remain neglected. The distribution of expenditures by object classification. The main expenditure of the education sector is related to staff salaries and pensions is shown in Table 3.5. In addition, a very small share of total expenditure is attributed to repairs and maintenance, which remained below 2 percent. Expenditure figures on employee salaries and pensions show that due to strengthened teacher recruitment processes in recent years and the reduction in ghost teachers, the share of employee costs fell from 83 to 79 percent in FY15. This has created fiscal space for increased spending on asset creation and school development. Table 3.5: Actual expenditures by object classification of expenditure (PRs million) FY 08-09 FY 09-10 FY 10-11 FY 11-12 FY 12-13 FY 13-14 FY 14-15 Employee costs (salaries and 32,546 38,981 53,443 47,711 80,693 86,919 92,433 pensions) Asset creation 1,688 4,038 5,241 8,951 4,957 5,701 7,066 Repairs and maintenance 459 337 444 325 220 235 439 Others 3,596 9,515 13,1664 10,577 12,181 12,664 16,754 Total 38,289 52,870 72,295 67,564 98,051 105,518 116,692 Source: Accountant General of Pakistan. The bulk of expenditure in the education sector is for pre-primary and primary education services, followed by secondary education. An overview of the distribution of education expenditure by sub- sector is shown in Figure 3.13. Almost three-quarters of total education spending is attributable to pre- primary, primary, and secondary schools, with universities, colleges and other professional institutes accounting for only 17 percent of expenditure. 60 Source: Accountant General of Pakistan. 68 Sindh: Public Expenditure Review 2017 Figure 3.13: Sub-sectoral share of expenditure, FY15 Source: Data from Accountant General of Pakistan, FY15. There has been no major shift in the various shares of expenditure in the sub-sectors over the 3 fiscal years to FY15. Table 3.6 shows a time-series analysis of the sub-sectoral distribution of budget allocations and expenditures. Sub-sectors vary in terms of the share of budget that they are actually able to spend. In FY15, 98 percent of the primary education budget and 81 percent of the secondary education budget were utilized. The main reason for the high budget execution rate is that teachers’ salaries account for the major share of the overall education budget. 61 Table 3.6: Original budget and expenditure by sub-sectors (PRs million) FY08-09 FY09-10 FY10-11 FY11-12 FY12-13 FY13-14 FY14-15 Pre- & primary Budget 17,855 22,305 26,956 1,509 42,644 55,450 57,512 education affairs & Expenditure 17,618 22,468 31,734 29,262 45,698 49,671 56,691 service Percent utilized (%) 99 101 118 1,939 106 89 98 Secondary education Budget 11,977 15749 17,755 628 27,228 31,811 36,973 Expenditure 10,965 13,743 17,896 14,355 25,777 27,478 29,987 Percent utilized (%) 92 87 78 2,286 95 86 81 General universities/ Budget 8,978 16,449 16,427 22,492 25,358 25,894 23,825 Colleges/ Prof. Expenditure 6,342 9,611 11,950 14,547 15,860 17,953 18,591 institutes Percent utilized (%) 71 58 73 65 63 69 78 Education & subsidiary Budget 821 593 501 581 1,743 1,800 2,398 education services level Expenditure 457 403 422 942 987 1,666 1,467 non-definable Percent utilized (%) 56 68 84 162 57 93 61 Sindh Education Reform Budget 11,369 15,150 14,167 12,822 14,741 18,521 25,083 Program + Expenditure 2900 6,617 9,971 7,918 8,863 7,852 9,026 Administration Percent utilized (%) 26 44 70 62 60 42 36 Source: Data from Accountant General of Pakistan. 69 Sindh: Public Expenditure Review 2017 Overall, budget execution rates have varied over the past few years. In the 2 years to FY15, budget execution rates for the Sindh Education Reform Program and Administration were particularly low, which carries the potential to seriously weaken governance and monitoring oversight of the E&LD. Budget utilization rates by economic classification of expenditure show unspent development budget, particularly at the higher education level. Table 3.7 provides additional details on budget utilization rates by economic classification and sub-sector. The high utilization rate in FY13 was at the primary level, with 108 percent utilization of the current budget. Notably, in FY15, the secondary education sub- sector spent only 30 percent of its development budget, while 59 percent of the tertiary education development budget was utilized. The education budget allocated to administration and SERP remained largely unutilized for FY15. This could be due to the late release of the school specific non-salary component of the budget (SSBs). Table 3.7: Budget utilization by level of education and economic classification (percent) Sub-sector FY13 FY14 FY15 Pre- and primary education affairs and service CE 108 95 100 DE 98 42 82 Secondary education RE 97 87 83 DE 30 58 30 General universities/ colleges/ prof. institutes RE 77 80 85 DE 35 45 59 Education and subsidiary education services RE 72 115 85 level non-definable DE 0 0 0 Administration + SERP RE 63 43 36 DE 39 31 6 Overall education sector RE 93 82 81 DE 43 43 60 Source: Data from Accountant General of Pakistan. RE = recurrent expenditure, DE = development expenditure. Education projects under the Annual Development Plan (ADP) are typically small in size, and often unapproved at the time of ADP preparation. Data on missing-facilities projects62 from the ADP show that in FY17 most projects were small-value schemes with estimated costs of less than PRs 100 million (Figure 3.14). Over time, the proportion of projects that were approved at the time of ADP submission has decreased, suggesting a lack of sufficient project appraisal, for example in FY17 when only 9 percent of projects were approved (Figure 3.15).63 62 These are development projects focused on the provision of basic school facilities, such as drinking water, boundary walls, toilets, and electricity. 63 Note: Statistics derived from ADP data, Planning and Development. 70 Sindh: Public Expenditure Review 2017 Figure 3.14: ADP project costs FY17 Figure 3.15: Percentage of ADP projects approved 35 FY10 100% 30 FY11 67% 25 FY12 75% Frequency 20 FY13 50% 15 FY14 81% 10 FY15 58% 5 FY16 92% 0 FY17 9% Estimated project cost (PRs million) Source: Data from Accountant General of Pakistan. Since FY10, the number of missing facilities- Figure 3.16: ADP project completion related projects has increased more than Estimated Cost spent tenfold, while the share of projects that are No. of Schemes close to completion has decreased. The analysis Linear (Estimated Cost spent) in Figure 3.16 estimates project completion for 70.0% 64% ADP schemes by utilizing total project cost and 60.0% 54% 74 82 expenditure information. It is evident that with 50.0% 46% relatively fewer projects to manage in FY12, the 62 40.0% project completion rate was relatively higher than in subsequent years. Over the past 4 years, 30.0% 24% 23% 22% 23% 42 project completion has ranged from 17 to 23 20.0% 17% 14 16 percent, indicating that the utilization of the 8 19 12 22 10.0% 6 development budget is low. 3 0.0% 2 In terms of recurrent expenditures, there are FY10 FY11 FY12 FY13 FY14 FY15 FY16 FY17 two interventions that directly transfer funds to Source: Data from Accountant General of Pakistan. schools: school management committee (SMC) grants and the school specific non-salary budget (SSB) reforms. Boxes 3.1 and 3.2 present an overview of both reforms, along with their implementation challenges. 71 Sindh: Public Expenditure Review 2017 Box 3.1: School management committee grants What are they? School management committee (SMC) grants are school-level grants that are directly transferred to public schools and are utilized at the discretion of the SMC . SMC grants are allocated to schools based on a formula that consists of three components: (i) school enrolment, (ii) classroom student ratio, and (iii) school level. The SMC is fully empowered to withdraw funds as and when needed, without seeking any authorization from education management authorities. How do they work? To receive a SMC grant, a functional public school has to submit an application to the Reform Support Unit (RSU) with the SMC’s bank account information, including details on the co-signatories (school head/teacher who serves as the SMC's secretary and one parent who serves as the SMC's chairperson) to the account. The SMC typically has three additional members who are parents or notable members of the community. A summary of the allocation process is outlined in the diagram below. SMC applications are received by the RSU and processed on a rolling basis; funds are released in batches as soon as approval and processing are complete. What can the funds be spent on? SMC guidelines dictate that expenditure should be according to three major categories and in accordance with the School Improvement Plan: (i) minor repairs and maintenance, (ii) cleanliness (including hiring of sweepers), and (iii) transportation and fares. What are the main positive and negative aspects of the program? The positives include:  Easy to comprehend the allocation formula.  SMC is empowered to withdraw funds when required without going through administrative hurdles.  Generally high utilization of SMC grant.  Ensures community involvement in financial management and school improvement. The negatives include:  School Improvement Plans can be more developed.  Lack of monitoring oversight of fund utilization. Functional Schools LSUs conduct Legend: SMC application workshop for Start shortlisted in SEMIS forms given to LSUs District Education SED database officials LSUs RSU Forms, supporting Schools return District Education documents are forms to RSU via officials distribute School scrutinized LSUs and DEOs forms to Schools Others Secretary Education Approved Funds released by approves applications sent to State Bank applications Finance Department SMC co-signatories Funds transferred to can withdraw funds Funds transferred to End each school s bank according to commercial banks account guidelines 72 Sindh: Public Expenditure Review 2017 SSB disbursement Overview Higher Secondary Legend: School defines need TEOs collect need DEOs consolidate Start requisition against Primary for all schools in a Schools need at district level SSB heads cost center Districts Cost center Vendor Cost center issues Recommendations Procurement Committee AG / DAO starts procurement work order to (vendor list) sent to process according to vendor cost center SPPRA rules Cost center submits Vendor supplies the documentation to AG releases check cost center District Accts to cost center Officer / AG Cost center delivers End check to vendor What are the main positive and negative aspects of the program? The positives include:  formula-based budgeting approach based upon a school specific formula.  greater oversight in school level expenditures.  less potential for political interference in fund allocation process. The negatives include:  highly restrictive in terms of fixed allocations against budget items.  delay in release of funds from finance department.  quarterly disbursement of funds leads to less money available for schools for expenditures in the beginning of the year.  funds are lapsable and cannot be reallocated between budget heads.  low levels of SSB utilization, reportedly due to limited procurement capacity of DDOs and complicated audit compliance processes at the district level. DDOs also experience difficulty in processing the expenditures through audit departments and at times allow the funds to lapse  AG documentation requirements are paper based and tedious. 73 Sindh: Public Expenditure Review 2017 Box 3.2: School specific budgets What are they? The school specific budget (SSB) is a non-salary, operational budget that is allocated to each public school every fiscal year. The SSB can be utilized by schools for day-to-day expenses based on specific budget items as defined in the SSB guidelines. How are funds allocated and what can the funds be spent on? The figure below outlines the allocation process for the SSB for each fiscal year. The SSB for each fiscal year is initially calculated by the cost centres by applying an incremental multiplier provided for the year by the Finance Department to the SSB for the previous fiscal year. The cost centre (or Drawing and Disbursing Officer [DDO]) is required to be a Grade 17 officer. For primary schools, the cost centre is typically the Taluka Education Officer, while for secondary schools, the school principal is the cost centre for his/her school. Once approved by the FD, this cost-centre SBB is broken down by the RSU’s SSB team to a school-level budget using a school specific allocation formula. Next, the budget for each school is then split down to seven expenditure categories. A basic outline of the formula used for splitting cost-centre allocation by school and budget item is outlined in the following tables: School-specific calculation Budget items breakdown Component Weightage Budget item Allocation percentage School category 20% Stationery 10% School enrolment 45% Travelling allowance 5% School size 35% In-class material and supplies 25% Library/lab/computer Up to 15% material Sports items Up to 10% Co-curricular activities 15% Furniture purchase 20% Approved by Legend: Cost center Compiled at District Section Officer SED, Start calculates annual level, approved by SED transmitted to SSB allocation Director FInance Districts Cost center No RSU SSB team uses Finalized in Budget Budget Finance formula to decide Book (Volume III) by Yes Approved by school-wise, head- cost center Finance? wise allocation Final budget disseminated to End schools via DEOs and TEOs How are funds disbursed? Schools can request disbursement against their allocated SSBs. School budgets are allocated and can be disbursed against any of the following budget heads: (i) in-class materials and supplies, (ii) library/laboratory, (iii) co- curricular activities, (iv) sports, (v) travelling allowance, (vi) stationery, and (vii) furniture. Need requisitions for spending are processed and approved by procurement committees at the regional, district or school levels (depending on the nature of procurement taking place). The following chart shows an overview of the SSB 74 Sindh: Public Expenditure Review 2017 disbursement process. How does expenditure: Schools can request disbursement against their allocated SSBs for specific budget heads. School budgets are allocated and can be disbursed according to any of the following budget heads: (1) in- class materials and supplies, (2) library/laboratory, (3) co-curricular activities, (4) sports, (5) travelling allowance, (6) stationery, and (7) furniture. Need requisitions for any spending are processed and approved by Procurement Committees at the regional, district or school level (based on school level or type of procurement). The following chart shows an overview of the SSB disbursement process. 3.7 Allocative Efficiency of Education Expenditures There is considerable inefficiency in primary school distribution. Figure 3.17 shows the total number of schools in a district by the total number of students in a district. Under an ideal scenario, there should be a strong positive relationship between the number of schools in a district and the number of students, reflecting that the establishment of schools is driven by the demand for schooling. In Sindh, this is not necessarily the case, as Mitthi ranks first in terms of the number of schools in a district (4,008 schools), but ranks 15th in terms of the number of students in the district (with an average of 37 students per school). Larkana, on the other hand, has the highest number of students per school (187). This suggests that there is a weak relationship between the number of schools and the number of students in a district, suggesting that schools are not optimally located to cater to demand. In fact, primary school placement in Sindh appears to be highly problematic. An analysis of primary school proximity indicates that 48 percent of boys’ primary schools and 41 percent of girls’ primary schools are located within 1 kilometer of the nearest neighboring primary school of the same gender (Figure 3.18). Figure 3.17: Number of enrolled students by number of schools in a district (FY15) 4,500 4,000 Mitthi 3,500 Karachi Number of schools 3,000 2,500 2,000 1,500 Larkana 1,000 500 Tando M. Khan 0 - 50,000 100,000 150,000 200,000 250,000 300,000 350,000 400,000 450,000 500,000 Number of students Source: SEMIS 75 Sindh: Public Expenditure Review 2017 Figure 3.18: Primary school proximity to the nearest primary school Boys Primary Schools Girls Primary Schools 35 2,482 1,467 30 25 20 Percent 15 10 5 0 0-0.5km 0.5 to 1km 1 to 2km 2-5 km 5-10 km 10-15 km >15 km Source: Statistics derived from SEMIS 2014 The relationship between the number of schools at the district level and per-student expenditure is weak, suggesting that many districts are operating under-capacity. Figure 3.19 shows the relationship between total expenditure in a district and student enrolment (all levels). Overall, there is a moderately positive relationship between the number of students and expenditure, whereas the relationship between the number of schools (all levels) and per-student expenditure in a district is weak (if the outlier is excluded it is evident that there is no apparent pattern that emerges with respect to the number of schools and per-student expenditure). This finding indicates that there is significant inefficiency in school placement across the province. Looking at individual districts, it appears that districts are being allocated funds in line with enrolment. However, there are a few notable exceptions: the districts of Sanghar, Khairpur, Dadu, Ghotki, Naushahro Feroz and Shaheed Benazir Abad have a relatively higher share of student enrolments compared with their annual expenditures. Figure 3.19: Number of schools and students (all levels) by total expenditure in a district 600,000 4,500 4,000 500,000 R² = 0.6438 3,500 R² = 0.095 400,000 3,000 Students Schools 2,500 300,000 2,000 1,500 200,000 1,000 100,000 500 0 0 0 10,000 20,000 30,000 40,000 0 10,000 20,000 30,000 40,000 Total expenditure (PRs million) Total expenditure (PRs million) Source Data from PSLMS and Accountant General of Pakistan. Per-student expenditure is skewed towards more urban, economically advantaged, districts due to higher numbers of teachers in these regions. Figure 3.20 shows per-student expenditure by district to evaluate the equity of expenditures in education. The data reflect that annual per-student expenditure 76 Sindh: Public Expenditure Review 2017 is skewed towards relatively few districts. Per-student expenditure is highest in the districts of Karachi64 and Hyderabad, due to the fact that the districts in these two cities have the lowest student-teacher ratios (STRs), at 18 and 19 students per teacher, respectively.65 In contrast, annual per-student expenditure in the districts of Ghotki, Badin and Tando Allah Yar is less than one-quarter of the per- student expenditure in Karachi. The district of Ghotki, for example, where education and other socioeconomic outcomes are poor, has the lowest per-student expenditure in the province, at just PRs 14,673 per student per year. This is reflected in Ghotki’s STR, which is the highest in the province, at 40 students per teacher. Figure 3.20: Annual per student expenditures pre-primary to higher secondary levels FY15 (PRs) Karachi 76,477 48,582 Thatta 47,783 34,613 Sukkur 26,586 26,002 Mirpur Khas 24,960 24,830 Matiari 24,285 23,259 Jacobabad 22,633 22,003 Umer kot 21,685 21,468 Dadu 20,604 19,972 Shahdadkot 19,963 19,394 Naushahro Feroze 19,239 17,732 Tando Allah Yar 16,564 15,561 Ghotki 14,673 Source: Data from Accountant General of Pakistan and SEMIS FY15. The majority of primary schools lack basic facilities, and there is an urgent need to increase expenditure for the provision of essential facilities. Schools in all districts across Sindh lack basic facilities. Figure 3.21 shows the percentage of primary schools with all four basic facilities.66 In the district of Tharparkar (Mitthi), only 2 percent of primary schools have all four basic facilities, followed by the district of Thatta (3 percent). Karachi and Hyderabad have the largest share of primary schools with all basic facilities, at 50 and 47 percent, respectively. Overall, there are 12 districts in Sindh where more than 75 percent of primary schools do not have all four basic facilities. 64 This includes all five districts of Karachi. 65 Much of the administrative structure of the E&LD is located in Karachi. Therefore, some of the expenditures reflect this. It is difficult to distill these expenditures from the overall spending. 66 The four basic facilities required by schools: (i) a fully enclosed boundary wall, (ii) functional toilets, (iii) electricity, and (iv) drinking water. 77 Sindh: Public Expenditure Review 2017 Figure 3.21: Availability of the four basic facilities in primary schools 100 % Primary schools with all four basic 80 60 50 45 45 47 41 36 36 facilities 40 32 35 24 25 28 15 16 17 20 7 10 11 12 13 14 2 3 0 Sanghar Mitiari Shikarpur Larkana Badin N. Feroze Mirpur Khas S. Benazirabad Ghotki Karachi K. Shahdadkot Umerkot Jamshoro Dadu Hyderabad Sukkur Tharparkar Thatta Kashmore Tando M. Khan Khairpur Jacobabad Tando Allah Yar Source: Data from Accountant General of Pakistan and SEMIS FY15. There is inefficient allocation of teachers both across and within districts. Given the low STR in Sindh, it is evident that the allocation of teaching resources across schools is inefficient. The share of single- teacher schools varies from 10 to 75 percent by district (Figure 3.22). As observed in Table 3.2 early, the STR in primary schools is 30:1 and has been stable for the last 5 years data are available. However, at the middle to higher school levels, the STR is low and has been declining over time. Given that there are a large number of schools with single teachers, this indicates a significant misallocation of teaching resources. Low STRs in urban districts indicate that many of these teaching resources could be shifted to rural areas. In the districts of Tharparkar, Ghotki and Umerkot, there are also a significant number of secondary schools that have only a single teacher per school. As explained in Box 3.3, even within the same union council there is scope for reallocating teachers from areas where there is a teacher surplus to areas where schools have a shortage of teachers. Box 3.3: Teacher allocation in Central Karachi UC 8-Faisal in the Central Karachi District is one example of a union council where teacher allocation is inefficient: KMC Markaz-e-Taleem-e-Balghan Boys Primary School has a total student enrolment of 36, with four teachers and an STR of 9:1. Within the same union council, there is another boys’ primary school, GBPS, J/M Siddqui-e-Akbar, which has a total enrolment of 83 students and only one teacher. Clearly, there is potential for teacher to be re-allocated to ensure a more equitable distribution of teachers within the same union council. Source: SEMIS FY15. 78 Sindh: Public Expenditure Review 2017 Figure 3.22: Percent of schools with single teacher, by school level % of secondary schools with single teacher % of primary school with single teacher 100 80 60 40 20 0 Tando… Tando M.… K.… S.… Shikarpur N. Feroze Badin Mirpur Khas Karachi Dadu Sanghar Ghotki Hyderabad Jamshoro Mitiari Sukkur Tharparkar Umerkot Larkana Kashmore Thatta Khairpur Source: SEMIS data FY15 Jacobabad The majority of teachers have a bachelor’s degree or higher, but the number of teachers with science backgrounds is extremely low. With the introduction of merit-based teacher recruitment in Sindh, it is evident that the average qualification of teachers has improved. Currently, 80 percent of teachers have a bachelor’s degree or higher. Figure 3.23 shows the distribution of teachers by academic qualification and by salary expenditure. As reflected below, the education system is spending a large portion of its budget on the salaries of teachers with higher qualifications. However, when this information is further disaggregated into teachers with science qualifications (i.e., bachelor’s or master’s degree in science), it is evident that there is a shortage of such teachers. Only 5 percent of public school teachers have a science qualification at the bachelor’s degree level or higher, and 5 percent of total teacher salary expenditures go towards paying the salaries of these science teachers.67 Figure 3.23: Total annual wage bill by academic qualification of teachers, FY14 Total Annual Salary Number of Teachers Salary Expenditure No. of Teachers 16.00 35,000 35.00 70,000 Salary expenditure in PRs billion 14.00 30,000 Salary expenditure in PRs billion 12.00 30.00 60,000 No. of Teachers 25,000 10.00 25.00 50,000 20,000 No. of Teachers 8.00 15,000 20.00 40,000 6.00 4.00 10,000 15.00 30,000 2.00 5,000 10.00 20,000 - - 5.00 10,000 - - Science No Science Background Background Academic qualification Source: SEMIS FY14, AG FY14 Average monthly teacher salaries range from PRs 35,000 to as high as PRs 95,000, depending on the pay scale. Teacher salaries by basic pay scale for public school teachers in Sindh are shown in Figure 67 This also includes teachers with MPhil and PhD degrees. The analysis is restricted to a sample of 69,248 teachers for which payroll data and teacher qualification data were available. 79 Sindh: Public Expenditure Review 2017 3.24. As expected, average salaries increase at higher pay scales, with the sharpest increase between Scales 16 and 17. 68 Figure 3.24: Average monthly teacher salary by basic pay scale 100,000 90,000 PST entry 80,000 Average monthly salary, PRs 70,000 60,000 50,000 40,000 30,000 20,000 10,000 - 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 BPS Source: SEMIS FY14, AG FY14 Subject-specialist teachers, on average, are the highest paid teachers. Overall, the average teacher salary in Sindh is about PRs 43,000 per month, while for just primary school teachers the average salary is PRs 36,000 per month. The average monthly salary of subject-specialist teachers is PRs 77,000. Overall, 60 percent of teachers’ salaries go towards primary school teachers, while 21 percent goes to high school teachers, indicating a relatively equitable distribution given that 65 percent of the total student enrolment is at the primary level. Figure 3.25: Average teacher monthly salary by designation 90,000 Average Monthly Salary 76,503 Average monthly salary (PRs) 80,000 68,453 70,000 63,951 60,000 51,177 50,806 50,000 41,073 40,150 43,404 42,512 36,263 39,089 36,458 40,000 30,000 20,000 10,000 - Teacher designation Source: Statistics derived from SEMIS FY15 and Accountant General of Pakistan. Public school teachers are generally well paid, even compared with other public sector professions. The average monthly salary for teachers starts at the same level as doctors, but at a higher level than engineers. However, it is also true that, on average, doctors work far longer hours with more working 68 PST Entry BPS has changed and is no longer BPS 7. 80 Sindh: Public Expenditure Review 2017 days per year than teachers. Figure 3.26 shows a comparison of the average monthly salary by the three public sector occupations by years of service. Figure 3.26: Average monthly salary by public sector occupation Teachers Doctors Engineers 180,000 160,000 Average monthly salary (PRs) 140,000 120,000 100,000 80,000 60,000 40,000 20,000 0 0-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 Years of service Source: Sindh AG Payroll data FY15. Note: The analysis includes, Doctors – Medical Officers; Teachers – BPS 17-20; Engineers – all job specifications. The presence of out-of-school children is not taken into account when expenditure allocations are made. The budget allocation is based on the number and level of schools, and the number of “in-school children�. Figure 3.27 show that the districts with the highest out-of-school children (OOSC) receive the lowest expenditures. For example, the district of Ghotki, which has the highest estimated OOSC population, has an average expenditure per student of less than PRs 15,000, while the district of Jamshoro has a low estimated OOSC population of less than 81,000 and a per-student expenditure of about PRs 35,000 per student. Figure 3.27: Regional variation in OOSC children and education expenditure per student, FY15 100,000 Expenditure OOSC 350,000 Expenditure/ student (PRs) 300,000 80,000 OOSC (numbers) 250,000 60,000 200,000 40,000 150,000 100,000 20,000 50,000 0 - S. Benazirabad Jamshoro Badin Sukkur Mirpur Khas Shikarpur Karachi Matiari Hyderabad Dadu Mitthi Sanghar Q. Shahdadkot Tando Allahyar Ghotki Thatta UmerKot Larkana Tando M. Khan Kashmore Khairpur Jacobabad Nau. Feroze Source: Data from Accountant General of Pakistan and PSLMS FY15. Given the existing number of teachers and school infrastructure, an additional 373,543 children could be accommodated in primary schools in Sindh. Another example of inefficiency in the sector is the sub- optimal utilization of school infrastructure and teaching resources. A school is classified as underutilized 81 Sindh: Public Expenditure Review 2017 if it has sufficient classrooms and also available teachers.69 This rather simplistic example shows how many additional students could be accommodated in existing schools to reach the E&LD’s STR limit of 40:1. Table 3.8 illustrates that with the existing level of teachers and available classrooms, the education system could accommodate an additional 317,000 students in primary schools at no extra cost. There is also the space and available teaching resources to accommodate an additional 56,487 students in higher level schools. Table 3.8: Estimated unutilized capacity in public schools by level and type of school Number of vacant capacity to accommodate these many students Boys 74,727 Girls 51,544 Primary Mixed 190,785 Total 317,056 Boys 18,122 Girls 17,326 Above primary Mixed 21,039 Total 56,487 Total 373,543 Source: Estimates based on SEMIS FY15. 3.8 Sector Effectiveness and Equity Analysis of Expenditure This section presents the results on the effectiveness and equity of public expenditure in the education sector. This is achieved by evaluating the relationship between public expenditure on education, education enrolment, and learning outcomes. Children from poor households are more likely to be out of school and this situation is more acute for girls. Data show that girls from the poorest income quintiles are the least likely to be enrolled in a public school (Table 3.9). The differentials between boys and girls are persistent across all income quintiles. The benefit of public schooling increases for students from higher income quintiles. Table 3.9: Public school attendance among 5 to 16-year population by gender and wealth quintile Public Private OOSC Total % % % % Male 38.3 5.5 56.2 100.0 Bottom 01-20% Female 24.5 3.8 71.7 100.0 Male 46.8 21.2 31.9 100.0 21-40% Female 33.7 19.4 46.9 100.0 Male 49.1 32.1 18.7 100.0 41-60% Female 41.7 26.6 31.7 100.0 Male 44.5 34.2 21.3 100.0 61-80% Female 34.8 31.0 34.3 100.0 69 It is assumed that an average-sized classroom can hold a maximum of 40 students. Similarly, a teacher can teach 40 students. Only those schools are used in the analysis that has both fewer than 40 students per classroom and per teacher at the same time. Once such a school is identified, the difference from 40 is added up to derive the aggregate figure of unutilized capacity. 82 Sindh: Public Expenditure Review 2017 Male 45.4 43.3 11.3 100.0 Top 81-100% Female 34.6 45.6 19.8 100.0 Source: PSLM 2014-15. Public school attendance varies by household income and urban/rural location. The share of children enrolled in public schools increases with income in rural areas and falls with respect to income in urban areas, both for boys and girls. This is due to the availability of more private schooling in urban cites for higher-income groups. However, girls are at most disadvantage in rural areas. Both poverty and location play a part in preventing girls from going to school, but the combination of both is certainly more detrimental to girls’ schooling than for boys. Figure 3.28 shows that gender differentials in public school attendance are higher in rural areas and virtually eliminated in urban areas. Figure 3.28: Public school participation among 5-16 year old, by urban/rural location, gender & wealth quintile Rural Urban 100 100 Male Female Male Female 80 80 60 60 40 40 20 20 0 0 Bottom 01- 21-40% 41-60% 61-80% Top 81- Bottom 01- 21-40% 41-60% 61-80% Top 81- 20% 100% 20% 100% Source: PSLMS FY15. There is no relationship between learning outcomes and public expenditure at the district level. Figure 3.29 shows total expenditure by district and learning outcomes of Grade 5 and 7 students on the Standardized Achievement Test (SAT).70 When combined with expenditure on education, SAT scores show no apparent relationship with education expenditure. For instance, the overall score for Grade 7 in Mirpur Khas is 28 percent with annual per-student expenditure of PRs 15,000, and in Sukkur the score is 21 with annual per-student expenditure of PRs 21,000. Even in Karachi, with highest per-student expenditure on education, the average score is 24 in the SAT FY15. 70 The Standardized Achievement Test (SAT) was initiated by RSUs of the E&LD in Sindh. The test is carried out annually at public schools to assess achievement of students in language, math, and science at Grades 5 and 7. 83 Sindh: Public Expenditure Review 2017 Figure 3.29: Average SAT score of districts for Grades 5 and 7, by expenditure per student 30 30 25 25 Grade 5 average score Grade 7 average score 20 20 15 15 10 10 5 5 0 0 0 20,000 40,000 60,000 0 20,000 40,000 60,000 Annual expenditure per primary student (PRs) Expenditure per above-primary student (PRs) Source: Data from SAT and Accountant General of Pakistan. The efficiency of education expenditure is declining. Figure 3.30 shows the effectiveness of public education expenditure by matching data from household surveys at the district level. It measures primary to higher secondary enrolment in public schools across four rounds of household surveys, and compares this with per-student annual public expenditure during the same time periods. Figure 2.30 clearly indicates that average enrolment in public schools in different districts has not improved over time, despite a four to fivefold increase in per-capita expenditures. This suggests that the real challenge in the education sector is the effective utilization of the available resources. Figure 3.30: Spatial distribution of public school attendance by annual per-student expenditure (aged 5-16) 100 2008-09 2011-12 2012-13 2014-15 90 % of 5-16 year olds attending pulic school 80 70 60 50 40 30 20 10 - - 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 Per-student public expenditure (in PRs) Source: PSLMS and Accountant General of Pakistan. 3.9 Recommendations Based on the preceding analysis of public expenditure on education and education sector outcomes, the following recommendations are made with a view to improving the efficiency and effectiveness of spending in the education sector: There is a need to strengthen the existing governance reforms in education to ensure more effective service delivery in the sector. The Sindh Schools Monitoring System initiative by the DGM&E is a critical initiative that should be strengthened and expanded across the province to ensure that there are 84 Sindh: Public Expenditure Review 2017 reliable and timely data to foster greater accountability and evidenced-based decision-making. The data that are generated from the monitoring system should be utilized at the district and sub-district levels, so that corrective actions can be taken in a uniform and timely manner. There is a need to focus on the effective utilization of the development budget. Given such low development budget execution rates, development projects should be actively monitored to ensure that work is taking place and meets certain standards, particularly given the poor state of school infrastructure and facilities in Sindh. There should be active follow-up on development projects at the provincial and district levels to ensure that problems are identified and addressed early on. There is a need to recruit teachers with science and math qualifications in both primary and secondary schools. The recruitment policy should continue to prioritize the recruitment of teachers with science and math backgrounds in order to keep up with the pace of teacher retirement. Teachers could be placed in schools where there is a shortage of teachers. Teacher rationalization could be implemented to ensure an adequate and equitable distribution of teachers, particularly at the primary level. Although at the macro level STRs seem adequate, within districts there is inequity in teacher placement. This warrants the re-allocation of teachers based on a needs-based formula. There may be additional scope for school consolidation. Given the fact that many primary schools are located within close proximity of each other, and that there is excess capacity in many primary schools in the province, there may be additional scope to consolidate schools, particularly in areas where there is an inequitable distribution of teachers. There is a need to ensure that education expenditure is outcome-based. This means investing greater resources in districts and areas where education outcomes are particularly poor, and where there is potential for additional enrolment gains. There is a need to strengthen budget planning in the education sector. In the past 2 years, there has been a reduction in the original budget, while prior to that there was an increase. Similarly, a large share of missing facilities ADP projects were included in the budget unapproved at the time of ADP submission, this indicates that the project appraisal process needs strengthening. There is a need to review the SSB and SMC interventions to ensure that funds are available in a timely manner and are effectively utilized. Both the SSB and SMC are important initiatives that need continued support moving forward. However, the SSB reform, as currently designed, may be overly prescriptive. While school-level allocations can and should be formula-based, there should be greater flexibility for school leaders to utilize funds across budget items based on the needs of the school. 85 Sindh: Public Expenditure Review 2017 Chapter 4: How to accelerate improvement in public health sector outputs and outcomes? 4.1 Introduction Health outcomes in Sindh have not improved as much as hoped in recent years and there is serious concern over the particularly poor health indicators among the rural population. At just over 55 million, Sindh is the second-most-populous province of Pakistan after Punjab. Its annual population growth rate of 2.8 percent is slightly higher than the average national growth rate of 2.6 percent. Just over half of Sindh’s population is rural, at 53.1 percent, with 43.7 percent urban. This large urban share of the province’s population, together with the political climate in Sindh, directly affects health systems in terms of needs and administrative requirements. However, while it is necessary to consider average provincial health indicators, it should be remembered that these data tend to conceal the needs of rural areas and the very large urban-rural disparity. On closer scrutiny, it is apparent that rural health indicators are far worse than those for urban areas, in some cases worse by a factor of two. Health outcomes relating to infant and maternal mortality for Sindh have seen little progress over the past few years. There has been no reduction in the infant mortality rate (IMR) of 81 deaths per 1,000 live births, while the neonatal mortality rate (NMR) has actually increased from 44 to 53 deaths per 1,000 live births since 2000. The under-5 mortality rate has also seen scant progress, remaining at about 105 deaths per 1,000 since 2000. Although there has been a slight decline in the maternal mortality rate (MMR) in Sindh, the province’s MMR of 314 per 100,000 is still far higher than the national average of 170. Meanwhile, the total fertility rate (TFR) in Sindh declined from 5.1 live births in FY91 to 3.9 in FY13, although the steep initial rate of decline shows signs of tapering off. Figure 4.1: Stunting and malnutrition are at crisis levels in Sindh, and getting worse Stunting Wasting Sindh 48 Stunting 43.3 41.8 41.6 43.7 - 2001: 44.1% 36.3 - 2011: 49.8% 14.3 15.1 11 10.8 11.8 8.6 Maternal Anemia (2011) Pakistan: 50.4% 1965 1979 1985-87 1994 2001-02 2011 Source: Data from various rounds of MICS. Under-nutrition is a serious issue in Sindh, with higher levels of maternal anemia and underweight children than seen in other provinces. Among under-5 children, 42 percent are underweight, while stunting accounts for 49.8 percent (2011). The incidence of low weight at birth is 30 percent, indicating 86 Sindh: Public Expenditure Review 2017 serious nutritional problems affecting expectant mothers. This is confirmed by the incidence of maternal anemia, at 62.0 percent in 2011, significantly higher than the national average of 50.4 percent. These indicators are worrying and indicate that emergency levels of poor health standards are already prevalent. Similar to the rest of Pakistan, Sindh is undergoing a demographic transition. This means that there are increasing levels of non-communicable diseases (NCDs), particularly in the urban areas, in addition to more traditional communicable diseases. These NCD diseases can also strike at an earlier age, undermining the economically productive population. At present, there is no strategy to address NCDs and most NCD interventions are concentrated at the costly tertiary level of health care. While Sindh has made some progress in improving a small number of health outcomes, in general its performance is far from adequate. The performance of the health sector remains broadly inadequate, with unacceptably low indicators in several areas, such as the level of institutional deliveries (64 percent); immunization coverage (35 percent); the use of contraceptives (29 percent); and the high unmet needs for contraception (21.7 percent71). Indeed, estimates suggest that 88 percent of induced abortions are due to unmet needs for contraception. Use of public sector services in Sindh is 22 percent, compared with 29 percent in rest of the country. The level is low because the population predominantly chooses to use private health providers in both urban and rural areas, including for promotive services, such as maternal and childcare. Immunization coverage has seen a modest increase, with existing BCG coverage at 76 percent, measles at 51 percent, and polio 3 at 84 percent. However, a gap of 20 percentage points exists between the survey data and administrative data levels. Contraceptive prevalence remains very low and sub-optimal at 26.7 percent, and this low level has remained largely stagnant since 2005. Communicable diseases are still prevalent. Malaria incidence is on the rise and has coincided with outbreaks of dengue fever, a situation that calls for a comprehensive vector control strategy. Tuberculosis is believed to be endemic. However, the precise caseload of TB for Sindh is unavailable, and case detection of 57 percent needs further improvement, although the treatment success rate of 87 percent is above the national target. Blood-borne diseases such as Hepatitis B and C have a high prevalence of 3 to 4 percent. However, vaccination coverage is at best 14 to 15 percent in the best- performing districts. HIV has a low overall prevalence of just 0.1 percent, but it is highly concentrated at 28.2 percent among injectable drug users. Out-of-pocket health spending is high, accounting for two-thirds of all health expenditure. Out-of- pocket (OOP) expenditure on health is 66 percent of total health expenditure. It is also regressively distributed, with poorer income quintiles spending a higher proportion of household income on health expenses. Medicines account for the largest share of OOP, while consultation fees account for the second-largest component across both public and private sector facilities. This PER of the health sector analyzes the performance and adequacy of public health expenditure in Sindh. This chapter presents a Public Expenditure Review of the health sector in Sindh. The main objective is to analyze the performance and adequacy of public expenditure to meet the health policy objectives of the province. Data are derived from the public accounts system, the Project to Improve Financial Reporting and Auditing (PIFRA), the Multiple Indicator Cluster Survey (MICS) 2014, and the Pakistan Social and Living Standards Measurement Survey (PSLSM, 2015). As private sector institutions are not part of the reporting system, analysis of the private health sector is excluded from this analysis. Furthermore, given the data limitations of the financial system in Sindh, it is not possible to adequately analyze the extent and the efficiency of public service provision. 71 Sindh MICS FY14. 87 Sindh: Public Expenditure Review 2017 This chapter presents three analyses to meet the main objective: (i) a profile and trend analysis of health indicators to present a health sector profile and understand the adequacy of expenditures; (ii) a profile and trend analysis of health sector expenditure, particularly in relation to the economic, functional, and objective classifications of health expenditure and allocative efficiency; and (iii) an analysis of the effectiveness and equity of expenditure by linking expenditure to health outcomes. This analysis also highlights geographic and gender-based inequalities of health expenditure. 4.2 Sindh Health Sector Overview The public health sector of Sindh comprises of four tiers of service delivery and related ancillary institutions. The four tiers are: (i) medical education and tertiary or specialized hospitals; (ii) secondary care hospitals (district/Taluka headquarters); (iii) primary healthcare institutions (rural health centers/basic health units/maternal and child-care health centers); and (iv) community level services (female health workers [the Lady Health Worker (LHW) program]/community midwives/vaccinators). The Health Department (HD) follows a centralized administrative structure, with as many as 29 subordinate/attached offices directly reporting to the secretary of health, excluding the district health offices (Figure 4.2). Figure 4.2: Organogram of the Health Department 88 Sindh: Public Expenditure Review 2017 4.3 Context and Policy Framework 4.3.1 Sindh Government’s Sector Strategy and Reform Plans A situation analysis was undertaken in 2011, leading to the development of a provincial health strategy. In the post-devolution environment, as the administrative and fiscal space of provinces increased, there was a need to adjust healthcare delivery systems, governance structures, and financial allocations in the provinces to improve health outcomes, while maintaining the equity, quality, and efficiency of health care. With the assistance of the Technical Resource Facility (TRF), the Sindh government conducted an evidence-based situation analysis (SA) of the health sector in 2011. This was followed by the development of a health strategy that picked up on the key areas identified in the SA report and also made use of feedback from consultative sharing with the Sindh HD and development partners. The subsequent Sindh Health Sector Strategy is a broad-based long-term strategic plan (2012-20) accompanied by an M&E framework and a financial framework. The strategy calls for a total outlay of US$169.271 million over the 5-year implementation period from FY15 to FY20. These funds will be required over and above the existing funds made available to the Sindh government for the health sector. The 5-year plan in turn is the basis for detailed annual plans that link to the Annual Development Plan (ADP) and Operational Expenditure Forecast for the Sindh HD. The strategy focuses on three major areas as a way forward: (i) prioritized health responses; (ii) cross-cutting sectoral measures; and (iii) stewardship and re-structuring. The Sindh Health Sector Strategy is based on eight strategic directions: (i) strengthening district health systems; (ii) strengthening urban primary health care; (iii) regulating the private sector; (iv) streamlining human resources; (v) special areas of emphasis; (vi) integrated action on drugs; (vii) improving governance and accountability; and (viii) increased and more efficient funding. The HD had also prepared a detailed implementation plan for the strategy. However, for various reasons the implementation of the strategy has remained patchy. While there has been progress in some areas, other areas have yet to be addressed. The Sindh government has already initiated several reforms and others are being prepared. There is an initiative for large-scale contracting to be carried out that includes contracting out management of primary health care (PHC) and the contracting of district and Tehsil hospitals. Reforms in the district health system are under consideration and a revised policy direction regarding health insurance and regulation is being prepared. Nutrition has received special attention and the province has prepared an action plan that will be led by a high-level nutrition task force. The nutrition plan is multi-sectoral in nature. The province has also prepared a Costed Implementation Plan (CIP) for Family Planning (i.e., provision of contraception), and steps are currently being undertaken to implement the CIP. However, ta lack of leadership or administrative structure for taking reforms forward, the absence of a mechanism to ensure the continuity of new initiatives and reforms, and the current reporting arrangements, are all hindering the smooth functioning of the HD. The frequent changes of the secretary of health necessitate the creation of a mechanism whereby reforms are carried forward and reporting arrangements are simplified in order to enable the efficient utilization of administrative resources. While the direction of reforms is sound, there are doubts over implementation. The underlying basis of the direction of the reforms is sound and identifies the critical areas in which changes are required if progress is to be made. However, implementation is the key challenge, as the conversion of plans into actual results will require regular follow-up and close monitoring, along with strong strategic leadership. 89 Sindh: Public Expenditure Review 2017 4.3.2 Issues and Challenges in the Health Sector in Sindh 4.3.2.1 Sector Performance While Sindh has made some progress in improving health outcomes, the pace of change has been slow and disappointing. The yawning rural-urban divide is illustrated by the poverty headcount ratio in rural areas being almost double that of urban areas. As a result of political and ethnic frictions, the overall socioeconomic environment has deteriorated over the past 20 years, leading to increasing disparities, sharpening economic and social differentials between urban and rural areas, and slowing down much needed improvements in education, health, and family planning outcomes. The Lady Health Worker (LHW) program is the flagship program of the HD for community interventions. Despite its prominence, the LHW program’s coverage was found to be only between 20 and 43 percent in at six districts.72 The LHW program is also in urgent need of new leadership and strategic direction, in addition to tighter field-based supervision across all districts. At present, there is no comprehensive strategy for technical training, such as laboratory or operating theater skills for assistants, nurses, and other ancillary staff. With regards to authority within the public health sector, there are paradoxical instances of undue decentralization over budgeting issues and undue centralization over staffing issues. There is a lack of investment in disaster preparedness systems to mitigate morbidity and mortality. Sectoral planning, regulation, public private partnerships, and M&E are new areas that are neither structurally reflected, nor for which prior training and experience have been imparted to HD staff. 4.3.2.2 Key Health Issues and Challenges in Sindh High rates of population growth: One of the key challenges facing Pakistan is unbridled population growth. For Sindh the effect of its own internal population growth with migration from rural areas is multiplied by migration from other provinces. The high rate of population growth overshadows the capacity and ability of public health infrastructure in Sindh to provide services and creates issues of access. In areas where the private sector has stepped in to provide health services this also leads to issues of financial access for the poor. In the absence of a comprehensive social safety net program, the availability of public health services for the poor has become more constrained over time. Increasing rural-urban divide: As a result of political and ethnic frictions, the overall socioeconomic environment has gradually deteriorated. With a poverty headcount ratio in rural areas almost double that of urban areas, the healthcare delivery system urgently needs strengthening to increase coverage, especially in high-risk districts. Governance issues: Compared with neighbouring provinces, Sindh faces more frequent transfers of personnel, political interference in decision-making, a lack of accountability, a lack of transparency in procurement processes, and inadequate M&E systems. All of these issues impact adversely on staff motivation and the provision of incentives for staff to perform. This poor performance and low level dynamism in the health sector are factors behind the diminishing share of development-partner support for health funding in Sindh. Weak stewardship functions: Provincial departments lack much of the structure, the skilled staff, adequate resources, and even the time, to carry out their core stewardship functions. These functions include: financing health care; regulation and accreditation; M&E, including public health surveillance; strategy-setting and choosing priorities; and human resource planning for the sector. 72 USAID/JSI baseline study. Mid-term Evaluation of the USAID Deliver Project, February 2013. 90 Sindh: Public Expenditure Review 2017 Major gaps in PHC coverage for the urban poor: Sindh has an unusual composition compared with the rest of Pakistan, with 47 percent of the population residing in urban areas. Karachi has the highest population growth rate in Pakistan, at 3.2 percent per year, mainly driven by a constant influx of poor rural migrants. While wealthy households living in urban areas can afford private health care, the poor living in slums need publically-financed support if they are to have access to primary health care. With increasing urbanization (and burgeoning urban slums) and a disproportionate dilution of existing health infrastructure and services, major gaps in the coverage of primary health care services have developed, resulting in poor MDG indicators and polio outbreaks in low-income (i.e., slum) urban areas. Human-resource deployment, retention, and capacity are sub-optimal in rural areas, particularly for female health staff: At present, there is no comprehensive strategy for the basic training of technical and ancillary staff. There is also a chronic shortage of all categories of female staff and specialists in rural districts, and several “ghost� providers, while there are an excessively large general cadre and support staff at all levels, resulting in efficiency losses. 4.4 Sindh Health Sector Profile Sindh reported a total 642,848 inpatients in 2015. The HD oversees tertiary hospitals (9), secondary care facilities (91), and primary care facilities (1,791), and provided healthcare services to about 40 million clients in 2015. There are about 16,000 beds in the public sector and 11,000 beds in 362 private hospitals. The Bureau of Statistics in Sindh reports that in 2015 the total number of inpatients was 642,848, of which 47 percent were male, while there were 39.22 million outpatients, of which 44 percent were male, who were provided with services at health institutions across the province. Young children from poor households are particularly exposed to Sindh’s poor performance in health indicators. Table 4.1 and Figures 4.1, 4.3 and 4.4 below show the dire state of health indicators affecting young children in Sindh. Average infant mortality in the province is 82 per 1,000 live births, but 117 in the poorest income quintile, while average under-5 mortalities is higher, at 104 per 1,000 live births, and 139 in rural areas. This means that the chances of a child from a poor household in a rural area not surviving to its fifth birthday is 14 percent. Meanwhile, stunting affects 42 percent of children across the province on average, but 60 percent in the poorest income quintile, while the share of underweight children has remained almost static, from 40 percent FY04 to 42 percent in FY14. Full immunization coverage for 12-23 month old infants is low, especially in rural districts, with the average for the province at 43 percent, and 32 percent for rural areas. The total fertility rate (TFR) declined from 5.25 in FY04 to 4.0 in FY04, while skilled birth attendance rose from just 38 percent in FY04 to 65 percent in FY04. Table 4.1: Sindh health indicators Total Urban Rural Richest Poorest Karachi Larakana Division Division Infant mortality rate (per 1,000 live births) 82 57 106 30 117 52 109 Under-5 mortality rate (per 1,000 live births) 104 69 139 34 154 62 104 Under-5 stunting (%) 42 33 50 23 60 27 46 Source: MICS 2014. 91 Sindh: Public Expenditure Review 2017 Figure 4.3: Nutrition status of under-5 children in Sindh 80 Underweight Stunted 70 60 50 40 30 20 10 0 Source: Data from MICS. Figure 4.4: Immunization coverage of 12-23 month old children for full immunization 80 70 60 50 40 30 20 10 0 Source: Data from MICS. 4.5 Public Health Expenditure Public health expenditure in Sindh has been increasing over the years, both in nominal and real terms. In nominal terms the health budget rose from PRs 14.9 million in FY09 to PRs 51.1 million in FY15 (a 350 percent increase), while in real terms the budget increased about 200 percent to PRs 30.4 million over the same period. As a percentage of provincial GDP, the health budget has increased from 0.47 percent to 0.98 percent, almost doubling from FY09 to FY15. This increase in health expenditure was also affected by the 2010 devolution following the 7th NFC Award, whereby the financing of vertical programs was booked at the provincial level. The budget execution rate for health is around 80 percent, with the blip seen in FY11 due to the federal grants for the vertical programs being budgeted at the federal level and expenditure booked at the provincial level. Although development expenditures remained about the same—except for an increase seen in the federal transfer of programs—recurrent expenditures maintained annual growth of about 30 92 Sindh: Public Expenditure Review 2017 percent, only slowing somewhat in the 2 years from FY14 to FY15 to around 10 percent. Per capita expenditure increased from PRs 388 to PRs 1,135 in nominal terms, and to PRs 676 in real terms, during the period FY09 to FY15. Figure 4.5a: Sindh’s health budget (PRs million and Figure 4.5b: Sindh health budget execution rates (PRs percent) million and percent) nominal real % of GDP Budget Expenditure Execution rates 60000 1.2 70 130% 50000 1.0 60 110% 50 90% PRs billion 40000 0.8 40 70% 30000 0.6 30 50% 20000 0.4 20 30% 10000 0.2 10 10% 0 0.0 0 -10% FY09 FY10 FY11 FY12 FY13 FY14 FY15 FY09 FY10 FY11 FY12 FY13 FY14 FY15 Source: Finance Department, GoS. Figure 4.6a: Recurrent and development expenditures Figure 4.6b: Distribution of health budget and (PRs million) spending by minor functional classification (percent) 076 - Health Administration Current Expenditure Development Expenditure 074 - Public Health Services 45000 073 - Hospital Services 40000 071 - Medical Products, Appliances and Equipme 100% 35000 18% 10% 14% 11% 13% 11% 15% 11% 13% 10% 11% 12% 14%11% 4% 9% 4% 9% 80% 11% 8% 11% 11% 11% 12% 15% 15% 4% 11% 30000 60% 25000 40% 70% 81% 75% 78% 75% 77% 82% 80% 85% 79% 81%79% 70% 74% 20000 20% 15000 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Budget Expenditures Budget Expenditures Budget Expenditures Budget Expenditures Budget Expenditures Budget Expenditures Budget Expenditures 10000 5000 0 FY09 FY10 FY11 FY12 FY13 FY14 FY15 FY08 FY09 FY10 FY11 FY12 FY13 FY14 Source: Data from Accountant General of Pakistan. An analysis of the budget by salary and non-salary components exposes an important caveat. Most basic health units (BHUs) in Sindh are contracted out to the President’s Primary Health Care Initiative (PPHI). Therefore, the salary component of these BHUs is reflected as provincial grants, thereby inflating the non-salary budget. This means that the non-salary component of expenditure, after accounting for the grants, increased from 35 to about 40 percent during the review period. Meanwhile, the repair and maintenance budget remained at around 1 percent throughout the period. 93 Sindh: Public Expenditure Review 2017 Figure 4.7a: Expenditure share by sub-sector (percent) Figure 4.7b: Distribution health expenditure by inputs (percent) Health Administration Salaries and Pension Operating Expenses Public Health Services Grant and subsidies Physical Assets Hospital Services Civil works Reapir and maintanance Medical Products, Appliances and Equipme Others 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% FY09 FY10 FY11 FY12 FY13 FY14 FY09 FY10 FY11 FY12 FY13 FY14 FY15 Source: Data from Accountant General of Pakistan. 4.5.1 Sindh’s Health Sector Budget FY17 The BSP for the current fiscal year, FY17, notes the weak performance of the health sector. The Budget Strategy Paper (BSP) for the current fiscal year, FY17, with regard to the health sector states that the “health system faces many challenges that are generally crosscutting in nature. Health-poverty-illiteracy nexus, coupled with gender inequality and regional disparities, results in uneven health coverage; inadequate supply of pure drinking water and sanitation services trigger the spread of communicable diseases�. While Sindh’s major health indicators have shown some improvement since FY07, they still remain well short of the MDG targets set for 2015: the infant mortality rate stood at 82 per 1,000 live births (Brass Method, MICS 2014), compared with an MDG target of 40; and the maternal mortality rate stood at 314 per 100,000 live births in the same year vis-à-vis the MDG target of 140. The BSP identifies four key challenges: (i) system challenges, including high out-of-pocket expenditures; (ii) governance challenges, including inadequate M&E, a shortage of female staff, and low levels of accountability; (iii) service delivery challenges, including slow progress in health service delivery indicators, for example low coverage of LHWs and the inadequate availability of drugs; and (iv) financial constraints, including the capping of federal transfers for vertical programs. The BPS then identifies three key principles for the health development program. These three key principles are identified as: (i) prioritizing the completion of ongoing projects, while non-performing projects may be terminated; (ii) making 20 percent of the available resources from the ADP for allocation to new projects; and (iii) making 12 to 14 percent of the available resources from the ADP for allocation to projects at the district level, while the remaining resources will be allocated for projects at the provincial level. Health and population welfare sector financing. The original estimate for health and population welfare for FY16 was 12.18 percent of the total recurrent budget, although this was later revised up to 12.42 percent (PRs 62.29 billion). For the current fiscal year, FY2016-17, the recurrent budget is PRs 68.55 billion, or 12.2 percent of the total recurrent expenditures budget. This means that while the allocation for health in the recurrent budget increased by about 10 percent in nominal terms, it has seen no significant increase as a share of the overall recurrent budget. For development expenditures, the original ADP was revised down by 16 percent, from PRs 142 billion to PRs 119.37 billion, in FY16. Health 94 Sindh: Public Expenditure Review 2017 and population welfare comprised 9.5 percent (PRs 11.44 billion) of the total ADP. For FY17, 8.9 percent (PRs 14.12 billion) of the total ADP has been allocated to health and population welfare. Although this represents a 23 percent increase on FY16 as a share of total development expenditures, in nominal terms it is less by 0.6 percent. This is somewhat mitigated by the fact that the Sindh government has identified nutrition as a special initiative and allocated PRs 1.0 billion in the FY17 budget (or PRs 3.0 billion over the next 3 years). With this proposed allocation, the total allocation for health, nutrition, and population welfare has increased to 9.6 percent of the ADP, which represents an increase of almost 32 percent of the revised allocation in FY16. 4.5.2 Performance of the Health Sector ADP Composition of the Health ADP. The Health ADP consists of 147 projects out of which 89 (60.5 percent) are ongoing and the remaining 58 projects (39.5 percent) are new. Of the total 147 projects, 16 projects (PRs 4.9 billion and 31 percent of the allocation) are for primary health and preventive care, excluding infrastructure. The total allocation for infrastructure is PRs 7.5 billion, which is about 55 percent of the total ADP. The focus is on upgrading existing health facilities in the province and adding required capacity to hospitals, in addition to the construction of 27 new BHUs and four new maternal and child health (MCH) centers. Preventative priorities of the ADP include the LHW program and surveillance of communicable diseases, together with immunization. The LHW program has been shifted from the Health ADP to the regular budget, such that LHW expenditure is no longer included in the ADP. The preventive priorities include: (i) expanding the LHW program with an allocation of PRs 250 million; and (ii) strengthening surveillance of malaria; and (iii) immunization activities, which are included as new ADP projects. There is also a special allocation of PRs 1.0 billion shown under special initiatives for nutrition activities. The population welfare ADP consists of a total of six projects, two ongoing and four new. One project is to implement the Costed Implementation Plan (CIP) for Family Planning and another to establish an m-Health initiative (using mobile/smart phones to collect data). There are no new projects to address private sector involvement or innovations to increase contraceptive utilization in the province. The health sector was allocated PRs 15.0 billion for the annual development program in FY17,73 an increase of 15.4 percent on the previous year. However, despite the government’s claims of prioritizing the health sector, the share of the health sector in the ADP has declined from 10.2 percent in FY14 to 7.6 percent in FY17. Within the HD, the Sindh government has allocated about 43.4 percent of the health sector annual development program financing for the renovation, modernization, and addition of facilities at various hospitals across the province for the coming fiscal year. Meanwhile, the provisions for the preventive care program stand out, with a 19 percent share of the Health ADP. In addition, around PRs 4.3 billion was allocated to teaching hospitals and for medical education, an increase of 15.2 and 13.6 percent, respectively. 73 Including PRs 145 million of a foreign-aided project. 95 Sindh: Public Expenditure Review 2017 Figure 4.8a: Share of the health sector in the ADP, Figure 4.8b: ADP Allocation to the health sector, FY17 FY12-FY17 (percent) (percent) 12% 50% 43.4% 10.2% 45% 10% 9.4% 40% 9.2% 35% 7.6% 30% 8% 25% 6.8% 18.9% 6.2% 20% 15.2% 13.6% 6% 15% 10% 7.6% 5% 1.4% 4% 0% 2% 0% FY12 FY13 FY14 FY15 FY16 FY17 Source: P&DD, GoS. Analysis of the provincial share of the Health ADP shows high regional disparities. Per capita ADP allocations for FY17 show that the largest share of health sector projects goes to Sujawal and Matiari, which stand out at PRs 1.2 billion and PRs 1 billion per capita allocations, respectively, followed by Kahirpur, Jamshoro, Hyderabad, and SBA. In contrast, districts including Kashmore, Thatta, Badin, Ghotki, Sanghar, and Shikarpur receive the lowest per capita allocations. The top 22 percent of provincial population districts receive around 40 percent of total Health ADP allocations, whereas the bottom 25 percent receive only 6 percent of the ADP allocations. Figure 4.9: Per capita ADP allocations, FY17 (PRs billion) 1.4 1.2 1.2 1.0 1.0 0.7 0.8 0.6 0.6 0.5 0.5 0.6 0.5 0.4 0.4 0.4 0.3 0.3 0.3 0.3 0.3 0.4 0.1 0.1 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.0 Source: Planning and Development Department, GoS 96 Sindh: Public Expenditure Review 2017 The size of the health sector portfolio has Figure 4.10: Allocations to health sector projects declined over the past 3 years. For instance, (percent and PRs billion) there were a total of 147 health sector projects New On-going Total Allocation (Rs. Billion) in FY17, down from 210 projects in FY15, with 58 new projects. However, the share of new 90% 83.9% 18 79.3% 77.4% projects in the ADP allocation has increased 80% 16 68.0% from 16 to 37.5 percent in the past 2 years, 70% 62.5% 14 while allocations to ongoing projects have 60% 12 declined significantly, from 84 to 62.5 percent in 50% 10 the total allocations for FY17. This expansion of 40% 32.0% 37.5% 8 the portfolio—adding new projects while 30% 22.6% 6 continuing with all existing projects—dilutes 20.7% 16.1% 20% 4 scarce resources from existing projects in the portfolio, contributing to increased costs and 10% 2 time overruns. As a result, more than 78 0% 0 FY13 FY14 FY15 FY16 FY17 percent of projects take more time than envisaged at the planning stage, while cost Source: Planning and Development Department, GoS overruns affect 4.1 percent of completed projects. The average cost of projects increased faster than the average allocation size, although projects are still expected to be completed within 5 years. For instance, projects were allocated only 17.7 percent of the average project cost in FY17. Moreover, 45.6 percent of projects were allocated less than PRs 50 million in FY17. However, this percentage has improved over the previous 2 years, from 71.4 percent in FY15. In addition, 25 percent of total projects had allocations of PRs 100 million or above in FY17 up from just 13.3 percent in FY15. Consequently, almost 82 percent of the health projects in the ADP portfolio should be completed within 5 years, while 14 percent (21 projects) are expected to take 6 to 10 years to complete based on current levels of allocation. Figure 4.11A: Distribution of portfolio allocation (% of Figure 4.11B: Project completion time at current total projects) levels of allocation (percent) Less than Rs. 50 million Rs. 50-100 million Over Rs. 100 million less than 1 yr 1-5 yrs 6-10 yrs 11-20 yrs 80% more than 20 yrs 69.2% 71.4% 70% 70% 63.5% 61.9% 57.5% 60% 60% 54.2% 49.4% 50% 45.6% 50% 40.1% 40% 40% 23.3% 28.0% 30% 24.5% 30% 26.3% 24.8% 15.4% 13.3% 19.7% 20% 13.5% 20% 16.7% 13.8% 10% 10% 0% 0% FY13 FY14 FY15 FY16 FY17 FY13 FY14 FY15 FY16 FY17 Source: Planning and Development Department, GoS. 97 Sindh: Public Expenditure Review 2017 The decline in overall allocations to the health Figure 4.12: Throw forward FY13 to FY17 (percent of sector, coupled with fragmentation of the ADP total project cost) diluting resources for existing projects, has 90% contributed to substantial throw forward of 80.58% 78.50% 77.92% 80% commitments. For instance, the throw forward 70.73% in the health sector for the current fiscal year, 70% 62.71% FY17, has increased by 27 percent compared 60% with FY16. Throw-forward in FY17 has reached 50% PRs 59 billion (71 percent of total project 40% costs) compared with PRs 46 billion in FY16, 30% placing resource liabilities on future development programs. The Sindh 20% government needs to focus on the efficient 10% completion of projects, as well as rationalizing 0% the size of the portfolio by eliminating non- FY13 FY14 FY15 FY16 FY17 performing projects. Source: Planning and Development Department, GoS 4.6 Allocative Efficiency of Health Expenditures The total allocation for health on the current budget is PRs 61.759 billion, of which about 50 percent is for Karachi, and 54 percent of the total is allocated to non-salary spending. The non-salary allocation varies from highs of 69 percent, 64 percent and 63 percent for Karachi, Sujawal, and Khairpur, respectively, to lows of 13 percent and 17 percent for Kashmore and Naushero Feroze, respectively. Excluding Karachi, Sujawal and Khairpur, the average allocation of non-salary spending is 35 percent. The current budget allocation for medical education is PRs 4.129 billion, of which 79 percent is non- salary and includes grants to autonomous medical institutions. For population welfare, the current budget allocation is PRs 3.7 billion, of which 57 percent is for non-salary allocations. There is an allocation of PRs 1.026 billion under a code A03970, which is most likely for logistics, which includes contraceptives. Figure 4.13: Recurrent expenditures by district 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Source: Data from Accountant General of Pakistan. 98 Sindh: Public Expenditure Review 2017 Figure 4.14: Childbirth at public health facility and per-capita expenditure, by district (percent and PRs) Delivery Public Sector Exp per capita 35 5000 30 4500 4000 25 3500 Percent 20 3000 PRs 2500 15 2000 10 1500 1000 5 500 0 0 Source: Data from Accountant General of Pakistan. In Figure 4.14, by plotting childbirth in a public health facility and per capita expenditure by district we see that there is no discernable relationship between expenditures and the utilization of public facilities. For instance, if we look at the relationship between underweight children and per capita health expenditure, we find that allocations by district do not match population needs (Figure 4.15). In view of this, the HD should consider undertaking a serious rethink of the focus of health expenditure and whether spending is achieving the government’s health-related goals. Figure 4.15: Malnutrition and health expenditure, by district, FY15 (percent and PRs) Underweight Exp per capita 80 5000 70 4500 4000 60 3500 50 Percent 3000 PRs 40 2500 30 2000 1500 20 1000 10 500 0 0 Source: Data from MICS and Accountant General of Pakistan. 99 Sindh: Public Expenditure Review 2017 The relationship between health expenditure and indicators is Scoring: Districts were ranked based on the confounded by many other factors, such as education and social following indicators: (low to high) the share norms. As such, an examination of whether expenditure is being of deliveries at a public health institution; directed appropriately is warranted. We do this by plotting per the share of fully immunization of children capita expenditure against an aggregate score developed from aged 12 to 23 months; and (high to low) the three health indicators: (i) the share of deliveries at a public health share of underweight children under-5. The institution, (ii) the share of fully immunized children aged 12 to 23 maximum possible score was 69 and months, and (iii) the share of underweight under-5 children. districts were ranked by per aggregate achievement. This score was then plotted Excluding the outlier, Hyderabad, we find that the score increases against per capita expenditure to offset the by 0.005 for every PRs 1.0 increase in per capita expenditure. effect of population size. However, the R-squared value reflects that this relationship only explains 32 percent of the variation. Figure 4.16: Per capita expenditure and health outcome performance score Score Linear (Score) 70.00 y = 0.0056x + 52.366 65.00 R² = 0.3291 Performance score 60.00 55.00 50.00 45.00 40.00 0.0 200.0 400.0 600.0 800.0 1000.0 1200.0 1400.0 1600.0 1800.0 2000.0 Per capita expenditure (PRs) Source: Data from MICS and Accountant General of Pakistan. 4.7 Recommendations While a small number of health indicators in Sindh may have seen some modest improvement in recent years, the pace and breadth of that improvement is disappointing. This slow pace and narrow breadth of improvement has been further exacerbated by inefficiencies in budget allocations for health spending, which are not responsive to public health needs. The huge urban-rural disparity in health indicators in Sindh is also a major concern and one that requires immediate attention, as it adversely affects roughly half the population of the province. While the Sindh Health Sector Strategy is certainly a step in the right direction, its success will depend largely on its implementation. The Sindh government is currently considering restructuring of the HD and this could offer a unique opportunity to address some of the inequities in the sector by improving the effective implementation of the provincial health strategy. Sindh’s high population growth rate needs to be curbed if the provision of basic health services is to keep up with demand. The current rate of population growth is unsustainable and creates continual pressure on the health service delivery system to expand its capacity. Without concurrent growth in spending this leads to the system being unable to provide basic requirements to the entire population. The focus provided by the Costed Implementation Plan (CIP) on Family Planning is one first step to 100 Sindh: Public Expenditure Review 2017 address unsustainable population growth. However, whether the requisite prioritization and financing will be forthcoming to make the CIP effective remains to be seen. Stunting is a major issue and one that in the longer term will have an adverse impact on economic growth of the province. A plan to address the issue of stunting has been initiated by the Sindh government, but once again it remains to be seen whether the plan can gain the necessary size and momentum to have a significant impact. Governance is another serious issue that needs to be addressed if health outcomes are to improve. The HD in Sindh is faced with the key issues of the over-centralization of authority, too many vertical programs, and the lack of mechanisms to enhance accountability and transparency. There is also weak oversight of service delivery and no clear M&E framework. This is now all the more important in the post-devolution environment, given that the focus of the mandate has shifted from just service delivery to include the new challenge of the stewardship of the entire health sector. One key problem is the lack of strong and stable leadership in the HD, which would benefit from reducing the frequency of staff transfers and the building of capacity in sectoral governance. Public sector spending on the health sector is low in real terms, and there are major inefficiencies and highly regressive OOP expenditures. About 66 percent of total health expenditure in Sindh comes from households’ out-of-pocket (OOP) expenditures, with the rest mainly accounted for by public sector sources. OOP spending is regressively distributed, with lowest income quintile spending on average 7 percent of monthly household income on health care, compared with the provincial average of 5.2 percent across all quintiles. Medicines account for largest share of OOP, mainly due to their inappropriate use by the private sector and low availability in the public sector. Within the public sector, the non-salary operational budget at the district level is particularly inadequate, ranging from 17 to 25 percent. Meanwhile, allocations for development expenditure at both the provincial and district levels are heavily tilted towards facility construction, which has high operational cost implications for coming years. Although primary health care provides a cost-efficient means for disease control and management, this has not been the focus of funding increases. There are also serious inefficiencies within the public sector, due to the proliferation of parallel vertical programs, redundant posts, an overly ambitious but sub-optimal district health system, and an input-based financial system. There are several positive programs underway. The introduction of subsidized user charges has been successfully piloted by the Dow University of Health Sciences (DUHS), drawing in large numbers of patients. Results-based financing has been introduced through the Norway-Pakistan Partnership Initiative (NPPI) in the form of two pilot schemes each using vouchers and contracting in disadvantaged districts, with support from independent monitors. However, these pilots have been slow to roll out and as a result the evaluation results are still awaited. Progress by PPHI in terms of BHU infrastructure and functionality have been seen. However, there is little difference in terms of preventive and promotive services, and the technical aspects of the services supplied are questionable. Contracting needs modification in terms of focusing the scope on low- performing facilities, rather than blanket contracting of both weak and better-performing facilities. The opening up of the bidding process would draw in more NGOs with expertise in health and also encourage the introduction of performance indicators. The LHW program has been successful but needs to be expanded. Successful evidenced-based pilots of the LHW program to provide neonatal care, childcare, and nutritional support have been implemented in rural Sindh, but the lessons learned have yet to be mainstreamed. Moreover, alternative outreach strategies are needed for remote areas not covered by LHWs. 101 Sindh: Public Expenditure Review 2017 There have also been successful pilot programs run by GPs in urban areas. Pilot programs using GPs in urban areas have been launched by NGOs covering NCDs, TB DOTS, family planning, and STIs and have proved successful. However, once again the lessons learned from these pilots have yet to be scaled up into a systematic strategy. Private sector participation needs to be encouraged but also better regulated. Sindh has the highest concentration of private sector health facilities and the highest utilization of the private sector for health in Pakistan. However, the care provided is largely unregulated. This unregulated health sector manages over 80 percent of the province’s health infrastructure and is utilized by 78 percent of households, including the poor. There are uneven standards of care within the private sector and reports of widely practiced quackery, shadow pharmacies, sub-standard diagnostics, and a lack of reporting linkages, with the poor most at risk of abuse. There has been little effort to increase consumer awareness and, although comprehensive consumer protection exists in Sindh, this needs to be reinforced in the case health care. Although regulatory needs are considerable across the sector, there is an absence of a regulatory authority or framework. A shift in approach is also needed to move away from a traditional punitive government-dominated approach to regulation towards a mix of incentives and legislation implemented in coordination with multiple stakeholders. The important issue of regulating pharmaceuticals is overlooked. Sindh currently lacks a comprehensive pharma strategy and this sub-sector has traditionally been overlooked in planning for health systems and health financing. Pressing regulatory issues include the proliferation of shadow pharmacies, inappropriate prescriptions by medical practitioners, and the relatively low use of recommended and cost-effective generic medication. There are also dire supply management issues, such as a high frequency of stock-outs, low quality parameters in purchasing, and a lack of transparent checks in the management of inventories. Health financing measures are also needed to reduce the high OOP expenditures on medication. Above all, there is a lack of adequate and skilled human resources, together with the absence of a central regulatory body to act as the hub for pharma issues and functions. Blood banks should be developed, with effective monitoring by the regulator, while a new regulator will need to be established to monitoring other uncovered facilities. Regulatory work has been initiated by the Sindh Blood Transfusion Authority in the area of registering blood banks and this is a promising first step. Similar work also needs to be extended to diagnostic facilities, private health service providers, and shadow pharmacies, and will require the setting-up of a new regulatory authority. Human resource deployment, retention and capacity are sub-optimal in rural areas, particularly among female health staff. At present, there is no comprehensive strategy for staff training and improvement. The annual increase in the number of doctors in Sindh is about 2,500, compared with 900 nurses and public health nurses) (LHVs), while at the same time there is unplanned growth of medical colleges in the province, which now number 19. There is a chronic shortage of all categories of female staff and specialists in rural districts, and several “ghost� providers, creating an urgent need to develop alternative staffing plans. Meanwhile, the excessive size of the general cadre and support staff at all levels results in efficiency losses. Existing administrative posts at both district and provincial levels lack mandatory management qualifications and there are frequent staff transfers and politicized postings that undermine performance. 102 Sindh: Public Expenditure Review 2017 Chapter 5: Need for a coherent policy: protecting the poor and vulnerable through social protection 5.1 Introduction While Sindh has made progress in poverty reduction, in recent years it has only fared better than one other province, namely Baluchistan. Although Sindh reported a 7.4 percent decrease in its poverty incidence rate over the 3-year period FY11-14, if this improvement is compared nationally the province only fares better than Baluchistan. It is less than one-third the decline seen in Punjab and almost one- fifth of the decline in Khyber Pakhtunkhwa (KP). In addition, Sindh still has high levels of inequality across various dimensions of human capital development, such as education, health, and labor opportunities. Based on the Sindh government’s social protection (SP) expenditure classification, on average SP spending has grown significantly since FY09. However, scrutiny of the data reveals a significant misclassification of expenditures, rendering real SP spending much lower. In addition, the analysis suggests that Sindh social protection and labor (SPL) initiatives are conceived, planned, and executed in a piecemeal fashion across several line departments, with little or no consideration of pre- identified SP risks and vulnerabilities. In the case of Sindh, SPL initiatives are characterized by a lack of careful program design and prudent resource allocation. To achieve its goals, a unified policy framework supported by institutional and implementation arrangements is essential. In order to achieve the SPL strategic goals of increasing resilience, equity, and opportunity, the Sindh government needs to consolidate, coordinate, and monitor SPL activities across government departments and line agencies. To this end, a unified policy framework duly supported by responsive institutional and implementation arrangements is essential. Furthermore, more equitable economic growth should be the focus of Sindh’s SP policy reform agenda going forward. Programs and schemes that aim to improve access to social services and productive inclusion, such as CCT for health, nutrition, and education, and better access to skills training, financial markets, job networks, and life skills (basic numeracy, negotiating skills, and work ethics for those without formal education) for the bottom 40 percent—especially women among the bottom 40 percent—should all be prioritized. This chapter presents a Public Expenditure Review of the social protection and labor (SPL) sector. For the purposes of the analysis in this chapter, SP initiatives are broadly categorized as: (i) social assistance (cash and in-kind transfers); (ii) social insurance (primarily workers’ welfare initiatives, excluding formal sector pensions that are covered separately under the World Bank’s analytical work on pensions); (iii) labor-market programs (skills and entrepreneurship development, e.g., vocational training); (iv) social care services (programs for widows, orphans, the elderly, free medical treatment, etc.); and (v) subsidies (covering only provincial spending on subsidies aimed at providing relief to the poor and vulnerable, such as subsidies for transport, agricultural inputs, etc.). This PER is limited to public sector expenditure, and so excludes private sector expenditure where beneficiary contributions are included in the benefits. Similarly, expenditure on the contributory Government Provident Fund (GPF) of government employees is also excluded. Thus, the scope of this PER is limited to non-contributory benefit transfers. 103 Sindh: Public Expenditure Review 2017 The primary aim of this chapter is to enhance the efficiency and effectiveness of public spending in the SPL sector. The ultimate goal of the analysis in this PER is to explore and reveal the reasons behind low efficiency in the public sector SPL initiatives, and suggest ways forward to improve service delivery and obtain better value for money from SPL spending. Box 5.1: Definition and goals of SPL systems SPL systems are defined in the World Bank’s 2012-22 Social Protection and Labor Strategy as policies and programs that help individuals and societies to manage risk and volatility, and protect them from poverty and destitution, through instruments that improve resilience, equity, and opportunity. SPL systems have three intertwined goals: a. Resilience for the vulnerable through insuring against the adverse impacts to wellbeing from a range of shocks; b. Equity for the poor through protecting against destitution and promoting equality of opportunity; and c. Opportunity for all through promoting human capital in children and adults. 5.2 Social Protection Policy in Sindh There is a lack of long-term planning for social protection policy in Sindh. There is an absence of any long-term SP vision clearly articulated through an SP policy, with a resultant SP strategy for Sindh. This has had the effect of leading to short-termism with regards to issues such as human capital development, when in fact such complex fields require long-term investment and, therefore, a long- term strategy and planning. This has resulted in piecemeal and “needs-based� budgetary allocations that, while being insufficient in the first place, have remained largely under-utilized due to weak institutional and administrative capacities. The Sindh government does not have an overarching body entrusted to coordinate, consolidate, and oversee the provision of SPL programs and services in the province. In the absence of a policy framework, Sindh’s SPL interventions are mostly ad hoc, and are managed under several stand-alone, uncoordinated, and poorly targeted programs.74 These programs include zakat, livelihood and educational stipends, healthcare assistance, general subsidies, social welfare activities, the Rural Support Program for Poverty Reduction, grants of state land to landless peasants, the Benazir Bhutto Shaheed Youth Development Program (BBSYDP), and the Employees Social Security schemes aimed at mitigating income shocks and reducing vulnerability among the destitution. While there are several national SPL interventions in the province, such as pension schemes and minimum wages, the federal 74 Several analyses have concluded that SP programs in Pakistan are generally poorly targeted. The following are some examples: Zakat targeting efficiency: https://www.microfinancegateway.org/sites/default/files/mfg-en-paper-targeting-efficiency-of- poverty-reduction-programs-in-pakistan-2006_0.pdf The paper comments on the nuances in targeting efficiency of zakat programs and concludes that there are leakages in targeting and, in at least one-fifth of cases, zakat funds go to third and fourth quintile households. A lack of targeting is the main concern in all social assistance programs. Zakat distribution and Baitul Maal schemes do not have any transparent and accountable method of targeting. In fact, these programs are aimed to target the “deserving needy�, but no objective targeting tool (e.g., proxy means testing) is used. According to the World Bank (2007), “around 27 percent of guzara (monthly cash allowance) beneficiaries and 37 percent of those receiving rehabilitation grants are not poor, accounting for 32 and 45 percent of the resources distributed under each modality�. Source < http://www.spdc.org.pk/Data/Publication/PDF/RR81.pdf> “Targeting Problems: Targeting based on means-testing is the prevalent form of social welfare schemes in the case of zakat distribution, as well as Baitul Maal schemes. Both these schemes do not have any transparent and accountable method for targeting.� Source < http://www.researchcollective.org/Documents/Social_Protection_Way_Forward.pdf> 104 Sindh: Public Expenditure Review 2017 zakat and Baitul Maal programs, and the national flagship social safety net program, the Benazir Income Support Program (BISP), there is little coordination between federal and provincial programs. 5.3 Impact of the 18th Constitutional Amendment and the Role of the Province in Social Protection Following the 18th Amendment to the Constitution in 2010,75 SP service delivery has largely become a provincial responsibility. The added obligations necessitate major policy and institutional reforms in an otherwise weak and, in some cases, non-operational SP sector at the provincial level that is characterized by limited planning, coordination, execution, and monitoring capacities. Currently, the Sindh government’s SPL initiatives aim at managing risks and vulnerabilities, mitigating the worst effects of poverty, avoiding detrimental coping strategies, and helping the poor and vulnerable to increase their resilience. However, the SPL initiatives that have so far been undertaken have largely proved ineffective. Instead, the non-existent policy framework and inappropriate budgetary allocations have created a de facto policy of regressive general subsidies, inadequate coverage, and inefficient and non-inclusive SPL interventions, as explained in more detail in the latter half of this chapter. 5.4 Poverty, Inequality and Vulnerability in Sindh Province Between 2001 and 2014, Sindh witnessed a considerable reduction in poverty incidence and vulnerability. However, despite this improvement, Sindh still has the second-highest poverty rate in Pakistan after Baluchistan, based on the newly adopted poverty line in 2016.76 While roughly half of Sindh’s population is urban, urban areas host only 28 percent of the poor. As a result, Sindh also has the highest gap between rural and urban poverty rates (Figure 5.1). Within the rural poor there is a Figure 5.1: Poverty rates comparison across provinces (percent) sub-group of ultra-poor, particularly in the southern districts of Sindh. This ultra-poor 60 Overall Urban Sindh Rural Sindh subgroup faces additional hardships as a result of recent natural disasters due to devastating 50 flooding and adverse soil conditions, such as 40 waterlogging and high soil salinity. In urban areas, growing peri-urban communities that have 30 migrated from rural parts of Sindh and other regions of the country (including Khyber 20 Pakhtunkhwa [KP] and the Federally Administrated Tribal Area [FATA]) in search of 10 livelihood opportunities, make up the bulk of the poor population. In addition, there are high 0 2010-11 2011-12 2013-14 levels of inequality across various dimensions of human capital development, such as education, Source: HIES and World Bank staff calculations. health, and labor-force participation. 75 th The 18 Amendment of the Constitution, among others, has abolished the concurrent legislative list, thus giving provinces more power over several subjects of public interest and service delivery, including social sector service delivery. 76 World Bank Poverty Team calculations using the new poverty line. 105 Sindh: Public Expenditure Review 2017 While there have been important gains in shared prosperity, in relative terms the progress in Sindh has been slower than in Punjab or KP. Sindh reported a 7.4 percent (FY11 to FY14) decrease in overall poverty incidence, but this is disappointing when measured against Punjab’s impressive poverty decrease of 25.5 percent and KP’s even more striking reduction of 34.7 percent. Sindh’s share of the national bottom 40 percent actually increased as a result of these varying relative reductions in poverty by province. The devastating floods of recent years, followed by droughts and the loss of livestock, together with the dependency of the rural population on agriculture, livestock, and fishing along the coastal belt, mean that the rural population is far more vulnerable than the urban population. Given that Sindh is so vulnerable to natural disasters (particularly floods and droughts), there is a high risk of shocks stemming from these events pushing near-poor and vulnerable households into poverty. According to Sindh’s Provincial Disaster Management Authority (PDMA) Multi-Hazards Contingency Plan 2013, 12 out of the 29 districts in Sindh are vulnerable to natural disasters owing to a “very high� to “high� risk of drought, while 14 districts are in the category of “very high� to “high� risk of flooding/severe rainfall. In addition, Karachi, Thatta, and Badin are in the “very high� to “high� risk category for cyclones and tsunamis (Annex Table 1).77 The rural population is also more vulnerable to changes in the cost of basic needs. When a vulnerability analysis is conducted using the new poverty line, a 20-percent increase in the cost of basic needs almost doubles the poverty rate in rural Sindh. From the 3.1 million individuals in FY14 in rural Sindh who were below the poverty line, this figure increased to 5.2 million after resetting the poverty line—an increase of 1.7 times. There is a sizeable difference in consumption between households in urban and rural areas in Sindh. Rural Sindh fares very poorly compared with both national and, to a lesser extent, urban Sindh consumption numbers. Figure 5.2 shows the differential in consumption expenditure across regions and between income quintiles. On a national basis, a household in the bottom quintile consumes on average only one-third as much as a household in the top quintile. In Sindh, the difference between rural and urban consumption is slightly less than nationally, with both the wealthiest urban and rural quintiles consuming double the poorest urban and rural quintiles. Figure 5.2: Average monthly consumption expenditure (PRs) National Urban Sindh Rural Sindh 45000 40000 35000 30000 25000 20000 15000 10000 5000 0 1st Quintile 2nd Quintile 3rd Quintile 4th Quintile 5th Quintile Source: HIES and World Bank staff calculations. 77 See Annex Table 1 for detailed table of risks from the PDMA Multi-Hazards Contingency Plan 2013, Government of Sindh. 106 Sindh: Public Expenditure Review 2017 Access to social services, particularly education, is highly unequal between urban and rural Sindh. This leaves individuals in rural areas at a huge disadvantage in terms of human capital development and their resultant opportunities to share in growth and prosperity (for a detailed analysis see Chapter 2 on education and Chapter 3 on health). Figure 5.3 shows that the difference in attendance ratios between primary school children in the first and fifth quintiles is almost four times, while for children in secondary school the difference is almost eight times. These huge differentials between primary and secondary school attendance signal very high levels of educational inequality, which in turn is a precursor for the transmission of intergenerational poverty. In addition, with such huge inequalities of opportunity, the poorest are destined to remain at a perpetual disadvantage in accessing the benefits of economic growth and new job opportunities. Figure 5.3: Primary and secondary school attendance ratios, by income quintile 90 Poorest Second Middle Fourth Richest 80 70 60 50 40 30 20 10 0 Primary school attendance and out of school children Secondary school attendance and out of school (Net Attendance Ratio) children (Net Attendance Ratio) Source: SEMIS According to the Sindh Multiple Indicator Cluster Survey (MICS) conducted in 2014, almost 42 percent of under-5 children are underweight and 17 percent are severely underweight. About 48 percent of children under 5 are stunted or short for their age, and one-quarter are severely stunted. The differential in rates of stunting and malnutrition is magnified across quintiles in each category. Almost three times as many children under 5 are malnourished in the poorest quintile compared with the richest quintile, while for stunting the difference between the richest and poorest quintiles is more than five times. Likewise, a comparison of full vaccination rates based on income quintiles shows a marked difference between the first and fifth quintiles. A child born into a poor household (bottom quintile) is one-third as likely to receive full vaccinations compared with a child born into a wealthy household (top quintile). 107 Sindh: Public Expenditure Review 2017 Figure 5.4: Percentage of children under 5 by nutritional status and income quintile 70 Poorest Second Middle Fourth Richest 60 50 40 30 20 10 0 Malnutrition (weight for age) Stunting (height for age) Source: Multiple Indicator Cluster Survey 2014. 5.5 Labor-Force Participation in Sindh Participation in the labor force in Sindh has remained at about 45 percent for almost a decade. Despite the stagnation in the overall level of labor participation, there has nonetheless been an encouraging increase in rural female labor-force participation, rising to almost 20 percent since 2005. However, this is in stark contrast to the participation of urban women in the labor force: urban female labor-force participation has remained stagnant at less than 2 percent, despite an overall increase in female education in the past 10-15 years (Figure 5.5). 78 Figure 5.5: Labor-force participation in Sindh (percent) 100 90 80 70 60 50 40 30 20 10 0 1992 1999 2005 2007 2008 2010 2012 2013 2014 Overall Male-Urban Female-Urban Male-Rural Female-Rural Source: Pakistan Labor Force Surveys. 78 Pakistan Social and Living Standards Measurement < http://www.pbs.gov.pk/content/pakistan-social-and-living-standards- measurement> 108 Sindh: Public Expenditure Review 2017 However, despite the modest progress in the rural areas, female labor-force participation in general is still extremely low in Sindh. Increasing labor-force participation, especially among women, is key to alleviating poverty and improving the prospects for economic growth. In Sindh, with more than 50 percent of its working age population out of work primarily due to very low female labor-force participation, eliminating the risks and vulnerabilities associated with poverty and achieving growth targets will remain a huge challenge. In addition, as shown in Figure 5.6, a considerable proportion of those rural women who do work are unpaid employees (80 percent). This share is higher than the share of rural women in unpaid employment in any other province in Pakistan. Figure 5.6: Composition of employment by type and gender (percent) Composition of Employment, Female Composition of Employment, Male Wage Employee Other Paid Wage Employee Other Paid Non-paid employee Employer Non-paid employee Employer Self-employed Self-employed 120 120 100 100 80 80 60 60 40 40 20 20 0 0 Rural Urban Rural Urban Rural Urban Rural Urban Rural Urban Rural Urban Rural Urban Rural Urban KP Punjab Sindh Balochistan KP Punjab Sindh Balochistan Source: Pakistan Labor Force Survey, LFS 2014/2015 Similar to Pakistan as a whole, Sindh is currently experiencing a youth bulge, with a large number of young people reaching working age and entering the labor force. This demographic dividend could represent a major opportunity for economic growth in Sindh if it can be harnessed effectively. Almost 27 percent of the entire population of Sindh is between 15 and 29 years of age.79 However, owing to limited investment in human capital, labor-market initiatives to harness this potential—especially aimed at addressing the gender gap and critical supply-side issues—should be undertaken as an urgent priority. These initiatives include: pre-natal and early childhood interventions aimed at reducing stunting and boosting cognitive development; pre-school learning and development and family support for schooling for the extremely poor; skills training programs for vulnerable youth; entrepreneurship programs promoting access to finance; and job placement for vulnerable groups. Besides the number of individuals in work, the quality and productivity of jobs is another important factor in boosting economic growth and helping to reduce poverty-related risks and vulnerabilities. The employment-to-population ratio in Sindh was 42.7 percent in FY11, implying that only 42.7 percent of the working-age population in the province (10 years and older)80 was actually employed.81 Currently, 79 Sindh Youth Policy 2016 < http://www.youthaffairs.gos.pk/wp- content/uploads/2016/03/YOUTH_POLICY_FRESH_TITLE_FINAL-9-12-2015.pdf> 80 In Sindh government surveys the working age is defined as 10 years and older. Child labor is allowed in some instances where stringent conditions (number of workings hours, breaks, the nature of work, etc.) are followed. 81 Sindh Employment Trends 2013 < http://www.ilo.org/wcmsp5/groups/public/---asia/---ro-bangkok/---ilo- islamabad/documents/publication/wcms_222494.pdf> 109 Sindh: Public Expenditure Review 2017 a large share of the working population in Sindh (about 85 percent) is employed in jobs in the informal sector, with consequently lower levels of earnings and productivity relative to formal sector jobs. Among employed women, unpaid work is the norm and many are employed in farm household enterprises. Among employed men, a minority are employed as formal sector wage workers, while the majority is self-employed in informal jobs. Despite an increase in overall educational attainment, there has only been very slow improvement in the quality of jobs over time, while the pace of the structural transformation (workers moving out of agriculture and going into services and industry) has also been sluggish. 5.6 Social Protection Expenditure in Sindh The Sindh government’s budget analysis of the previous fiscal year, FY16, focuses on the improvement of overall social sector service delivery and on increasing the share of budget allocations going towards priority policy areas. 5.6.1 Social Protection Spending in Sindh Social protection spending has seen significant variations in recent years, indicating ad hoc policy interventions. Although the share of SP expenditure in overall spending remained, on average, about 5 percent of Sindh’s budget in the period FY09 to FY15, there was significant variation between years. While education and health expenditures have both seen persistent increases in their shares since FY12, the overall share of SP expenditure declined from a high of 9 percent in FY12 to 5 percent in FY15. The large variations are indicative of incoherent and ad hoc SP interventions. Despite the volatile nature of the share of SP expenditures over the years, the overall share of SP expenditure grew significantly in the period from FY09 to FY15, expanding by about 32 percent on average in real terms (Figure 5.7). Figure 5.7: Share of total expenditures, FY09 to FY15 (percent) Health Education Social Protection Others 100% 90% 80% 70% 64% 65% 69% 72% 71% 67% 60% 74% 50% 40% 30% 2% 5% 4% 5% 5% 6% 9% 20% 21% 21% 21% 21% 15% 17% 14% 10% 8% 5% 5% 6% 9% 9% 9% 0% FY09 FY10 FY11 FY12 FY13 FY14 FY15 Source: Data from Accountant General of Pakistan. However, disaggregated data of this SP expenditure paint a very different picture. Disaggregated spending on SP components listed in Table 5.1 shows the department data of expenditures that are classified under SP. There appears to be a significant misclassification of these expenditures under the SP functional classification, while several departments listed have data showing SP spending, without having any SP-related activities. 110 Sindh: Public Expenditure Review 2017 Table 5.1: Disaggregated social protection spending (PRs) PRs million FY12 FY13 FY14 FY15 S26 - GOVERNOR'S SECRETARIAT/HOUSE - 570 - - S31 - FINANCE DEPARTMENT & COOPERATION DEPART 268 216 948 961 S33 - BOARD OF REVENU 20,097 6,628 2,132 1,611 S35 - AUQAF, RELIGIOUS AND MINORITIES AFFAIRS 213 782 602 711 S36 - PLANNING & DEVELOPMENT DEPARTMENT 42 521 112 2 S38 - C & W (WORKS AND SERVICES) DEPTT. 9,873 6,535 4,277 15,145 S39 - LOCAL GOVERNMENT DEPTT. 9,334 6,526 6,644 5,098 S40 - EDUCATION & LITERACY DEPARTMENT 70 210 298 349 S41 - CULTURE & TOURISM 1,338 2,109 2,386 1,854 S42 - HEALTH DEPRATMENT 3 97 100 206 S43 - POPULATION WELFARE 1,137 1,409 1,966 2,437 S44 - LABOUR, TRANSPORT, INDUS. & COMMERCE 168 12 13 15 S46 - ENVIRONMENT &ALTERNATE ENERGY - - - - DEPARTMENT S49 - IRRIGATION & POWER DEPTT. - 67 130 13 S53 - MINES & MINERAL DEVELOPMENT 8 9 9 11 S54 - LIVESTOCK AND FISHERIES DEPARTMENT - - - - Total 42,551 25,691 19,617 28,413 Source: Data from Accountant General of Pakistan. In view of this discrepancy, the World Bank collected data from FY13 to FY16 from Sindh government departments under the ASPIRE82 classification. The analysis below is based on the data collected from different line departments with active SP roles and interventions. There was a nominal increase in SPL spending from FY13 to FY14. However, a sharp increase in SPL spending is witnessed in FY15, owing to massive subsidies to the transport sector. Figure 5.8a shows overall SPL spending from FY13 to FY16. Figure 5.8b shows individual trend lines representing the rest of the SPL categories (social assistance, social care services, labor-market programs and social insurance), and highlights the negligible expenditure increases if general subsidies are stripped out. Figure 5.8a: SP spending over FY13 to FY16 (PRs Figure 5.8b: SP spending disaggregated (PRs million) million) General Subsidies Social Assistance 30,000 Labor Market Programs Social Care services Social Insurance 30,000 25,000 25,000 20,000 20,000 15,000 15,000 10,000 10,000 5,000 5,000 - FY13 FY14 FY15 FY16 - FY13 FY14 FY15 FY16 Source: Data from the Government of Sindh. 82 The Atlas of Social Protection Indicators of Resilience and Equity: A database of comparable indicators to monitor SPL performance over time and across countries 111 Sindh: Public Expenditure Review 2017 5.6.2 Share of SP Sub-categories in Overall SP Spending in FY16 In FY16, the bulk of SP spending went towards untargeted or weakly targeted subsidies, which consumed almost 70 percent of total SP spending, followed by social assistance spending, at 18 percent. The lowest spending areas were labor-market programs, social care services, and social insurance. Most significantly, the spending on social-care services constituted less than 1 percent of overall SP spending in FY16. Figure 5.9: Social protection sub-categories of spending in FY16 (percent) 0.6% 3.6% 1.8% 18.9% General Subsidies Social Assistance Labor Market Programs Social Care services 75.1% Social Insurance Source: Data from the Government of Sindh. 5.6.3 Budget Utilization Pattern in SP over 5 years from FY13 to FY16 Besides its low budgetary allocation, the SPL sector has been further weakened by poor utilization over the past few years due to weak institutional and administrative capacities. The overall budget utilization pattern against yearly allocations for social protection is shown in Figure 5.10. The lowest spending was 29.7 percent in FY13, which gradually improved in subsequent years to 51.8 percent in FY14 and 81.7 percent in FY16. In FY15, overspending of 165.4 percent was witnessed against the allocated budget, due to an unprecedented increase in transport subsidies. In addition, other than the year of overspending due to transport subsidies, there is a consistent pattern underspending in SP sub- categories. Figure 5.11 shows the disaggregated budget allocations and spending for the same period. Figure 5.10: Social protection budget allocations and Figure 5.11: Budget allocations and spending in SP sub- spending categories 112 Sindh: Public Expenditure Review 2017 Labor Market Programs Social Assistance 30,000 27,089 24,612 Social Care services General Subsidies 25,000 25,000 22,123 Social Insurance 20,000 20,000 14,879 15,000 15,000 9,669 10,000 10,000 7,563 5,000 2,876 3,917 5,000 - Spending Spending Spending Spending Budget Budget Budget Budget - Budget Budget Budget Budget Spending Spending Spending Spending FY13 FY14 FY15 FY16 FY13 FY14 FY15 FY16 Source: Data from the Government of Sindh. 5.6.4 Social Protection Expenditure Summary Both in budget allocations and actual spending, general subsidies far outweigh all other items. The budget allocations in the various SP areas and actual spending over the years are shown in Table 5.2. This confirms that there is very low utilization of allocated funds (apart from the overspending in FY15) and those general subsidies far outweigh all other items. Table 5.2: Analysis of budgeted and actual spending (PRs million) FY13 FY14 FY15 FY16 SP areas Budget Spending Budget Spending Budget Spending Budget Spending General subsidies 3,000 1,090 3,300 1,293 7,237 21,805 18,888 15,581 Labor-market 104 150 239 150 605 174 368 programs Social assistance 3,981 903 2,330 837 3,857 718 6,351 3,918 Social care services 52 3 147 0 738 188 192 124 Social insurance 2,636 776 1,636 1,547 2,896 1,296 1,484 2,133 Grand total 9,669 2,876 7,563 3,917 14,879 24,612 27,089 22,123 Source: Data collected from GoS departments. 5.7 Institutional Arrangements for Social Protection in Sindh Currently, Sindh SPL initiatives are conceived, planned, and executed in a piecemeal fashion across several line departments. There is no central entity to plan, consolidate, coordinate, and monitor these SPL interventions. As a result, the SPL sector is replete with duplication, inadequate coverage, and programmatic inefficiencies. There is little or no meaningful consideration of pre-identified SP risks and vulnerabilities, nor are the SPL programs sustainable at current levels of funding. Sindh’s SPL initiatives are scattered across a host of different departments and autonomous governing bodies. The focus departments include the Community Development Department, the Social Welfare Department, the Department of Zakat and Ushr, the Department of Food and Agriculture, the Health Department, the Sindh Education Foundation, the Education Department’s Endowment Fund, the Benazir Bhutto Shaheed Youth Development Program (BBYSDP), and the Provincial Disaster Management Authority. In addition, SP interventions are also being carried out through different institutional arrangements, such as the Sindh Province Pension Fund, the Sindh Social Relief Fund, the 113 Sindh: Public Expenditure Review 2017 Viability Gap Fund, the People’s Housing Cell Fund, the Sindh Government Employees Group Insurance Fund, the Sindh Agriculture Supplies Organization Pensioner’s Fund, the Sindh Flood Relief Fund, Endowment Fund for Peoples’ Primary Healthcare Initiative, and the Investment Fund for Sindh Civil Servant Housing Foundation. 5.8 Inventory and Analysis of Sindh SPL Programs Besides untargeted general subsidies that account for the lion’s share of Sindh’s SPL spending, the Sindh government has also initiated several SPL programs to address the needs of vulnerable and at- risk segments of the population. These programs are generally aimed at poverty alleviation, women’s development and empowerment, improvements in socioeconomic conditions and welfare, skills development, the protection of children, and emergency responses, particularly in flood-affected areas of Sindh. However, closer scrutiny of these programs reveals that only a fraction of Sindh’s vulnerable and at- risk population, in a few geographical areas, is actually covered under these programs. In addition, the selection of beneficiairies is too few and the services/benefits offered are too too little to have any meaningful impact. In FY15, the most recent year for which data are available, about 1.0 to 1.3 million beneficiaries benefited from SPL intervetions. This is a mere 2.5 percent of Sindh’s population and only 7.5 percent of Sindh’s poor population living below the poverty line (34 percent of the population of Sindh). Sindh’s overall SPL undercoverage notwithstanding, owing to non-objective targeting and poor coordination, it is difficult to estimate the percentage of the bottom two quintiles (bottom 40 percent) of the population that benefits from these programs, or the movement of eligible beneficiaries from one progarm to others based on their specific needs. Table 5.3 summarizes the major SPL interventions, the line departments involved, the funding sources, and the stated goal of the programs. Table 5.3: Major social protection programs in Sindh Program Department Funding Overall goal of program source Social Assistance Guzara allowance Zakat and Ushr Provincial Provision of assistance to deserving families having Department budget destitute, widowed and old-aged persons for daily dietary needs. Education stipend Zakat and Ushr Provincial Provision of educational stipends to deserving students. Department budget Stipend to madrassas Zakat and Ushr Provincial Monthly scholarships to students of Deeni Madaris for Department budget religious education. Marriage assistance Zakat and Ushr Provincial Financial assistance to deserving girls for marriage Department budget purposes. Healthcare Assistance Zakat and Ushr Provincial Funds provided to 7 government hospitals in the Department budget district to facilitate the poor patients. Support limit per patient is PRs 3,000 for outdoor and PRs 5,000 for indoor treatment. Rehabilitation Zakat and Ushr Provincial One-time support of PRs 5,000 to rehabilitate Allowance Department budget businesses. Provision and Agriculture Provincial Hand pumps were provided to the drought-affected Installation of Hand Department budget area of Kohistan for water supply. Pumps in Kohistan area UCBPRP - Community P&DD - SGRRP Provincial Construction of low cost housing, water schemes, Investment Fund (CIF) budget education and health facilities for women. 114 Sindh: Public Expenditure Review 2017 Scholarships to workers Worker’s Non- Provision of merit-based scholarships to worker’s children Welfare Board government children to continue their education at college and funds university level. Financial aid for Worker’s Non- Financial assistance to workers for daughter's marriage daughter marriage Welfare Board government by providing an amount of PRs 100,000 per daughter. funds Death grant to widows Worker’s Non- Financial assistance to the heirs of deceased workers by of workers Welfare Board government providing PRs 500,000 in case of his/her death. funds Sindh Social Relief Fund Finance Provincial Provide relief to the vulnerable and disadvantaged Department budget people of Sindh. Social Insurance UCBPRP - Micro Health P&DD - SGRRP Provincial UCBPRP is a women targeted program providing insurance (MHI) budget income-generating grants, interest free loans, construction of low cost housing, training, water schemes, and education and health facilities. Sindh Govt. Employee All Govt. Provincial GoS has started this fund for the provision of insurance Group insurance Departments budget facility to its employees. Government All Govt. Provincial A service for GoS employees to avail themselves of contribution to health Departments budget medical facilities from private hospitals in case of the insurance need for specialized medical treatment. Labor Market Programs Sindh Skills BBSYDP Provincial Addresses poverty and unemployment through youth- Development Project, budget based human resource development. component-I Benazir Bhutto Shaheed BBSYDP Provincial Addresses poverty and unemployment through youth- Youth Development budget based human resource development in the province. Program (BBSYDP) UCBPRP - Vocational P&DD - SGRRP Provincial Provides vocational training to women. training programs (VTP) budget UCBPRP - Income P&DD - SGRRP Provincial Provides income-generating grants and interest-free generating grant budget loans to women. Social Care services Strengthening of Darul Social Welfare Provincial Rehabilitation of street children. Atfal and Rehabilitation Department budget of Street Children Strengthening of Darul Social Welfare Provincial To give shelter to the women victims of violence. Aman at Mirpurkhas Department budget and Jacobabad Strengthening of 17 Social Welfare Provincial The centers rehabilitate disabled persons such as the existing rehabilitation Department budget blind, the deaf and dumb, crippled and polio affected. center for physically handicapped persons Establishment of Women’s Provincial To initiate social protection programs/ activities for Women Development Development budget women for their sustenance. Complex at Shaheed Department Benazir Abad and Sukkur 3 Day care centers in Women’s Provincial To support working women. Karachi: one each at Development budget Divisional level Department Mediation for women in Women’s Provincial To help women in settling their marital disputes. 115 Sindh: Public Expenditure Review 2017 all Divisional HQs in Development budget Sindh through Department institutionalized Alternate Dispute General Subsidies Wheat subsidy Agriculture Provincial Department budget Subsidy to sugarcane Agriculture Provincial Support small sugarcane growers. growers Department budget Subsidy on fertilizers Agriculture Federal and Subsidy on urea fertilizer to farmers. Department provincial budget Provision of Assistance Agriculture Provincial A facility for farmers to buy tractors at subsidized rates to Farmers on Purchase Department budget to improve the agricultural production. of 11,000 Wheel Type Tractors Provision of Solar Agriculture Provincial Provision of solar water pumps/ tube wells on Pumps Subsidy to Department budget subsidized rates to farmers in Sindh. Farmers (50%) Provision of Power Agriculture Provincial To enhance agriculture production by provision of Sprayer Subsidy to Department budget machinery at subsidized rate. Farmers (50%) UCBPRP - Low cost P&DD - SGRRP Provincial UCBPRP is a woman targeted Program providing housing scheme (LCHS) budget income-generating grants, interest free loans, construction of low cost housing, training, water schemes, education and health facilities. Food Subsidies Food Provincial Provide wheat to the flour mills at low cost to maintain Department budget an affordable price of flour for consumers. Transport Subsidies Transport Provincial Provide low cost transport services to the general Department budget public. 5.9 Efficacy of Sindh Social Protection Initiatives Sindh’s SPL sector is primarily driven by untargeted regressive subsidies (i.e., benefiting the rich more than the poor, due to the former’s greater purchasing power). Unfortunately, the SPL sector is not guided by any umbrella policy or supported by auxiliary social sector policies in a coherent manner. Most SPL programs either do not target the poor at all, or they rely on very loosely defined community- targeting mechanisms. The absence of a central coordinating agency and robust analysis of SP-related risks and vulnerabilities has resulted in a weak understanding of the role of SP interventions. This in turn has led to a plethora of inadequate and inequitable SP programs that have little impact. Below is a brief synopsis of Sindh’s SP sectoral effectiveness. 5.9.1 Program Objectives, Target Populations and Adequacy SPL sector program objectives and target populations are not clearly defined under a coherent risk and vulnerability assessment framework. The lack of any fundamental clarity on SPL goals in promoting resilience, equity, and opportunity, and the limited objective comprehension of the composition and needs of target populations, have resulted in ad hoc SPL programs that make little if any meaningful contribution in reducing risks and vulnerabilities among the poor. With the exception of zakat, which does have a semblance of community targeting, most SPL interventions in Sindh are not specifically targeted towards the poor and involve extremely low coverage. 116 Sindh: Public Expenditure Review 2017 5.9.2 Inequity and the Lack of Consideration of Risks and Vulnerabilities The coverage of most SPL programs is limited to a few districts, Union Councils, and areas or groups without any objective classification. In addition, despite very poor indicators, concerted efforts to promote gender equity, in particular through female-targeted SPL programs, are lacking. Due to the limited number of beneficiaries and the inadequate level of the benefits, even in targeted areas, both horizontal and vertical equity is compromised. 5.9.3 Delivery Mechanisms and Coordination Most SPL programs in Sindh are implemented through government departments. Other than government departments, a few programs are implemented through NGOs/civil society organizations (CSOs), but these usually have even more limited reach and capacity than government programs. The delivery mechanisms vary between programs and geographical areas. For example, zakat is delivered through the provincial Zakat and Usher Department, whereby beneficiaries are selected by local zakat committees. In addition, most social care services, such as women’s crisis centers and childcare centers, etc., are available only in a handful of districts. The lack of institutional capacity to consolidate, coordinate, and monitor SP programs has resulted in additional programmatic and service delivery inefficiencies. 5.9.4 Sustainability The absence of any long-term SPL vision clearly articulated through SPL policy, together with the lack of a resultant strategy, has led to short-termism with regard to issues such as human capital development that in fact require long-term investment. This has resulted in a piecemeal, ad-hoc, and “needs-based� budgetary allocation process and, despite the benefits being insufficient in the first place, allocations have remained largely under-utilized due to low capacity. 5.10 Recommendations In order to achieve improved allocative and distributive efficiencies and to achieve SPL goals of increasing resilience, equity, and opportunity, the Sindh government needs to consolidate, coordinate, and monitor SPL activities across government departments and line agencies. Only in this way can it help to control the duplication of effort and the fragmentation of interventions, thereby saving precious financial resources. Some of specific recommendations and policy actions for achieving this are presented in the matrix below. For strategic steering, program design and oversight, consolidate and coordinate the SP sector under a unified policy framework, duly supported by a responsive institutional and implementation arrangement. Under an appropriate legal framework, establish a central SP entity, such as an SP Authority or Board, entrusted to consolidate, coordinate and monitor SP portfolio in Sindh. Initiate the SP policy formulation process, either through the newly formed SP entity, or in the interim, under the P&DD. The policy formulation process needs to be evidence-based, ensuring adequate stakeholder consultation, analysis of risks and vulnerabilities, clear articulation of SP vision; identification of key policy objectives (such as equity, resilience, social cohesion, gender equity and opportunity) and prioritized intervening areas. Equitable growth (lessening the capability deprivation) should be the focus of Sindh SP policy reform agenda. Existing SP programs/interventions should be reviewed for an increased focus on the marginalized groups (i.e., rural poor, women) and alignment with key policy objectives. New interventions aimed at reducing inequality and promoting equity, such as CCTs for health and education, 117 Sindh: Public Expenditure Review 2017 and labor-market programs designed for the most marginalized, should be introduced with a particular focus on human capital development. All cash-based SP interventions should be brought under a unified targeting and disbursement mechanism. Review and gradually rationalize general subsidies and, in time, possibly provide ONLY targeted subsidies to the most vulnerable, while using the freed-up resources towards core SP programs. Review the distributional impact of untargeted subsidies across economic group quintiles. Develop an action plan to phase out untargeted subsidies based on outcomes of the analytical review. Develop mechanisms for better targeting of subsidies that cannot be phased out, as mentioned in point 1. Explore possibilities for targeting through the PMT score database (NSER). Fully utilize existing national poverty database to steer provincial programs towards increasing coverage to the poorest of the province. There is need to review and analyze the targeting mechanisms of existing SP interventions and introduce unified targeting through the national poverty database (NSER) Moreover, develop data-sharing and coordination mechanisms between the provincial and federal governments and, subsequently, with provincial line departments. 118 Sindh: Public Expenditure Review 2017 Annexes Annexes-I: Vulnerability in Sindh Province Source: PDMA Multi-Hazards Contingency Plan 2013, Government of Sindh. Scoring keys: I. Very High 5; II. High 4; III. Medium 3; IV. Low 2; V. Very Low 1; VI. Dots NA. 119 Sindh: Public Expenditure Review 2017 Annexes-II: Salient Social Protection Initiatives in Sindh A brief review of some of the major SPL programs in Sindh is presented below. The Union Council-based Poverty Reduction Program (UCBPRP) The Union Council-based Poverty Reduction Program (UCBPRP) was initially started in four districts of Sindh for an estimated cost of PRs 3.0 billion. The second phase called the Sindh Union Council and Community Economic Strengthening Support Program (SUCCESS) was expanded in another 216 Union Councils (out of about 1,400 Union Councils in Sindh) of eight additional districts for a total cost of PRs 4.9 billion. Major program components include: social mobilization; income-generating grants; community investment funds; technical and vocational skills training; and micro health insurance covering emergency expenses in case of hospitalization. The Benazir Bhutto Shaheed Youth Development Program (BBSYDP) The Benazir Bhutto Shaheed Youth Development Program (BBSYDP) was started in FY08 to address poverty and unemployment in Sindh, in particular by providing skills training to youth. To date, the BBSYDP program has trained about 292,000 youth in 389 trades covering 89 employment sectors. The program was started at an estimated cost of PRs 1,724.2 million over a period of 3 years from FY11 to FY14. The Sindh Technical and Vocational Training Authority (STEVTA) Besides the BBSYDP, STEVTA is implementing different programs under the Sindh Skills Development Project. These include: the introduction of special training programs for marginalized and vulnerable groups; the introduction of 70 market-driven courses; and improvements in the infrastructure of 50 institutions. Incomplete projects from the E&LD have been transferred to STEVTA with an allocated amount of PRs 5440.0 million, of which spending was PRs 1,229.9 as of June 2015. Zakat Zakat has traditionally been the flagship social assistance program of the Sindh government for the poor and destitute, particularly orphans and widows, the handicapped, and the disabled. In Sindh, zakat alms are disbursed directly through direct cash assistance and also indirectly through educational institutions, public hospitals, charitable institutions, Deeni Madras, and healthcare institutions. About 85,000 beneficiaries receive zakat allowances. Figure 5.12 shows the ways in which zakat is spent, by category. Figure A.1: Category spending of zakat funds (PRs million) 200 179 180 160 140 120 100 80 50.4 60 33.6 40 30 25 40 20 0 Educational Stipends Stipends to Students Health Care Social Welfare / Marriage Assistance Total of Deeni Madaris Rehabilitation to un-married women ZAKAT SPENDING FOR THE YEAR 2014-15 (Million) Source: http://www.sindh.gov.pk/dpt/Zakar_Usher/detailrelease.htm 120 Sindh: Public Expenditure Review 2017 The Shaheed Benazir Housing Scheme The Shaheed Benazir Housing Fund was established in FY08 to facilitate the construction of 6,000 low cost houses across Sindh. Besides completing the initial 6,000 houses in September 2015, an additional PRs 2,966 million has been released for the construction of a further 6,000 houses. Girls’ Stipend The Sindh government provides educational stipends to girls going the secondary schools in rural and unprivileged areas of the province. A total of PRs 1,132 million has been disbursed across 415,868 beneficiaries in FY12, while PRs 947 million was disbursed against an allocated budget of PRs 1,200 million in FY14. In addition, free textbooks are provided to students at primary and secondary schools. The spending on free textbooks was PRs 1.6 billion in FY16 and PRs 1.8 billion has been allocated for free textbook distribution in the current fiscal year, FY17. Centers for the Physically Handicapped, Street Children and Women Victims of Violence The Social Welfare Department (SWD) allocated RPs 314.5 million for strengthening of Darul Atfal and Rehabilitation of Street Children in FY11, out of which PRs 93.5 million had been spent by FY15. In FY13, spending on the strengthening of Darul Aman in Mirpurkhas and Jacobabad stood at PRs 69.8 million, against an allocated sum of PRs 89.0 million. The major initiative of SWD in social care services was the strengthening of 17 existing rehabilitation centers for physically handicapped people, with an allocation of PRs 199.9 million in FY15. However, spending as of June 2015 remained at zero. The Sindh Tractor Scheme The Sindh government provides a subsidy to farmers for the purchase of a total of 11,000 tractors. Under the scheme, a subsidy of PRs 200,000 and PRs 300,000 is provided for small and large tractors, respectively. In the period FY14 to FY16, PRs 1,502 million was spent on this scheme. However, the true nature of the relationship between the tractor scheme and social protection remains unclear. Wheat Subsidies The Sindh government Food Department provided a wheat subsidy of PRs 4.7 billion in FY16 and RPs 1.5 billion in FY15. The wheat subsidy is provided to maintain the price and reduce price fluctuations. The Sindh government provides wheat to flour mills at a subsidized low cost to maintain below-market flour prices for consumers, poor and rich alike. As such, the wheat subsidy is a poorly targeted and therefore regressive government program. 121 Sindh: Public Expenditure Review 2017 Annexure III: Sindh - Provincial GDP Estimates Background Sindh is the second largest province of Pakistan in terms of size of population, but geographically significantly smaller province than Punjab and Baluchistan. According to 1998 census, Sindh comprises of 23 percent of Pakistan’s population, 18 percent of the land area and generates about 33 percent of its economic activity. With almost half of its population living in urban areas, Sindh is the most urbanized province in Pakistan. The provincial capital, Karachi, is not only the largest city in the country; it is country’s industrial, commercial and financial hub, along with being the major seaport. As such, Karachi contributes heavily to the province and country’s GDP. Table A.1: Basic characteristics of provinces in Pakistan Area Population Population Percentage Sq. Km Number Proportion Density Urban Rural Pakistan 796,096 132,352,279 100.00 166.3 32.5 67.5 Punjab (incl. Islamabad) 206,251 74,426,525 56.23 360.9 31.7 68.3 Punjab 205,345 73,621,290 55.63 358.5 31.3 68.7 Islamabad 906 805,235 0.61 888.8 65.7 34.3 Sindh 140,914 30,439,893 23.00 216 48.8 51.2 Khyber Pakhtunkhwa (incl. FATA) 101,741 20,919,976 15.81 205.6 14.7 85.3 Khyber Pakhtunkhwa 74,521 17,743,645 13.41 238.1 16.9 83.1 FATA 27,220 3,176,331 2.40 116.7 2.7 97.3 Balochistan 347,190 6,565,885 4.96 18.9 23.9 76.1 Source: Pakistan Bureau of Statistics. Note: Population is that of the 1998 Population Census. Some key characteristics of Sindh’s economy are as follows: The concentration of Sindh’s economy in one city: As mentioned above, Sindh represents a standard dual economy with distinct areas of modern and traditional economies. The ethnic configuration reinforces this duality as there are very few and weak linkages between the urban and rural economies. This has not only created social and political friction but has also constrained growth in the provincial economy. Our GDP estimates shows that during FY06 to FY15, for most years Sindh’s economy grew at a faster rate than the national economy, but Table A.2: Provincial labor force participation and this did not benefit the poor in the unemployment rates in Pakistan (Percent) province’s slums and rural areas, thus Labor Force creating social and economic inequality. Participation Unemployment Rate Rate Urbanization and labor force: Sindh has Pakistan 45.70 6.24 the most urbanized population in Pakistan. Punjab 48.33 6.38 Half of the Sindh population is urban and Sindh 45.30 5.26 most of it resides in the provincial capital Khyber-Pakhtunkhwa 36.85 8.57 Balochistan 42.34 3.93 Karachi. The businesses are intense in Source: Pakistan Bureau of Statistics, “Pakistan Labor Force Survey, 2010-11�. Karachi, leaving rural of Sindh with a scanty labor force for production and development of the area. Investor Confidence: Prolonged political and sectarian unrest has badly impacted investors’ confidence, and Karachi has seen some of the lowest levels of new investment over the past two decades. Some businesses have reportedly relocated themselves to other areas within the country and even outside the country. However, some of government’s recent actions have led to a noticeable improvement in Karachi’s law and order situation. As a result, foreign direct investment (FDI) has grown positive from 122 Sindh: Public Expenditure Review 2017 negative from 2015 and onwards. With China’s investment under China Pakistan Economic Corridor (CPEC) – US$213.4 million, and Netherland’s investment of 51 percent shares in Engro Foods (produces dairy products, beverages, frozen desserts) or US$467.4 million, has increased the FDI to US$1080 million during July-December FY17 as compared to US$979 million in the same period last year. Though the CPEC investment in the first six months of FY17 is less than the last six months of the year by half, China has become the biggest bilateral donor by providing US$847 million in loan to Pakistan. The size, growth and composition of the economy: Economic analysis of, and as such economic planning, at provincial level is hampered by the lack of official regional accounts. While the preparation of sub-national accounts has been a key element of the work plan of Pakistan Bureau of Statistics, to date there has been no apparent progress in this direction. As such, to analyze the provincial economies and to help the provincial governments to plan better, different agencies and researchers have attempted to estimate Gross Domestic Product (GDP) at provincial level. For the purpose of this note, an attempt is made to update and improve the World Bank methodology to estimate GDP for Sindh. The Methodology for Estimation of the Provincial GDP, 1999/00-2014/15 As in most other developing countries of the world, in Pakistan too, the calculation of Gross Domestic Product (GDP) contains some methodological weaknesses. Two of the most obvious shortcomings are the periodicity (the national accounts are compiled on annual basis only) and aggregate nature of these accounts. While the National Strategy for Development of Statistics calls for preparation of quarterly GDP at the national level and of annual regional accounts at provincial level, it would be quite some time when the work on these initiatives would start. Until such time, these shortcomings would continue to hamper monitoring of national aggregates at national level and regional planning in the country. Until 1971 Pakistan had a centralized system of administration—the “one unit� system—under which no distinction was made between the various geographical and cultural entities within the region called West Pakistan, which later came to be known post-1971 as Pakistan. Hence, no need was felt to estimate Regional Accounts. In 1971, the “one unit� was broken into four separate administrative units or provinces of Punjab, Sindh, Khyber Pakhtunkhwa (NWFP at that time) and Balochistan. However, even under this new system relatively little autonomy was given to the provinces, as major decision- making authority de-facto remained at the Federal level. Until the early eighties, Pakistan’s GDP was estimated with the base year of FY60. In the early 1980s, the base year was changed to FY81, and in early 2000’s the base was again changed to FY00, which had continued until 2013 when the base year changed to FY06. Economic analysis of the Sindh, as that of any other province, is seriously compromised by the non- availability of regional accounts. While the Pakistan Bureau of Statistics (PBS) computes, on annual basis, the national accounts of Pakistan, only cursory efforts have been made to estimate the regional accounts. Most of the attempts made at official levels were aborted because lack of interest at the relevant segments of the government and weak technical capacity of the provincial Bureaus of Statistics to undertake this exercise. On the other hand, exercises undertaken by autonomous researchers failed to get the required recognition from the government. The construction of provincial GDP series is a large and complex undertaking; as disaggregated data are needed for a very large number of province-specific variables. Moreover, the exercise is also extremely time intensive, as a large volume of data needs to be analyzed to derive consistent and robust estimates. In the past, the World Bank has endeavored to estimates provincial GDPs by using some broad “allocators� to derive the shares of each province in sectoral value added and the national income. In doing this, provincial GDP was estimated using a combination of the three traditional approaches—production, expenditure, or income. More specifically, wherever detailed provincial data 123 Sindh: Public Expenditure Review 2017 were available, for example, agriculture, mining and quarrying, manufacturing, and wholesale and retail trade, sectoral value-added were estimated using the production approach. The expenditure approach was used to compute value-added of the construction, ownership of dwellings, electricity and gas distribution, and public administration and defense sub-sectors, whereas a variant of the income approach was applied to estimate value-added in the transport, storage, and communication, banking and insurance, and services sub-sectors. In almost all of these instances, the direct and indirect methods were combined to compute sectoral value-added by apportioning the national income to the province using appropriate allocators. While these allocators were applied only where disaggregated data at the provincial level were not available, clearly the present exercise represents a “second-best� approach to estimate sectoral value-added at the provincial level. While these estimates give a fairly good idea of the trends and composition of economic activity at provincial level, lack of government ownership undermines their importance as it makes it very difficult to build a case for their use as an instrument for sub-national policy and planning. More importantly, these estimates were calculated to meet the need of a specific task and were therefore “time specific� no attempt was made to regularly maintain and update these estimates. The present estimation of regional accounts is also undertaken to make use in the Sindh Public Expenditure Review 2017. Estimation Methodology: In order to estimate provincial GDPs, the following data sources have been used. - Agricultural Statistics of Pakistan, of various years; - Pakistan Statistical Yearbook 2015; - Economic Survey of Pakistan various editions; - Labor Force Survey 2015; - Household Integrated Economic Surveys (HIESs) various issues - Pakistan Energy Year Book, 2015; - National Institute of Population Studies; Estimates/forecast of annual provincial population; and - Livestock Census of 1996 and 2006. The Agricultural Statistics of Pakistan was used as the main source for actual production of the major and minor crops. The crops have been selected in compliance with the national rebasing exercise of 1999/00 done by the then Pakistan Bureau of Statistics (Pakistan Bureau of Statistics as of now). The Pakistan Statistical Year Book of 2011 along with its previous years’ publications, and the Household Integrated Economic Surveys (HIESs) for FY06, FY08 and FY11 were used as the main databases. While the Statistical Year Books provides most of the actual data in the analysis, the HIES provides sectoral sources of household income. The Labor Force Surveys was useful to extract data on labor force and sectoral employment etc. The Energy Year Book provides information on the Energy sector production and the prices/cost of generation. Information on National GDP, its deflators, Consumer Prices, etc. is pulled together from the Economic Survey of Pakistan. The National Institute of Populations Studies provided the Bank few years back, the estimates of provincial population and it forecast for the future years, based on the 1998 population census. On requesting the newer set of data on Population estimates, it was stated that since there is no new census after 1998, the same estimates are being used for various analyses. The sector wise methodology is discussed below: Agriculture: Crop sector: Value-added in agriculture in all provinces has been estimated by applying provincial allocators to the national value-added estimate for this sub-sector. Overall, the computation of value- added in crop agriculture is based on 12 major crops and 8 groups of minor crops. The groups are 124 Sindh: Public Expenditure Review 2017 classified as pulses, vegetable, fruits, condiments, oilseed, flower and foliage, fodder, and other miscellaneous crops. The crop output data for the provinces and for Pakistan was obtained from published sources. Due to the non-availability of input costs of agriculture, production allocators were used to estimate value-added for the crop sub-sector. Every province share of total crop production in national output was assumed to remain the same as its share of total national value-added. These output shares (allocators) were used to estimate value-added of wheat, rice, cottonseed, sugarcane, maize, gram, barley, jawar, bajra, rapeseed and mustard, sesame seed and tobacco for major crops for the years 1999/00 to 2010/11. The simple addition of the value-added of each crop represents the total value-added of major crops. Similarly, value-added of minor crops was estimated using lentils (mash, moong, masoor), potatoes and onions, oilseeds, fruits etc. as the main minor crops. Livestock: The livestock sub-sector includes the value-added of actual livestock population, including the newborn. The census of livestock population was available for 1996 and 2006 while the other years were estimated by the provincial share of livestock value of the national value of livestock in the census year under consideration. Actual data on slaughtering, poultry and milk, was also taken into account while calculating Livestock value of each province. The weight of each product in total value-added was derived from the national accounts data. Fishery: The provincial value added in the fishery sector is derived from the national value-added in the sector by using the provincial share in total fish production as weight. Forestry: The major component of forestry is timber and firewood production. The value added of this sector was assumed to be proportionate to the timber and firewood productions shares of Pakistan production of timber and firewood. The constant share of FY00 is used to draw the final estimates. In other words, the provincial value-added in the forestry sector was derived from the national (sectoral) value-added by using the production share of Timber and firewood to the national shares multiply with the 1999/00 prices of these productions. Manufacturing: The manufacturing sub-sector is sub-divided into large-scale and small-scale manufacturing. Large-scale manufacturing includes units that employ 20 or more workers on any working day during the year and use power, while the small-scale manufacturing includes those units that employ 10-19 workers in any day of the year. Value-added in the provincial large scale manufacturing sub-sector is estimated on the pattern of national GDP calculations by PBS based on the Census of Manufacturing Industries data and the Quantum Index of Manufacturing (QIM) estimates. The provincial Bureau of Statistics conducts monthly surveys of industrial production and employment in their respective provinces and estimate QIM based on these data. The survey reports cover about 90 percent of the total industrial units. The methodology used to estimates the monthly index was to measure the percentage increase in the current month’s production of any item from the average monthly production of FY00. The weighted percentage increase was calculated with the assigned weights based on FY00 CMI data for each item. The benchmark estimates (FY00) of large scale manufacturing value-added are derived on the basis of the provincial shares of value-added from the CMI for the census years. The CMI data have their deficiencies of time lag, considerable non-response, under-reporting of output value and over-reporting of input cost (supposedly due to fear of taxation), and under-coverage of non-registered companies. However, despite these shortcomings the CMI data were used assuming that data problems in the all provinces were no different to those at the national level. The benchmark data points obtained from the CMI were then interpolated for the years where the CMI was not conducted using the annual percentage increase in the provincial QIMs. 125 Sindh: Public Expenditure Review 2017 The small scale value-added, which includes value added on Slaughtering, has been estimated based on the Census of Small Household Manufacturing Industries (SHMI). The growth for the next year was given in the survey as 5.31 percent at the national level. The subsequent study by the Quaidian Economic Consultants Quaid-i-Azam University, Islamabad in FY00 estimated the growth rates of 6.86 and 7.51 percent for the years FY99 and FY00, respectively. Due to unavailability of such surveys at provincial levels, the current analysis of SSM is based on estimating employed labor force shares of each province to the national employed labor force. The slaughtering is estimated using provincial shares of monthly expenditures on milk, all kind of meat, and eggs. The data on percentage distribution of consumption of these products has been obtained from HIES of various year. Construction: Ideally, the construction sub-sector’s value-added should be calculated through the expenditure approach, as is done by PBS at the national level. However, since the provincial expenditure data on construction was not available, the provincial share of construction workers was used instead to derive estimates for FY00 to FY15. The construction worker’s data was missing for some years, which were then interpolated to get the full series. The data on percentage distribution of employment in construction industry in Pakistan and the provinces were obtained from PBS’s Labor Force Surveys of various years. Electricity and Gas: In the Electricity and Gas sub-sector, value-added was estimated on the basis of electricity generation, electricity consumption, and natural gas consumption. The data were taken from various Energy Year Books (FY00 to FY15) published annually by the Hydrocarbon Development Institute of Pakistan. Data were obtained for electricity generation for hydel and thermal electricity; these figures were multiplied with a single year (i.e., constant) price to get constant rupee value of electricity generation. The value of electricity consumption was then taken a proxy for value of electricity transmission and distribution. Similarly, the value of gas consumption was taken as a proxy for the value of gas distribution. These three constant (i.e., at a single year price) values of the three components were then added for both Pakistan and the provinces to get constant value generated in the sub-sector at the national and provincial level. Finally, the provincial value-added in electricity and gas was derived by applying to the national value-added estimate the derived ratio of the provincial value to national value. To estimate the value added for all other sectors (Transport and Storage, Finance and Insurance, and, Community, Social and Personal Services), the income approach has been used. The average household income from each sector and the number of surveyed household were used to get an estimate of total sample income derived from each sector in each province. The HIES uses a stratified random sampling methodology. However, in order to get more robust estimates, the survey over- samples the two smaller provinces (SINDH and Balochistan) and under-samples Punjab and Sindh. As such, the total household income derived from HIES needs to be adjusted by “raising� factors to make sample estimates in line with national population shares. To make these estimates more meaningful, two additional factors have to be kept in mind. First, the HIES distinguishes between “earned� and “unearned� income. While the former includes income that a household receive as an employer, employee of through self-employment, whereas the latter includes income from housing (i.e., rent), and as transfer payments. Rents, both actual and imputed, are incomes from “ownership of dwellings� and is a part of Pakistan’s national accounts classification. Transfers payments could be discarded for calculation of provincial GDP because if these payments originate from outside the province (i.e. from other provinces or abroad) they can be considered as “factor income from abroad�, which is part of GNP and not GDP. If these payments originate from within the province, it would imply “double counting� of income, as the source of these payments is already accounted in earned or rental income. 126 Sindh: Public Expenditure Review 2017 Second, as some of the earned income could be originating from outside the province (i.e. from other provinces), there is a need to make necessary adjustments. It is assumed that earned income originating outside the province is likely to be small on the net basis and therefore could be ignored for the purpose of calculating provincial GDP. The adjusted total household income (derived from each sector) for each province, and for the country, was used to derive provincial shares in each income-generating sector. These shares are then applied to the national value-added in each sector to get the valued-added estimates for each province, which total up to give an estimate of provincial GDP. On the basis of this estimation, it is clear that Sindh’s economy has been very volatile over the last decade. There have been wide fluctuations in GDP growth (Figure A.2), which although is a reflection of happenings in the national economy, but also shows the reaction of the provincial economy some key local developments. Figure A.2: Trends in GDP growth, Pakistan and Sindh National Sindh 10.0% 8.0% 8.3% 6.0% Annual Change 5.3% 5.5% 5.0% 4.0% 3.6% 3.8% 3.9% 4.0% 3.7% 2.6% 2.0% 0.0% 0.4% FY/05 FY/06 FY/07 FY/08 FY/09 FY/10 FY/11 FY/12 FY/13 FY/14 FY/15 -2.0% These trends, point to three important features of the provincial economy. First, there is a strong correlation (correlation coefficient of 0.84) between provincial and national growth supporting the fact that the national economy is dependent upon the growth of Sindh, as most of the commercial and industrial growth comes from the metropolitan city of Karachi. Second, Sindh’s diversified economy also comprises of a well-developed agricultural sector that has an effective irrigation system. Around 14 percent wheat, 30 percent rice, 30 percent sugar cane, 25 percent cotton and 30 percent vegetable crops grown in Pakistan are from Sindh. This provides immense opportunity for setting up export based agri-processing industry in the province. Reasonably strong agriculture growth in Sindh is evident in the last three years (FY13, FY14, and FY15), which depicts higher growth of agriculture at the provincial level than national growth. Growth is attributed to sugarcane and cotton for having adequate water availability, more fertilizer off take, and relief in the prices of input. This is also clear from the spikes in growth in FY13, and FY14, where a good performance by provincial agriculture sector, which subsequently have a positive impact on value added of the wholesale and retail trade and transport and storage sectors, drive the Sindh GDP growth to more than twice the national average. 127 Sindh: Public Expenditure Review 2017 Figure A.3: Trends in agriculture growth, Pakistan and Sindh National Sindh 0.09 0.08 0.07 7.1% 0.06 Annual Change 0.05 0.04 3.4% 3.5% 3.6% 0.03 2.9% 2.7% 2.7% 0.02 1.8% 2.0% 0.01 0.2% 0 0 0 -0.01 FY/05 FY/06 FY/07 FY/08 FY/09 FY/10 FY/11 FY/12 FY/13 FY/14 FY/15 Steps to improve the business environment in Sindh, say by maintaining law and order situation, overcome energy crises, and reducing bureaucratic hurdles in exports, will not only encourage conducive local market, it will also improve access to the international markets. It will have a major impact on GDP growth of provincial economy by earning from local production as well as from exports markets. Having a 37 percent of provincial share of manufacturing in national growth, Sindh has a strong manufacturing correlation of 0.94 with the manufacturing growth of national GDP and the provincial value added growth is more or less the same as national value added of this sector for the last decade. Figure A.4: Trends in industrial growth, Pakistan and Sindh Pak Sindh 10.0% 9.0% 8.6% 8.0% 6.0% 6.1% 4.6% 4.5% Annual Change 4.0% 3.2% 2.0% 2.5% 2.1% 1.4% 0.0% FY/06 FY/07 FY/08 FY/09 FY/10 FY/11 FY/12 FY/13 FY/14 FY/15 -2.0% -4.0% -4.2% -6.0% Although the national and Sindh’s employed labor has marginally declined in FY14 (-4 percent, in Sindh, according to Labor Force Survey), the percentage distribution of construction workers has increased from 5.3 percent in FY13 to 7.3 percent in the following year in Sindh. This led to an increase in Sindh’s construction value added by 40 percent in FY14. (According to HIES and Labor force Survey, Balochistan’s construction value added also increased by 45% in the same period due to a significant increase in the percentage distribution of construction workers in Balochistan). In FY11, a favorable NFC Award provided Sindh (and other provinces) with a substantial increase in federal government’s fiscal transfers. This was accompanied by a 50 percent increase in salaries of government employees (much larger increase than inflation). This led to a 12 percent growth in real value-added of the community and social services sector. Similarly, the increase is fiscal resources led to 128 Sindh: Public Expenditure Review 2017 a 30 percent increase in provincial development expenditure, which prompted a 36 percent increase in (real) value added in construction sector. A 30 percent increase in value-added in mineral sector was primarily due to large increase in gas production from three gas fields (Nashpa showed a growth of 1,016 percent in production, Manzalai 67 percent and Makori 124 percent; production of Mela gas fields declined by 23 percent but the production was too small to have a significant impact on overall production). This extra production of gas, coupled with coming of line of small hydel power plants (set up the Sindh government), caused an 89 percent increase in value added of electricity and gas distribution sector. Finally, unlike the overall national economy, which shows a downward trend in economic growth; the underlying growth in Sindh has an upward trend until FY14. The underlying growth over the last twelve years has increased from 3.7 percent in FY02 to 4.8 percent in FY14. However, it came down to 3.2 percent in FY15. The concern however is that this escalation is not fast enough to absorb the increase in population in the province. The per capita GDP of the province (at constant prices) has been more than the national average and the gap has somewhat been increasing overtime (Figure A.5). Figure A.5: Trends in per capita GDP, Pakistan and Sindh Pak Sindh 75,000 70,000 Rupees per person per annum 65,000 60,000 55,000 50,000 45,000 40,000 35,000 FY/00 FY/01 FY/02 FY/03 FY/04 FY/05 FY/06 FY/07 FY/08 FY/09 FY/10 FY/11 FY/12 FY/13 FY/14 FY/15 Moreover, the GDP growth of 3.2 percent in FY15 is significantly less than a sustained growth of 7- 7.5 percent required by the province to absorb all the persons entering the labor force in future. Given the slow acceleration in trend growth, it will take the province 30 more years to achieve the growth rate, which is consistent with full employment levels. 129 Sindh: Public Expenditure Review 2017 Figure A.6A: The Composition of Sindh's GDP in FY01 Rental 2% Minining Agri 0% Service 18% 22% Finance Manufacturing 3% 14% Transport 18% Energy 6% Commerce 14% Construction 3% The composition of the Sindh’s economy is changing, although very gradually (Figures A.6A and A.6B). Figure A.6A reflects the composition of the economy in FY01 and Figure 5B presents the composition 15 years later (i.e., in FY15). It is quite apparent that while the share of agriculture has been stagnant, the manufacturing sector has been increasing due to the government’s various measures to increase the industrial base of the country. Figure A.6B: The composition of Sindh's economy in FY15 Rental 5% Agri Service Minining 18% Finance 15% 6% 2% Transport 14% Manufacturing Energy Commerce 18% 1% 19% Construction 2% 130 Sindh: Public Expenditure Review 2017 Table A.1: Gross provincial value-added at constant factor cost, at 1999/00 prices (PRs million) Whole Finance Community Ownersh Agricultur Mining & Electricity Sale & Transpor , Real , Social, ip of e, Fishing Quarryin Manufacturi , Gas, Constructio related t& Estate Personal Dwelling & Forestry g -ng Water n trade Storage etc. Srv. s TOTAL PAKISTAN 2005/06 1,775,346 254,345 1,065,323 110,109 186,380 1,523,067 959,499 282,912 1,054,046 504,743 7,715,770 2006/07 1,836,125 273,032 1,161,551 96,066 210,436 1,612,086 1,025,694 308,673 1,094,377 524,929 8,142,969 2007/08 1,869,310 281,635 1,232,430 131,767 242,768 1,703,741 1,082,452 328,071 1,131,024 545,950 8,549,148 2008/09 1,934,691 274,710 1,180,964 115,812 218,777 1,652,874 1,137,003 296,427 1,200,815 567,947 8,580,020 2009/10 1,939,132 282,269 1,197,164 135,098 237,034 1,682,466 1,170,627 286,775 1,280,143 590,725 8,801,433 2010/11 1,977,178 269,798 1,227,091 221,379 216,754 1,718,014 1,198,896 274,674 1,402,092 614,460 9,120,336 2011/12 2,048,794 283,727 1,252,670 224,490 223,429 1,746,511 1,254,126 279,171 1,518,334 639,003 9,470,255 2012/13 2,103,600 294,727 1,310,522 165,275 225,840 1,808,124 1,304,697 304,252 1,636,493 664,542 9,818,072 2013/14 2,160,223 299,588 1,369,003 174,482 242,203 1,880,004 1,355,575 315,032 1,714,669 691,095 10,201,875 2014/15 2,222,337 311,095 1,412,453 177,866 259,271 1,943,499 1,421,242 334,513 1,808,597 718,662 10,609,535 PUNJAB 2005/06 976,107 32,021 487,125 50,114 102,258 784,533 521,868 124,218 546,108 291,604 3,915,958 2006/07 1,007,737 32,129 532,281 45,003 122,068 828,188 560,947 143,447 580,851 305,457 4,158,108 2007/08 1,001,605 32,998 563,254 61,619 150,567 859,555 594,657 160,771 613,890 319,515 4,358,431 2008/09 1,032,162 31,235 540,177 53,644 133,084 834,135 625,718 152,617 653,855 326,358 4,382,985 2009/10 1,028,212 30,973 549,586 62,056 149,851 846,410 645,314 154,558 698,907 333,097 4,498,963 2010/11 1,062,537 29,749 563,907 95,423 125,577 872,041 661,982 154,416 767,147 339,802 4,672,580 2011/12 1,137,425 31,418 575,363 99,976 131,209 906,084 699,385 163,144 782,481 370,122 4,896,609 2012/13 1,172,399 31,573 603,600 76,812 135,390 940,572 714,086 184,215 864,712 398,990 5,122,351 2013/14 1,203,858 30,680 634,690 83,736 130,254 979,386 730,659 196,999 928,610 426,895 5,345,768 2014/15 1,230,001 29,877 655,829 83,527 140,904 1,008,341 752,136 215,398 1,002,548 455,073 5,573,635 SINDH 2005/06 389,543 164,003 407,281 30,648 41,638 427,229 254,672 51,917 305,502 162,188 2,234,619 2006/07 397,804 175,004 444,212 24,260 41,424 452,818 260,079 55,546 305,560 163,403 2,320,110 2007/08 416,656 180,013 471,847 34,390 46,368 488,042 262,842 57,987 303,287 164,537 2,425,969 2008/09 427,110 179,139 449,875 29,979 38,316 465,246 276,465 51,556 323,286 174,692 2,415,663 2009/10 451,140 179,204 455,171 35,598 38,071 486,191 285,015 49,177 345,834 185,336 2,510,737 2010/11 454,256 162,637 466,070 59,624 38,153 493,446 292,265 46,535 379,900 196,527 2,589,414 2011/12 442,906 168,057 472,925 60,894 40,136 484,485 301,781 46,827 422,230 189,815 2,630,055 2012/13 479,671 174,781 493,264 37,396 39,675 515,268 336,130 50,631 434,111 176,160 2,737,087 2013/14 514,997 173,519 511,513 42,588 57,942 546,817 367,808 52,117 438,019 165,238 2,870,559 2014/15 531,292 177,251 525,475 45,022 59,046 565,047 404,906 55,119 444,321 154,424 2,961,903 K.P. 2005/06 129,593 9,089 94,989 19,847 35,281 120,413 132,724 77,377 134,294 34,724 788,329 2006/07 137,428 15,920 102,234 16,880 37,689 128,886 136,452 79,410 136,373 38,321 829,593 2007/08 126,778 18,669 109,153 22,005 36,404 129,594 136,511 79,087 135,966 42,095 836,263 2008/09 132,835 17,223 105,529 18,731 38,957 126,454 144,191 66,709 147,721 46,282 844,631 2009/10 127,689 27,189 105,783 21,524 41,782 125,246 149,276 60,025 161,063 50,847 870,425 2010/11 134,582 35,222 108,226 40,045 45,419 130,185 153,718 53,281 180,332 55,834 936,844 2011/12 138,750 39,956 110,680 38,766 43,703 131,951 150,886 50,013 192,059 55,994 952,757 2012/13 143,434 45,216 116,346 31,237 41,679 137,580 167,174 50,171 222,966 63,690 1,019,494 2013/14 153,894 51,694 121,465 30,618 40,909 146,683 181,680 47,668 238,860 70,956 1,084,425 2014/15 164,015 55,226 126,667 30,919 46,892 155,426 198,640 46,306 257,251 78,759 1,160,100 BALOCHISTAN 2005/06 280,104 49,233 75,929 9,500 7,202 190,892 50,235 29,399 68,142 16,227 776,864 2006/07 293,155 49,978 82,824 9,923 9,255 202,194 68,216 30,270 71,593 17,748 835,158 2007/08 324,270 49,955 88,175 13,753 9,429 226,550 88,443 30,226 77,881 19,803 928,485 2008/09 342,585 47,113 85,383 13,458 8,421 227,040 90,629 25,545 75,954 20,615 936,741 2009/10 332,090 44,903 86,623 15,921 7,330 224,619 91,022 23,015 74,339 21,445 921,307 2010/11 325,803 42,190 88,888 26,287 7,605 222,342 90,931 20,442 74,713 22,296 921,498 2011/12 329,713 44,296 93,702 24,854 8,381 223,991 102,073 19,188 121,564 23,072 990,833 2012/13 308,096 43,157 97,312 19,830 9,095 214,705 87,307 19,234 114,704 25,702 939,141 2013/14 287,475 43,694 101,335 17,540 13,098 207,118 75,428 18,248 109,180 28,006 901,123 2014/15 297,030 48,741 104,482 18,398 12,429 214,686 65,560 17,689 104,477 30,406 913,897 Source: World Bank staff calculations. 131 Sindh: Public Expenditure Review 2017 Table A.2: Gross provincial value-added at current prices (PRs million) Transp Mining Electric ort Finance, Agriculture and ity, Whole Sale and Real Community, Ownershi , Fishing & Quarryin Manufacturin Gas, Constructio and related Storag Estate Social, p of Forestry g g Water n trade e etc. Personal Srv. Dwellings TOTAL PAKISTAN 2005/06 1,775,346 254,345 1,065,323 110,109 186,380 1,523,067 959,499 282,919 1,054,046 504,743 7,715,777 2006/07 2,014,317 297,183 1,226,012 101,231 221,791 1,720,385 1,080,163 342,757 1,171,971 559,956 8,735,766 2007/08 2,393,527 324,258 1,572,886 146,125 269,106 2,201,667 1,065,682 401,060 1,343,970 636,974 10,355,255 2008/09 2,998,651 413,260 1,679,088 146,985 293,910 2,479,782 1,693,864 481,308 1,648,265 707,268 12,542,380 2009/10 3,461,311 475,371 1,943,862 209,938 302,555 2,824,169 1,834,496 474,733 1,933,034 789,229 14,248,697 2010/11 4,592,720 494,739 2,527,651 406,156 318,451 3,568,178 1,923,433 536,345 2,393,510 886,370 17,647,553 2011/12 4,753,075 642,205 2,809,684 439,637 378,140 4,006,835 1,905,704 570,503 2,871,580 984,148 19,361,511 2012/13 5,334,975 696,976 3,030,650 368,040 423,367 4,369,465 2,311,796 522,327 3,346,334 1,092,749 21,496,679 2013/14 5,984,046 743,088 3,333,007 429,214 490,260 4,809,962 2,474,826 600,731 3,800,129 1,229,113 23,894,376 2014/15 6,575,204 709,850 3,188,746 456,478 548,911 4,873,913 3,105,727 626,418 4,191,238 1,371,421 25,647,906 PUNJAB 2005/06 976,003 32,021 487,125 50,114 102,258 784,533 521,868 124,221 546,108 291,604 3,915,857 2006/07 1,103,614 34,971 561,276 47,423 128,655 883,826 590,736 159,287 622,035 325,839 4,457,661 2007/08 1,308,335 37,992 719,586 68,333 166,902 1,110,763 585,444 196,540 729,471 372,787 5,296,152 2008/09 1,653,846 46,989 767,614 68,083 178,788 1,251,441 932,171 247,804 897,496 406,415 6,450,648 2009/10 1,883,401 52,162 888,239 96,432 191,272 1,420,775 1,011,275 255,859 1,055,360 445,029 7,299,805 2010/11 2,578,308 54,551 1,152,793 175,069 184,496 1,811,159 1,062,043 301,521 1,309,596 490,171 9,119,705 2011/12 2,650,440 71,114 1,277,865 195,791 222,063 2,078,733 1,062,749 333,395 1,479,883 570,036 10,005,425 2012/13 2,995,302 74,666 1,380,718 171,048 253,807 2,272,961 1,265,291 316,252 1,768,181 656,086 10,977,168 2013/14 3,379,075 76,099 1,527,716 205,986 263,656 2,505,745 1,333,938 375,655 2,058,026 759,233 12,043,288 2014/15 3,664,398 68,174 1,467,158 214,366 298,313 2,528,721 1,643,585 403,361 2,323,304 868,414 13,212,951 SINDH 2005/06 389,578 164,003 407,281 30,648 41,638 427,229 254,672 51,919 305,502 162,188 2,234,656 2006/07 435,517 190,484 468,602 25,564 43,659 483,238 273,890 61,680 327,224 174,307 2,484,166 2007/08 527,255 207,257 605,860 38,137 51,398 630,675 258,770 70,888 360,389 191,969 2,942,598 2008/09 650,540 269,488 643,395 38,048 51,474 698,001 411,868 83,712 443,749 217,545 3,507,820 2009/10 799,151 301,799 744,595 55,318 48,595 816,115 446,648 81,408 522,214 247,615 4,063,457 2010/11 1,038,694 298,234 971,044 109,391 56,054 1,024,848 468,891 90,868 648,527 283,494 4,990,045 2011/12 1,013,946 380,390 1,073,019 119,253 67,927 1,111,502 458,570 95,693 798,551 292,340 5,474,685 2012/13 1,202,747 413,325 1,149,722 83,274 74,377 1,245,182 595,590 86,922 887,678 289,671 6,006,395 2013/14 1,408,928 430,391 1,255,781 104,764 117,284 1,399,024 671,495 99,381 970,759 293,876 6,589,746 2014/15 1,555,424 404,448 1,183,868 115,544 125,009 1,417,026 884,808 103,218 1,029,666 294,688 7,229,752 K.P. 2005/06 129,606 9,089 94,989 19,847 35,281 120,413 132,724 77,379 134,294 34,724 788,344 2006/07 151,596 17,328 108,345 17,788 39,722 137,544 143,698 88,178 146,042 40,878 891,120 2007/08 162,235 21,494 136,570 24,403 40,353 167,468 134,396 96,682 161,566 49,114 994,280 2008/09 203,862 25,909 147,781 23,773 52,335 189,715 214,810 108,315 202,764 57,635 1,226,902 2009/10 225,203 45,789 170,088 33,448 53,332 210,237 233,931 99,366 243,208 67,934 1,382,535 2010/11 302,860 64,588 219,762 73,468 66,729 270,384 246,616 104,039 307,845 80,542 1,736,833 2011/12 324,925 90,440 245,898 75,919 73,965 302,721 229,279 102,204 363,236 86,238 1,905,517 2012/13 367,346 106,928 269,643 69,560 78,134 332,472 296,216 86,132 455,926 104,730 2,090,583 2013/14 426,869 128,220 297,157 75,318 82,806 375,285 331,687 90,897 529,373 126,194 2,293,624 2014/15 487,431 126,013 293,311 79,352 99,276 389,777 434,071 86,714 596,154 150,295 2,516,384 BALOCHISTAN 2005/06 280,159 49,233 75,929 9,500 7,202 190,892 50,235 29,400 68,142 16,227 776,919 2006/07 323,590 54,399 87,789 10,457 9,755 215,777 71,839 33,612 76,669 18,932 902,820 2007/08 395,703 57,515 110,871 15,252 10,452 292,761 87,073 36,950 92,545 23,104 1,122,225 2008/09 490,402 70,874 120,297 17,080 11,312 340,624 135,015 41,477 104,256 25,672 1,357,010 2009/10 553,556 75,621 140,940 24,740 9,356 377,043 142,641 38,100 112,252 28,651 1,502,901 2010/11 672,858 77,366 184,052 48,228 11,173 461,787 145,884 39,917 127,543 32,163 1,800,970 2011/12 763,764 100,262 212,903 48,673 14,185 513,879 155,105 39,211 229,910 35,533 1,975,884 2012/13 769,580 102,058 230,566 44,157 17,050 518,849 154,699 33,021 234,549 42,263 2,167,784 2013/14 769,175 108,378 252,352 43,147 26,513 529,907 137,707 34,798 241,970 49,809 2,378,323 2014/15 867,951 111,215 244,409 47,216 26,313 538,390 143,263 33,125 242,113 58,023 2,609,309 Source: World Bank staff calculations. 132 Sindh: Public Expenditure Review 2017 Table A.3: Annual growth in gross provincial value-added at constant factor cost, at 1999/00 prices (percent) Whole Sale Finance, community, Agriculture, Mining Electricity, and Transport Real Social, Ownership Fishing & and Gas, related and Estate Personal of Forestry Quarrying Manufacturing Water Construction trade Storage etc. Srv. Dwellings TOTAL PAKISTAN 2005/06 7.05 4.62 8.57 -26.55 10.19 -2.36 3.96 42.89 9.94 3.51 5.28 2006/07 3.42 7.35 9.03 -12.75 12.91 5.84 6.90 9.11 3.83 4.00 5.54 2007/08 1.81 3.15 6.10 37.16 15.36 5.69 5.53 6.28 3.35 4.00 4.99 2008/09 3.50 -2.46 -4.18 -12.11 -9.88 -2.99 5.04 -9.65 6.17 4.03 0.36 2009/10 0.23 2.75 1.37 16.65 8.35 1.79 2.96 -3.26 6.61 4.01 2.58 2010/11 1.96 -4.42 2.50 63.87 -8.56 2.11 2.41 -4.22 9.53 4.02 3.62 2011/12 3.62 5.16 2.08 1.41 3.08 1.66 4.61 1.64 8.29 3.99 3.84 2012/13 2.68 3.88 4.62 -26.38 1.08 3.53 4.03 8.98 7.78 4.00 3.67 2013/14 2.69 1.65 4.46 5.57 7.25 3.98 3.90 3.54 4.78 4.00 3.91 2014/15 2.88 3.84 3.17 1.94 7.05 3.38 4.84 6.18 5.48 3.99 4.00 PUNJAB 2005/06 6.98 10.66 9.74 -24.61 0.77 -2.12 9.28 31.62 12.81 4.37 6.12 2006/07 3.24 0.34 9.27 -10.20 19.37 5.56 7.49 15.48 6.36 4.75 6.18 2007/08 -0.61 2.70 5.82 36.92 23.35 3.79 6.01 12.08 5.69 4.60 4.82 2008/09 3.05 -5.34 -4.10 -12.94 -11.61 -2.96 5.22 -5.07 6.51 2.14 0.56 2009/10 -0.38 -0.84 1.74 15.68 12.60 1.47 3.13 1.27 6.89 2.07 2.65 2010/11 3.34 -3.95 2.61 53.77 -16.20 3.03 2.58 -0.09 9.76 2.01 3.86 2011/12 7.05 5.61 2.03 4.77 4.48 3.90 5.65 5.65 2.00 8.92 4.79 2012/13 3.07 0.49 4.91 -23.17 3.19 3.81 2.10 12.92 10.51 7.80 4.61 2013/14 2.68 -2.83 5.15 9.01 -3.79 4.13 2.32 6.94 7.39 6.99 4.36 2014/15 2.17 -2.62 3.33 -0.25 8.18 2.96 2.94 9.34 7.96 6.60 4.26 SINDH 2005/06 12.50 3.44 4.32 -29.02 37.07 -1.87 19.53 44.50 17.61 3.70 7.99 2006/07 2.12 6.71 9.07 -20.84 -0.52 5.99 2.12 6.99 0.02 0.75 3.83 2007/08 4.74 2.86 6.22 41.76 11.94 7.78 1.06 4.39 -0.74 0.69 4.56 2008/09 2.51 -0.49 -4.66 -12.83 -17.37 -4.67 5.18 -11.09 6.59 6.17 -0.42 2009/10 5.63 0.04 1.18 18.74 -0.64 4.50 3.09 -4.62 6.97 6.09 3.94 2010/11 0.69 -9.24 2.39 67.49 0.22 1.49 2.54 -5.37 9.85 6.04 3.13 2011/12 -2.50 3.33 1.47 2.13 5.20 -1.82 3.26 0.63 11.14 -3.42 1.57 2012/13 8.30 4.00 4.30 -38.59 -1.15 6.35 11.38 8.12 2.81 -7.19 4.07 2013/14 7.36 -0.72 3.70 13.88 46.04 6.12 9.42 2.93 0.90 -6.20 4.88 2014/15 3.16 2.15 2.73 5.71 1.91 3.33 10.09 5.76 1.44 -6.54 3.18 K.P. 2005/06 10.99 52.71 12.74 -29.24 23.99 1.36 -19.74 57.98 0.55 3.63 3.23 2006/07 6.05 75.17 7.63 -14.95 6.82 7.04 2.81 2.63 1.55 10.36 5.23 2007/08 -7.75 17.27 6.77 30.36 -3.41 0.55 0.04 -0.41 -0.30 9.85 0.80 2008/09 4.78 -7.75 -3.32 -14.88 7.01 -2.42 5.63 -15.65 8.64 9.95 1.00 2009/10 -3.87 57.87 0.24 14.91 7.25 -0.95 3.53 -10.02 9.03 9.86 3.05 2010/11 5.40 29.55 2.31 86.05 8.70 3.94 2.98 -11.24 11.96 9.81 7.63 2011/12 3.10 13.44 2.27 -3.19 -3.78 1.36 -1.84 -6.13 6.50 0.29 1.70 2012/13 3.38 13.16 5.12 -19.42 -4.63 4.27 10.79 0.32 16.09 13.75 7.00 2013/14 7.29 14.33 4.40 -1.98 -1.85 6.62 8.68 -4.99 7.13 11.41 6.37 2014/15 6.58 6.83 4.28 0.98 14.63 5.96 9.33 -2.86 7.70 11.00 6.98 BALOCHISTAN 2005/06 -1.01 -0.89 21.25 -22.25 -18.49 -6.53 -24.98 57.11 -16.20 -11.38 -3.59 2006/07 4.66 1.51 9.08 4.46 28.51 5.92 35.80 2.96 5.07 9.37 7.50 2007/08 10.61 -0.05 6.46 38.59 1.88 12.05 29.65 -0.15 8.78 11.58 11.17 2008/09 5.65 -5.69 -3.17 -2.15 -10.70 0.22 2.47 -15.49 -2.47 4.10 0.89 2009/10 -3.06 -4.69 1.45 18.30 -12.95 -1.07 0.43 -9.90 -2.13 4.02 -1.65 2010/11 -1.89 -6.04 2.61 65.11 3.75 -1.01 -0.10 -11.18 0.50 3.97 0.02 2011/12 1.20 4.99 5.42 -5.45 10.21 0.74 12.25 -6.14 62.71 3.48 7.52 2012/13 -6.56 -2.57 3.85 -20.22 8.51 -4.15 -14.47 0.24 -5.64 11.40 -5.22 2013/14 -6.69 1.25 4.13 -11.55 44.02 -3.53 -13.61 -5.13 -4.82 8.97 -4.05 2014/15 3.32 11.55 3.11 4.89 -5.11 3.65 -13.08 -3.06 -4.31 8.57 1.42 Source: World Bank staff calculations. 133 Sindh: Public Expenditure Review 2017 Table A.4: Annual shares of gross provincial value-added to national value-added using estimates at current prices (percent) Whole Sale Finance, Community, Agriculture, Mining Electricity, and Transport Real Social, Ownership Fishing & and Gas, related and Estate Personal of Forestry Quarrying Manufacturing Water Construction trade Storage etc. Srv. Dwellings TOTAL PAKISTAN 2005/06 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 2006/07 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 2007/08 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 2008/09 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 2009/10 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 2010/11 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 2011/12 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 2012/13 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 2013/14 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 2014/15 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 PUNJAB 2005/06 54.98 12.59 45.73 45.51 54.87 51.51 54.39 43.91 51.81 57.77 50.75 2006/07 54.79 11.77 45.78 46.85 58.01 51.37 54.69 46.47 53.08 58.19 51.03 2007/08 54.66 11.72 45.75 46.76 62.02 50.45 54.94 49.01 54.28 58.52 51.14 2008/09 55.15 11.37 45.72 46.32 60.83 50.47 55.03 51.49 54.45 57.46 51.43 2009/10 54.41 10.97 45.69 45.93 63.22 50.31 55.13 53.90 54.60 56.39 51.23 2010/11 56.14 11.03 45.61 43.10 57.94 50.76 55.22 56.22 54.71 55.30 51.68 2011/12 55.76 11.07 45.48 44.53 58.73 51.88 55.77 58.44 51.54 57.92 51.68 2012/13 56.14 10.71 45.56 46.48 59.95 52.02 54.73 60.55 52.84 60.04 51.06 2013/14 56.47 10.24 45.84 47.99 53.78 52.09 53.90 62.53 54.16 61.77 50.40 2014/15 55.73 9.60 46.01 46.96 54.35 51.88 52.92 64.39 55.43 63.32 51.52 SINDH 2005/06 21.94 64.48 38.23 27.83 22.34 28.05 26.54 18.35 28.98 32.13 28.96 2006/07 21.62 64.10 38.22 25.25 19.68 28.09 25.36 18.00 27.92 31.13 28.44 2007/08 22.03 63.92 38.52 26.10 19.10 28.65 24.28 17.68 26.82 30.14 28.42 2008/09 21.69 65.21 38.32 25.89 17.51 28.15 24.32 17.39 26.92 30.76 27.97 2009/10 23.09 63.49 38.30 26.35 16.06 28.90 24.35 17.15 27.02 31.37 28.52 2010/11 22.62 60.28 38.42 26.93 17.60 28.72 24.38 16.94 27.10 31.98 28.28 2011/12 21.33 59.23 38.19 27.13 17.96 27.74 24.06 16.77 27.81 29.70 28.28 2012/13 22.54 59.30 37.94 22.63 17.57 28.50 25.76 16.64 26.53 26.51 27.94 2013/14 23.54 57.92 37.68 24.41 23.92 29.09 27.13 16.54 25.55 23.91 27.58 2014/15 23.66 56.98 37.13 25.31 22.77 29.07 28.49 16.48 24.57 21.49 28.19 K.P. 2005/06 7.30 3.57 8.92 18.02 18.93 7.91 13.83 27.35 12.74 6.88 10.22 2006/07 7.53 5.83 8.84 17.57 17.91 7.99 13.30 25.73 12.46 7.30 10.20 2007/08 6.78 6.63 8.68 16.70 15.00 7.61 12.61 24.11 12.02 7.71 9.60 2008/09 6.80 6.27 8.80 16.17 17.81 7.65 12.68 22.50 12.30 8.15 9.78 2009/10 6.51 9.63 8.75 15.93 17.63 7.44 12.75 20.93 12.58 8.61 9.70 2010/11 6.59 13.05 8.69 18.09 20.95 7.58 12.82 19.40 12.86 9.09 9.84 2011/12 6.84 14.08 8.75 17.27 19.56 7.56 12.03 17.91 12.65 8.76 9.84 2012/13 6.89 15.34 8.90 18.90 18.46 7.61 12.81 16.49 13.62 9.58 9.73 2013/14 7.13 17.26 8.92 17.55 16.89 7.80 13.40 15.13 13.93 10.27 9.60 2014/15 7.41 17.75 9.20 17.38 18.09 8.00 13.98 13.84 14.22 10.96 9.81 BALOCHISTAN 2005/06 15.78 19.36 7.13 8.63 3.86 12.53 5.24 10.39 6.46 3.21 10.07 2006/07 16.06 18.30 7.16 10.33 4.40 12.54 6.65 9.81 6.54 3.38 10.33 2007/08 16.53 17.74 7.05 10.44 3.88 13.30 8.17 9.21 6.89 3.63 10.84 2008/09 16.35 17.15 7.16 11.62 3.85 13.74 7.97 8.62 6.33 3.63 10.82 2009/10 15.99 15.91 7.25 11.78 3.09 13.35 7.78 8.03 5.81 3.63 10.55 2010/11 14.65 15.64 7.28 11.87 3.51 12.94 7.58 7.44 5.33 3.63 10.21 2011/12 16.07 15.61 7.58 11.07 3.75 12.83 8.14 6.87 8.01 3.61 10.21 2012/13 14.43 14.64 7.61 12.00 4.03 11.87 6.69 6.32 7.01 3.87 10.08 2013/14 12.85 14.58 7.57 10.05 5.41 11.02 5.56 5.79 6.37 4.05 9.95 2014/15 13.20 15.67 7.66 10.34 4.79 11.05 4.61 5.29 5.78 4.23 10.17 Source: World Bank staff calculations. 134