Kingdom of Lesotho Public Expenditure Review Improving expenditure efficiency for inclusive development and growth Report Number: 127317 June 2018 GGODR Africa Region Document of the World Bank i Standard Disclaimer: This document 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, e-mail pubrights@worldbank.org. Regional Vice President: Makhtar Diop Country Director: Paul Noumba Um Senior Global Practice Director: Deborah L. Wetzel Practice Manager(s): Hisham Waly Task Team Leader(s): Shiho Nagaki, Aslı Şenkal ii Acknowledgements This report was prepared by a World Bank team led by Shiho Nagaki (Senior Public Sector Specialist) and Aslı Şenkal (Economist). The team was comprised of Tuan Minh Le (Lead Economist), Waewnet Sukkasem (Consultant), Servaas Van der Berg (Professor, Stellenbosch University), Kirk David Schmidt (Governance Specialist), Christopher Bender (Consultant), Gert Johannes Alwyn Van Der Linde (Lead Financial Management Specialist), Harisoa Danielle Rasolonjatovo Andriamihamina (Senior Education Specialist), Omer Ramses Zang Sidjou (Health Economist), Patricio V. Marquez (Lead Health Specialist), Kanako Yamashita-Allen (Senior Health Specialist), Lucilla Maria Bruni (Senior Economist), Masato Tokuda (Consultant), Marius Tolmie (Consultant), and Sashana Whyte (Young Professional). The team would like to extend their gratitude to the Government of Lesotho for the close collaboration and contributions, mainly from the following ministries and institutions: Ministry of Finance, Ministry of Public Service, Ministry of Development Planning, Ministry of Education and Training, Ministry of Health, Lesotho Revenue Authority, and the Bureau of Statistics. The team benefitted from the overall guidance of Paul Noumba Um (Country Director), Janet Entwistle (Country Representative for Lesotho), James Brumby (Governance Director), Guenter Heidenhof (former Practice Manager), Hisham Waly (Practice Manager), Sebastien Dessus (Program Leader), Paolo Belli (Program Leader), Mark Roland Thomas (Director), Mathew A. Verghis (Practice Manager) and Marek Hanusch (Senior Economist). The team is also grateful for invaluable inputs and advice from peer reviewers and International Monetary Fund (IMF) colleagues: Serdar Yilmaz (Lead Public Sector Management Specialist), Fernando Blanco (Lead Economist), Theo David Thomas (Economic Adviser), Furqan Ahmad Saleem (Lead Public Sector Specialist), Fernando Gabriel Im (Senior Economist), Toby Linden (Lead Education Specialist), Christine Dieterich (Mission Chief for Lesotho, IMF), Jiro Honda (Deputy Division Chief, Africa Department, IMF), and Aidar Abdychev (Economist, IMF), Alexander Massara (Economist, IMF), and Christian Alanso (Economist, IMF). iii Table of Contents Acknowledgements ...................................................................................................................................... iii Acronyms and Abbreviations ...................................................................................................................... ix Executive Summary ...................................................................................................................................... 1 Chapter 1: Macroeconomic Overview and Recent Expenditure Trends ....................................................... 9 I. Macroeconomic Overview .................................................................................................................... 9 II. Revenue Performance ......................................................................................................................... 10 I. Government Expenditure ............................................................................................................... 13 II. Expenditure Performance in the Social Sectors ............................................................................. 17 III. Implementation of Expenditure Controls and Revenue Management ........................................... 20 IV. Public Financial Management........................................................................................................ 22 Chapter 2: Taxation..................................................................................................................................... 26 I. Revenue Performance .................................................................................................................... 27 II. Tax Policy Issues ........................................................................................................................... 31 III. Tax Administration ........................................................................................................................ 47 Chapter 3: Public Sector Employment and Wage Bill Management .......................................................... 51 I. Public Service Employment and the Wage Bill ............................................................................. 51 II. Civil Service Pension ..................................................................................................................... 69 Chapter 4: Public Investment Management ................................................................................................ 73 I. Trends in Public Investment .......................................................................................................... 73 II. Brief Overview of PIM Institutional Mapping .............................................................................. 78 III. Assessment of PIM System Performance ...................................................................................... 80 IV. Ranking Priorities in PIM Reforms: The Way Ahead ................................................................... 89 Annex 1: Composition of Tax Revenue Performance from IFMIS and MTFF .......................................... 97 Annex 2: Tax Expenditures: A Primer ........................................................................................................ 98 Annex 3: TADAT - Performance Outcome Areas...................................................................................... 99 Annex 4: Staff Number per Ministry ........................................................................................................ 100 Annex 5: Composition of Expenditure for Staff Compensation (FY2015/16) ......................................... 101 Annex 6: Wage Bill Projection Model ...................................................................................................... 102 Annex 7: Summary of Diagnostic of the PIM Performance ..................................................................... 105 Annex 8: Old Age Pension ....................................................................................................................... 108 Bibliography ............................................................................................................................................. 114 iv Boxes Box 1. SACU Revenue-sharing Formula .................................................................................................... 27 Box 2. Structure of Comparator Civil Services .......................................................................................... 53 Box 3. HRM Perceptions of Civil Servants in Lesotho .............................................................................. 61 Box 4: History of the PAC .......................................................................................................................... 79 Box 5: PIM and SOEs in Lesotho: Transparency and Accountability Issues ............................................. 88 Box 6: Formal Project Appraisal in Ireland ................................................................................................ 92 Figures Figure 1. Contributions to Growth, Production .......................................................................................... 10 Figure 2. Contributions to Growth, Expenditure ........................................................................................ 10 Figure 3. Revenue Sources (percentage of GDP) ....................................................................................... 11 Figure 4. Overall Expenditures, Revenues, and the Budget Deficit (percentage of GDP) ......................... 11 Figure 5. Net Fiscal Impact of All SOEs .................................................................................................... 13 Figure 6. Net Fiscal Contribution (percentage of GDP) ............................................................................. 13 Figure 7. Government Spending in 2015 (percentage of GDP) .................................................................. 14 Figure 8. UGS in 2015 (percentage of GDP) .............................................................................................. 14 Figure 9. Composition of Recurrent Spending (as a percentage of GDP) .................................................. 15 Figure 10. UGS (as a percentage of GDP) .................................................................................................. 15 Figure 11. Composition of Government Spending (as a percentage of GDP) ............................................ 17 Figure 12. IMR............................................................................................................................................ 18 Figure 13. MMR ......................................................................................................................................... 18 Figure 14. Projected OAP Benefit Costs (percentage of GDP) .................................................................. 20 Figure 15. Gross International Reserves (US $ million and months of imports) ........................................ 21 Figure 16. Public Debt, External and Domestic (as a percentage of GDP) ................................................ 21 Figure 17. BOOST ...................................................................................................................................... 23 Figure 18. Tax, Revenue Collection, and Total Public Expenditures, 2004–2017 ..................................... 26 v Figure 19. Tax Mobilization in SSA, FY2014/15 a..................................................................................... 29 Figure 20.VAT Productivity in Lesotho, FY2007/08–FY2015/16 ............................................................. 34 Figure 21. C-Efficiency Ratio in SADC Countries ..................................................................................... 35 Figure 22. Lesotho’s Cost of Revenue Collection for FY2010/11 to FY2014/15 (Maloti, millions and as a percentage of total tax revenues) ................................................................................................................ 48 Figure 23. Cost of Revenue Collection as a Percentage of Total Tax Revenues in Selected SSA Countries, 2008a ........................................................................................................................................................... 49 Figure 24. Wage Spending, 2015 (as a percentage of GDP) ...................................................................... 51 Figure 25. Wage Spending in Lesotho (as a percentage of GDP) .............................................................. 51 Figure 26. Staff Number by Category......................................................................................................... 52 Figure 27. Staff Distribution and Vacant Positions per Grade.................................................................... 53 Figure 28. Civil Service Structure in Zambia (Left) and Cameroon (Right) .............................................. 53 Figure 29. Wage Bill (as a percentage of GDP) and Employment (as a percentage of the working-age population) .................................................................................................................................................. 55 Figure 30. Wage Spending, Average Government Wages, and Government Employment ....................... 56 Figure 31. Wage Spending Projections (percentage of GDP)..................................................................... 58 Figure 32. Age Distribution of Public Servants (2015/16) ......................................................................... 59 Figure 33. Number of Participants by Age Covered by the Legacy Pension Scheme ................................ 69 Figure 34. Benefit Costs for the Legacy Pension Scheme (percentage of GDP) ........................................ 70 Figure 35. Lesotho Public, Private, and Gross Capital Formation (as a percentage of GDP), 2007–2015 74 Figure 36. Public Investment in SACU Countries (as a percentage of GDP), 2011/12–2015/16............... 75 Figure 37. Sources of Capital Financing (FY2015/16) ............................................................................... 77 Figure 38. Schematic View of Monitoring at the Center Stage during Project Implementation ................ 94 Figure 39. Decision Tree in Selecting the Public Investment Modality for a Unified PIM Framework .... 96 Tables Table 1. Public Sector Revenues and Expenditures, 2015–2019 (percentage of GDP).............................. 12 Table 2. Comparison of Travel Allowance Between Lesotho and the World Bank .................................. 16 vi Table 3. Health Expenditures in Lesotho Compared to Other Countries in the Region ............................. 18 Table 4. NSDP Targets Set for 2016 and 2020, and 2016 Outcomes ......................................................... 21 Table 5: Lesotho: General Government Revenues, FY2011/12 to FY2017/18 .......................................... 28 Table 6. General Government Revenues (percentage of GDP), FY2014/15 .............................................. 30 Table 7: Lesotho: Tax Revenue, FY2011/12 to FY2017/18 ....................................................................... 31 Table 8: Tax Expenditures for Quarter 4 of FY2015/16 (in Maloti) .......................................................... 33 Table 9: VAT Productivity and C-efficiency in SADC Countries.............................................................. 35 Table 10: Cross-Country Comparison of Tobacco Prevalence and Compliance within SADC Countries 37 Table 11: Tobacco Excises and Retail Prices: Comparison within SADC Countries ................................ 37 Table 12: Excise Tax Performance, FY2012/13–FY2015/16 ..................................................................... 38 Table 13: Total, Recorded and Unrecorded APC, 2010 (in liters of pure alcohol; 15+ years population) in SADC .......................................................................................................................................................... 38 Table 14: VAT Calculations on Alcohol and Tobacco Sales (Maloti, millions) ........................................ 39 Table 15: Key Income Tax Rates................................................................................................................ 39 Table 16: Statutory Income Tax Rates ........................................................................................................ 40 Table 17: METR: Lesotho versus SSA Comparators ................................................................................. 41 Table 18: CITs in SADC Countries ............................................................................................................ 42 Table 19: Investment Incentives in Lesotho ............................................................................................... 42 Table 20: Forgone Revenues from the 10 Percent CIT inside the SACU, Exemption of WHT on Dividends, and 125 Percent Deduction for Training Costs......................................................................... 45 Table 21: Indicators on Paying Taxes in SACU, SSA, and OECD Countries ............................................ 49 Table 22: The Value of Regulations and Tax Indicators ............................................................................ 50 Table 23: COLA ......................................................................................................................................... 54 Table 24: Pay by Grade (Lesotho versus South Africa) ............................................................................. 55 Table 25: Annual COLA ............................................................................................................................. 59 Table 26: Attrition....................................................................................................................................... 60 Table 27: Temporary and Contractual Workers.......................................................................................... 60 vii Table 28 Estimated Fiscal Savings ............................................................................................................. 63 Table 29: Health Employment and Wage Bill, 2015/16 ............................................................................. 64 Table 30: Status of Clinical and Non-Clinical Staff in Health Centers, 2016 ............................................ 65 Table 31: Human Resource Positions for Doctors/Medical Officers in District Hospitals, 2016............... 66 Table 32: In-Position Personnel in the Public-Private Partnership Facilities ............................................. 66 Table 33: Number of VHWs Active at the Time of Survey, 2016 ............................................................. 67 Table 34: Benefit Indexation ...................................................................................................................... 71 Table 35: Sector Investment as a Percentage of Revised Total Capital Expenditures in Lesotho (FY2010/11–FY2016/17) ........................................................................................................................... 76 Table 36: Comparative Infrastructure Investment Outcomes ..................................................................... 77 Table 37: Comparison of Budgets and Outturns for Capital Expenditures (in Maloti, millions) ............... 78 viii Acronyms and Abbreviations APC Alcohol Per Capita Consumption ASYCUDA Automated System for Customs Data ATRN Africa Tax Research Network BFP Budget Framework Paper BOS Lesotho Bureau of Statistics CBA Cost-Benefit Analysis CHAL Christian Health Association of Lesotho CIT Corporate Income Tax CMP Customs Modernization Program COLA Cost-of-Living Adjustment DPCM Department of Project Cycle Management DTA Double Taxation Agreement DTEP Distance Teacher Education Program ECCD Early Childhood Care and Development ECF Extended Credit Facility ECOL Examinations Council of Lesotho ERSA Economic Research Southern Africa ETPM Enterprise Taxation and Policy Management FBT Fringe Benefit Tax GDP Gross Domestic Product GoL Government of Lesotho HIV Human Immunodeficiency Virus HRM Human Resource Management ICR Implementation Completion and Results Report IDM Institute of Development Management IEMS Institute of Extra-Mural Studies IFC International Finance Corporation IFMIS Integrated Financial Management Information System IMF International Monetary Fund IMR Infant Mortality Rate IRMS Integrated Revenue Management System IT Information technology JC Junior Certificate LDTC Lesotho Distance Teaching Center LGCSE Lesotho General Certificate of Secondary Education LLA Lesotho Liberation Army LM Line Ministry LNDC Lesotho National Development Corporation LRA Lesotho Revenue Authority METR Marginal Effective Tax Rate MDG Millennium Development Goal MMR Maternal Mortality Rate MTEF Medium-Term Expenditure Framework MTFF Medium-Term Fiscal Framework MoDP Ministry of Development Planning MoET Ministry of Education and Training MoF Ministry of Finance ix MoH Ministry of Health MPU Ministerial Planning Unit NMDS National Manpower Development Secretariat NSDP National Strategic Development Plan NUL National University of Lesotho OAP Old Age Pension OECD Organisation for Economic Co-operation and Development PAC Project Appraisal Committee PAG Project Appraisal Guidelines PAYE Pay-As-You-Earn PBM Performance-Based Management PCM Project Cycle Management PCN Project Concept Note PEFA Public Expenditure and Financial Assessment PER Public Expenditure Review PETS Public Expenditure Tracking Survey PIM Public Investment Management PIT Personal Income Tax PFM Public Financial Management PFMA Act Public Financial Management and Accountability Act PM Prime Minister PODCPF Public Officers’ Defined Contribution Pension Fund PPAD Procurement Policy and Advisory Department PPP Public-Private Partnership PSD Private Sector Development PSC Public Sector Commission PSIC Public Sector Investment Committee PSID Public Sector Investment Database PSIP Public Sector Investment Program PSLE Primary School Leaving Examinations PSMP Public Sector Modernisation Project QMMH Queen Mamohato Memorial Hospital QSDS Quantitative Service Delivery Survey SACU Southern Africa Customs Union SACMEQ Southern and Eastern Africa Consortium for Monitoring of Education Quality SADC South African Development Community SDR Social Discount Rate SNG Sub-national Government SOE State-owned Enterprise SOP Standardized Operating Procedure SSA Sub-Saharan Africa TADAT Tax Administration Diagnostic Assessment Tool TB Tuberculosis TSA Treasury Single Account TSC Teaching Service Commission TSD Teaching Service Department TVET Technical and Vocational Education and Training UGS Use of Goods and Services UNAIDS Joint United Nations Programme on HIV/AIDS x UNESCO United Nations Educational, Scientific and Cultural Organization VAT Value-Added Tax VHW Village Health Worker WDI World Development Indicators WEO World Economic Outlook WFP World Food Programme WHO World Health Organization WHT Withholding Tax xi Executive Summary I. Introduction 1. This Public Expenditure Review (PER) was prepared in response to a request from the Ministry of Finance (MoF) and is designed to inform Lesotho’s fiscal consolidation due to a narrowing of its fiscal space. Lesotho is facing a tough macro-fiscal outlook due to a sharp decline in Southern African Customs Union (SACU) revenues. This situation necessitates a significant adjustment in the current fiscal stance to ensure longer-term fiscal sustainability. However, the adjustment should be tailored to minimize any adverse growth and poverty impacts. Thus, this PER is intended to support the government’s efforts to adjust its policies to better address Lesotho’s current macro-fiscal circumstances. 2. Lesotho is one of the poorest and most unequal countries in the world, despite a relatively good growth performance over the past 15 years. Lesotho’s per capita gross national income is about US$1550. Lesotho’s poverty rate is 59 percent (US$1.90 purchasing power parity [PPP] per day), its Gini coefficient is 0.541, and about 59 percent of the population now lives below the international poverty line of $1.90/day. Both poverty and extreme poverty disproportionately affect the rural population, and the bottom 40 percent of Lesotho’s population experienced a decline in consumption each year between 2002 and 20112. This compares to increases, albeit meager, for the remaining 60 percent of the population over the same period. Lesotho’s gross domestic product (GDP) grew at an annual average rate of 4 percent between 2000 and 2016, whereas its GDP per capita grew at an average rate of 2.8 percent during the same period. 3. Despite the high level of government spending, Lesotho faces challenges in addressing inclusive growth and providing access to quality services for the poor — while also operating in a highly fragile environment. After political turmoil, the new government with a fragile coalition of 7 parties was established in June 2017. The government is facing a significant challenge to improving access to and the quality of public services. It is also seeking to invigorate the domestic private sector to diversify the growth sources of its economy. The level of unemployment is very high, with a low employment-to- working-age population ratio, which limits prospects for social mobility and poverty reduction. II. Macro-Fiscal Profile 4. Lesotho maintained an average growth rate of 4 percent between 2000-2016. On the production side, services have been the most resilient contributor to growth, except during 2016. However, the contribution of agriculture and industry remained volatile. Lesotho has a very small agricultural sector. Due to climatic variability and lack of good irrigation, the contribution of agriculture can be very volatile. Likewise, the contribution of mining has been unpredictable. When the expenditure side is considered, the contributions varied from one year to the next. Household consumption has contributed positively to growth over this period. In addition, government consumption and investment have also made significant, positive contributions to output growth. However, given the narrowing fiscal space it will be hard for the Government to stimulate growth in the near term. As such, Lesotho’s growth performance is expected to 1 World Bank, Macro-Poverty Outlook, Spring 2017. 2 Systematic Country Diagnostic, Lesotho, World Bank, 2014 1 be driven by industry, mainly mining and construction — especially with the start of the second phase of the Lesotho Highlands Water Project (LHWP2). 5. Lesotho is facing a tough macro-fiscal outlook in the near-to-medium term due to a sharp decline in Southern African Customs Union revenues. Economic growth is expected to slow to 3.6 percent in FY 2017/18 as the government’s ability to drive economic growth is limited due to the decline in SACU revenues. The SACU revenues declined from 25 percent in FY2014/15 to 13.2 percent of GDP in FY2017/18 because of the economic slowdown in South Africa. Lesotho’s fiscal deficit registered 6.4 percent of GDP in FY2016/17, which was mostly financed through a fast drawdown of government deposits at the Central Bank. Indeed, Government deposits at the Central Bank declined to 4.2 months of import coverage at the end of FY2016/17. The fiscal balance is projected to register a deficit of 8 percent of GDP at the end FY2017/18, and the reserve coverage is expected to further decline to 3.6 months of imports due to lower tax collection, with no significant adjustments in expenditure and lower SACU revenues.3 If the current macro-fiscal framework is maintained without further adjustments —and given the current SACU revenue projections — Lesotho can find itself in a position with a reserve coverage of below 3 months of imports by the end of FY 2019/20. This would in turn place significant pressure on its exchange rate sustainability. Lesotho’s debt-to-GDP ratio stood at 35.4 percent of GDP at the end of FY2016/17 and is assessed with low risk. However, an induced devaluation would lead to high inflation and debt service. It would also affect the poor and vulnerable households, and reduce the government’s financial capacity to invest in development projects. Given its low export supply response and inelastic demand for foreign products, Lesotho could fall into a low growth-high poverty cycle from which it would be difficult to rebound. 6. Lesotho’s fiscal outcomes are mainly driven by volatile SACU revenues. The SACU revenues are highly volatile and unpredictable. In the absence of a mechanism to counteract this volatility, Lesotho’s economy is prone to amplified boom and bust cycles due to its highly procyclical fiscal policy. Lesotho has experienced similar episodes of decline in SACU revenues that were followed by sharp fiscal adjustments. However, when SACU revenues rebound, fiscal spending again increases—with significant fiscal consequences due to procyclical capital investments that go with long gestation. As SACU revenues are projected to remain low in the medium term, Lesotho will need to consolidate its fiscal spending to maintain macroeconomic and exchange rate sustainability through preserving buffers over the next three years. 7. There is an urgent need for reform to bring Lesotho to a higher and sustainable growth path. Sharp and temporary adjustments to close the financing gap in the short-run are not sufficient to reach the medium-term policy goals stated in the National Strategic Development Plan (NSDP) and to ensure sustainable growth and poverty reduction as the deeply rooted structural issues remain. Ensuring better revenue and expenditure management while safeguarding public service delivery remain at the center of policy concerns and cannot be addressed through sharp and temporary fiscal adjustments. III. Efficiency and Effectiveness of Public Expenditures 8. Lesotho’s macro-fiscal challenges stem from high Government spending that are fueled by large but volatile SACU revenues. Lesotho has one of the highest government spending rates as a percentage of GDP in Sub-Saharan Africa (SSA) that is, at 47.3 percent of GDP in FY 2016/17, where recurrent spending accounted for 37.4 percent of GDP. Government revenues generally demonstrate a high dependency on SACU receipts, which are highly volatile. Government spending and, especially current expenditures, rise whenever SACU revenues increase. However, when SACU revenues decline sharply, the 3 Data as of January 2018 and it is subject to change. 2 government is forced to implement large fiscal cuts leading to a boom and bust cycle. The large inflows are absorbed mostly by government consumption, particularly the public-sector wage bill. The public-sector wage bill reached 17.8 percent of GDP in FY 2016/17, whereas the use of goods and services accounted for almost 10 percent of GDP. These figures are among the highest in the world and represent a significant challenge to fiscal consolidation. 9. High spending in the public sector has not resulted in pro-poor growth, with health and education results far below those of other comparator countries. Lesotho has the world’s third highest human immunodeficiency virus / acquired immunodeficiency syndrome (HIV/AIDS) prevalence, at 23.6 percent. However, HIV incidence in Lesotho decreased over the past 20 years.4 Lesotho’s level of maternal mortality is also among the highest in Sub-Saharan Africa. Lesotho is currently not on track to meet the health-related Millennium Development Goals (MDGs). Although education spending is one of the highest in the world as a percentage of GDP (13.5 percent of GDP) and universal access to primary education enshrined in law, poor educational outcomes persist across the country. Lesotho is also not on track to achieve the two education-related MDGs, with both primary enrollment (82 percent) and primary completion (65 percent) rates lagging far behind the goal of 100 percent by the end 2015. 10. Although the resource allocation, including allocations to the social sectors, is broadly in line with the NSDP and the country’s development needs, resource management remains a challenge. Spending on social sectors is above 20 percent of GDP, and the allocation of resources is broadly adequate. In terms of capital expenditures, compared to its SSA and world peers, Lesotho has sustained a relatively high level of capital expenditures thanks to the strong level of external financing. When fiscal consolidation is required, budget-financed public investment is subject to larger cuts, which again exacerbates the boom and bust cycles. In addition, although investment has been concentrated in key areas such as public infrastructure, finance, water and energy, the outcomes (such as access to electricity or water for agricultural activities) appear to have remained low. Overall, public investment management (PIM) is a major problem and requires a strengthening in the oversight function, as well as improvements in the project approval, financing and monitoring process. Since the mid-2000s, the Government of Lesotho has embarked upon PIM reforms, including the creation of the Public Sector Investment Committee (PSIC) in 2013 as part of its broader public financial management (PFM) reforms. 11. Outcomes fell short of the targets set out in the First NSDP, despite the high level of government spending. The first NSDP set inclusive growth as a key development objective. A target growth rate of 7 percent was established to assist in halving the poverty rate in Lesotho. Through support to the private sector, the NSDP envisaged creating 50,000 private sector jobs, mainly in agriculture, manufacturing and mining. HIV prevalence would also be halved. The NSDP called for optimal use of public resources in order to address the following objectives: (i) improving resource allocation and cost efficiency; (ii) promoting fiscal consolidation and ensuring fiscal discipline; (iii) improving tax/revenue administration; (iv) increasing the absorptive capacity of the ministries to spend budget allocations more efficiently; (v) sustaining the real value of capital expenditures; and (vi) containing the wage bill and increasing public sector efficiency. However, the review of the NSDP5 highlighted the under-performance of the Plan due to a lack of implementation of key policies, including those related to macro-economic stability, adequate resource allocation and spending, programs for targeted people, and monitoring of developments results — combined with chronical political instability. The latest Public Expenditure and Financial Accountability (PEFA) Assessment in 2017 indicates that the upstream processes of budget formulation and process perform relatively well. However, policy objectives are too weak or undeveloped to have an impact on the allocations in the downstream processes of in-year budget execution. The ambitious targets set in the NSDP for 2016 have not been achieved and are unlikely to be achieved by 2020. 4 Lesotho Health PER (2017) World Bank. 5 Assessment of the NSDP 1 and Scoping of the NSDP 2, Government of Lesotho, December 2016. 3 Although the NSDP set a 7 percent growth rate as a target to achieve by 2016, the mechanisms were neither identified nor set in place to reach these targets. 12. Lesotho’s macro-fiscal challenges mostly stem from high recurrent expenditures, which at 38 percent of GDP. The level of recurrent expenditures is high, with a wage bill of 18 percent of GDP (FY2016/17) and around 10 percent of GDP spent on the Use of Goods and Services (UGS) in FY2016/17. Spending items such as travel subsistence and expenses related to the government vehicle fleet for civil servants are particularly high, at 2 percent of GDP. This requires redefining the policy for management and practice of such UGS. 13. The wage bill in Lesotho is its largest expenditure item and one of the highest in the world as a share of GDP. Due to lack of private sector employment opportunities and the country’s high unemployment rate, the public service has absorbed employment demand. In this context, the number of public servants has increased by 60 percent since 2002. The high wage bill is attributable to the following factors: (a) an increase in net hiring, (b) the high wage scale, compared with neighboring countries, including South Africa (adjusted in purchasing power), and (c) the high pace of real annual wage increases at 7-9 percent on average for the past five years. The wage increase is comprised of two dimensions: Cost- of-living Adjustment (COLA) and automatic notch increase (2.7 percent on average). It is noteworthy that the same wage increase is applied across the board, regardless of staff performance. Also, the COLA exceeded the inflation rates in 2013 and 2014. 14. Human Resource Management (HRM) reform is critical to improving not only the wage bill management, but also the performance of the public sector as a whole. Potential fiscal savings of reforms are estimated in the rage of 0.2 to 1.9 percent of GDP. Due to uncontrolled hiring and a lack of establishment controls, the civil service today has a bottom-heavy structure with more than 50 percent of total staff serving as support staff. Nevertheless, the vacancy rate for the lowest grade staff remains as high as 57 percent, whereas the overall vacancy rate is 23.3 percent. In short, a significant number of support staff — with fewer technical staff and middle management staff with high wages —are not conducive to improving public services. Since 2015, the Government has initiated a major HRM reform program with the support of the Public Sector Modernization Project (PSMP) financed by the World Bank. The Ministries of Finance and Public Service migrated human resource, payroll, and pension records into one integrated HRM system. The recent initiative for the cleaning up of human resource/payroll records has so far identified approximately 7 percent of ‘non-identified workers’ (or potential ghost workers).6 The elimination of such workers would create a positive fiscal impact. The operation will be followed by a biometric census that is expected to start in early 2018. The authorities also recognize the importance of performance-based management (PBM), which is underpinned by the Public Service Act. Chapter 3 provides a detailed analysis with reform options that could create positive fiscal savings. 15. The volatile macroeconomic situation calls for an improvement of public financial management and better expenditure controls and revenue management. Effective and transparent PFM systems and institutions are critical preconditions for effective macro-fiscal management, fiscal policy and budgeting. This is especially the case for a country like Lesotho that depends heavily on volatile revenues and that has a fixed exchange rate regime with limited monetary policy options. The government has undertaken PFM reforms since 2012, supported by the PFM Reform Action Plan 2012/13 - 2017/18. However, most of the scores in the latest PEFA 2017 assessment have deteriorated or are rated D or D+ at best. The PEFA (2017) concludes that the budget process does not have a strong policy or strategic focus. Expenditure control, including payroll and procurement controls, and financial reporting are deficient, and the coverage of audits and the independence of the supreme audit institution are questionable. 6 The result has yet to be finalized. 4 IV. Revenue Performance 16. Historically, Lesotho had a strong revenue performance because of high SACU revenues, as well as a relatively high tax collection rate in comparison to its SSA peers over the past decade. Lesotho’s main revenue sources are SACU revenues, tax revenues, grants, and other non-tax revenues. The first two categories constitute the largest two components of Lesotho’s income, with 16 and 21 percent of GDP, respectively, in FY2016/17 respectively. As part of SACU, it receives an important portion of its revenues from the union. However, these revenues are primarily driven by South Africa’s economic growth and are highly volatile. SACU revenues averaged 22 percent of GDP between 2000/11 and 2016/17, with a maximum of 32 percent of GDP in FY2006/07, and a minimum of 13.3 percent in FY2011/12.7 Although tax collection stands above the average of SSA peers over the past decade, there is room to improve the complex tax policy and administration procedures. Combined with the ambiguity in provisions of tax benefits, the cost of tax collection and compliance remain high, and exert a negative impact on the business environment and formality. A tax collection and public expenditure comparison shows unfavorable developments since 2010, when public expenditures dramatically rose above total revenues from all sources. Simplifying the tax policy and administration procedures can enhance revenue performance. 17. The current setting of the value-added tax (VAT) regime is complex. Although the threshold is comparable to a modern VAT elsewhere, the structure of the VAT regime is overly complex with four different rates, including the zero rating of certain non-export commodities. It is also riddled with multiple exemptions. The potential revenue loss resulting from the zero rating of non-export sales, reduced rates, and exemptions is expected to be significant, collectively amounting to almost 19 percent of the total actual VAT collection — or 6 percent of the total tax revenues. A standard policy recommendation would call for Lesotho to rationalize both the rate and exemption structure to enhance the efficiency of the tax. To strike a balance between VAT efficiency and equity, and given the relative high productivity of the regime, it is recommended that policy reform be focused on unifying the VAT rate structure. During the short-to- medium term, it is proposed that the VAT policy reforms focus on streamlining and unifying the statutory rate structure at the standard rate of 14 percent. 18. With the new bill on tobacco and alcohol taxation, additional fiscal revenues are expected. Compared with its SSA peers, Lesotho is a country with a high rate of smoking prevalence. However, the country’s total tax rate on tobacco is lower than 7 other countries within the 15-country South African Development Community (SADC). Compared to its neighboring countries, Lesotho’s alcohol per capita consumption (APC) falls in the midrange. Excise revenues from alcohol accounts for the majority of total excise collection. The Government plans to introduce the new bill on alcohol and tobacco with technical assistance support from the World Bank. The new bill is expected to incorporate a levy of 30 percent on tobacco products and 20 percent on alcohol. Under such provisions, the expected additional fiscal revenues to be generated amount to about R 163 million (US$ 11.6 million) or about 0.45 percent of GDP according to 2017/18 projections. 19. The statutory rates as applicable to the Personal Income Tax (PIT) and Corporate Income Tax (CIT) are comparable to the global averages. A closer look at the CIT rate structure indicates, however, that the country grants preferential treatment to manufacturing over non-manufacturing. In addition, Lesotho provides a zero rate on profits from exports outside of the SACU. The multi-rate structure and other underlying provisions that affect the determination of the corporate taxable income lead to cross sector and cross-asset distortions. Some may argue that tax incentives can be used as part of the overall package to support manufacturing and/or exports due to the market and coordination failures that cannot be readily and easily resolved over the immediate terms.8 The underlying distortionary impact of incentives 7 The SACU revenue-sharing mechanism is discussed in detail in Chapter 2. 8 See, for example, World Bank January 2017: South Africa Economic Update. 5 needs to be acknowledged at the tax policy design level. Restructuring of the set of tax incentives can be considered to make them more effective, given the strategic decision already made on industrial development (e.g., shifting away from reduced or zero rates or other profit-targeting instruments to those that help directly reduce the costs of investment or the operational costs in running a business. Such simple measures would help drive a firm from a loss making to a profit-making position and hence are more likely to induce marginal investments). 20. Given some procedural inefficiencies in tax administration, the Lesotho Revenue Authority (LRA) will need to undertake a comprehensive assessment of tax administration to identify key areas for reforms. Although Lesotho has long embarked on the concept of comprehensive revenue administration, the outcomes have not significantly improved over time. Problems remain in key areas of business processes (especially in the core performance function of auditing and investigation) and in the use of information technology to enhance the quality of voluntary tax compliance. It is critical that the LRA now take steps to review and ascertain what has been done, what challenges remain, and how to prioritize and sequence further reform actions. As an immediate step, the revenue agency may consider conducting a detailed assessment of tax administration. The review and diagnostics would help the LRA to establish the baseline for better targeting and investing of its tax administration reforms. V. Policy Recommendations 21. Beyond the urgent short-term objective of restoring fiscal stability, Lesotho faces important long-term development challenges that will require the Government to reassess how public spending contributes to advancing its development goals. There is significant scope to improving macro-fiscal planning and policies, increasing the efficiency and effectiveness of government spending, and improving PFM systems and budget management for a more efficient, accountable, and transparent government. 22. It is recommended that the Government of Lesotho (GoL) regularly conduct Public Expenditure Reviews (PERs) for reducing and addressing fiscal risks, and encouraging better strategic use of resources. Timely decision-making on fiscal consolidation and/or strategic allocation and use of resources will be key to achieving the development goals set out in the NSDP, particularly when the fiscal situation deteriorates as is the case today. It is highly recommended to regularly conduct PERs using the expenditure analysis tool, including BOOST, which was developed as part of the current PER. To this end, it is necessary to improve statistical capacities and strengthen/make use of data from the information systems, such as the Integrated Financial Management Information System (IFMIS), the Human Resource Management Information System (HRMIS) (for Human Resource and payroll/pension management), and the Education Management Information System (EMIS) for strategic purposes. 23. The government should fast track PFM reforms. Although the government has been undertaking PFM reforms since 2011/12, it is critical that it fast track the priority reforms. These include: the enhancement of the IFMIS; and the introduction of the Treasury Single Account (TSA) to improve budget credibility, procurement and internal controls, cash management, the reliability of the IFMIS information and reporting, and auditing. These functions are key to improving spending efficiency and revenue management, while also reducing the risks to financial integrity. 24. Certain recurrent expenditures should be subject to rationalization. Recurrent expenditure and public asset management require a clear rule for use and official records. The PER recommends rationalizing certain expenditure items in the category of UGS, which are not directly contributing to the efficiency and performance of public services. The GoL can particularly consider cuts to its governmental vehicle fleet spending and international and local travel subsistence. It can achieve this by reducing per 6 diems, the number of non-essential trips and providing subsistence in South African rand, instead of US dollars, at least for SACU countries. The government should also clarify the rule for its use, entitlement, and generally, management of public assets, and record all detailed expenditures in a clearer manner in the PFM system. 25. In the short term, constraining the growth of public wages can be accomplished by reforming the compensation practices. Given the significant wage increase in real terms every year —regardless of the country’s economic situation — limiting the annual wage increases that are comprised of COLAs and notch increases can be considered as a measure to contain the wage bill. The government should continue its efforts to clean up the human resource records and suspend any payments for identified ghost workers. Addressing the size of the public-sector workforce is another option. A hiring freeze or limitation on new hiring in the non-essential categories of staff (Grades A to C) for a limited time, or attrition-based downsizing, can be considered. However, given the social impact, the option and associated risks should be carefully studied and managed during implementation. 26. In the medium-to-long term, the introduction of an evaluation system and the restructuring of the civil service are necessary to enhancing the performance of public services. To this end, it will be recommended to first conduct a functional/institutional review, determine job descriptions, revise the establishment lists or introduce the establishment list for the education sector, and devise the strategic hiring plan accordingly. Once this preparatory phase is completed, as defined in the Public Service Act, a staff performance evaluation system can be introduced on a pilot basis, and then rolled out across the ministry. In addition, given the current distorted structure of the civil service, some supporting services such as cleaning and building maintenance services can be outsourced to the private sector. This will help to reduce staff costs, and create private sector jobs. Age distribution analysis suggests that there will be a batch of retirees in 2020/21, 2030/31, and 2040/41 when a major restructuring and/or rejuvenation of the public service could be considered. 27. To achieve the NSDP development goals, improvement of efficiencies and the quality of public investment projects will be essential. Although the government has undertaken some critical reforms — including the creation of the Public Sector Investment Committee (PSIC) as an oversight mechanism with an aim to improving the quality and impact of public investment — there is scope for improvement. In order to ensure that the PSIC is fully functional and empowered to play the role of a ‘gatekeeper’, the government needs to develop the public-sector investment program as part of the NSDP, improve the project appraisal process and develop a unified PIM database. These actions are complementary and are aimed at forging the link between the outcomes of the PSIC project appraisal review and the financing decision by the MoF, or as part of the budget preparatory process. Once an effective mechanism is instituted, the program should incorporate public investments from all modalities—including those financed by donors—to ensure a cohesive and effective response to the national development priorities, as well as consistency with the GoL planning and budgeting processes. 28. Strengthening project monitoring and enhancing a link with onward financing are key to ensuring the completion of quality investment projects. Lesotho could feasibly make project monitoring more effective. Some mechanism is to be established for organizing the MoF/MoDP joint monitoring (both financial and physical) with line ministries, as well as for holding the latter accountable for its own monitoring reports. The PSIC could be easily extended to include the mandate and convening authority to collaborate with the MoF and the MoDP, with the support from the Cabinet of Ministers and effective communication with the line ministries. The goal would be to establish an institutionalized mechanism that creates a direct link between monitoring outcomes and onward financing. For this purpose, it is also necessary that the MoDP/MoF jointly consider reviewing the entire process involved in management and monitoring of project execution, and then prepare a manual/uniform set of guidelines on Standardized Operating Procedures (SOPs). 7 29. The harmonization of the Public-Private Partnership (PPP) into a unified PIM framework is another element required to enhance cost-benefit analyses of public investments, and to improve the quality of public services. To date, Lesotho has experienced both successes and failures in managing PPPs. The time for ‘trials and errors’ is to come to an end, and efforts are to be made to harmonize PPPs into a unified PIM. It would clearly imply that at the current stage, when PIM is to be strengthened, the Government not rush. Rather, it should move with care in the approval of any PPP proposals. Although the country’s PIM capacity remains limited, Lesotho may use external expertise to help review each PPP proposal with regard to appraisal, risk allocation in contracting, approval, and design of appropriate monitoring mechanism. 30. In terms of revenue collection, simplifying the tax policy and administration procedures can enhance its performance. Decreasing the ambiguity in the provisions of tax benefits can lead to lower costs for tax collection, as well as higher tax compliance. More specifically, the current VAT regime with four different rates is overly complex. The potential revenue loss from the zero rating of non-export sales and reduced rates are significant. During the short-to-medium term, a policy reform should be focused on the tax rate structure to streamline and unify the statutory rate at the standard of 14 percent. It is a welcome development that the Government of Lesotho has announced the VAT will be raised to 15 percent to align with that of South Africa to prevent smuggling.9 31. Over the medium term, the income tax reforms could focus on a few key steps. First, keep the standard rate of CIT at 25 percent unchanged over the mid-term. Second, consider harmonizing the various different PIT and CIT rate income tax regimes to avoid unintended arbitrage. Currently, the statutory PIT and CIT rates are not harmonized. Third, consider introducing inheritance and gift taxes. Fourth, conduct a thorough review of the current structure of the tax incentives and explore options for reform. The qualification for incentives should be systematically applied when any investor satisfies the specified conditions as required by the governing laws and legislation. Over time, the GoL should consolidate all incentives within the purview of tax legislation under the authority of a single government body (ideally the Ministry of Finance). 32. In addition, Lesotho can introduce a unified levy of 30 percent on both tobacco and alcohol products. Under this scenario, the expected additional fiscal revenues to be generated can amount to about R 200 million (US$14.2 million) or about 0.6 percent of GDP respectively, based on 2017/18 projections. 9 See MOF Budget Speech 2018/19, available at: http://www.finance.gov.ls/documents/budget%20speeches/Budget%20Speech%202018-19%20Final.pdf. 8 Chapter 1: Macroeconomic Overview and Recent Expenditure Trends 33. This Public Expenditure Review (PER) was prepared in response to a request from the Ministry of Finance (MoF) and is designed to inform Lesotho’s fiscal consolidation in the context of declining Southern Africa Customs Union (SACU) revenues. Lesotho is facing a tough macro-fiscal outlook due to a sharp decline in SACU revenues. This situation necessitates an adjustment in the current fiscal stance to ensure fiscal sustainability. However, the adjustment should be tailored to minimize any adverse growth and poverty impacts. Lesotho is a lower-middle-income country with a gross national income of US$1,280 (as of 2015), where an estimated 59 percent of the population is still trapped in extreme poverty. Government spending in Lesotho accounts for 50 percent of gross domestic product (GDP). As such, improving the efficiency and effectiveness of government spending can have a significant, positive impact on GDP growth and poverty reduction, while also decreasing uncertainty about fiscal sustainability. 34. From 2000-2016, Lesotho’s GDP grew at an average annual rate of 4 percent, and GDP per capita grew at an average annual rate of 2.8 percent. However, growth slowed to 2.3 percent in 2016. Although an average growth rate of 4 percent is relatively high when compared to other countries in the region, the growth rate falls short of the National Strategic Development Plan (NSDP) target growth rate of 7 percent. Poverty remains extremely high at 59 percent, with the international poverty rate defined as US$1.9 per day. Micro-simulations indicate that with the current macro-projections, the poverty rate would decline to 57 percent by 2019,10 falling short of the target of 10 percent set in the NSDP. 35. The services sector was the most resilient sector from 2000–2016. On the production side, services have been the most resilient contributor to growth. However, the contribution of agriculture and industry remained volatile. Lesotho has a very small agricultural sector. Due to climatic variability, the contribution of this sector has been very volatile. The contribution of industry remained mostly positive for this period in its entirety, although performance was also unpredictable (figures 1 and 2). Likewise, the contribution of mining has been volatile. When the expenditure side is considered, the contributions varied from one year to the next. Household consumption has contributed positively to growth over this period. In addition, government consumption and investment have also made significant, positive contributions to output growth. However, given the narrowing fiscal space it will be hard for the Government to stimulate growth in the near term. As such, Lesotho’s growth performance is expected to be driven by industry, mainly mining and construction — especially with the start of the second phase of the Lesotho Highlands Water Project (LHWP2). 36. Lesotho is facing a tough macro-fiscal outlook in the near-to-medium term due to a sharp decline in SACU revenues. SACU revenues have declined from 25 percent in FY2014/15 to 13.6 percent of GDP in FY2016/17 — mostly due to lower growth in South Africa. The decline in SACU revenues narrowed the fiscal space and has led to a considerable decline in fiscal buffers. In this context, the Government of Lesotho (GoL) missed an opportunity to make a gradual adjustment in FY2016/17.11 Lesotho’s fiscal deficit reached 7 percent of GDP in FY2016/17. Due to limited borrowing opportunities, the GoL financed the deficit through a drawdown of government deposits at the Central Bank. The reserve coverage declined from 6.1 months of imports in FY2015/16 to 4.2 months of imports at the end of FY2016/17. To maintain fiscal sustainability and preserve buffers, Lesotho will have to considerably adjust its spending downwards over the next three years. Given the current projections of SACU revenues 11 Data as of January 2018 and is subject to revision. 9 provided by the MoF, Lesotho’s reserve coverage can go below 3 months of imports by the end of FY2019/20 — thereby risking macroeconomic stability and exchange rate sustainability. 37. There is an urgent need for reform to bring Lesotho to a higher and sustainable growth path. Sharp and temporary adjustments to close the financing gap in the short-run are not sufficient to reach the medium-term policy goals stated in the NSDP and to ensure sustainable growth and poverty reduction as the deeply rooted structural issues remain. Ensuring sustainability of fiscal policy without adhering to short term policy changes only enhance the boom-bust cycles. Maintaining a sustainable and a higher quality of government spending through safeguarding better public service delivery remain at the center of policy concerns and cannot be addressed through sharp and temporary fiscal adjustments. Figure 1. Contributions to Growth, Production Figure 2. Contributions to Growth, Expenditure 8% 20% 6% 10% 4% 0% 2% 2008 2009 2010 2011 2012 2013 2014 2015 2016 -10% 0% 2000 2002 2004 2006 2008 2010 2012 2014 2016 -2% -20% Discrepancy Imports -4% Exports Changes in inventories Taxes (net of subs) Services GFCF Mining Industry (excl. mining) HH consumption Govt consumption Agriculture GDP @MP GDP growth Sources: Lesotho Bureau of Statistics, National Accounts, and Sources: Lesotho Bureau of Statistics, National Accounts, World Bank staff calculations and World Bank staff calculations. Note: GDP= gross domestic product; MP= Market Prices Note: HH= household. GFCF: Gross fixed capital formation. 38. Historically, Lesotho had a strong revenue performance, albeit one fueled by volatile SACU revenues. High revenues enabled Lesotho to increase public expenditures substantially over the past decade, when government spending accounted for 50 percent of GDP in FY2016/17. Lesotho’s revenue sources can be classified into four main categories: SACU revenues, tax revenues, grants, and other nontax revenues. The first two categories constitute the largest two components (figure 3). As part of SACU, Lesotho receives an important portion of its revenues from the Union. However, these revenues are primarily driven by South Africa’s economic growth and are highly volatile ( figure 3). SACU revenues averaged 22 percent of GDP between 2000/11 and 2016/17, reaching a maximum of 32 percent of GDP in FY2006/07 and a minimum of 13.3 percent in FY2011/12.12 Furthermore, the volatility of SACU revenues has increased over time following the introduction of the current revenue-sharing agreement in 12 The SACU revenue-sharing mechanism is discussed in detail in Chapter 2. 10 FY2005/06.13 The volatility of SACU revenues renders macroeconomic management and budgeting difficult (as the outer three years are difficult to forecast when the overall fiscal envelope is determined). However, high SACU revenues enabled increased government expenditures, which then become difficult to adjust downward when revenues go down. This in turn can mask underlying structural deficits when SACU revenues increase again. Rises in SACU revenues, or revenues in general, also correspond to increases in government spending (figure 4). Figure 3. Revenue Sources (percentage of GDP) Figure 4. Overall Expenditures, Revenues, and the Budget Deficit (percentage of GDP) SACU revenues constitute a large proportion SACU revenues mask the underlying of the revenues and are highly volatile deficits 60% 70% 40% 60% 30% 50% 50% 20% 40% 40% 10% 30% 0% 30% 20% 10% -10% 20% 0% -20% 10% 0% Budget deficit (RHS, % of GDP) 2002/03 2005/06 2008/09 2011/12 2014/15 2000/01 2001/02 2003/04 2004/05 2006/07 2007/08 2009/10 2010/11 2012/13 2013/14 2015/16 2016/17 non-SACU budget deficit (RHS, % of GDP) Revenue (% of GDP) Tax revenue Grants Other revenue SACU Overall expenditure (LHS, % of GDP)) Sources: Ministry of Finance (MoF) Medium-Term Fiscal Sources: MoF MTFF and World Bank staff calculations. Framework (MTFF), and World Bank staff calculations. Note: LHS= Left Hand Side; RHS=Right Hand Side. 39. Tax revenues have increased in the last 15 years contributing to a rise in revenues, whereas the contribution of grants has declined over time. Lesotho has improved its tax collection over the last 15 years and receives sizeable revenues through taxes as compared to other Sub-Saharan Africa (SSA) countries.14 This rise in tax revenues also enabled Lesotho to increase government expenditures. In FY2016/17, tax revenues as a percentage of GDP declined to 21 percent, from 22.6 percent in FY2015/16, due to institutional issues at the Lesotho Revenue Authority (LRA). Although donor grants provided a sizeable contribution during 2011/12 and 2012/13, they have been declining. The slow pace of reforms has led donors to withdraw financing for budget support operations in Lesotho. Revenues are expected to be low as a percentage of GDP in the near term (figure 3 and table 3). 13 World Bank (2016), SACU Receipts and Fiscal Policy in Southern Africa, Report No. AUS13682. 14 A more detailed analysis is provided in Chapter 2. 11 Table 1. Public Sector Revenues and Expenditures, 2015–2019 (percentage of GDP) 2015/16 2016/17 2017/2018 2018/2019 2019/2020 Actual Actual Projected Projected Projected Real GDP growth 2.5% 3.1% 3.4% 4.0% 3.0% Gross International Reserves (months of imports) 6.1 4.2 3.6 3.1 2.5 As a % of GDP 2015/16 2016/17 2017/2018 2018/2019 2019/2020 Actual Actual Projected Projected Projected Tax Revenues 20.3 20.4 17.7 20.8 20.6 Grants 3.0 2.5 2.1 2.0 2.0 Other Revenues 4.2 4.8 3.6 3.9 3.9 SACU 19.7 13.2 16.3 18.8 19.0 TOTAL REVENUE AND GRANTS 47.2 40.9 39.7 45.5 45.4 Expense 37.9 37.4 37.7 33.9 32.5 o/w Compensation of employees 17.1 17.8 17.4 16.8 16.7 o/w Use of goods and services 9.5 9.7 10.0 9.2 8.9 o/w Interest Payments 0.8 0.7 0.9 1.1 1.1 o/w Subsidies 0.8 1.2 0.8 0.9 0.8 o/w Grants 3.4 3.7 3.7 2.8 2.7 o/w Social benefits 2.8 2.6 2.8 2.7 2.6 o/w Other expenses 2.7 1.9 2.2 2.1 2.0 Non-financial assets 11.4 9.9 10.0 10.0 10.0 TOTAL EXPENDITURE 49.3 47.3 47.7 43.9 42.5 Overall Fiscal Balance -2.1 -6.4 -8.0 1.7 2.9 Primary Balance -1.3 -5.7 -7.1 2.7 4.0 Non-SACU Fiscal Balance -21.8 -19.6 -24.3 -17.1 -16.1 Primary balance (excluding SACU receipts) -21.0 -18.8 -23.4 -16.1 -15.1 Gross Public Debt 41.3 35.4 35.4 38.8 43.2 Sources: MoF and World Bank staff calculations/projections. 40. Procyclical non-tax revenues also contribute to the boom-bust cycle observed in Lesotho. Lesotho receives non-tax revenues through water and diamond royalties, as well as dividends from state- owned enterprises (SOEs) and parastatals. Water royalties averaged 2.4 percent of GDP between FY2010/11 and FY2016/17. Similarly, dividends averaged 1.39 percent of GDP, but have been very volatile. The SOE Policy Note (2016) prepared by the World Bank finds that the net fiscal impact of SOEs has been positive. It fluctuated between 2 to 5 percent of GDP when all the flows to and from the government are accounted for (figure 5). It is important to reinvest these revenues from the extractive industries through soft or hard investments to ensure inter-generational equity. Figure 6 shows that the SOEs that have a positive net fiscal impact are highly procyclical. Indeed, SOEs constitute an important portion of the production base of the Lesotho economy, contributing to the government’s ability to effectively manage the boom and bust cycles. 12 Figure 5. Net Fiscal Impact of All SOEs Figure 6. Net Fiscal Contribution (percentage of GDP) 6% 1 Standard 5% LPB 0.8 Bank 4% 0.6 3% LEC Letseng 0.4 LNIG 2% LNDC 0.2 1% 0 0% -1.0% -0.5% 0.0% 0.5% 1.0% 1.5% 2.0% -0.2 WASCO 2009 2010 2011 2012 2013 2014 -1% -0.4 Econet -2% -0.6 BEDCO SOEdebt arrears -3% Total dividends -0.8 Loti Brick taxes grants -1 change capital grants subventions net fiscal impact (% of GDP) Source: World Bank SOE Policy Note (2016). Source: World Bank SOE Policy Note (2016). Note: BEDCO=Basotho Enterprises Development Corporation; LEC= Lesotho Electricity Corporation; LNDC=Lesotho National Development Corporation; LNIG= Lesotho National Insurance Group; WASCO=Water and Sewerage Company. I. Government Expenditure 41. Lesotho has one of the highest government spending rates as a percentage of GDP in SSA. In fact, it reached almost 50 percent of GDP in FY2015/16. Lesotho’s government spending as a percentage of GDP is very high when compared to other countries in SACU, such as Botswana and Swaziland. Lesotho also spends a significant proportion of its budget on capital spending, which could be in line with the lower- middle-income countries’ developmental needs. However, recurrent spending is one of the highest when compared to other SSA countries. Lesotho is also one of the biggest spenders on wages in the world, with a wage bill of 18 percent of GDP (FY2016/17).15 Similarly, Lesotho spent around 10 percent of GDP on Use of Goods and Services (UGS) in 2015, which is also high by regional standards. (figures 7 and 8). 15As seen in figure 7, Lesotho has a very high wage bill in terms of spending as a percentage of GDP. Chapter 3 addresses the details of the wage bill and human resource management (HRM). 13 Figure 7. Government Spending in 2015 Figure 8. UGS in 2015 (percentage of GDP) (percentage of GDP) 70 14 60 12 50 10 40 8 30 6 20 4 10 2 0 0 Angola Botswana Chad Zambia The Gambia Mali Burundi Cabo Verde Swaziland Mauritania South Africa Namibia Lesotho Guinea-Bissau Gabon Liberia Djibouti Zimbabwe Equatorial Guinea Angola Botswana Zambia Comoros Cape Verde Chad Mali Swaziland Burundi Mauritania Namibia Lesotho Guinea-Bissau South Africa Gambia Djibouti Gabon Liberia Equatorial Guinea Sources: World Economic Outlook (WEO) and MoF. Sources: WEO and MoF. Current Expenditure 42. Lesotho’s macro-fiscal challenges stem mostly from high recurrent expenditures, which are linked to improper inter-temporal revenue management. This is not the first time that Lesotho is facing a similar decline in SACU revenues. In FY2009/10 Lesotho’s recurrent spending reached a record high of 50 percent of GDP, an increase of 8.7 percent of GDP from FY2009/10. Although 2.3 percent of the increase was due to increased capital spending, the wage bill rose from 15 percent of GDP to 19.4 percent of GDP — accounting for half of the increase. With the onset of the global financial crisis, SACU revenues declined from 30 percent of GDP in FY2009/10 to 14.4 and 13.3 percent of GDP in FY2010/11 and FY2011/12, respectively, thereby leading to a call for sharp fiscal consolidation. Because of the drop in SACU revenues, the GoL had to cut recurrent spending by 10 percent of GDP in FY2010/11 (figure 4). Lesotho embarked on a fiscal consolidation program, supported by the International Monetary Fund (IMF) Extended Credit Facility (ECF), in response to rising fiscal pressures. It managed to cut expenditures by 11 percent of GDP in 2010.16 During the fiscal consolidation, most of the cuts came through the UGS component (5 percent in FY2010/11) and the wage bill (2 percent). However, capital spending was mostly preserved. The fiscal consolidation efforts were also supported by the World Bank through a series of Development Policy Operations aimed at promoting growth, competitiveness, and public-sector efficiency. The first operation of Social Discount Rate (SDR) 13.4 million (US$20 million equivalent) was approved by the World Bank’s Board of Directors on June 3, 2013. However, as soon as SACU revenues rose again in FY2012/13, expenditures also started to rise again. 43. Although human immunodeficiency virus (HIV) and tuberculosis (TB) prevalence might explain the high share of UGS — as health care materials and services constitute an important portion of the UGS component of the budget — countries that are prone to the same epidemics spend 16See the Implementation Completion and Results Report (ICR) for the First Growth and Competitiveness Development and Policy Grant, World Bank, 2016. 14 less on UGS. Spending on health care materials and services accounted for 3.5 percent of GDP in FY2016/17. This is partly due to the increasing transfers to the referral hospital, Queen Mamahato Memorial Hospital (QMMH), in the last five years (since FY2012/13). According to the World Bank Health PER (2017), the government paid 53 percent of its recurrent budget to outsourced health services in FY2015/16, as reflected in figure 9. High expenditures are partially attributable to the high HIV and TB incidences observed. However, UGS is still high compared to the other countries in the region that are prone to the same epidemics, such as Swaziland. Figure 9. Composition of Recurrent Spending (as Figure 10. UGS (as a percentage of GDP) a percentage of GDP) Wage bill and UGS constitute a large 10% proportion of spending 60% 8% 50% 6% 40% 30% 4% 20% 10% 2% 0% 0% 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 COE UGS Others Interest payments Subsidies Local travel Food and Beverages Grants Social Benefits Rent of buildings Other International travel Utilities Source: MoF. Source: MoF. Note: COE=Compensation of Employees 44. Expenditures pertaining to the governmental vehicle fleet, international and local travel, and subsistence cost more than 2 percent of GDP (figure 10). Due to the unavailability of detailed, disaggregated data, such as number of trips, travel class, days spent on each trip, mileage, and so on, it is not possible to quantify any potential savings. However, anecdotal evidence suggests that civil servants have a high incentive to travel. and do so very often taken given the high rates of travel subsistence provided for missions and training. This leads to both higher spending and a loss of government service productivity, as staff are frequently occupied in training and workshops abroad. Because travel subsistence is provided as a lump-sum payment covering both accommodations and per diems, travelers are not required to reimburse the difference between the actual costs and government-provided subsistence. If the expenditures are less than the reimbursement, the difference is retained by the civil servant. Travel subsistence is denominated in U.S. dollars for all the countries, including SACU countries. Travel provisions have not been revised since 2009, despite the depreciation of the South African rand against the U.S. dollar. Furthermore, provisions are provided at a flat rate across all cities in a country. Table 4 displays the difference between the Lesotho per diems and the sum of World Bank hotel cap rates and per diems (that is, the actual expenditure can be lower for World Bank employees as the actual spending on hotels might 15 differ from the hotel cap rates).17 For example, per diems are almost 50 percent higher than the World Bank rates for Pretoria, South Africa 27 percent higher for Harare, Zimbabwe 30 percent higher for Lusaka, Zambia and 45 percent for Mbabane, Swaziland. The difference is even larger when compared to South Africa.18 Table 2. Comparison of Travel Allowance Between Lesotho and the World Bank Country City Name Difference Between Lesotho and World Bank Rates for selected cities (US$) Botswana Gaborone 40 Botswana Maun 110 Botswana Other or unknown 98 city Brazil Rio de Janeiro 107 Namibia Luderitz 84 Namibia Oshakati-Namibia 87 Namibia Windhoek 55 South Africa Bloemfontein 189 South Africa Cape Town 104 South Africa Durban 89 South Africa East London 177 South Africa Johannesburg 32 South Africa Pretoria 145 Swaziland Manzini 70 Swaziland Mbabane 103 USA Boston, MA 40 USA Washington, DC 73 Zambia Chipata 206 Zambia Lusaka 97 Zambia Mansa 208 Zambia Mfuwe 174 Zimbabwe Bulawayo 173 Zimbabwe Gweru 208 Zimbabwe Harare 95 Sources: MoF and the World Bank. Capital Expenditures 45. Although Lesotho increased the allocation of resources to capital spending after FY2007/08, as the fiscal space narrows, capital spending as a percentage of GDP declines (figure 11 and table 3). Lesotho, a lower-middle-income country with serious development needs, allocates a significant portion of its budget to recurrent spending. However, given the high level of government spending as compared to other SSA countries, capital spending as a percentage of GDP is also high.19 Evidence suggests that when fiscal consolidation is required in Lesotho (and in Swaziland), capital expenditures that are budget financed are subject to cuts, which exacerbates the boom and bust cycles.20 The lower execution of the capital budget compared to the recurrent expenditures also worsens this cycle and lowers public investment quality. As 17 The daily per diem for Lesotho is also expected to cover one-time payments such as airport taxes or visa fees. However, these payments could be reimbursed separately upon submission of the receipts for these expenditures. 18 The per diems for South African government officials can be found at http://www.dpsa.gov.za/dpsa2g/documents/rp/2017/17_10_p_29_01_2017_financial_manual.pdf. 19 A comparative perspective of public spending is discussed in detail in Chapter 4. 20 World Bank (2016) SACU Receipts and Fiscal Policy in Southern Africa, Report no: AUS13682. 16 the fiscal space continues to narrow due to the decline in SACU revenues, allocations to the capital budget decrease (table 3). Figure 11. Composition of Government Spending (as a percentage of GDP) The space for capital spending is narrowing 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% Recurrent Expenditure Capital Spending Source: MoF. II. Expenditure Performance in the Social Sectors 46. Lesotho’s governmental spending is broadly in line with the NSDP and the country’s development needs; the allocation to the social sectors is over 20 percent of GDP, but social outcomes are mixed. Despite the high level of social spending, more specifically, spending on education and health, the outcomes remain poor. Social spending is not very well targeted. However, this Public Expenditure Review will not address the equity of fiscal spending and service delivery in detail because the quality of the last household survey was poor. The upcoming household survey supported by the World Bank is expected to close this information/data gap. Health Sector 47. Lesotho’s total health expenditures stand at 10.6 percent of GDP, which is almost twice the SSA average as a percentage of GDP; however, results remain mixed. According to the Health PER (2017), although private expenditures are only 24 percent of the total, the Government finances 44 percent of the expenditures and the rest is financed by donors. However, the absolute level of health expenditure per capita is lower when compared to the other countries in the region except for Mozambique and Zimbabwe (table 5). The Health PER (2017) notes that HIV incidence in Lesotho has decreased over the past 20 years. In this context, Joint United Nations Programme on HIV/Acquired Immunodeficiency Syndrome (AIDS) (UNAIDS) data indicate that there was a reduction in HIV/AIDS incidence from 5.21 to 2.01 (that is, 2.01 new infections per 100 people) in 2014. However, maternal mortality rates (MMRs) and infant mortality rates (IMRs) are among the highest in the world, with MMRs and IMRs of 1,204 per 100,000 and 59 per 1,000, respectively (figures 12 and 13). The MMR and IMR rates in Lesotho are not consistent with the Millennium Development Goals (MDGs), which aimed to reduce IMRs by two-thirds and MMRs by three- 17 quarters from 2009–2015. Again, according to the Health PER (2017), key primary health service coverage has expanded in Lesotho, and infrastructure has improved at the primary and tertiary health care levels.21 Figure 12. IMR Figure 13. MMR 100 91 91 1,300 1,243 1,200 1,100 1,024 80 -35% 1,000 939 960 +9% 60 59 900 60 57 -17% 800 651 49 700 555 50 600 40 38 +17% 34 500 400 -31% 252 257 300 20 200 83 100 0 0 -67% 2004 2006 2008 2010 2012 2014 2004 2006 2008 2010 2012 2014 Les Zim SA Les Zim SA Source: Health PER, World Bank (2017). Source: Health PER, World Bank (2017). Note: Les= Lesotho; SA= South Africa; and Zim= Zimbabwe. Note: Les= Lesotho; SA= South Africa; and Zim= Zimbabwe. Table 3. Health Expenditures in Lesotho Compared to Other Countries in the Region Health expenditure, public Health expenditure per Health expenditure, public (% of government Health expenditure, total Country capita (current US$) (% of GDP) expenditure) (% of GDP) Lesotho 105.11 8.08 13.08 10.62 Botswana 385.31 3.19 8.84 5.41 Mozambique 42.00 3.94 8.81 6.98 Namibia 499.02 5.36 13.86 8.93 South Africa 570.21 4.24 14.23 8.80 Swaziland 247.90 7.00 16.58 9.25 Zimbabwe 57.71 2.47 8.49 6.44 Lower middle income 90.01 1.64 0.00 4.48 Sub-Saharan Africa 97.67 2.31 0.00 5.46 Source: Lesotho Health PER, World Bank (2017). Education Sector 48. Lesotho allocated almost 10 percent of its GDP to education spending, including the tertiary bursaries/loans in FY2015/16; however, educational outcomes are poor and inequitable. Lesotho’s budget allocation to the social sectors is in line with the NSDP and the country’s developmental needs, but the composition of educational spending does not seem to be poverty reducing. Lesotho allocates 42 percent to primary education, 28 percent to tertiary education, and only 23 percent to secondary education. Access to secondary education, a Sustainable Development Goal target, is much lower for poor children, thereby placing them at a disadvantage in the labor market. The education chapter notes that progression is slow, and the Southern and Eastern Africa Consortium for Monitoring of Education Quality (SACMEQ) scores 21For further analysis on health expenditures, see the Lesotho Health PER, World Bank (2017) and the Lesotho Public Expenditure Tracking Survey (PETS), World Bank (2017). 18 are poor.22 Further, progression and performance are uneven across districts for the highest school grades, and at the highest levels in the school-leaving examinations.23 49. The national bursaries/loans, which accounted for almost 3 percent of GDP between FY2010/11 and FY2011/12, are designed as a revolving fund; however, in practice, only 4 percent is recovered. National manpower bursaries/loans are the main source of funding for tertiary education. They are almost six-fold higher than the subventions/subsidies provided by the Government to the higher education system. Improving the sustainability of the system is also crucial to improving the quality of higher education. Although it is designed to be a revolving fund and a loan scheme, it is de facto a grant scheme as the recovery rate is around 4 percent. These loans are also interest free and repayments are not adjusted for inflation. Therefore, the incentive to pay the loans once the student finds a job is almost nonexistent. In addition, graduates who work in the civil service pay a smaller portion of the loan compared to graduates working in the private sector in Lesotho, thus incentivizing jobs in the public sector. To increase recovery, the Government can consider setting an interest rate on student loans, which could begin accruing after graduation It could create a list of beneficiaries who are in default, sign an agreement with the LRA for data matching, and follow up with beneficiaries by engaging the LRA who are experienced in debt collection. Further, the Government can ask beneficiaries to repay the full loan amount plus the accrued interest, as the current system incentivizes working in the public sector with lower repayment rates required from civil servants in Lesotho. If the Government wants to encourage expertise in some sectors, for example, to address the shortages in medical personnel, then it can set up nonrepayable bursaries for priority sectors. Old Age Pension 50. Launched in 2004, the old age pension (OAP) scheme, an important part of social spending in Lesotho, is provided to citizens ages 70 and older who are not beneficiaries of a larger civil service pension or social grant (with an objective of poverty reduction).24 The scheme is noncontributory, budget-financed, and eligibility is not contingent on means-testing (annex 8). Since the introduction of the scheme in 2004, the real (that is, inflation adjusted) value of its benefit has almost doubled from Maloti (M) 307 to M 580 (US$ 43.5 equivalent) as of March 2017. The number of OAP beneficiaries greatly exceeds the population of persons aged 70 and older. The magnitude of this discrepancy has grown consistently and substantially over time to the extent that an estimated 39.9 percent (or roughly two in five persons) of the 87,110 beneficiaries (as of March 2016) were ineligible for benefits based on the rules of the scheme. Paying the OAP benefits to large numbers of ineligible beneficiaries is expensive. The cost of the OAP scheme—which was initially 1.2 percent of GDP in 2004—has hovered around 1.8 percent of GDP since 2009. It would have trended even higher had GDP grown less rapidly. 51. Eliminating ineligible beneficiaries would reduce the cost of the OAP by an estimated 40 percent (to 1.1 percent as compared to 1.8 percent of GDP). Assuming that reforms can be implemented over a period of five years to accomplish this objective, the projected fiscal cost of the scheme should fall rapidly to roughly 1.0 percent of GDP by 2022, and then decline more gradually thereafter (figure 14). The projected benefit costs are estimated to decline gradually over the coming 25 years due to GDP growth, and as the population of persons ages 70 and older in Lesotho—which fell from 2008 to 2015—is expected to resume growing, but only very slowly (by about 0.5 percentage points, on average, per year). 22 International educational evaluation, the Southern and Eastern Africa Consortium for Monitoring of Education Quality (SACMEQ), 23A more detailed analysis on education is provided in Chapter 4. 24The OAP scheme also covers some 432 former members of the Lesotho Liberation Army (LLA), active during the 1970s, and some 1,692 former members (and their widows) of the African Pioneer Corps for service during the two World Wars. 19 Figure 14. Projected OAP Benefit Costs (percentage of GDP) Source: World Bank staff estimates based on data provided by the authorities. Note: The solid line shows the cost of the OAP scheme if benefits are only paid to persons eligible to receive them. The broken line shows the cost of the OAP scheme under reform measures that eliminate ineligible beneficiaries over the years 2018–2022. III. Implementation of Expenditure Controls and Revenue Management 52. A fragile political environment combined with highly volatile revenues and high expenditures make fiscal discipline hard to maintain; the current macroeconomic situation urgently requires better expenditure controls and revenue management. Intertemporal management of revenues that would decrease Lesotho’s dependence on volatile revenues would support the reduction of the procyclicality of fiscal policy in Lesotho, while maintaining macroeconomic stability. Lesotho’s revenue performance depends mostly on SACU revenues that are driven by South Africa’s economic performance, thereby amplifying the boom and bust cycles in Lesotho. Expenditure controls, better revenue management, combined with better targeted social protection programs, would protect the poor and the vulnerable from economic downturns. At the same time, it would keep Lesotho away from a low growth-high poverty equilibrium from which it would be difficult to rebound. 53. Avoiding the macro fiscal volatility can also create welfare gains. The rationale behind this finding is that the Government has a comparative advantage in smoothing shocks through the accumulation of precautionary buffers and the conduct of a counter-cyclical fiscal policy. However, evidence suggests that the expenditure controls are more likely to be functional where a strong institutional arrangement exits for example, a fiscal responsibility law and strong political commitment. Honda et al. (2017) stress that fiscal objectives should be addressed transparently, and a fiscal variable should be targeted with a clear link between the numerical target and the fiscal objective. In addition, mechanisms should be built in to deal with exceptional circumstances and ex-post deviations. Further, there should be a clear statutory basis with effective monitoring and accountability arrangements. 54. With overall government spending at 50 percent of GDP, most of the adjustments should come from the expenditure side — and preferably from recurrent expenditure. The Government spending accounts for half of the economy in Lesotho. As a result, fiscal consolidation will come at the cost of lower growth in the near term, but with longer-term benefits for macroeconomic stability. Although a more gradual adjustment was possible with the FY2015/16 and FY2016/17 budgets, Lesotho missed this opportunity. The budget deficit hit 7 percent of GDP in FY2016/17, which led to a rapid decrease in the GoL deposits at the Central Bank. Reserve coverage decreased from 6.1 months of imports in FY2015/16 20 to 4.2 months of imports at the end of FY2016/17 (figure 15). With the current projected budget deficit of 4.3 percent of GDP as of the end of FY2017/18, the reserve coverage is projected to decline below 3 months of imports by the end of FY2019/20, thereby placing the exchange rate under pressure. In future, Lesotho can either finance its deficits by drawing down its reserves and/or through a composition of domestic and external borrowing. However, Lesotho, having run its reserves down to 4.2 months of imports, does not have much room for debt financing either. Lesotho remains at moderate risk for debt distress, with a debt- to-GDP ratio of 35.4 percent in FY2016/17. It is also highly exposed to fluctuations in the South African rand, as 92 percent of the portfolio is composed of external debt (figure 16). Lesotho’s external borrowing is mostly limited to concessional financing and is linked to capital projects. Although the authorities believe that there is a market appetite for more domestic borrowing, domestic markets remain limited. In this context, the markets would demand assurances about the fiscal sustainability for domestic financing. Figure 15. Gross International Reserves (US $ Figure 16. Public Debt, External and Domestic (as a million and months of imports) percentage of GDP) 1200 7 50% 6 45% 1000 40% 5 800 35% 4 30% 600 3 25% 400 20% 2 200 15% 1 10% 0 0 5% 0% Gross international reserves (US$ million, LHS) Gross international reserves (months of External debt Domestic debt imports, RHS) total Source: IMF Source: IMF. 55. Despite the high spending levels, Lesotho has not achieved the macro-fiscal targets set out in the NSDP. The ambitious targets set in the NSDP have not been achieved for 2016, and are unlikely to be achieved by 2020. There needs to be a clear link between the NSDP and MTFF. However, the MTFF set out in the NDSP was based on an overly optimistic MTFF, with extremely high projected growth rates (table 6). In addition, although the NSDP has set some targets to be achieved by 2016, the mechanisms were neither identified nor were they put in place to reach these targets. A macro-fiscal framework needs to adhere to a consistent, credible, and a sustainable framework. Indeed, it should also guide the evaluation process, which is lacking in Lesotho. Table 4. NSDP Targets Set for 2016 and 2020, and 2016 Outcomes NSDP Targets Outcome 2016 Target 2016 Target 2020 GDP growth rate (%) 2.3 7 7 Consumer price index 8.5 6 6 Share of secondary sector (% of GDP) 26.0 70 70 Agriculture GDP growth rate (%) ‒7.2 6 5 Investment (% of GDP) 28.5 45 45 21 Total debt (% of GDP) 47.8 40 40 Percentage of population living below the poverty line 57.0 15 10 Reserves (months of import coverage) 4.2 6 6 Sources: NSDP, World Development Indicators (WDI), IMF, Lesotho Bureau of Statistics (BOS), and World Bank staff calculations. IV. Public Financial Management 56. Effective and transparent Public Financial Management (PFM) is a necessary condition for a better macro-fiscal management which is lacking in Lesotho (Im et al 2015). Strong PFM systems and institutions are crucial to macroeconomic management, fiscal policy, and budgeting, especially for a country with volatile revenues and fixed exchange rates. Although Lesotho has made some progress in PFM, the reform progress has been extremely slow. The integrated financial management information system (IFMIS) was introduced in FY2009/10, however, it is still not properly used for accounting for all central budget expenditures and revenues. There are considerable differences between IFMIS data and the Medium-Term Expenditure Framework (MTEF) both in revenues and expenditures. Government officials note that IFMIS data changes on a daily basis, even looking several years back. This is because the year- end closures are not done, and the data can change retroactively. While the main difference in expenditures is due to unrecorded donor spending, the revenue data is also underreported. This raises important concerns for fiscal transparency, and it also creates issues for proper macro-fiscal management. Similarly, Lesotho does not have a functional treasury single account (TSA). Month-end closures and bank reconciliations are not yet fully functional. As a result, Lesotho does not have a functioning cash management system. 57. The Public Expenditure and Financial Assessment (PEFA) (2016) finds several inadequacies in the PFM systems in Lesotho, and there have not been any significant improvements since the 2012 assessment. PEFA (2016) deems compliance and fiscal discipline at the component level to be inadequate. It notes that significant changes are needed to improve financial management systems that support the accounting and reporting functions. According to the Assessment, there are a host of outstanding concerns, including: (i) spending controls are deficient; (ii) reporting on budget outturns is late; (iii) timely accounting and reporting are problematic; (iv) annual financial statements are late and of low quality; (v) central control over the cash function is not working well; (vi) internal controls for payroll and procurement are inadequate; (vii) external controls are hampered due to late submission of financial statements; and (viii) the Auditor General’s reports are delayed. The Lesotho Development Policy Lending operation included an indicator on public accounts. The Implementation Completion and Results (ICR) report (2015) noted significant delays in the publishing of the annual audit report. The delay for FY 2013/14 was 24 months, and for the 2014/15 accounts the delay was 12 months. 58. The connection between policy priorities and budgeting and strategic allocation of resources is weak. Table 6 shows the targets set for 2016 in the NSDP, which have mostly not been met. Further, they are unlikely to be met by 2020. According to the PEFA (2016), although the upstream processes of budget formulation and processing are performed well, policy objectives are either inadequate or undeveloped. As such, they do not have the necessary impact on allocations in the downstream process in terms of in-year budget execution. Most budget allocations remain incremental, and sharp cuts are necessitated. This then leads to policy inconsistencies. Similarly, the current PFM system in Lesotho hampers efficient service delivery because the budget does not have a strong policy or strategic focus, and controls are inadequate. The budget process is ineffective because budget information is not transparent and comprehensive. This hampers the ability of the citizenry to follow the policies, allocations and service delivery. Policy-based budgeting with a multiyear perspective is also lacking, and governmental procurement systems remain inefficient. In this regard, both its performance and controls are deficient. 22 59. As part of the public expenditure review exercise, the World Bank introduced the BOOST tool to the GoL. Figure 17 displays the five-step BOOST delivery model used to enhance budget transparency. The tool enables the government to classify the IFMIS data by analyzing and identifying the spending trends according to administrative unit, functions or economic classifications (which are not yet available for Lesotho). To increase the transparency of fiscal spending, the GoL can consider publishing the data on the open budgets portal.25 Figure 17. BOOST 1. Country 2. BOOST 3. Applying 4. Sustaining 5. Budget engagement construction BOOST BOOST Transparency 60. In summary, there remain significant PFM system issues with regard to budget credibility and resource allocation, fiscal discipline, transparency, financial accountability, and the inability to ensure good service delivery. Furthermore, the management of fiscal risks from SOEs, subnational governments, public-private partnerships (PPPs), and other contingent liabilities are not calculated and addressed in the budgeting process. The SOE Policy Note (2016) documents that some wholly-owned SOEs were not able to pay the on-lent loans. This has in turn increased the pressures on the budget as these risks were not accounted for. Similarly, the budget analysis does not include a sensitivity analysis about the public debt, although most public debt is denominated in foreign currency. Table 7. Fiscal Risk Monitoring PI-10 Dimension Score Justification for 2016 score 2016 10 Fiscal Risk Reporting D 10.1 Monitoring of Public Corporations D The MoF’s Private Sector Development (PSD) does not currently receive any financial reports from Public Corporations. However, the majority of SOEs submit reports to their line ministries. Fiscal risks are not monitored. 10.2 Monitoring of Sub-National Governments D There is no systematic monitoring of SNG (SNG) fiscal positions and audits have been late due to late submissions of financial statements. 10.3 Contingent Liabilities and other Fiscal D Contingent liabilities are disclosed in a Risks limited way that leaves uncertainty about the volume of these risks, particularly contingent liabilities arising out of PPPs. Source: PEFA (2016). 61. Lesotho does not have TSA, and is unable to keep track of accounts that are outside of the treasury system. The GoL has four main treasury accounts, including a: (i) Main Revenue Account, (ii) Recurrent Expenditure Account, (iii) Trust Account, and (iv) Capital Account. In addition to these four main accounts, there are 254 government bank accounts outside the treasury system that were consolidated from other accounts. Most of these accounts were used for donor-funded projects, but were not closed after project completion/closure. Some accounts serve other purposes. According to the PEFA (2016), the 25 World Bank, Open Budgets Portal. See http://wbi.worldbank.org/boost/boost-initiative 23 consolidation of the treasury accounts is done monthly. However, the other 254 accounts remain outside the system — without any reconciliation. 62. The lack of a TSA and the attendant lack of knowledge of the amount of available reserves for short-term cash management renders monetary policy difficult. Understanding the availability of earmarked reserves is crucial to understanding the fiscal risks, both to manage the short-term cash balances and reserve management. Despite the availability of cash balances, the lack of proper cash management means that the Government has had to borrow from financial markets. According to the IMF (2010) a TSA lowers liquidity reserve needs and can prevent the overuse of Treasury-bills. It also reduces the volatility of cash flows from the treasury, allowing for a lower cash reserve/buffer in meeting unexpected volatility, which would in turn enhance reserve management. The availability of timely and accurate cash/reserve information is extremely crucial, as Lesotho has a fixed exchange rate regime and needs to maintain enough reserves to support it. 63. Introducing a TSA would decrease the interest rate-related costs and help maintain financial integrity; it would also minimize other transaction costs related to budget execution and bank fees . According to the IMF (2010), a TSA would contribute to minimizing transaction costs during budget execution through controlling the delay in the remittance of government revenues by collecting banks and making rapid payments of government expenses. It would also facilitate reconciliation between banking and accounting data, and lead to more efficient control by monitoring funds allocated to the various government agencies. Further, the daily monitoring of funds allocated to various government agencies would decrease risks to financial integrity by enhancing transparency in the usage of available funds. With daily and proper usage of the IFMIS, timely information on government cash resources would be available in real time. 64. Lesotho’s macro-fiscal challenges stem from intertemporal revenue mismanagement that has led to high, procyclical government spending. SACU revenues, which are extremely volatile, constitute an important portion of government revenues. This has enabled Lesotho to adjust its spending, especially current spending, upwards when revenues are high. However, this also makes it extremely difficult to adjust current spending downwards, which is usually downward sticky. It can also lead to inefficient cuts in capital spending and exacerbate boom-bust cycles. 65. Lesotho allocates a large amount of spending in UGS when compared to other SSA countries. The GoL spends more than 2 percent of GDP on vehicle-related expenditures, as well as international and local subsistence for travel. Targeted cuts in this component have the potential to produce savings, while having limited impact on growth and poverty alleviation. Revisiting the regulations on travel allowances and denominating the travel allowances in the SACU region in Rand can generate fiscal savings. At the same time, it can also reduce incentives for unnecessary travel and contribute to better civil service productivity. Similarly, Lesotho spends a very large amount on vehicle-related expenditures. Taking stock of the vehicles and entitlements could increase fiscal transparency while also leading to an identification of cost-cutting measures. 66. The cyclical nature of government spending urgently requires better expenditure controls and revenue management. Implementation of measures that would limit Lesotho’s dependence on volatile SACU revenues that are flexible enough to preserve the efficiency of capital spending, can increase the resilience of the economy to external shocks through the accumulation of precautionary buffers. Expenditure controls and better revenue management that is supported by strong institutional arrangements, for example, a fiscal responsibility law could also limit the political pressures that lead to higher and inefficient spending. 24 67. Sustaining macroeconomic stability can only be successful with effective and transparent PFM systems, which Lesotho currently lacks. The PFM reform program in Lesotho aims to improve budget credibility, procurement and internal controls, cash management, the reliability of the IFMIS information and reporting, and auditing. All are key to improving spending efficiency and reducing risks to financial integrity. However, progress in this area has been extremely slow. The World Bank is also supporting PFM reforms, which focus on improving the reliability of IFMIS information for better cash management and financial reporting. A continued focus and strong leadership at the MoF and Cabinet are essential to ensuring the achievement of project development objectives. In particular, a call for regular publication of in-year budget information from the IFMIS can assist in improving its reliability. 25 Chapter 2: Taxation 68. Total revenues in Lesotho have been historically high and extremely volatile due to its dependence on receipts from the SACU. Total intake from both tax and non-tax revenues have been high, and reached almost 60 percent of GDP in 2012/13. However, they have been on a decline since then due to a drop in SACU revenues. A tax collection and public expenditure comparison shows the unfavorable developments since 2010 when public expenditures dramatically rose above total revenues from all sources (figure 18). Tax collection has been relatively stable and stands above the average of SSA peers for the same income group. Nonetheless, it has been lagging tax efforts in other SACU comparators. 69. Simplifying the tax policy and administration procedures can enhance revenue performance. Lesotho has been engaged in comprehensive tax policy and administration reforms since the mid-2000s. However, outcomes of the revenue performance reforms have been mixed. The complexity of tax policy and administration procedures—combined with ambiguity in the provisions of tax benefits—continues and makes the cost of collection and compliance high. At the same time, it causes a negative impact on the business environment and formality. Tax incentives, embedded in all major direct income and indirect consumption taxes (notably the value-added tax [VAT]), tend to induce significant revenue losses. Efforts to simplify the fiscal incentives and structure of the main taxes would significantly increase the efficiency of the tax administration, as well as tax intake. The analysis indicates that by reaching the level of the highest productivity in VAT and Corporate Income Tax (CIT) within SACU, Lesotho would expect to gain an additional 2-3 percentage points of GDP from the respective taxes. In addition, the enactment of the new bill on alcohol and tobacco—with the uptake in rates—would raise almost 1 additional percentage point of GDP (Marquez and others 2017). Figure 18. Tax, Revenue Collection, and Total Public Expenditures, 2004–2017 70% 60% 50% 40% 30% 20% 10% 0% 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 Public Expenditure total Tax Revenue Total Tax Revenue + SACU + Other revenue Source: Lesotho Medium-Term Expenditure Framework (MTEF) and World Bank staff calculations 70. To compensate for the decline in SACU revenues over the medium term, the Government has been focusing on furthering the reforms of its tax system. The Mining Tax Bill and the Bill on Tobacco/Alcohol Taxation have been revised and submitted to the Parliament. Other tax policy measures— such as the unification of the rate structure, the rationalization of the exemptions in VAT, and the streamlining of the income tax incentives — are to be explored in tandem with continued tax administration reforms. All reform measures will contribute to enhancing both revenue and efficiency. 71. The chapter provides an analysis of Lesotho’s tax system and highlights the main issues in tax policy design and administration, with implications for further reforms over the short to medium term. It focuses on the major tax policies, specifically VAT, taxation of tobacco and alcohol, and income 26 taxes, which collectively account for 96 percent of the total tax collection. To the extent that these tax regimes are appropriately restructured, they will exert the most tangible impact on the overall collection and efficiency of the entire revenue system. The mechanics of SACU is outside the scope of this chapter’s analysis. As the saying goes—tax administration is tax policy in most tax systems (Casanegra de Jantscher 1990). The tax policy analysis within the scope of this chapter is complemented by a brief assessment of the ongoing revenue administration reforms, as well as the long-standing challenges that need to be overcome. 72. The chapter is organized as follows: Section I touches on revenue performance. Section II presents a brief overview of the fiscal revenue and tax performance. Section III analyzes the structure of major tax policies, including: VAT, excise taxes on tobacco and alcohol, and the income taxes. Section IV reviews and assesses the administrative aspect of taxation. I. Revenue Performance 73. To ensure the consistency of analysis within the scope of the chapter, and because of major discrepancies between the IFMIS and the MTEF, the data from the MTFF was used for this analysis. It was supplemented with the administrative data from the LRA. Annex 1 presents the composition of revenues from these two different sources. While major taxes are included in both sources, grants from foreign governments are included in the MTFF, but excluded from the IFMIS (annex 1). 74. Lesotho’s revenue consists of four main components: taxes, grants, transfers from SACU, 26 and other sources.27 Revenues from SACU, combined with taxes, predominate. Box 1 summarizes the SACU revenue-sharing mechanism. Box 1. SACU Revenue-sharing Formula The SACU is a customs union among five countries of Southern Africa: Botswana, Lesotho, Namibia, South Africa, and Swaziland. It was established in 1910 and is the oldest customs union in the world. SACU countries pool their customs and excise revenues and distribute them according to a revenue-sharing formula described in the 2002 Agreement which has been in place since 2005 (Cuevas 2015). As the lower-middle-income countries in the customs union, Lesotho and Swaziland receive large portions of their government revenues in the form of transfers from the SACU. The revenue share accruing to each member state is calculated from three basic components: a share of the customs pool, a share of the excise pool, and a share of a development component. These three different components are distributed as follows. The customs component is allocated according to each country’s share of to tal intra-SACU trade, including reexports. The excise component, net of the development component, is allocated on the basis of the country’s share of SACU GDP. The development component, fixed at 15 percent of the total excise pool, is distributed to all SACU members according to the inverse of each country’s per capita GDP. Thus, the revenue-sharing formula is as follows: 𝑀𝑖𝑖𝑛𝑡 𝐺𝐷𝑃𝑖 𝐺𝐷𝑃𝑖𝑃𝐶 1 1 𝑅𝑖 = 5 𝑖𝑛𝑡 𝑥 𝐶 + 5 𝑥 𝐸 𝑥 0.85 + [1 −𝑃𝐶 𝑥 10] 𝑥 5 𝐸 𝑥 0.15 ∑𝑖 𝑀𝑖 ∑𝑖 𝐺𝐷𝑃𝑖 𝐺𝐷𝑃𝑖 ∑5 𝑖 5 where Ri is total revenue received by country i, C, and E are the total customs and excise revenue collection in the pool, GDPi is the country i GDP, GDPi PC is country i’s GDP per capita (Cuevas 2015). Source: 26 SACU’s five member states maintain a common external tariff, share customs revenues, and coordinate policies and decision making on a wide range of trade issues. 27 Other sources of revenue include items such as interest, dividends, rents, and diamond and water royalties. 27 Note: Total revenue intake has been high —and largely dependent on SACU revenues. The revenue pattern as shown in table 1 reached its peak at more than 58 percent of GDP in 2012/13. While tax collection has been stable at 20–21 percent of GDP during the period of 2011/12–2015/16, SACU revenues fell sharply from more than 26 percent of GDP in 2012/13 to less than 21 percent of GDP in 2015/16. The grant component has decreased as well, from about 7 percent of GDP in 2011/12 –2012/13 to less than 2 percent of GDP in 2014/15. It then rose to just 3 percent of GDP in the following year. Revenue projections for 2016/17 and 2017/18 show that total revenue continues to decrease. This trend is being driven by the expected receipts from SACU. During the same period, tax collection is projected to rise, and partially compensate for the sharp reduction in SACU receipts. Table 5: Lesotho: General Government Revenues, FY2011/12 to FY2017/18 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 2017/18 Act. Act. Act. Act. Act. Budget Budget In Maloti Revenue 9,627,185 13,144,907 13,274,487 14,582,544 15,322,029 14,346,355 16,869,641 1. Tax Revenue 4,282,813 4,596,303 5,147,646 5,785,431 6,600,236 7,392,700 8,256,947 Income Tax 2,395,012 2,348,631 2,680,218 3,058,589 3,643,701 3,891,270 4,373,121 Personal 1,469,101 1,445,721 1,563,319 1,651,905 1,701,595 1,905,721 2,144,738 Corporate 555,774 453,121 665,262 878,079 1,121,628 1,152,573 1,268,722 Unallocable 370,137 449,789 451,637 528,604 820,479 832,975 959,661 Taxes on Property 125,074 149,830 127,700 139,000 - 347,620 196,455 Taxes on goods and services 1,589,924 1,818,773 1,886,558 2,343,625 2,705,333 2,845,214 3,312,381 Value added tax 1,369,552 1,640,425 1,732,300 2,116,794 2,210,184 2,323,396 2,578,686 Excise taxes 200,605 163,324 132,601 206,661 469,775 495,610 710,583 Others 19,767 15,024 21,657 20,170 25,373 26,208 23,112 Import and export duties 151,424 277,408 451,129 242,738 250,044 307,394 373,482 Stamp taxes 21,378 1,661 2,041 1,479 1,159 1,203 1,508 2. Grants 1,436,781 1,703,459 1,047,460 495,700 966,114 975,500 1,072,100 Recurrent/Budget Support 289,928 196,400 149,100 4,200 14,514 4,876 - Capital 1,146,853 1,507,059 898,360 491,500 951,600 970,625 1,072,100 3. Other revenue 1,154,942 878,819 1,024,831 1,267,344 1,357,059 1,459,244 1,386,396 4. SACU 2,752,650 5,966,327 6,054,551 7,034,069 6,398,620 4,518,911 6,154,198 In percentage of GDP Revenue 46.5 58.2 52.7 52.1 50.0 42.2 49.6 1. Tax Revenue 20.7 20.3 20.4 20.7 21.5 21.8 24.3 Income Tax 11.6 10.4 10.6 10.9 11.9 11.4 12.9 Personal 7.1 6.4 6.2 5.9 5.6 5.6 6.3 Corporate 2.7 2.0 2.6 3.1 3.7 3.4 3.7 Unallocable 1.8 2.0 1.8 1.9 2.7 2.5 2.8 Taxes on Property 0.6 0.7 0.5 0.5 - 1.0 0.6 Taxes on goods and services 7.7 8.0 7.5 8.4 8.8 8.4 9.7 Value added tax 6.6 7.3 6.9 7.6 7.2 6.8 7.6 Excise taxes 1.0 0.7 0.5 0.7 1.5 1.5 2.1 Others 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Import and export duties 0.7 1.2 1.8 0.9 0.8 0.9 1.1 Stamp taxes 0.1 0.0 0.0 0.0 0.0 0.0 0.0 2. Grants 6.9 7.5 4.2 1.8 3.2 2.9 3.2 Recurrent/Budget Support 1.4 0.9 0.6 0.0 0.0 0.0 - Capital 5.5 6.7 3.6 1.8 3.1 2.9 3.2 3. Other revenue 5.6 3.9 4.1 4.5 4.4 4.3 4.1 4. SACU 13.3 26.4 24.0 25.2 20.9 13.3 18.1 Memorandum Items Act. Act. Act. Act. Act. Est. Est. Nominal GDP (in Maloti) 20,699,363 22,603,906 25,177,659 27,967,984 30,650,855 33,988,964 33,989,964 Sources: GoL, MTEF, and author’s calculations. 75. Tax collection is relatively high compared to a broad set of select countries in SSA. However, it is the lowest among the SACU members. Figure 19 confirms the established observation: countries with 28 higher income per capita tend to collect more in taxes.28 Almost all SSA countries raised 10–20 percent of their GDP in taxes in 2014. Lesotho stayed above the trend line, indicating the country’s better tax performance compared with other countries with a similar level of GDP per capita. Table 8 presents a more detailed snapshot of comparative tax and revenue performance in selected SSA countries. It shows that in 2014/15, Lesotho collected almost 21 percent of GDP from taxes and 52 percent from total revenues. This is indicative of the country’s better tax and revenue effort as compared to the average in the same group of SSA lower-middle-income countries (15 percent and 27 percent, respectively). In a direct comparison between Lesotho and the other SACU countries, however, it is evident that Lesotho collects the highest revenue as a share of GDP, but has the lowest intakes from taxes (last panel in table 8). Figure 19. Tax Mobilization in SSA, FY2014/15 a AGO NAM 30 SYC SWZ ZWE BWA MOZ ZAF Tax Revenue (% of GDP) LSO 20 LBR SEN MRT MUS GIN CPV MWI KEN NER GMB BEN CIV GHA TGO BFA ZMB CMR MLIRWA STP GAB ETH UGA COG GNQ BDI TZA COM TCD MDG 10 ZAR ERI SLE NGA GNB SDN CAF 0 5 6 7 8 9 10 Ln GDP per capita (Constant 2010 US$) Sources: MTFF, WEO, WDI, and author’s calculations. Note: a. GDP per capita (Constant 2010 U.S. dollar) in 2014 is used for all countries, except Eritrea (2011 data are used). 28The observation is well-established in tax analysis literature. For an example, see Organisation for Economic Co-operation and Development (OECD) 2016: Revenue Statistics in Africa. The report looks at the pattern of tax-to-GDP ratio and GDP per capita of OECD countries as compared with Latin American and Caribbean countries, as well as African countries. It shows that tax-to- GDP ratios tend to be higher in high-income countries. 29 Within table 8, make following change: insert hyphens upper-middle-income; lower-middle-income. Table 6. General Government Revenues (percentage of GDP), FY2014/15 *GNI pc Total Tax Excise Property Selected Countries No. of countries in group Range in US$ VAT Trade CIT PIT + SSC Revenue Revenue Tax Tax 1/ High Income 79 > 13160 26.64 13.95 3.34 1.18 0.33 0.85 1.52 7.47 Upper middle income 56 4030 - 17360 30.29 16.89 3.49 0.89 1.64 0.37 1.86 2.83 Lower middle income 52 1020 - 4380 27.15 15.19 2.19 0.98 0.96 0.06 1.33 1.54 Low income 31 < 1040 19.46 12.48 0.79 0.45 0.51 0.01 0.29 0.39 of Which, SACU: Botswana Upper middle income 7030 40.42 25.19 4.02 - 11.06 0.04 5.28 5.91 Namibia Upper middle income 5680 34.25 32.54 7.30 0.11 12.80 0.20 5.00 7.20 Lesotho Lower middle income 1470 52.14 20.69 7.57 0.74 0.87 0.50 3.14 5.91 South Africa Upper middle income 6810 28.24 24.02 7.04 2.50 1.05 1.30 5.17 8.74 C:\Users\wb479719\AppData\Local\Microsoft\Windows\INetCache\Content.Outlook\P151AD6U\Data for Tax Lesotho_050217.xlsx Sources: WDI, IMF Fiscal Affairs Department (Revenue Mobilization), IMF Article IV, MTFF, 2015 Tax Statistics (South Africa), and author’s estimates and calculations. Note: GNI= Gross National Income; PIT = Personal Income Tax; SSC= Social Security Contributions; VAT= value-added tax. *1/ Per capita gross national income measured according to the Atlas Method. WDI *32.23 percent of GDP in 2014, of which oil revenues are 23.4 percent of GDP and non-oil revenues are 8.9 percent of GDP. 76. The composition of domestic tax collection in Lesotho indicates that income and consumption taxes dominate (table9). In 2015/16 (the latest year for which actual revenue data are available), the collection from income taxes accounted for more than half of the total taxes or 55 percent, and consumption taxes generated more than one-third of total taxes or 41 percent. The combination of these main taxes accounted for 96 percent of total tax revenues. The major direct income taxes are the Personal Income Tax (PIT) and Corporate Income Tax (CIT). With regard to taxes on goods and services, the VAT accounts for the largest share. It contributes 34 percent to the total taxes or 85 percent to the tax intake from goods and services. Also, excise taxes account for more than 7 percent of total taxes, or 17 percent of the taxes on goods and services. It should be noted that the high collection of VAT is partly attributable to the high consumption share of GDP, as well as to the high share of the government UGS. Household consumption alone accounts for approximately 82 percent of GDP (2014/15). By comparison, the lowest and the second highest share in SACU is less than 46 percent in Botswana and 74 percent in Swaziland, respectively. Regarding the government UGS, Lesotho attains the highest share of GDP compared with other SSA countries (almost 13 percent of GDP in 2015/16). 30 Table 7: Lesotho: Tax Revenue, FY2011/12 to FY2017/18 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 2017/18 Act. Act. Act. Act. Act. Budget Budget In millions of Maloti Tax Revenue 4,283 4,596 5,148 5,785 6,600 7,393 8,257 Income Tax 2,395 2,349 2,680 3,059 3,644 3,891 4,373 Personal 1,469 1,446 1,563 1,652 1,702 1,906 2,145 Corporate 556 453 665 878 1,122 1,153 1,269 Unallocable 370 450 452 529 820 833 960 Taxes on Property 125 150 128 139 - 348 196 Taxes on goods and services 1,590 1,819 1,887 2,344 2,705 2,845 3,312 Value added tax 1,370 1,640 1,732 2,117 2,210 2,323 2,579 Excise taxes 201 163 133 207 470 496 711 Others 20 15 22 20 25 26 23 Import and export duties 151 277 451 243 250 307 373 Stamp taxes 21 2 2 1 1 1 2 In percentage of Total Tax Revenue Tax Revenue 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Income Tax 55.9 51.1 52.1 52.9 55.2 52.6 53.0 Personal 34.3 31.5 30.4 28.6 25.8 25.8 26.0 Corporate 13.0 9.9 12.9 15.2 17.0 15.6 15.4 Unallocable 8.6 9.8 8.8 9.1 12.4 11.3 11.6 Taxes on Property 2.9 3.3 2.5 2.4 - 4.7 2.4 Taxes on goods and services 37.1 39.6 36.6 40.5 41.0 38.5 40.1 Value added tax 32.0 35.7 33.7 36.6 33.5 31.4 31.2 Excise taxes 4.7 3.6 2.6 3.6 7.1 6.7 8.6 Others 0.5 0.3 0.4 0.3 0.4 0.4 0.3 Import and export duties 3.5 6.0 8.8 4.2 3.8 4.2 4.5 Stamp taxes 0.5 0.0 0.0 0.0 0.0 0.0 0.0 Sources: GoL MTEF, and authors’ calculations. II. Tax Policy Issues 77. This section analyzes the design of major direct and indirect taxes and presents a broad outline of policy implications for further reforms. Value Added Tax 78. The VAT Act No. 9 of 2001 and the Amendment Act No. 6 of 2003 govern the implementation of VAT. 79. Policy determinants of the VAT base. From the policy perspective, the VAT base is determined by both the threshold level and exemptions. With regard to the VAT threshold, if business turnover exceeds M 850,000.00 per year (equivalent to US$65,000), it must register for the VAT. For businesses with turnover below this threshold, registration is voluntary. The Minister of Finance can change the level of threshold by a notice published in the Government Gazette. The tax is applied to all goods and services, except for those that are exempt or for businesses with turnover below the threshold. Although the established threshold is comparable to regional and international standards, the list of exemptions is excessive. Some are embedded under the broad categories of the so-called ‘standard exemption’ in 31 international practice. The following groups are entitled to VAT-exempt commodities: education services, financial services, medical and dental services, public transportation, postal offices, unimproved land, selected types of leases of immovable property, water, specified imports from South Africa, and others (for example, exemptions related to amateur sporting organizations). In addition, several other VAT exemptions are available for diplomats, contractors, and certain organizations. 80. Complexity of the rate structure. There are four VAT rates. The standard rate is 14 percent, slightly lower than the regional average of 15 percent. Three other rates are also applicable: 15 percent on all alcoholic beverages and tobacco; 5 percent on electricity and telephone services; and 0 percent on exports (as conforming to a destination VAT) and selected goods and services. Section 6A of the VAT (Amendment) Act No.6 of 2003 provides that the following non-export commodities are granted a zero rating: (a) agricultural inputs—fertilizers, seeds, and pesticides; (b) beans; (c) breads; (d) lentils; (e) livestock and poultry feed; (f) maize (grain); (g) maize meal; (h) milk; (i) paraffin intended for use as fuel for cooking, illumination or heating; (j) peas; (k) sorghum meal; (l) unmalted sorghum grain; (m) wheat (grain); and (n) wheat flour. As such, the current rate structure is overly complex. It is driven by two concerns: first, how to relieve the VAT burden on the poor (equity consideration in the VAT design), and second, how to collect more from ‘sin’ goods (alcohol and tobacco). Neither of these concerns is deemed to be well-targeted in the current setting of the VAT policy. 81. The revenue and equity impact of reduced statutory VAT rates, zero rating of non-exports, and exemptions. Cross-country experience indicates that the reduced statutory rates of VAT are a poor income redistribution tool (IMF 2001). The reduced rate on electricity or telephone services does not tend to provide any/ targeted relief to the poor. To the contrary, it can exacerbate inequities. In the case of electricity, (Mokoatle 2016) highlights that “the reduced VAT rate on electricity does not bring any benefit to the rural households as their incomes are too low to be ignited by low prices resulting from low VAT rate. The only beneficiaries to the policy are the mostly well-off urban households and the government ministries�. Multiple rates also increase the risk of noncompliance and tax evasion, and raise the cost of administration.29 Similarly, the impact of exemptions on the poor may not be evident. Exemptions can provide some tax relief only if they are applied exactly on the final value. However, in the case of an exemption being applied to any intermediary stage of the production-distribution chain, the regime becomes inefficient: total revenue rises due to cascading (that is, the VAT would mimic a turnover tax). By comparison, zero rating of non-export commodities presents the most severe form of complexity and creates an opportunity for abuse and evasion. Unlike exemptions, which would break the VAT chain at the exemption stage, zero rating would trigger the need for taxpayers to file for refunds with the tax administration (LRA) to assess and audit the refund claims. The benefit to final consumers (particularly the poor—as intended by the design of the zero-rate regime) is not always clear-cut: if this special regime is applied to any stage of the production-distribution chain before final sales, the net impact of the effective VAT burden on the final consumers will be exactly the same as if no zero rating were being granted. 82. The VAT regime can potentially result in significant revenue losses. The current setting of the VAT regime is riddled with multiple rates, exemptions. The non-export zero rating is prone to abuse and amounts to a revenue loss. The results of the estimates of the fiscal loss due to VAT expenditures are shown in table 10. The revenue loss is estimated for the three types of VAT expenditures for the fourth quarter of FY2015/16: that is, a zero rating of non-exports, exemptions, and a reduced rating. 83. The mechanism for estimation is simple. First, the full VAT expected to be collected at the standard rate of 14 percent is estimated by multiplying that rate with the value of the final sales. The data on final sales are provided by the LRA authorities and represent the differential between the gross sales and the 29According to a recent LRA study, 6 percent of the total VAT audits in FY2011/12 were caused by taxpayers applying the wrong rate. 32 deemed intermediate sales. Second, the expected VAT collection at the statutory preferential rates or due to exemption is similarly calculated. The difference between the expected full collection at the standard rate and the one with the exemption or at reduced rates shows the estimate of the revenue loss. The outcomes of the estimation can be interpreted as the upper bound of the extent of the revenue loss. 84. The estimation is based on the implicit assumption that the exemption was to be applied exactly at the point of final sales. As such, it does not consider the cascading effect—if in fact, certain exemptions are applied, unintentionally, to the middle stage(s) of the production-distribution chain. The value of zero rating supplies of the M 416.9 million would have generated a VAT of M 58.4 million, if they were taxable at 14 percent. Similarly, the value of exempt supplies at M 108.5 million would lead to a loss of M 15.2 million. A 5 percent rate was granted to telecommunications services and electricity supply valued at M 342.34 million. This would amount to a potential revenue loss of M 30.8 million, compared to the scenario whereby the standard rate of 14 percent was applied. The total revenue loss is significant: M 104.4 million, or about 19 percent of the total VAT collected during the same period. Table 8: Tax Expenditures for Quarter 4 of FY2015/16 (in Maloti) Sales Values of VAT at 14 VAT at Revenue Loss Sales Percent Reduced % of VAT % of Tax Rate In Maloti Collection a Revenue b Zero-rated 416,905,749.84 58,366,804.98 — 58,366,804.98 10.56 3.54 sales Exemptions 108,517,174.23 15,192,404.39 — 15,192,404.39 2.75 0.92 Five percent 342,336,270.18 47,927,077.83 17,116,813.51 30,810,264.32 5.58 1.87 sales Total 867,759,194.25 121,486,287.20 17,116,813.51 104,369,473.69 18.9 6.3 Sources: LRA: Report on Tax Expenditure and author’s calculations. Note: a. For comparison, the base refers to the quarterly total of VAT revenues as prorated from the annual figure in FY2015/16 and amounts to M 2,210,184,082. b. The percentage of prorated quarterly tax revenues in FY2015/16 is M 6,711,036,386. 85. VAT productivity as an indicator of VAT aggregate performance. Some conceptual measurement of VAT productivity can be used as a complement to indicate the extent of efficiency loss due to a mixture of factors, both policy and administrative. The VAT efficiency is the ratio between VAT collection (as a share of GDP) and the standard (statutory) VAT rate. The VAT C-efficiency is defined as the ratio of VAT revenues to the product of the standard VAT rate and household consumption.30 In relative terms, VAT efficiency measurement is less relevant to the indicator of C-efficiency due to the issues inherent in the mechanics of GDP accounting in a country such as Lesotho with its high share of informality, as well as the masking of the share of consumption (a more appropriate proxy of the VAT base) of GDP. Figure 20 shows the trends in VAT efficiency and C-efficiency. Both are upward trending. However, C- efficiency is steadier, whereas VAT efficiency evidences a bumpier pattern. 30Technically, the denomination of the VAT C-efficiency indicator could be the total of household and government consumption. However, for cross-country comparative analysis, which deals with the common practice of exempting government consumption from the VAT base, it is more suitable to use household final consumption. 33 Figure 20.VAT Productivity in Lesotho, FY2007/08–FY2015/16 80 0.6 70 66.2 66.9 0.5 60 58.4 56.4 55.4 VAT Productivity Ratio (%) C-efficiency ratio (%) 53.6 48.8 50.9 49.7 0.4 50 40 0.3 30 0.2 20 0.1 10 - - 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 VAT Productivity C-efficiency Sources: MTFF, WEO, WDI, and author’s calculations. 86. The productivity of VAT in Lesotho compares favorably to the broad set of SSA countries, but it stands below that of other SACU countries, except for Swaziland. Table 11 presents the detailed information and calculation of both VAT efficiency and C-efficiency for the South African Development Community (SADC) countries. Figure 21 shows the C-efficiency in cross-country comparison, indicating that most SADC countries—except for the Democratic Republic of the Congo, Madagascar, Malawi, Tanzania, and Uganda —have a C-efficiency higher than that of the average SSA. The C-efficiency in Lesotho is lower than that of other SACU countries, particularly Botswana, Namibia, and South Africa, and implies ‘lost’ revenues in relative terms. Specifically, if Lesotho strived to achieve a similar efficiency level of the ‘best’ SACU performer (South Africa), it would have collected 9.5 percent of GDP from VAT, or raised the intake from this tax by a full 2 percentage points of GDP. 34 Table 9: VAT Productivity and C-efficiency in SADC Countries (4) (1) (2) (3) = (2)/(1) Household Final (5) = (2)/((1)*(4)) SADC Country Categories Statutory VAT VAT Revenue (% VAT Productivity Consumption (% C-efficiency Rate of GDP) of GDP) Angola Upper middle income No VAT - - - - Congo, Dem. Rep. Low income 16% 3.30 0.21 72.20 28.57 Madagascar Low income 20% 1.63 0.08 85.30 9.58 Malawi Low income 16.5% 4.70 0.29 80.90 35.24 Mauritius Upper middle income 15% 6.75 0.45 76.70 58.67 Mozambique Low income 17% 8.90 0.52 70.00 74.79 Seychelles High income 15% 10.00 0.67 52.10 127.99 Uganda Low income 18% 3.58 0.20 74.80 26.59 Zambia Lower middle income 16% 5.80 0.36 53.90 67.25 Zimbabwe Low income 15% 6.80 0.45 88.10 51.46 Botswana Upper middle income 12% 4.02 0.34 45.90 73.03 Namibia Upper middle income 15% 7.30 0.49 61.70 78.88 Lesotho Lower middle income 14% 7.57 0.54 81.67 66.19 South Africa Upper middle income 14% 7.04 0.50 60.30 83.35 Swaziland Lower middle income 14% 5.50 0.39 73.70 53.30 Sources: MTFF, WEO, WDI, and author’s calculations. Note: Household Final Consumption (percentage of GDP) is drawn from 2014 data. Figure 21. C-Efficiency Ratio in SADC Countries 140 120 100 C-efficiency ratio (%) 80 60 40 20 0 Namibia Zambia Uganda Mauritius Swaziland Madagascar Botswana Seychelles Mozambique Zimbabwe Malawi South Africa Congo, Dem. Rep. Lesotho C-Efficiency Ratio Average in Africa Region Sources: MTFF, WEO, WDI, and author’s calculations. Policy Implications 87. The current setting of the VAT regime is complex—with a multi-rate structure and an extensive list of exemptions well beyond the notion of ‘hard-to-tax’ sectors or commodities. In particular, the VAT benefits granted under the zero-rating risk opens loopholes for unnecessary complexity, abuse, avoidance, and evasion. However, to strike a balance between VAT efficiency and equity given the relative high 35 productivity of the regime, it is recommended that the policy reform be focused on unifying the rate structure. 88. As the VAT rate was increased to 15% in South Africa, it is an welcome development that the Government of Lesotho has decided to do the same for the VAT rate harmonization..31 Further efforts should be made to focus on streamlining and unifying the statutory rate structure. Electricity and telecommunication services that are subject to the reduced rate of 5 percent should be brought into the standard rate list. In the immediate term, if need be, direct compensation—or charge-free supplies—could be granted. However, it would be limited exclusively to low-income households for a yet-to-be-set minimum consumption of utilities (electricity). This dual policy measure would enhance efficiency, revenue productivity, and equity. It is also worth noting that Lesotho does not levy an excise tax on mobile phone equipment. This omission, combined with a reduced VAT rate, would double the benefit to the sector and further the tax-induced distortion across sectors. As presented in table 13, unifying the zero rating of non-exports and a 5 percent rate into the standard rate of 14 percent, could generate a gain of approximately 5.5 percent of the total tax intake. However, the higher rate of 15 percent on liquor and tobacco should also be brought down to the standard rate of 14 percent in concurrence with revamping and raising the excise on such commodities (taxation on alcohol and tobacco is discussed further in the following sections). Tobacco and Alcohol Taxation Assessment 89. The recent global financial crisis and the subsequent fall in SACU revenues have wreaked havoc on Lesotho’s economy. In an effort to bring the budget deficit back to sustainable levels in the medium term, the Government has begun to explore a series of expenditure-cutting measures and has also searched for ways to increase its domestic tax collection. As such, the Government is considering raising the excise tax on alcoholic beverages and tobacco. A new tax regime on tobacco and alcohol could be a relatively easy reform for the Government to implement because it would address both fiscal needs and social concerns (associated with over-consumption, such as alcoholism, drunk-driving fatalities, and an increase in the Government’s health bill). 90. Tobacco. Lesotho has significant potential to raise revenues from tobacco taxation, particularly, with regard to excise taxes. First, compared to its peers in SSA, the country is among those with the highest tobacco users as a share of the total population. As of 2013, the daily smoking prevalence in Lesotho of an adult aged 15 and over is 20 percent, the second highest in SSA after Sierra Leone (23 percent) among the countries that report the data (WHO African Region 2015). 91. Second, Lesotho’s effective rate on tobacco continues to be relatively low by regional comparison. Many countries with lower prevalence rates apply more stringent policies, including taxation, than those in Lesotho. The country’s total tax rate is lower than 7 other countries within the 15 country SADC. Countries such as Botswana, the Democratic Republic of Congo, Mauritius, and Swaziland—all with lower smoking prevalence—apply significantly higher taxes on tobacco than Lesotho (46 percent). By comparison, Seychelles, the country with a smoking prevalence close to that of Lesotho (17 percent), applies 80 percent excise taxes on tobacco products (table 12). Currently, cigarettes are subject to a specific excise of US$34.31 per 1,000 cigarettes in Lesotho. As with the majority of its SADC peers, the country does not 31 See the Budget Speech 2018/19, available at: http://www.finance.gov.ls/documents/budget%20speeches/Budget%20Speech%202018-19%20Final.pdf 36 apply any value-added taxes on tobacco, leaving the risk of any excise being eroded due to inflation (table 13). Table 10: Cross-Country Comparison of Tobacco Prevalence and Compliance within SADC Countries Country Adult Taxes as a % Price of the Most Tax Administration Daily Sold Brand a Smoking Specif Ad VAT/ Import Total Tax Bans or If Duty-free Prevalence ic Valorem Sales Duties Tax Stamps Limited Imports are * Excise Excise Tax Applied on Limited, Duty- on Duty-free free Allowance Tobacco Imports (number of Products by cigarette Travelers sticks) Angola n.a. 0 0 23 0 23 n.a. None n.a. Congo, Dem. 14 Yes n.a. n.a. 4.4 10 14 10 48 Rep. Madagascar 17.6 0 64 17 0 81 No None n.a. Malawi 12.4 15 0 4 2 21 No Limited n.a. Mauritius 16.0** 59 0 13 0 72 Yes Limited 200 Mozambique 16.7 16 0 15 0 31 No None n.a. Seychelles 17.3 67 0 13 0 80 No Limited 200 Uganda 7.0** 35 0 10 0 45 No None n.a. Zambia 5.0 0 20 1 0 21 No Limited n.a. Zimbabwe n.a. 23 24 13 0 60 No Limited n.a. Botswana 16.2 42 9 11 0 62 Yes Banned n.a. Namibia 19.0** 29 0 4 0 33 No None n.a. Lesotho 20.4 33 0 15 0 46 No Banned n.a. South Africa 16.0** 37 0 12 0 29 n.a. None 200 Swaziland 5.9 33 0 20 0 53 n.a. n.a. n.a. Source: WHO African Region. 2015. Report on the status of major health risk factors for noncommunicable diseases. ** WHO. 2015. Report on the Global Tobacco Epidemic: Raising taxes on tobacco. 2015. Note: a. Tax rates appear here are not at the statutory rate. All data were collected between June 2014 and January 2015 by WHO regional data collectors. The two main inputs used to calculate the share of total and excise taxes were prices and tax rates and structure. WHO used its own methodology to calculate prices and tax rates. Table 11: Tobacco Excises and Retail Prices: Comparison within SADC Countries Country Current excise rates Retail prices as at July 2014 (US$ per 1,000 cigarettes) ** (US$ per pack of 20 cigarettes) *** Marlboro or equivalent lowest-cost brand Angola No excise 2.06 1.54 Congo, Dem. Rep. Mixed excise 2.81 0.60 Madagascar Ad valorem excise 4.06 0.63 Malawi 20.00 3.01 1.00 Mauritius 61.94 5.09 2.96 Mozambique 8.84* 2.62 0.98 Seychelles 184.32 7.56 6.09 Uganda 14.00 (Soft Cap)/22.22(Hinge Lid) 2.29 0.76 Zambia 8.08* 4.89 0.98 Zimbabwe 15.00* 2.00 1.00 Botswana Mixed excise 3.08 n.a. Namibia 3.15 3.55 1.59 Lesotho 34.31 n.a. 2.80 South Africa 50.84 3.14 1.68 Swaziland 39.60 4.48 n.a. Sources: * IMF. 2016. Malawi: Strategy for Comprehensive Tax Reform to Raise Revenue and Promote Growth. ** Based on exchange rate as of March 3, 2017 Other Sources: Angola: WHO. 2015. Report on the Global Tobacco Epidemic: Raising taxes on tobacco. Madagascar: WHO. 2015. Report on the Global Tobacco Epidemic: Raising taxes on tobacco. 37 92. Alcohol. Excise revenues from alcohol account for the majority of total excise collection in an approximate range from 60-90 percent during FY2012/13–FY2014/15 (table 14). Compared to neighboring countries, Lesotho’s alcohol per capita consumption (APC) falls in the mid-range. However, when compared to SACU countries, Lesotho’s alcohol sales tax/VAT is higher than in Botswana, South Africa, and Swaziland (table 15). Table 12: Excise Tax Performance, FY2012/13–FY2015/16 2012/13 2013/14 2014/15 2015/16 % of Tax Revenue Tobacco 0.11% 0.11% 0.09% 0.05% Alcohol 2.43% 2.29% 2.32% 2.45% Electronics 0.20% 0.06% 0.09% 0.11% % of Excise Revenue Tobacco 3.22% 4.46% 2.64% 0.68% Alcohol 68.29% 89.01% 64.86% 34.47% Electronics 5.49% 2.40% 2.44% 1.49% Sources: LRA, MTFF, and author’s calculations. Table 13: Total, Recorded and Unrecorded APC, 2010 (in liters of pure alcohol; 15+ years population) in SADC Country Total APC Distribution of Recorded National Any % Sales Excise Tax as % of the Excise (recorded + Consumption Monitoring Duty Tax/VAT Retail Prices (2012) as % of unrecorded) % % % % for Alcohol Paid, (2012) Beer Win Spirit the Consumption Excise Retail Be Win Spirit Othe e s or Tax Price er e s r* Stamps for 1 or litre of Labels Pure n.a. Alcohol (2008) Angola 7.5 64.3 13.7 17.4 4.7 No 9.3 10 1 1 1 n.a. Congo, 3.6 24.0 0.7 2.0 73.3 No 1.5 16 n.a. n.a. n.a. 9.3 Dem. Rep. Madagascar 1.8 56.0 9.5 34.5 0.0 No n.a. 20 n.a. n.a. n.a. 1.5 Malawi 2.5 9.1 1.2 13.4 76.2 No n.a. 16.5 9 9.5 9.5 n.a. Mauritius 3.6 66.2 12.3 21.3 0.2 No 36.6 15 n.a. n.a. n.a. n.a. Mozambiqu 2.3 63.0 7.3 25.4 4.3 No n.a. 17 n.a. n.a. n.a. 36.6 e Seychelles 5.6 67.0 22.2 10.8 0.0 No 5.4 No n.a. n.a. n.a. n.a. Uganda 9.8 9.4 0.1 1.9 88.6 No 37.2 18 60 n.a. 60 n.a. Zambia 4.0 22.7 2.9 13.6 60.7 No 31.1 16 n.a. n.a. n.a. 37.2 Zimbabwe 5.7 23.7 1.7 6.8 67.7 Yes 30 15 n.a. n.a. 40 31.1 Botswana 8.4 56.0 11.8 11.5 20.7 No n.a. 12 40 40 40 30 Namibia 10.8 96.7 0.3 0.9 2.1 No 4.1 16.5 n.a. n.a. n.a. n.a. Lesotho 6.5 51.3 0.2 18.9 29.6 No 23.9 15 n.a. n.a. n.a. 4.1 South Africa 11.0 48.1 17.8 16.7 17.4 Yes 10.2 14 35 33 48 23.9 Swaziland 5.7 33.6 0.8 0.7 65.0 No No 14 24 n.a. n.a. 10.2 Source: WHO. Global status report on alcohol and health 2014, and WHO database. Note: * Other alcoholic beverages (for example, fortified wines, rice wines, or other fermented beverages made of sorghum, millet, or maize). 93. Currently the VAT rate on alcohol and tobacco is set at 15 percent, rather than the standard rate of 14 percent. The aim is to raise more revenue and discourage consumption. The 15 percent rate has generated extra revenue but has led to higher compliance costs. Table 16 shows the amount of VAT calculated at 15 and 14 percent. The gain from a high rate of 15 percent, which would mean additional revenues, would be forgone if alcohol and tobacco were taxed at 14 percent. In addition to the reduction in revenue loss stemming from the multiple VAT rates, a standard rate would allow for a reduction in administrative costs. It would also deter smuggling, and increase compliance. Further, such a revenue loss would be more than 38 compensated by an increase in excise taxes. As reported by the Africa Tax Research Network (ATRN) (Mokoatle 2016), the excise duty would be an effective instrument to control alcohol and tobacco consumption while also raising revenues. Table 14: VAT Calculations on Alcohol and Tobacco Sales (Maloti, millions) 2010/11 2011/12 2012/13 2013/14 VAT at 15 percent 96.86 106.31 120.44 132.49 VAT at 14 percent 90.40 99.22 112.41 123.66 Gain from a high rate 6.46 7.09 8.03 8.83 Source: LRA. Policy Implications 94. The Government plans to introduce the new bill on alcohol and tobacco with technical assistance support by the World Bank. As reported by the Economic Research Southern Africa (ERSA) (Moshoeshoe 2012), the demand for alcohol and tobacco is inelastic. This implies that raising taxes on these commodities would be an effective (and efficient) revenue-enhancing instrument. The new bill is expected to incorporate a levy of 30 percent on tobacco products and 20 percent on alcohol. Marquez and others (2017) estimate that under such provisions, the expected additional fiscal revenue to be generated is about R 163 million (US$11.6 million) or about 0.45 percent of GDP (2017/18 projections). It is further projected that if VAT/other levy revenues are added, this amount would increase to R 341 million or about 1.07 percent of GDP (Marquez and others 2017). 95. The World Bank also recommends that in the next round of reforms, Lesotho introduce a unified rate of 30 percent on both tobacco and alcohol products. Under this scenario, the expected additional fiscal revenue to be generated is about R 200 million (US$14.2 million) or about 0.6 percent of GDP respectively (2017/18 projections). If the VAT/other levies are to be added, the amount of revenue increase would reach R 428 million or about 1.17 percent of GDP (Marquez and others 2017). Income Tax Regimes Assessment 96. The income tax regime has undergone some changes in recent years. On June 13, 2014, the GoL enacted a new Income Tax Act by replacing the rate structure of the pay-as-you-earn (PAYE) system of 22 percent and 35 percent (at the threshold at M 40,368) with dual rates of 20 percent and 30 percent respectively (at the threshold at M 51,670). The fringe benefit tax (FBT) rate increased from 35 percent to 40 percent. The main income taxes and their key rates are summarized in table 17. The income tax regime exempts gifts and inheritances. Table 15: Key Income Tax Rates 2013 2014a Corporate tax 10% (manufacturing), 25% (non-manufacturing) PAYE 22% or 35% b 20% or 30% c Withholding Tax (WHT) 25% for non-residents (dividends, interest, royalties, natural resource payments, trustee's fees). If dividends from manufacturing: no WHT applicable. For dividends to countries within the Double Taxation Agreement (DTA), the rates are 10 or 15%, all others: 10%. Fringe Benefit Tax (%) 35 40 Source: LRA and Deloitte Guide to Fiscal Information: Key Economies in Africa 2014/15. a. Years of assessment commencing on or after April 1, 2014. b. At a rate of 22 percent for incomes below the M 40,368 threshold, 35 percent for incomes over M 40,368. 39 c. 20 percent for incomes below the M 51,670 threshold, 30 percent for incomes over M 51,670. 97. The statutory rates as applicable to the PIT and CIT are comparable to the global average. Table 18 provides a snapshot of the cross-country comparative standard income tax rates. The CIT rate, with the highest standard rate of 25 percent, is similar to that of Organisation for Economic Co-operation and Development (OECD) countries and the global average. However, it is lower than the regional average. The standard and highest statutory PIT rate of 30 percent in Lesotho is within the same range as both global and regional averages, but lower than that of OECD countries. Table 16: Statutory Income Tax Rates Country CIT PIT Angola 30 17 Botswana 22 25 Burundi 35 30 Congo, Dem. Rep 35 40 Eritrea 34 30 Ethiopia 30 35 Gambia 31 30 Lesotho 10/25 30 Liberia 35 25 Madagascar 20 20 Malawi 30 30 Mozambique 32 32 Namibia 33 37 Rwanda 30 30 Sierra Leone 30 30 South Africa 28 41 Sudan 35 15 Swaziland 30 33 Tanzania 30 30 Uganda 30 40 Zambia 35 35 Zimbabwe 25.8 50 Kenya 30 30 Africa region average 30 31 Source: PWC (various years): Worldwide tax summaries. 98. Some efforts have been made to relieve the PAYE (penalty and interest regime) burden on the lower-income group. The increase of the threshold from M 40,368 (equivalent to US$3,140) to M 51,670 (or US$4,019) would bring additional individual income earners down to the lower bracket at a rate of 20 percent.32 Also, it should be noted that there is no exemption threshold from the PAYE regime. This implies that those workers — even earners at the minimum wage—would all fall into the first income tax bracket and continue to be subject to a PAYE, albeit at slightly lower rate. However, the highest income earners are expected to gain substantially due to the reduced marginal rate from 35 percent to 30 percent. 99. The CIT regime creates and magnifies cross-sector distortions. A closer look at the CIT rate structure indicates that the country grants preferential treatment to manufacturing over non-manufacturing. 32 The exchange rate used is as of June 7, 2017. 40 In addition, Lesotho provides a zero tax rate on profits from exports outside of the SACU. The multi-rate structure and other underlying provisions that affect the determination of the corporate taxable income lead to differentiated marginal effective tax rates (METRs) or the extent of cross sector/cross-asset type distortions. 100. The CIT for manufacturing in Lesotho is one of the lowest, whereas the CIT for all other sectors combined (including the service sector), is the highest in SSA. METR is one of the indicators that would provide some guide as to how investors would view the tax burden when making business decisions.33 Mintz and Chen (2015), not surprisingly, estimate the METRs at the lowest rate of 12.6 percent for manufacturing and at the highest rate of 27.4 percent for the service sector (table 19). Compared to other select SSA countries, the METR in Lesotho is the highest for all sectors combined, as well as for the service sector (24 percent and 27 percent, respectively). However, the rate for manufacturing is the fourth lowest in the sample (almost 13 percent). The study highlights the significant extent of cross-sector distortion in Lesotho (with the sector gap METR reaching the highest level — almost 15 percent) — as compared with those in other countries in the sample). Table 17: METR: Lesotho versus SSA Comparators Statutory Company Marginal Effective Tax Rate Country Income Tax Rate Overall Manufacturing Services Sectoral gap Botswana 15 12.6 8.6 13.1 -4.4 Ghana 25 14.6 14.5 14.6 -0.1 Lesotho 10 24.2 12.6 27.4 -14.8 Madagascar 20 12.6 16.7 11.7 5 Mauritius 15 8 8.5 7.9 0.7 Rwanda 30 18.5 26 17.7 8.3 Sierra Leone 15 16.4 11.3 16.7 -5.4 South Africa 28 14.2 15.5 13.9 1.6 Tanzania 30 18 13.2 18.8 -5.7 Uganda 30 13.4 8 14 -6 Zambia 35 17.4 22.9 16.6 6.4 Source: Mintz and Chen 2014. 101. The CIT productivity is an aggregate indicator used to gauge the overall effectiveness and efficiency in both tax policy and administration. Table 20 shows the estimates of such an indicator for Lesotho and the 15-member states of the SADC. The indicator is calculated as the ratio between the CIT collection as a share of GDP and the highest CIT standard rate. Lesotho’s CIT efficiency ratio ranks below the average, ranking 9 of the 15 SADC countries. If Lesotho strives to achieve the level of CIT productivity of Botswana (0.24) at the unified statutory rate of 25 percent, it would raise 6 percent of GDP — or almost 3 percentage points of GDP difference from the actual collection of 3.14 percent in 2014/15. The relatively low productivity of the CIT can also be explained by examining the fiscal incentives embedded in the tax regime and nontax provisions. 33The METR measures the difference between the pre- and post-tax rate of returns at the margin, where the return on the last dollar invested just covers the cost of capital. 41 Table 18: CITs in SADC Countries (1) (2) (3) = (2)/(1) Efficiency Ratio SADC Standard CIT CIT Revenue Efficiency Ratio Ranking Rate (% of GDP) High income Seychelles 30.00 4.87 0.16 5 Upper middle income Angola 30.00 9.22 0.31 1 Mauritius 15.00 2.32 0.15 6 Botswana 22.00 5.28 0.24 3 Namibia 33.00 5.00 0.15 7 South Africa 28.00 5.17 0.18 4 Average 25.60 5.40 0.21 Lower middle income Zambia 35.00 4.60 0.13 8 Lesotho 25.00 3.14 0.13 9 Swaziland 30.00 1.90 0.06 12 Average 30.00 3.21 0.11 Low income Congo, Dem. Rep. 35.00 1.70 0.05 13 Madagascar 20.00 0.78 0.04 14 Malawi 30.00 3.17 0.11 10 Mozambique 32.00 8.30 0.26 2 Uganda 30.00 0.68 0.02 15 Zimbabwe 25.75 2.60 0.10 11 Average 28.79 2.87 0.10 Sources: WEO, WDI, PwC Worldwide tax summaries, and author’s calculations. 102. Investment incentives. Lesotho grants multiple tax and non-tax incentives for investment. Table 21 provides a summary of such incentive provisions. Table 19: Investment Incentives in Lesotho Investment Incentives Non-tax Incentives • Financial incentives to manufacturing companies established in Lesotho, for example, unimpeded access to foreign exchange and long-term loans (especially for exporters). • Reduced municipality rates applicable to industrialists. • Free repatriation of profits derived from manufacturing companies. • Bank-administered foreign currency accounts are permissible. Tax Incentives CIT • Reduced statutory rate: A maximum manufacturing corporate tax rate of 10 percent on profits for intra-SACU trade. • Zero percent CIT on manufacturing profits generated from the export of manufactured goods outside of the SACU. • No advanced corporate taxes are paid by companies on the distribution of manufacturing profits. • DTAs available. a 42 • Deduction of start-up costs (a four-year write-off period) is provided for investments incurred in starting a business to produce income subject to the CIT. • Tax exemption and additional allowances for capital and other expenditures with respect to new buildings, machinery and plants, electric power, transportation, and water for pioneer industries (under the Pioneer Industries Encouragement Act of 1969). • Training costs are allowable at 125 percent for tax purposes. WHT • No WHT on dividends distributed by manufacturing firms to local or foreign shareholders. • Payments made for external management skills and royalties related to manufacturing operations are subject to a reduced rate of WHT of 10 percent. Sources: LRA, Deloitte: Guide to Fiscal Information - Key Economies in Africa 2014/15. Note: Currently DTAs are in effect with Mauritius, South Africa, and the United Kingdom. DTAs with Botswana, Malaysia, Namibia, Seychelles and Swaziland are under negotiation. 103. The key income tax incentives amount to significant tax expenditures or revenue losses. Basically, tax expenditures induced by fiscal incentives can be calculated using one of three methods: initial revenue loss, final revenue loss, and outlay equivalence. A primer on tax expenditures is shown in annex 2. For this chapter and based on the limited data provided, the authors attempt to estimate the size of the CIT expenditures using the initial revenue loss method. This method does not take into account the behavioral change by taxpayers in response to the incentives granted. The estimate is applied to representatives in the manufacturing sector, the SOEs of Maluti Mountain Brewery (for the group and for the company), and the Lesotho Flour Mills. Table 22 provides detailed data and an estimation for the three businesses in the sample. Some qualifying notes on the estimated revenue loss are worth noting, as follows: (a) Revenue losses due to a zero percent CIT rate on profits from exports outside the SACU. Revenues forgone from this exemption could not be calculated due to the lack of availability of financial statements regarding exported manufactured goods outside the SACU. An important portion of exporters outside the SACU are in the textile industry. Many of Lesotho's garment companies are subsidiaries of larger global textile and garment corporations with operations around the world. Hence, the LRA is facing a substantial risk from transfer pricing. Where such practices are rampant, the domestic tax rate matters little. (b) Forgone revenues from the reduced CIT rate (10 percent) for corporate manufacturing. For Maluti Mountain Brewery (group), the reduced CIT rate led to an increased effective tax rate to a range from 10 percent in 2009/10 to 13.5 percent in 2014/15.34 The annual average revenue forgone stands at M 12,469 million (US$ equivalent), relative to the case whereby a standard rate of 25 percent was to be applied. The effective CIT rate for Maluti Moutain Brewery (company) dipped to less than 7 percent in 2014. In the case of Lesotho Flour Mills, the effective rate decreased to 10.6 percent and 10.7 percent in 2010 and 2012, respectively. (c) Forgone revenues from WHT exemptions on dividends distributed by manufacturing companies to local or foreign shareholders35. The revenue loss due to exemption of the WHT on dividends from the manufacturing sector (as against the dividend rate of 25 percent applicable to non-residents from non-manufacturing profits) granted to Maluti Mountain 34The effective tax rate is simply defined as the rate of tax payment in total pre-tax income. 35 The estimate of the revenue forgone as shown indicates the higher side of the revenue loss magnitude. This does not capture the fact that part of the dividends from manufacturing would go to the government as the equity owner. Thus, the removal of exemptions of the withholding tax on dividends would also reduce government’s non-tax revenues. 43 Brewery (group and company) reaches its peak of M 3.8 million (US$ equivalent) in 2014/15 and M 16.7 million (US$ equivalent) in 2014, respectively. (d) Forgone revenues from the combined reduced CIT rate and exemption of dividends for manufacturing. For the Maluti Mountain Brewery (group), the combined revenue loss due to both the reduced rate of CIT and exemption of distributed dividends is significantly high— reaching a maximum of almost 185 percent of the total tax actually collected in the year of 2013/14. For the Maluti Mountain Brewery (company), the revenue loss is expected to peak at almost six times the total tax paid by the company in 2014. (e) Forgone revenues from the 125 percent special deduction for training costs. With the Lesotho Flour Mills’ available data on the 125 percent special deduction for training costs, the authors estimate that the forgone revenues from this special deduction at the prevalent CIT rate of 10 percent (as applied to manufacturing) and at the standard rate of 25 percent (for non-manufacturing) are substantial. The revenue loss can be as high as M 2.5 million (US$ equivalent) at the 10 percent CIT rate, but could reach M 6.3 million (US$ equivalent) at the 25 percent CIT rate. 44 Table 20: Forgone Revenues from the 10 Percent CIT inside the SACU, Exemption of WHT on Dividends, and 125 Percent Deduction for Training Costs Maluti Mountain Brewery (for group) M ‘000 FY 2009/10 FY 2010/11 FY 2013/14 FY 2014/15 Profit before tax 68,431 98,207 86,538 111,965 taxes paid* 7,552 10,395 8,330 15,132 Effective Tax Rate 11.0% 10.6% 9.6% 13.5% Standard CIT rate in lesotho 25.0% 25.0% 25.0% 25.0% Tax charge at standard corporate tax rate 17,108 24,552 21,635 27,991 Revenue forgone from 10% CIT inside SACU (B) 9,556 14,157 13,305 12,859 Total dividends paid 7,552 10,395 8,330 15,132 Revenue forgone from WHT exempt (C) 1,888 2,599 2,083 3,783 Total revenue foregone (% of Taxes Paid) 1.52 1.61 1.85 1.10 Maluti Mountain Brewery (for company) M ‘000 2010 2011 2014 2015 Profit before tax 65,754 92,982 73,570 86,117 taxes paid* 6,905 9,084 5,083 8,185 Effective Tax Rate 10.5% 9.8% 6.9% 9.5% Standard CIT rate in lesotho 25.0% 25.0% 25.0% 25.0% Tax charge at standard corporate tax rate 16,439 23,246 18,393 21,529 Revenue forgone from 10% CIT inside SACU (B) 9,534 14,162 13,310 13,344 Total dividends paid 30,990 52,888 66,987 64,567 Revenue forgone from WHT exempt (C) 7,748 13,222 16,747 16,142 Total revenue foregone (% of Taxes Paid) 2.50 3.01 5.91 3.60 Lesotho Flour Mills M ‘000 2009 2010 2011 2012 2013 2014 Profit before tax 10,794 13,910 22,179 34,628 (3,847) (11,865) taxes paid* 6,149 1,478 2,716 3,700 N/A N/A Effective Tax Rate 57.0% 10.6% 12.2% 10.7% N/A N/A Standard CIT rate in lesotho 25% 25% 25% 25% 25% 25% Tax charge at standard corporate tax rate 2,699 3,478 5,545 8,657 N/A N/A Revenue forgone from 10% CIT inside SACU (B) (3,451) 2,000 2,829 4,957 N/A N/A 125% Additional Training Deduction 8,131 15,499 126,856 29,602 14,461 6,731 100% Standard Training Deduction 6,505 12,399 101,485 23,682 11,569 5,385 Additional expensed from special deduction (D) 1,626 3,100 25,371 5,920 2,892 1,346 Revenue forgone at CIT 10% (D) 163 310 2,537 592 289 135 Revenue forgone at CIT 25% (B) + (D) 407 775 6,343 1,480 723 337 Total revenue foregone (% of Taxes Paid) N/A 1.88 3.38 1.74 N/A N/A Sources: SOEs: A Country Policy Note and author’s calculations. 104. In addition to the significant potential revenue losses induced by multiple incentives, such provisions are being provided across different tax and investment laws, as well as other legislation. The ambiguity would entail a risk that investors would be required to go through complex and discretionary procedures for qualification. Currently, institutional responsibility for the discretionary granting and administration of tax incentives is divided among finance, trade, investment, or other ministries, thereby increasing the risk of corruption and rent seeking. In sum, it would risk being detrimental to the overall business environment, rather than enhancing it. Policy Implications 105. Some key policy implications can be derived from an analysis of the income tax regime in Lesotho, as follows: 45 • Keep the standard CIT rate of 25 percent unchanged over the medium term. The rate is comparable to the international average. The policy focus would be on streamlining the multi- rate structure of the CIT and rationalizing the complex fiscal incentives. • Consider harmonizing the various income tax regimes. Currently the statutory PIT and CIT rates are not harmonized, and this may lead to unintended arbitrage consequences. Some policy options are available for consideration. • If the CIT dual rate structure of 10 percent and 25 percent and the dividend tax of 25 percent are to be kept unchanged, the implied highest marginal PIT rates should be 32.5 percent and 43.75 percent, respectively36. • It is unclear why the FBT rate has been increased from 35 percent (equivalent to the highest marginal rate of the PAYE) to 40 percent, whereas the PAYE rate is reduced to 30 percent. This mismatch could potentially incentivize tax avoidance. • Consider introducing inheritance and gift taxes. An inheritance tax is typically levied on a beneficiary who inherits money, property, or some other asset from a person who has died. To avoid tax planning, an inheritance tax, if ever introduced, needs to be complemented with a gift tax. Some may argue that instituting an inheritance tax amounts to double taxation. Others (particularly tax economists) would, however, raise the prospect that the tax can be a direct means of achieving income redistribution and equality of opportunity. This could be a response to the question: Why tax salaried or non-salaried income earners who work hard, but exempt those who do not make any effort to earn the income they receive from an inheritance or gift? If the tax is to be considered, it has to be as comprehensive as possible (to promote equity and avoid tax planning and avoidance). It should also be levied directly on the recipients rather than on the donors.37 As proven in other developed countries, the tax is highly political. As such, the introduction of this tax may be considered over the longer-term. • Conduct a thorough review of the current structure of tax incentives and explore reform options. The use of tax incentives in developing countries has long been a controversial issue. Despite the controversy, given the current practice that every SADC country offers investment tax incentives of one form or another, it would be more sensible and pragmatic to ensure that the incentives are well-targeted, effective in inducing marginal investment, and transparent. This would be preferable to cancelling all fiscal incentives in a blanket manner. Qualifications for incentives should be systematically applied when any investor satisfies the specified conditions as required by the governing laws and legislation. Over time, the GoL should consolidate all incentives in the confines of tax legislation under the authority of a single government body (ideally the MoF). Such a consolidation would increase transparency and help avoid any unintended overlap and inconsistencies in the design of incentives. It would also limit discretionary power and enable policy makers to address problems that may arise 36 Although certain countries continue to retain a double taxation system (under the classical scheme) whereby capital income is subject to CIT and the dividend is taxed at a personal level, efforts are also made to relieve such distortions with various integration schemes. A dual tax regime — one that taxes labor income with a progressive rate structure subjects others (basically passive or capital income such as interests, dividends) to a flat, lower rate —can be used as a way to mitigate the classical distortionary impact of double taxation. In the case of Lesotho, this applies mainly to the high dividend tax. As such, it is recommended that some degree of harmonization in income taxation be made to avoid any unintended distortions in the decision of whether to incorporate or not. For that, (tc)+(1-tc)*td=tp, where tc is standard CIT rate, td = dividend rate, and tp= marginal PIT rate. 37 See, for example, IFS 2011 (Institute for Fiscal Studies): Tax by Design-the Mirrlees Review. 46 with the governance of tax incentives. The estimation of the magnitude of tax expenditures due to certain types of tax benefits is shown in Table 22. It provides the basis for some policy implications for Lesotho in considering how to streamline its current tax incentive system. Although the chapter does not aim to provide rigid ‘prescriptive’ recommendations, suggestions are offered for Lesotho to look into international experience with certain types of incentives that can provide more ‘bang for the buck’. In particular, those incentives aimed at lowering the cost of investment are often more effective to attracting marginal investments— than those targeted at increasing firms’ profits.38 III. Tax Administration Assessment 106. The Government has implemented revenue administration reforms for more than a decade, and recently launched its Modernization Agenda. This chapter focuses on an analysis of the main domestic tax policies, and includes some snap-shot descriptions of the tax administration with the aim of providing a more complete picture of the overall tax system.39 The objectives of the LRA’s modernization agenda are to enhance the quality of taxpayer service, reduce the burden of compliance, and lower administration costs.40 The reform agenda consists of three major programs: administration reform of inland taxes; the Customs Modernization Program (CMP); and the modernization of the core support processes of revenue administration. Although a comprehensive tax administration diagnostic assessment will be required, the recent PEFA (2016) concludes that the LRA is equipped with a minimum set of tools for modern tax administration, such as: a functional website that provides information about taxpayer obligations and rights; a dedicated service offices for large taxpayers, as well as small and medium taxpayers; modern tax management software and information systems; a revenue risk management unit; and revenue performance reports. 107. In FY2012/13, the LRA began implementation of the Integrated Revenue Management System (IRMS) that aims to integrate customs and tax administration. In 2014, the LRA launched a pilot project for Enterprise Taxation and Policy Management (ETPM) and started the process of reengineering management of the tax debt. Since late 2014, the new Automated System for Customs Data (ASYCUDA) world system has been introduced. In this regard, it should be noted that the administration reforms are in line with the review of the legal framework for taxation covering, among other legislation, the Tax Administration and the Income Tax Amendment Bills. However, the tax administration system remains inefficient and ineffective. For example, the 2016 draft PEFA for Lesotho points out several bottlenecks in the tax system. • Poor links between two sets of tax management software (ASYCUDA for customs and the ETPM for VAT and income tax); • Disconnection of the tax management software from the IFMIS; • Lack of capacity to detect tax evasion; 38 Cost-based tax incentives would involve specific allowances linked to investment expenses (for example, accelerated depreciation schemes and special tax deductions and credits). On the other hand, profit-based tax incentives generally reduce the tax rate applicable to taxable income (for example, tax holidays and preferential tax rates). 39 At the time of this writing, the African Development Bank (AfDB) has confirmed its commitment to support the Lesotho tax administration reforms. 40 Source: LRA Annual Report 2014/2015 (Draft). 47 • Insufficient revenue audit and investigation of small and medium taxpayers; and • Inadequate frequency of performance reporting (currently, reporting is done quarterly). 108. LRA senior officials also shared their concerns about the shortage of audit staff. Resources are not even close to optimum due to the lack of rigorous risk management tools, techniques, and planning. 109. The cost of tax collection, relative to regional and international best practices, has been persistently high, and compliance has been unduly cumbersome. As figure 22 shows, the collection costs41 in Lesotho steadily rose over the past five years, reaching approximately 6 percent of the total tax intake in 2014/15. The increased rate of collection costs during this period exceeds that of tax revenues (the first stands at around 1.22 percent, with the latter is at 1.18 percent). In fact, the cost of revenue collection is high compared to the regional average, and even more so in relation to the world’s best performers. Figure 23 shows the cost of revenue collection in selected SSA countries in 2008 (OECD 2011).42 Whereas Benin, Botswana, and South Africa evidenced a stellar performance, Lesotho presents itself as the worst in the selected group. The discrepancy becomes even more evident when the cost of collection in Lesotho is compared with the one of the world’s best systems. (The cost of collection in selected OECD countries ranges from the lowest of 0.38 percent in Switzerland to the highest of more than 1.7 percent in Japan, according to the OECD database). Figure 22. Lesotho’s Cost of Revenue Collection for FY2010/11 to FY2014/15 (Maloti, millions and as a percentage of total tax revenues) 7,000 7.0% 5.8% 5.9% 6,000 6.0% 5.2% 4.8% 4.7% 5,000 5.0% % of total tax revenue Million Maloti 4,000 4.0% 3,000 3.0% 2,000 2.0% 1,000 1.0% 0 0.0% 2010/11 2011/12 2012/13 2013/14 2014/15 Total Tax Revenue Cost of Collection Source: LRA Annual Report, 2014/15, page 29. 41 The cost of collection is denoted simply as the ratio between the operating costs and the total collected revenue. This definition would eliminate the problem with accounting for investment costs, which would highly distort the cost of collection especially during the reforms (for example, building new infrastructure and/or the new information technology system). 42 The countries in the sample are Benin, Botswana, Burundi, Ethiopia, Ghana, Kenya, Lesotho, Malawi, Mauritius, Rwanda, Senegal, Sierra Leone, South Africa, Tanzania, Uganda, and Zambia. 48 Figure 23. Cost of Revenue Collection as a Percentage of Total Tax Revenues in Selected SSA Countries, 2008a Lesotho Source: OECD 2011. Note: a. Carter, Alan, and Ana Cebreiro. 2011. Africa's Tax System: A survey. OECD Center for Tax Policy and Administration. OECD Observer No 284. 110. Tax administration ruling and practice when combined with the unnecessary complexity of the tax regime translates into high compliance costs to taxpayers and becomes detrimental to the overall business environment and formality.43 The Doing Business (2017)44 rankings for comparator economies and the regional average show that Lesotho fares very poorly compared to its peers. The survey ranks Lesotho at the bottom among the SACU countries in the ease of paying taxes. Compared with all 190 economies in the Doing Business sample, the country’s ranks 91st. As shown in table 23, on average, firms in Lesotho spend 324 hours a year filing, preparing, and paying three major taxes (profit taxes, labor taxes, including mandatory contributions, and consumption taxes). This compares with an average of 304 hours for SSA and 163 hours for the OECD high-income countries. However, the tax burden, as measured by the Total Tax Rate,45 is at less than 14 percent of total commercial profits. This is significantly lower than all the other SACU countries, and even lower in comparison with the SSA average (at 47 percent) or with the average in the OECD high-income countries (at 41 percent). Table 21: Indicators on Paying Taxes in SACU, SSA, and OECD Countries Indicator Lesoth Botswa Namibi South Swazilan SSA OECD High o na a Africa d Income Payments (number per 32 34 27 7 33 38.8 10.9 year) Time (hours per year) 304. 324 153 302 203 122 163.4 2 Total tax rate (% of 13.6 25.1 20.7 28.8 35.1 47.0 40.9 profit) Post-filing index (0– 78.9 89.9 79 58.6 72.5 53.4 85.1 100) 43 Lesotho’s shadow economy is estimated to be 33 percent of GDP and may employ close to 200,000 people. Schneider 2012 and IMF 2013. 44 World Bank. Doing Business 2017: Equal Opportunity for All, Economic Profile 2017, Lesotho. 45 The indicator of the Total Tax Rate as a share of commercial profits measures the amount of taxes and mandatory contributions payable by the business in the second year of operations. 49 Source: Doing Business database. 111. The World Bank’s Enterprise Survey (2016) shows that Lesotho’s tax administration imposes a significant adverse burden on the investment climate and formality. Tax administration ranks at 9 out of the top 10 constraints from a list of 15 business environment obstacles (as measured by the mean score). Table 24 shows that more than 39 percent of firms surveyed consider tax administration in Lesotho a major constraint, compared to less than 29 percent in SSA, or 20 percent in an all-country comparison. It is worth noting the negative perception of businesses on the overall tax rate: 38 percent of firms surveyed regard tax rates in Lesotho as a major constraint, whereas the share is just 35 percent and 30 percent in SSA and the all-country comparison, respectively. This is in stark contrast with the fact that the overall tax burden on corporate profits in Lesotho is substantially below that of the SSA average (14 percent versus 47 percent, respectively). One legitimate conjecture might be that the high cost of compliance has significantly distorted the business perception of the effective tax rate in the confines of the design of the Enterprise Surveys. Such a negative perception is even more undesirable when put in the context of the worsening overall competitiveness of the country. (The World Economic Forum’s Global Competitiveness Report 2016/17 shows Lesotho is poorly rated, at 120 out of 138 sampled countries. In addition, the ranking fell from the preceding year).46 Table 22: The Value of Regulations and Tax Indicators All Indicator Lesotho SSA Countries Percentage of firms identifying tax rates as a major constraint 38.1 35.2 30.1 Percentage of firms identifying tax administration as a major 39.3 28.6 20.1 constraint Percentage of firms identifying business licensing and permits as 29.3 17.5 12.9 a major constraint Source: Enterprise Surveys, World Bank 2016. Policy Implications 112. Although Lesotho has long been engaged in comprehensive revenue administration, the outcomes have not significantly improved. It is commendable that the authorities are making efforts to review both tax policies and administration as part of a concerted, coherent reform program. Tax administration reforms, however, are highly politically charged, fiscally costly, and long term in nature. As such, it is critical that the LRA step back to review and ascertain what has been done, what challenges remain, and how to prioritize and sequence further reforms. 113. As an immediate step, the revenue agency will need to conduct a detailed assessment of the country’s tax administration. The assessment would encompass the legal framework for administration, the existing set of rulings and processes, and the supporting information and communications technology systems. The review and diagnostics would help the LRA to establish a baseline for better targeting and investing in tax administration reforms. The newly developed and tested Tax Administration Diagnostic Assessment Tool (TADAT 2015) could be highly relevant to Lesotho.47 The overall framework of TADAT— including its performance outcome areas and associated measurement indicators — are summarized in annexes 2 and 3. 46 World Economic Forum. The Global Competitiveness Report 2016 –2017. 47 The Field Guide for TADAT (2015) is available at: http://tadat.org/files/IMF_TADAT-FieldGuide_web.pdf. 50 Chapter 3: Public Sector Employment and Wage Bill Management I. Public Service Employment and the Wage Bill Overview of the Wage Bill and Composition of the Public Service 114. The wage bill in Lesotho has been growing since 2001, and wage spending in terms of the share of GDP is now one of the highest in the world (figure 24). The government wage bill in Lesotho has increased over the past 15 years (figure 25) and is estimated at 18 percent of GDP or over 37 percent of government spending in 2016/17. It is the highest even among the small countries of open economies, such as Cape Verde, Comoros, and Kosovo, as well as the low-middle-income countries and low-income countries. Figure 24. Wage Spending, 2015 (as a percentage of GDP) 25.00 20.00 15.00 10.00 5.00 0.00 Botswana Angola The Gambia Benin Tunisia Capo Verde Comoros Kosovo Afghanistan Maldives Marshall Islands South Africa Namibia Swaziland Kuwait Iraq Lesotho Guinea Malawi Bosnia Morocco Micronesia Brunei Madagascar Zimbabwe Mauritius Liberia Saudi Arabia Mozambique Gabon Source: World Bank. Figure 25. Wage Spending in Lesotho (as a percentage of GDP) 60.0 50.0 40.0 30.0 20.0 10.0 0.0 2000 2002 2004 2006 2008 2010 2012 2014 2016 Wage Bill Government expenditures Source: World Bank. 51 115. The number of public sector staff has steadily increased over the past 15 years and has contributed to the high wage bill. The number of teachers and police/army staff has almost doubled, and the number of civil servants has increased by 60 percent since 2002. The current size of the public service is about 44,000 staff in 2016/2017—excluding the parastatals—of which civil servants constitute 15,729, police and military 10,566, and teachers 14,603. Together, they represent 3.3 percent of the working-age population. Figure 26. Staff Number by Category Source: MoF/Public Service. 116. In terms of the civil service staff composition, Lesotho has a bottom-heavy structure with a high vacancy rate, and the establishment control is not functioning. The establishment list, which is the list of established and funded positions across the Government, is a critical instrument for Human Resource Management (HRM) and hiring control. However, although Lesotho is managing human resources based on the establishment list, the analysis suggests a significant gap between the establishment list and the actual head-count, as well as a high vacancy rate of 23.3 percent (figure 27). In addition, whereas the current staff composition is concentrated in the lowest grades (A–C: supporting staff), most of the vacant positions are still concentrated in Grade A. There are relatively few middle management and technical staff. In the past, the Establishment Committee comprised of Ministers of Public Service, Development Planning, and Finance regularly reviewed and revised the list according to real public service needs. The Establishment Committee played a critical role, but several years ago, it ceased to exist. No mechanism is in place to review the establishment list from the strategic perspective. Only the Public Sector Commission (PSC) is involved in the hiring process. The PSC’s role is to ensure that consistency exists between new hiring and/or reassignment of staff and vacant positions. However, it does not seem to play any strategic role in readjusting the establishment lists. In addition, the hiring of Grade A staff has been exempted from the PSC approval process for a long time and has instead been done at the prerogative of each ministry. Uncontrolled hiring in Grade A appears to be one of the main reasons for the inflated number of Grade A staff. 52 Figure 27. Staff Distribution and Vacant Positions per Grade Source: MoF/Public Service. Box 2. Structure of Comparator Civil Services Although there is no best model for staff distribution across grades, mid-career and technical staff in charge of operational and technical work often constitute the core part of any public service. In the case of Zambia,48 most of the staff are distributed across the grades with a concentration on the range of F to L grades. In the case of Cameroon,49 the civil service has a top-heavy mechanism (Grade A1/2) with a large number of contractual workers (Category 8). In Cameroon, in the wake of a structural adjustment program that limited and/or froze new staff hiring and cut the base salary by half, many technical staff (mainly teachers) who were hired as contractual workers in the late 1980s and 1990s, now constitute one-third of the work force. To compensate for low wages, many mid-career staff were promoted to a higher grade. This has led to a situation in which the number of managerial positions (Grade A) is almost as high as that of technical positions (Grades B and C) in the civil service structure. Hence, low base wages, an inverted civil service structure, and a high reliance on poorly paid contract workers without clear career prospects, has resulted in low staff morale, ineffectiveness, and poor performance in public administration. This example suggests an important lesson for well-planned recruitment, career management of staff, and an adequate remuneration policy. Figure 28. Civil Service Structure in Zambia (Left) and Cameroon (Right) 48Grade A is the lowest grade, and Grade R is the highest. 49Grade D is the lowest grade, and Grade A2 is the highest for permanent staff. Category 1 is the lowest, and category 12 is the highest for contractual staff. 53 Analysis of Wage Spending 117. The major challenge regarding the high wage bill is partly attributable to the fast pace of wage increases in real terms. The compensation of employees is mainly comprised of the base wage, allowances, and benefits. The current grading system for civil servants, police, army, and teachers was established with a circular notice of the Ministry of Public Service in 2013. The notice also made an adjustment in the number of grades and notches for each category of service. On this occasion, salary adjustments were provided to certain grades and appear to have increased the wage bill. In line with the new grading system, the Ministry of Public Service reviews and revises the wage scale each year. With the new system, all public employees benefit from the two parts of salary increases: a Cost-of-Living Adjustment (COLA) and automatic notch increases every year. COLAs are provided to all employees to ensure that wages rise by at least the expected value of inflation. In this context, it should be noted that the COLA exceeded the inflation rate in some years (table 28). In addition, all employees are automatically promoted to a higher notch in their grade, which results in average wages rising even faster, that is, by an estimated 2.7 percent per year (2.8 percent before 2013). 118. The Public Service Act stipulates that promotion and salary increases be based on employee performance; however, these increases are not currently linked with performance evaluations, thereby undermining workforce incentives, a sense of fairness, and the morale of public service employees. Table 25shows that the Government has been providing 7-9 percent of annual salary increases across the board. Holding the size of the public-sector workforce constant, Lesotho’s real (that is, inflation-adjusted) economy must grow by roughly 2.7 percent every year just to prevent the wage bill from rising further as a share of GDP. Table 23: COLA Inflation Real Term Increase (on average) (%) COLA (%) (CPI) (%) 2007 10.0 8.0 12.8 2008 15.0 10.7 17.8 2009 8.5 7.2 11.3 2010 3.5 3.8 6.3 2011 5.0 6.0 7.8 2012 3.0 5.0 5.8 2013 6.0 5.0 8.7 2014 6.0 5.0 8.7 2015 4.0 5.0 6.7 2016 6.0 7.0 8.7 2017 4.0 6.6 6.7 Sources: MoF, World Bank, IMF. 119. International comparisons suggest that Lesotho’s wage bill is high. Figure 29 shows that although Lesotho’s wage bill relative to its GDP is the highest among a broad sample of countries, public employment expressed as a share of the working-age population is slightly below average. The contrast is even starker when Lesotho is compared directly with South Africa. Despite having a much lower per capita GDP, the average government salaries in Lesotho are roughly 75 percent of those in South Africa; for workers in the lowest pay grades, the ratio is smaller—38 percent for Grade A and 43 percent for Grade B—whereas it exceeds 90 percent for workers in Grade F through Grade I (table 26). 54 Figure 29. Wage Bill (as a percentage of GDP) and Employment (as a percentage of the working-age population) Source: IMF staff estimates. Table 24: Pay by Grade (Lesotho versus South Africa) Ratio of pay Pay relative to GDP Grade (Lesotho to South Africa) per capita South PPP South Lesotho Unadjusted Lesotho Africa adjusted 1/ Africa A 1 0.32 0.38 1.8 1.1 B 2 0.36 0.43 2.4 1.3 C 3 0.45 0.54 3.6 1.5 D 4 0.59 0.71 5.5 1.8 E 5 0.66 0.79 7.4 2.2 F 6 0.80 0.97 10.8 2.6 G 7 0.79 0.96 13.2 3.2 H 8 0.78 0.94 16.1 4.0 I 9 0.78 0.95 19.3 4.8 J 10 0.73 0.88 22.4 6.0 K 11 0.58 0.70 28.2 9.4 L 12 0.56 0.68 32.4 11.2 Average 0.62 0.75 13.6 4.1 Source: IMF staff estimates. 120. A disaggregation of wage spending reveals that although public employment as a share of the working-age population has fallen, average wages as a share of per capita GDP have risen (figure 30). This implies that the increase in wage spending observed over the past five years in Lesotho is attributable to policies that relate to the compensation of public employees rather than to policies that affect the overall size of public employment. The pay structure review conducted in 2012/13 was intended to increase equity within the public sector and better attract and retain a skilled work force. However, it added nearly 2 percentage points to the magnitude of the wage bill the following year. 55 Figure 30. Wage Spending, Average Government Wages, and Government Employment Source: IMF staff estimates. 121. A country’s fiscal capacity is constrained by its economic output; Lesotho’s GDP per capita is only 25 percent of that in South Africa. Average wages in the public sector, expressed relative to per capita GDP, are only 4.1 percent in South Africa but amount to 13.6 percent in Lesotho. This implies that the average civil servant in Lesotho is almost 3.5 times as costly as the average civil servant in South Africa when measured against a proxy for each country’s budgetary resources. Given the fluidity of the labor supply between South Africa and Lesotho, paying relatively high salaries may be necessary to attract and retain skilled personnel in the public sector. However, such high salaries risk making Lesotho’s wage bill unaffordable. 122. Although the payment of comparatively high wages can help attract and retain skilled workers, high wages do not necessarily lead to better public service outcomes. The outcomes in the education and health care sectors in Lesotho tend to lag those of its counterparts across a number of key dimensions — despite the fact that public spending on wages in both sectors is substantially higher. Therefore, the challenge of managing the wage bill in Lesotho is not only an issue of limiting the fiscal burden, but also one of improving the efficiency and effectiveness of public wage spending. Also, with the decline in SACU revenues and heightened fiscal pressures, the growing wage bill may directly affect capital spending and, ultimately growth. 123. Improving the efficiency and effectiveness of wage spending—and managing the overall fiscal burden—are challenges faced by virtually all countries around the world. The path chosen to navigate this difficult terrain must necessarily reflect national choices about the nature of the government’s role in society. In addition, policy makers must choose the path that respects local conditions and constraints related to a country’s level of economic development, budgetary resources, institutional capacity, labor market structure, demography, and geography. The geographic factor is particularly important for Lesotho, given that it is physically enveloped by the larger and much richer economy of South Africa. Many of the choices facing policy makers embody trade-offs that can be difficult, if not impossible, to evaluate. Furthermore, wage and employment policies are often shaped by forces unrelated to the core objective of improving public service outcomes. Such forces can include political pressure to award ad hoc wage 56 increases or to expand the size of the public workforce in the run-up to elections. Alternatively, fiscal pressures can force governments to enact crude and untargeted cost-cutting measures in response to external economic shocks. In all such respects, Lesotho is no different than many other countries. 124. Inequality and a lack of clarity in the remuneration and benefit policies for different categories of staff appear to be another factor in the inefficiency of spending related to compensation. Although the share of the wage bill for statutory offices is insignificant as a part of the total wage bill, a disaggregation of wage expenditures indicates that they are have a rather significant impact on the wage and allowance spending for statutory offices. According to the Legal Notice 2016, for example, staff members of the PSC and Independent Electrical Committee (Category C) receive M 401,628 (US$ equivalent) per month, whereas a similar category of civil servants receives M 160,000 to M 300,000 (US$ equivalent). In addition, statutory staff members receive a good package of benefits, such as a rent-free government house or housing allowance (M 2,500–M 4,000 per month), a vehicle or car allowance, a one- time interest free loan, and tax-free utilities allowance. However, civil servants have no such allowances, and only have access to car and housing loans as benefits (except for staff at Grade F and above who are eligible for a government house). Furthermore, as part of the benefit policy for retention and attraction of doctors, rent-free housing has been provided to them. Another issue is the use of government vehicles. There is a reference in the Supplement of Public Service Act 2011. It states that a public officer is entitled to transportation by a government vehicle, if necessary, while performing official duties during working hours. However, the regulation and application of this policy does not appear to be clear with regard to entitlement and practice. Given the fact that government spending for the vehicle fleet represents a significant portion of the total budget, a clearer regulation will be required. In general, while the special allowances could be relevant to addressing specific needs and situations, they also raise questions about inequality among public servants. This may require a standardization of benefit policies, especially on housing (rent-free housing, allowances, and loan) and vehicles. 125. In the absence of ad hoc measures that materially impact wage levels or employment, Lesotho’s wage bill is projected to decline from 18.0 percent to 17.5 percent of GDP over five years. This expected modest decline in the wage bill is not attributable to an easing of fiscal pressure from wage spending, but rather to expectations for increasing robust, real economic growth. Even minor fiscal shocks—such as further declines in the revenue received from the SACU can increase the wage bill. So too, slightly slower-than-anticipated levels of economic growth, modestly higher-than-projected levels of recruitment in the public sector, and/or small ad hoc increases in wages could easily push the wage bill beyond 19 percent of GDP within a year — and prevent it from falling below its current level for the foreseeable future (figure 31). 57 Figure 31. Wage Spending Projections (percentage of GDP) Source: World Bank estimates. Note: Based on data provided by the authorities and the World Bank’s most recent fiscal framework. 1/ Assumes real GDP contracts by three percentage points in 2017 but grows as projected thereafter. 2/ Assumes nominal GDP grows by one percentage point less than is projected under the consultant’s baseline projection. 3/ Assumes recruitment is higher than projected by 500 workers for police and 250 for defense from 2017–2019. 4/ Assumes annual wage increases average 0.5 percentage points higher than the rate of inflation. Measures for Containing the Wage Bill and Increasing the Performance of the Public Service 126. In the short term, constraining the growth of public wages can be accomplished by changing the existing compensation policies and/or enforcement and implementation of the existing laws and policies. The two tiers of across-the-board salary increases representing 7-9 percent per year have not only created a significant fiscal burden but have also generously rewarded nonperforming staff. It would be advisable to limit the wage increase, move to the combined salary increase system, and limit notch increases to better performing staff. In addition, enforcement and implementation of the existing policies defined in the law, such as performance-based pay, would be urgent and critical not only to containing the wage bill, but also to improving the performance of the public administration. 127. Similarly, the Government will need to address the size of the public-sector workforce. As noted, the growing public sector resulted from uncontrolled hiring in the lower grades. Therefore, one option could entail implementation of a hiring freeze, limiting the hiring in nonessential positions or grades, or attrition-based downsizing. The age distribution analysis indicates the potential timing for restructuring and rejuvenation of the public service in a strategic manner. The Public Service Act defines the retirement age as 60 years for civil servants and 65 years for teachers. From the current age distribution (figure 32), there will be a batch of retirees in 2020/21, 2030/31, and 2041/2041. In light of the current bottom-heavy structure which does not seem to reflect real manpower needs, the timing could be an opportunity for the Government to consider an attrition-based resizing of the public service. Accordingly, it could address the fundamental structural and organizational problems, such as the duplication of services and institutions or staff assigned to nonqualified positions. From the strategic perspective, it could also be an opportunity to recruit competent and well-qualified young people. 58 Figure 32. Age Distribution of Public Servants (2015/16) Sources: MoF/Public Service, World Bank staff estimates. 128. This section examines savings estimates for measures that could be deployed in the short term to cope with lower-than-expected SACU revenues or other fiscal contractions. Savings are cumulative as a percentage of GDP, and relative to the baseline projection. It should be emphasized that these measures are not presented as recommendations. Rather, the 30 general, across-the-board measures are better suited to times of economic crisis and cannot be properly substituted for structural reforms aimed at wage levels or the ‘rightsizing’ of public employment. Instead, the measures are presented as illustrations of the savings that could accrue from the (limited) portfolio of tools now available to the Government for adjusting wage spending in the short term. • Annual COLA. Suspending the annual cost of living increases for one year would provide immediate and substantial fiscal relief. However, it would likely be highly contentious. Providing increases that lag inflation by some modest amount, though, would be less likely to be as contentious because real wages would still rise ( but not as much) for workers receiving annual notch increases. At the same time, the fiscal relief would be less immediate, compounding gradually over time (table 27). Table 25: Annual COLA Cost of Living Increases 2017/18 2018/19 2019/20 2020/21 Suspend for one year (2017) 1.08 1.08 1.07 1.07 One percentage point below inflation (2017-2020) 0.17 0.33 0.50 0.65 Two percentage points below inflation (2017-2020) 0.34 0.67 0.98 1.29 Three percentage points below inflation (2017-2020) 0.51 0.99 1.46 1.91 Source: World Bank staff estimates. • Attrition. In the absence of mechanisms to retrench redundant workers, attrition-based policies are the only mechanism available for reducing the size of the workforce in absolute numbers (table 28). 59 Table 26: Attrition Attrition 2017/18 2018/19 2019/20 2020/21 Fix the size of the workforce at its 2016 value 0.17 0.34 0.51 0.68 Replace two of three departing workers 0.23 0.47 0.70 0.95 Replace one of three departing workers 0.29 0.59 0.89 1.21 Source: World Bank staff estimates • Temporary and contractual workers. Reducing the number of temporary workers and choosing not to renew the appointments of contractual workers is permissible under the existing labor laws and regulations. However, the resulting fiscal relief would be modest given their small numbers (table 29). Table 27: Temporary and Contractual Workers Temporary and Contractual Workers 2017/18 2018/19 2019/20 2020/21 Reduce by one-quarter over three years 0.05 0.11 0.16 0.14 Reduce by half over three years 0.11 0.22 0.31 0.29 Source: World Bank staff estimates. 129. Across-the-board measures, such as those discussed earlier, may generate meaningful fiscal relief in times of economic crisis — but could be problematic from the labor policy perspective. • Across-the-board measures aimed at wage levels, for example, can introduce or exacerbate disparities in the relative compensation paid to workers of varying skills, experience, and performance. This could, in turn, potentially create incentives for workers filling critical positions and for higher-performing workers to leave public employment. It should be noted that the existence of a wage premium for the average public-sector worker does not necessarily imply that all such workers receive higher compensation than their private sector counterparts with comparable skills and experience. • Similarly, across-the-board measures aimed at employment (such as attrition-based policies) can result in the loss of critical expertise or skills, ultimately leading to a reversal or deterioration in the quality of essential public services. To some degree, these problems can be addressed by applying differential policies to different categories of workers or by targeting (that is, focusing on non-essential positions or exempting critical sectors). However, neither are a proper substitute for medium-term structural reforms in terms of improving the existing wage and employment policies. 130. Over the medium term, structural reforms are the way to improve the efficiency and effectiveness of wage spending. Although total public employment relative to the working-age population in Lesotho may be roughly comparable to a broad sample of developing countries, considerable room still exists for improving how labor is deployed across Lesotho’s public sector. As noted, reforms that empower the Government to easily retrench redundant workers and dismiss those who are underperforming, for example, would give it the tools it currently lacks to eliminate overstaffing. It would also help address the needless duplication of functions across the public sector. Easing restrictions on redeployment would enable the Government to better match personnel resources to overall staffing needs. Reforms that address wage- setting policies offer the potential to level the existing disparities in the relative compensation of workers with varying skills, experience, and performance while, at the same time, reducing average wage costs. 131. Structural pay reforms have the potential to offer both fiscal relief and improvement in workforce incentives and fairness, thereby improving morale. The goal behind all wage-setting policies, 60 of course, is to provide sufficient remuneration to attract and retain qualified workers, as well as to create incentives for them to perform to the best of their ability over the course of their careers. In the public sector, however, this goal is often easier stated than met. There are many reasons for this, but the most relevant include the strong job protections enjoyed by public sector workers in most countries. In Lesotho, as in many developing countries, a job in the public sector is effectively a guarantee of lifetime employment. There are, however, challenges involved in evaluating performance in the public sector when outcomes can be harder to measure. In this context, policy makers must contend with the difficulties inherent in the design and implementation of wage-setting policies that actually succeed in motivating civil servants to work harder and better. 132. Linking compensation to performance is a desirable option that is already underpinned by the Public Service Act. In many countries, the introduction of performance-based pay has fallen short of expectations. Sometimes, the causes have been country specific, but international experience suggests that performance-based pay has failed primarily because of: (i) insufficient financial support (performance- based pay is, by no means, cost-free); (ii) inadequate institutional buy-in and commitment; (iii) a lack of institutional HRM capacity; (iv) poorly designed mechanisms for performance appraisal; or (v) a lack of confidence in the scheme, or perceptions on the part of public sector workers that performance awards are inequitable or capricious. Building on the supplement No. 2 to Gazette No. 43 dated April 8, 2011 that stipulates the conditions of employment for public officers50, Lesotho once introduced Performance-based Management (PBM) in 2014 at the Minister and Permanent Secretary levels. However, its enforcement failed precisely because of the afore-mentioned reasons. 133. The result of a survey on the current HRM practice indicates a readiness for introduction of PBM; however, a cautious, step-by-step approach for its enforcement is required. In 2014/15, the Government attempted to introduce such a system for the ministers and permanent secretaries. Although a few ministers and permanent secretaries signed the performance contracts, PBM was ultimately not implemented or sustainable due to a lack of appropriate evaluation mechanisms, as well as performance monitoring and incentives. However, according to the recent survey (box 3), 70 percent of civil servants responded favorably to an introduction of PBM. If well-designed, applied and evaluated, PBM could increase motivations to perform better by all staff, including better performing and even nonperforming staff. If adopted, it is recommended that it first be introduced on a pilot basis with a simple and feasible evaluation system, and then implemented across the entire public administration. Box 3. HRM Perceptions of Civil Servants in Lesotho The Ministry of Public Service conducted a survey on HRM for randomly selected civil servants in April 2017. A total of 159 participants, including 59 percent of male staff and 41 percent of female staff, responded. The respondents included Grades A–E: 18 percent; Grades F–G: 40 percent; Grades H–I: 33 percent; Grades J–K: 8 percent; and Grade M: 1 percent. Nearly 60 percent of the respondents have a 10-year plus career. The results indicate the application of different human resource practices from those stipulated in the Public Service Act. Although the entitlement of allowances is clearly defined with circulars, 78 percent of respondents do not receive any. A total of 69 percent of respondents are entitled to career advancement, but it takes place at different times and with varying frequencies (36 percent responded whenever, 26 percent responded every 6 months, and 26 percent responded every 12 months). However, the Public Service Act stipulates annual automatic promotions. A total of 76 percent of staff joined the public service through a competitive process, and 24 percent joined through different paths, including personal contacts. Respondents particularly wish for changes regarding salary increases, continuous training, the promotion system, and the introduction of performance management. 50The issue of performance appraisal and pay was also discussed more recently in “Performance Management and Related Pay Policy,� dated June 2016, promulgated by the Ministry of Public Service. 61 Top 10 Agree and Strongly Agree Top 10 Disagree and Strongly Disagree 100% 50% 0% Vacant positions are filled efficiently in quick time frame My remuneration package compares favourably with what I can earn outside the GOL Benefits Access to company sponsored training and seminars My remuneration is based on how well I do my work I receive sufficient training to do my work well. Salary Connection between pay and performance Equality amoung employees is practised with regard to promotion opportunities The GOL is interested in developing employees 134. The introduction of performance-based pay in Lesotho needs to take a phased approach given the limited HRM capacities; it should be designed around existing wage-setting mechanisms. Structural pay reforms should abolish the existing de facto annual, automatic notch increases given across- the-board to all workers who are not yet at the highest point in their pay grade — and irrespective of their performance. Instead, notch increases should be rationed such that increases are awarded only to the highest-performing workers. Rationing notch increases will require: • The development of clearly defined performance criteria for all categories of workers across the entirety of the public sector; and 62 • The design and successful implementation of a formalized and—crucially—transparent process for periodically evaluating the performance of workers against these criteria. For performance-based pay to succeed, workers should be represented when designing the reforms; managers must be properly trained to conduct fair performance appraisals. Most importantly, performance appraisals should be designed to facilitate an ongoing dialogue between employees and their managers, rather than becoming a rigidly formalized process that only results in a performance rating being issued once per year. Introducing performance-based pay is a complex undertaking that will require considerable HRM capacity. It is not the sort of reform that lends itself to ‘second tries’ — and cannot be accomplished quickly. 135. Over a longer horizon, making performance-based pay succeed will also require all existing wage tables to be revised as part of the process of conducting periodic pay reviews. As noted, the objective of pay reviews is to establish pay-parity (that is, the principle that workers with comparable qualifications and experience ought to be paid comparably, regardless of where they work). As part of this process, the existing public-sector wage tables should be revised so that the amount of each notch increase is small enough to be consistent with the Government’s fiscal capacity, but still large enough to create meaningful incentives for workers to improve their performance. 136. Performance-based pay has the potential to generate significant and enduring fiscal relief, the magnitude of which depends on the degree to which notch increases are controlled. For illustrative purposes—assuming that the process of structural pay reform was complete and could be implemented in the current fiscal year—performance-based pay would generate estimated fiscal savings (cumulative, as a percentage of GDP relative to the baseline projection) as shown in table 30. Table 28 Estimated Fiscal Savings Notch Increases Awarded Each Year To: 2017/18 2018/19 2019/20 2020/21 One in five workers 0.35 0.69 1.02 1.34 Two in five workers 0.26 0.52 0.77 1.01 Three in five workers 0.18 0.35 0.52 0.68 Four in five workers 0.09 0.17 0.26 0.34 Source: World Bank staff estimates. 137. The Human Resource Management Information System and the PFM processes need to be strengthened. Inadequate information systems and PFM processes deprive the Government of the essential tools required to monitor and control wage spending, forecast staff needs, manage recruitment, and identify employment and payroll irregularities. 138. Since 2015, the Government has initiated a major HRM reform with support from the World Bank Public Sector Modernization Project (PSMP). The MoF and the public service transferred their human resource, payroll, and pension records into one integrated HRM system (Resources Link). The recent Rapid Results Initiative I supported the cleaning up of human resource/payroll records in 100 days and has so far identified approximately 7 percent of ‘non-identified workers’ (or potential ghost workers).51 This operation will be followed by a biometric census that is intended to start in 2018. It is important to note that during field operations for the biometric census, recruitment needs to be suspended to prevent registering illegitimate cases. The revision of establishment lists should be considered as a priority. It could be led by the Ministry of Public Service together with the human resource departments of the line ministries (LMs), 51 The result is yet to be finalized. 63 with support from the PSMP. It could be completed by August 2018 to reflect the new staff hiring plan in the FY2019/20 budget. To further strengthen payroll control, the interface between the Resources Link and the IFMIS should also be a priority action. Human Resource and Wage Bill Management in the Health Sector52 53 139. The Ministry of Health (MoH) has halved its human resource vacancy rate in recent years from 44 to 22 percent (table 31). This can be attributed, in large part, to the Nursing Education Partnership Initiative, which was established in 2011. It greatly increased the number and quality of qualified nursing graduates each year. Nevertheless, a shortage of qualified staff remains one of the major health sector bottlenecks in Lesotho. The ratio of doctors to the population is 0.9 per 10,000, whereas the ratio for nurses and midwives is 10.2 per 10,000. Both ratios remain below the World Health Organization (WHO) Regional Office for Africa’s regional average of 2.6 and 12.0, respectively. Lesotho remains a nurse-driven health service, with an average of 11.58 nurses per doctor54 across the government-run health system. Maseru comprises 747 human resources for health staff followed by the next highest number of 231 in Berea. Remote areas such as Mokhotlong, Qacha’s Nek, Quthing, and Thaba Tseka, are also less staffed on average, especially with doctors. The tertiary level of care absorbs the largest share of medical doctors in the country. The QMMH staff constitute slightly less than half of all the doctors in Lesotho, with 85 (or 46 percent of the total). Meanwhile, the government staff is equal to 62 doctors (34 percent of the total) and the Christian Health Association of Lesotho (CHAL) staff consist of 36 (20 percent of the total). Nurses are more evenly distributed across facilities due to the prevalence of a large number of nurse-staffed primary health centers across the country operated by both by the Government and the CHAL. Table 29: Health Employment and Wage Bill, 2015/16 Source: IMF. 140. Although 20 percent of the posts remain unfilled, more than half of the vacant positions are concentrated in the grades for non-essential staff, reflecting an ineffective establishment list that does not reflect real staffing needs. One-third of employees in the MoH constitute supporting staff, including office assistants, clerical assistants, and drivers (table 34). Another 360 positions out of 718 vacancies are 52 Extracted from the Health PER, World Bank 2017. 53 Human resource analysis in the education sector is found in Chapter 5. 54 2,120 nurses/183 doctors = 11.58. 64 for employees other than health care workers, of which 52 percent remain concentrated in the grades for supporting staff (Grades A to C). For health care workers, 16 posts for doctors and over 340 posts for nurses, remain unfilled. If the ratio of doctors-to-population is far below the WHO Regional Office for Africa’s regional average and Lesotho’s health care is largely nurse-driven, this number of vacant positions may not reflect the real staffing needs for health care services in Lesotho. Therefore, it will be critical to conduct a staffing needs and institutional assessment to examine if the current vacancies represent genuine needs for service delivery. Overall, filling these positions would require a 13 percent increase in employment in the MoH, and a substantial increase in the health sector budget. 141. According to facility-based data, the shortage of clinical staff is also confirmed to be significant in Lesotho. The MoH-run facilities have vacancies of nearly 43 percent for nurse clinician positions and one-fourth for nurse assistant positions. Also, vacancies are quite high in the case of nurse assistants in both categories of health centers. Table 32 assesses the shortage of clinical and non-clinical staff in the health centers based on the extent of data provided from the facilities during the survey. Table 30: Status of Clinical and Non-Clinical Staff in Health Centers, 2016 Authorize In- Vacancy Owner Designation d position (percent) Nurse 82 62 24 assistant Nurse Ministry of Health (MoH) 30 17 43 clinician Nurse 100 92 8 midwife Sub-total: Clinical 315 211 33 Sub-total: Non-Clinical 169 161 5 Total MoH 484 378 22 Nurse 79 60 24 assistant Christian Health Association of Lesotho Nurse 29 24 17 (CHAL) clinician Nurse 70 74 ‒6 midwife Sub-total: Clinical 192 164 15 Sub-total: Non-Clinical 252 257 ‒2 Total CHAL 444 421 5 Source: Public Expenditure Tracking Survey (PETS)/Quantitative Service Delivery Survey (QSDS) 2016. 142. With regard to staff performance assessments, the PETS/QSDS data reveal that they are not being conducted systematically at the primary health care level. However, they are being done internally for 53 percent of the health centers (44 out of 83 health centers) and externally for 40 percent of health centers (33 out of 83 health centers). During the survey, responses from 44 health centers were received, of which 12 confirmed that the internal performance assessments were linked with salary or incentive payments. The district-wide distribution reveals that 9 out of these 12 health centers fall within the World Bank’s Health Sector Performance Enhancement Project pilot area (that is, Quthing accounts for 5 health centers and Leribe for another 4 health centers). The project uses the Performance-Based Financing approach as a key feature to enhance efficiency. 143. At the secondary level of care, with some limited information collected, vacancies for positions of doctors/medical officers appear to be higher in the district hospitals under the MoH than those in the CHAL facilities (table 33). Vacancy rates of 18 to 26 percent of medical doctors are still observed. 65 Table 31: Human Resource Positions for Doctors/Medical Officers in District Hospitals, 2016 Owner Position Number MoH Authorized 19 In-position 14 Vacancy (%) 26 CHAL Authorized 22 In-position 18 Vacancy (%) 18 Source: PETS/QSDS 2016. 144. The responses received for each of the eight hospitals reflect inconsistencies in performance assessment practices. Only in the case of the CHAL-run Maloti and Scott District Hospitals were the staff performance assessments linked with salary/incentives. Three facilities—Quthing District Hospital (MoH), the CHAL-run Malute District Hospital, and Scott District Hospital—reported that facility assessments were also linked with the financing. The staff or facility assessments were not conducted for the Berea and Paray Mission District Hospitals (annex 3). 145. According to routine data at the tertiary level, with good staff retention of the clinical positions at the QMMH network, clinical staff strength increased slightly from 584 to 590 between 2012 and 2015. The number of physicians available also increased (from 82 to 87 in the three-year span). A 29 percent increase in the allied health professional positions was also noted. However, the reduction in the number of in-position staff in the non-clinical capacity is evident. It decreased by 17 percent points in this three-year period (table 34). Table 32: In-Position Personnel in the Public-Private Partnership Facilities % Change Between Position March 2014 March 2015 March 2016 FY2013/14 and FY2015/16 Physicians 82 80 87 6 Dentists 5 5 5 0 Nurses (RN and above) 312 300 308 ‒1 Nursing assistants 95 95 93 ‒2 Allied health professionals 24 30 31 29 Clinical support services 17 16 16 ‒6 Radiology 14 12 14 0 Pharmacy 35 35 36 3 Sub-total: Clinical 584 573 590 1 Administrative 16 16 16 0 Support staff (ward clerks) 139 125 133 ‒4 Catering 44 39 26 ‒41 All other (specify) 147 135 112 ‒24 Sub-total: Non-clinical 346 315 287 ‒17 Total 930 888 877 ‒6 Sources: PETS/QSDS 2016. Compiled from Tšepong Performance Reports for March 2014, March 2015, and March 2016. 146. Regarding performance assessments, during the PETS/QSDS, the QMMH network that it had undertaken internal staff performance assessments. It also confirmed links with financial gains. Further, the facility performance assessment was conducted on a regular basis and linked to financing. Apart from the QMMH, only Likotsi stated that the facility-level performance assessment was linked with financing. 66 147. Village health workers (VHWs) constitute an important human resource component for the health sector in Lesotho. They are trained and receive a fixed compensation (M 300 per month, approximately US$35) for the various community outreach activities they conduct. The VHWs are recruited by the GoL and some donor and implementing agencies (Global Fund, Partners in Health) with a defined mandate. Whereas the VHWs recruited under the GoL receive a fixed salary, donor-funded VHWs are paid a fixed amount upon completion of some verified performance indicators. Supervision arrangements depend on the policy of the relevant donor/implementing agency. For example, the VHWs recruited under Partners in Health are organized in such a way as they can be coordinated and supervised by relatively more senior staff located at the facility level. 148. The majority of facilities acknowledge that their staffing includes active VHWs. Of the 98 facilities surveyed during the PETS/QSDS, around three-fourths (74 percent, or 73 facilities) stated that the VHWs are active in the catchment areas of the facilities.55 Furthermore, the VHWs are reported to be active in 66 out of 83 health centers. The CHAL-run health centers (85 percent, or 29 out of 34) fare well as compared to the MoH-owned health centers (79 percent, or 35 out of 44) in engagement of VHWs in the catchment areas of their facilities (see table 35). Table 33: Number of VHWs Active at the Time of Survey, 2016 Number of Number of Active VHWs Number of active Type Facilities VHWs per Female Male Total Responded Health Facility District hospital (MoH) a 1 55 8 63 63 District hospital (CHAL)b 3 161 9 170 57 FC (PPP)c 3 83 7 90 30 Health center (MoH) 33 1,547 85 1,632 49 Health center (CHAL) 27 1,194 82 1,276 47 Health center (LRCS)d 2 80 3 83 42 Total 69 3,120 194 3,314 48 Source: PETS/QSDS 2016. Note: a. Quthing Hospital (Quthing district). b. 98 VHWs in Paray Mission Hospital (Thaba-Tseka), 54 VHWs Maloti Hospital (Berea), and 18 VHWs in Scott Hospital (Maseru) c. 33 VHWs in Likotsi FC (Maseru), 30 VHWs in Qoaling FC (Maseru), and 27 VHWs in Mabote FC (Berea) d. 38 VHWs in Thaba-Bosiu HC (Maseru) and 45 VHWs in Mapholaneng HC (Mokhotlong) The data in this table pertains to the health facilities with active workers in the catchment areas. Policy Recommendations for Human Resource and Wage Bill Management 149. The following reforms are of greatest importance for improving the efficiency and effectiveness of public wage spending and HRM: 150. In the immediate and short term (one to two years), the following policy options could be considered: (a) Wage setting and pay policy The findings for VHWs are covered appropriately in this section on ‘Health Centers’ as a majority of the responses received 55 were for the health centers. 67 • To limit the annual wage increases, including COLAs, move to the consolidated wage increase combining COLA and notch increases. (b) Public employment policy • To freeze or limit hiring of non-essential vacant positions, especially in Grades A to C, for one to two years until the revision of establishment lists based on a functional review is completed. This measure can be exempted in the health sector in which supporting staff constitute critical elements in hospital management. (c) Human resource/payroll control • To complete the cleanup of human resource records and suspend salaries of identified ghost workers. • To introduce the PBM with a simple evaluation mechanism that limits notch increases only to better performers (pilot phase). • To revitalize the Establishment Committee comprising the Ministers of Public Service, Finance, and Development Planning. • To interface the Resources Link and the IFMIS. 151. In the medium term (two to three years), the following policy options could be considered: (a) Wage setting and pay policy • To develop the performance-based evaluation/pay mechanism with the revised pay policy, including the allowance systems, and introduce it on a pilot basis before rolling it out across the administration. (b) Public employment policy • To conduct the government functional/institutional review, revise, and introduce in the case of the education sector, the establishment lists and draw the strategic hiring plan for 2020/21. Do not renew unnecessary positions and/or to reassign staff according to the strategic staffing plan. • Over the long term, shift from a lifetime model of public sector employment to a more flexible model that allows for adjusts to employment in response to changing needs, demographics, and technological developments. • To outsource some of the supporting services, such as cleaning provided by Grade A staff, to private companies through a competitive procurement process. (c) Human resource/payroll control • Improving the processes governing the creation, amendment, and verification of personnel service records, particularly with respect to the incorporation of national identification numbers and biometric census data into Resource Link (the payroll and HRM system) and the home affairs database, as supported by the PSMP. • Strengthening the quantity and quality of data available for fiscal analysis, budget formulation, and the monitoring of public sector employment and compensation. • Better integration of wage bill forecasting and staffing data into the process of budget formation and execution. 68 • Strengthening transaction controls, payroll reconciliations, and auditing, particularly with respect to wage-related payments made outside the Resource Link. • Improving human resource processes and establishing controls to enhance personnel management. II. Civil Service Pension 152. The wage bill projections include expenditures attributable to the legacy-defined benefit pension scheme for public sector workers (closed to new entrants as of 2008).56 Although the legacy pension scheme may have been closed almost a decade ago, it remains relevant to the wage bill in two respects. First, benefits will continue to be paid on an ongoing basis directly from the budget for many years to come. Second, a sizeable unpaid obligation still exists for workers transferred to the Public Officers’ Defined Contribution Pension Fund (PODCPF), when it was introduced as a replacement for the legacy scheme. 153. The legacy defined benefit pension scheme is noncontributory (that is, unfunded), so benefits are paid directly from the budget. When the scheme was closed to new entrants in 2008, the existing workers born after April 1, 1969—a cutoff subsequently changed to 1964—were transferred to the PODCPF. As a result, the scheme now covers only a relatively small number of older public-sector workers (age 53 years and above) and pays pensions to approximately 7,500 beneficiaries. When the last remaining teachers covered by the scheme reach their mandatory retirement age in April 2029 —retirement ages are 55 for the military, police, and correctional facilities; 60 for the civil service; and 65 for the teaching service—the number of the scheme’s beneficiaries will inexorably decline and eventually disappear as retirees die (figure 33). Figure 33. Number of Participants by Age Covered by the Legacy Pension Scheme Source: World Bank staff estimates. 56The projections also incorporate the cost of the contributions that the government makes as an employer for workers enrolled in the PODCPF, introduced in 2008 as a replacement for the legacy scheme. Contributions made by workers to the PODCPF, however, are deducted from gross wages. Therefore, from an accounting perspective, they have no additional budgetary impact. 69 154. At retirement, virtually all the workers still covered by the legacy scheme will be entitled to a ‘full’ pension, and most will receive the maximum benefits provided under the scheme . Their benefits will include a monthly pension, the annualized value of which is computed according to the formula given below; and a gratuity (a one-time lump-sum payment) equal to 12.5 percent of the full pension. Despite the provision of an accrual rate of 2.2 percent (that is, for each year of service, workers accrue a benefit equal to 2.2 percent of their final salary), benefits are capped at 400 months (or 33.3 years) of service. Benefits under the legacy scheme are effectively indexed to inflation (that is, adjusted so that their purchasing power remains roughly constant over time) and are payable until beneficiaries die. The legacy scheme does not provide disability or survivorship benefits. (𝑀𝑜𝑛𝑡ℎ𝑠 𝑜𝑓 𝑆𝑒𝑟𝑣𝑖�𝑒) × (𝐹𝑖𝑛𝑎𝑙 𝑆𝑎𝑙𝑎𝑟𝑦) 𝐹𝑢𝑙𝑙 𝑃𝑒𝑛𝑠𝑖𝑜𝑛 = 540 𝑀𝑜𝑛𝑡ℎ𝑠 155. Most workers still covered by the legacy scheme will ultimately receive a monthly pension equal to 74.1 percent of their final monthly salary — a level of income replacement that is generous by international comparison. Even in OECD countries, accrual rates are typically substantially less than 2 percent—where the mean and median accrual rates are about 1.7, using the most recently available data (OECD 2016). This results in substantially lower benefits in retirement relative to pre-retirement wages. Furthermore, most of OECD countries rely on a lifetime measure of wages rather than final wages when computing a worker’s benefits. This translates into even lower levels of income replacement when measured against the worker’s wages immediately before his or her retirement. 156. Without taking into account the unpaid debt attributable to workers transferred to the PODCPF, the cost to the budget of the legacy pension scheme is currently about 1.1 percent of GDP per year. The cost of the scheme will rise steadily for the next few years as older public-sector workers still covered by the scheme reach retirement. In 2023, its cost is projected to peak at about 1.4 percent of GDP, after which it will begin to decline as the last cohorts of teachers retire. The cost of the legacy pension scheme will then shrink gradually as beneficiaries die. However, it is not projected to disappear completely for another 45 years or so (figure 34). Figure 34. Benefit Costs for the Legacy Pension Scheme (percentage of GDP) Source: World Bank staff estimates. 70 157. The Government has little room to alter the future trajectory of benefit costs under the legacy pension scheme. Changing the rules of eligibility for older workers still covered by the scheme —or changing the formula by which their benefits are computed—would be unfair given that all of them are within thirteen years of retirement (and more than half are estimated to be within five years of retirement). It would also be unwise when considered through the lens of social policy (in general, the benefits provided by social insurance schemes should be fully predictable for persons close to retirement). In addition, it may be so contentious as to be politically infeasible. Lowering the mandatory retirement age for teachers might help address overstaffing in the education sector. However, unless the positions they vacate are mostly eliminated, it would generate little savings in total wage-related spending (and could actually increase total spending). 158. As part of a broader policy response to a severe economic downturn, the Government could choose to change how it indexes benefits for retirees covered by the legacy pension scheme. Such a move could be justified on the grounds that initial levels of income replacement are comparatively high. Also, benefits are indexed to inflation so that their purchasing power remains roughly constant as beneficiaries grow older. Indexing benefits at a rate that slightly lags inflation for a couple of years, for example, would generate modest fiscal relief without greatly reducing purchasing power. Such a policy could generate some savings but, again, should probably only be used in extreme economic duress as part of a broader strategy of fiscal consolidation. Estimates for the savings (cumulative, in percent of GDP relative to the baseline projection) are shown in table 36. Table 34: Benefit Indexation Benefit Indexation 2017/18 2018/19 2019/20 2020/21 Suspend for one year (2017) 0.06 0.06 0.05 0.05 One percentage point below inflation (2017-2020) 0.01 0.02 0.03 0.04 Two percentage points below inflation (2017-2020) 0.02 0.04 0.06 0.08 Three percentage points below inflation (2017-2020) 0.03 0.06 0.09 0.12 Source: World Bank staff estimates. 159. In addition to the issue of ongoing benefit expenses, the budget will eventually need to absorb the cost of an unpaid obligation attributable to workers transferred to the PODCPF in 2008 . When the PODCPF was introduced, roughly 20,000 public sector workers born before April 1, 1964 were transferred from the legacy pension scheme. Their initial investment account balances in the newly created defined contribution scheme were supposed to have been funded with an amount equal to the value of the pensionable rights they had accrued under the legacy scheme. This obligation, however, was only partially satisfied when the Government transferred M 1.1 billion (US$ equivalent) (in two tranches) to the PODCPF in 2009. It seems clear that the Government did not fully cover all of its financial obligations to workers transferred to the PODCPF in 2008. What remains unclear, however, is how the unpaid balance of this obligation should be valued now. 160. Assigning a current value to the Government’s unpaid historical obligation to the PODCPF is legally and methodologically problematic. The PODCPF’s enabling legislation—the PODCPF Act of 2008—seems to have envisaged that, at the time of the fund’s creation, two steps would be taken. First, an actuarial analysis would be conducted for each worker to determine the value of his or her ‘prior service liability’. Second, the Government would transfer sufficient funds to the PODCPF to establish individual investment accounts with initial balances that—in the aggregate—would fully satisfy its liability for prior service for all workers transferred to the newly created scheme. However, neither step was actually taken. 71 161. Conceptually, two very different approaches could be taken to value the Government’s unpaid historical obligation to the PODCPF. One approach—taken by the PODCPF’s actuaries in consultation with the fund’s Board of Trustees—is to presume that the PODCPF’s enabling legislation was intended to establish a minimum benefit guarantee for workers transferred to the scheme such that, at retirement, their benefits would be at least as large as the benefits they would have received under the legacy pension scheme.57 Using this approach, the fund’s actuaries estimated the value of the Government’s unpaid obligation to be worth approximately M 5.6 billion (US$ equivalent) as of March 2016. The problem with this approach, however, is that nowhere in the enabling legislation—or in the PODCPF’s published rules of operation—is the meaning of ‘past service credit’ defined, nor is there any discussion of any sort as to what constitutes a minimum benefit guarantee. 162. A second, and seemingly more reasonable, approach presumes that the legislation was concerned only with the value of rights accrued by workers at the date of their transfer. This reading of the legislation presumes that the Government’s obligation in 2009 was limited to funding initial investment accounts at the PODCPF with amounts equal in value to the pensionable rights that workers had accrued under the legacy pension scheme up until the date of their transfer. Using this approach, the Government’s unpaid historical obligation to the PODCPF would be significantly smaller because it does not include a contingent liability associated with any sort of benefit guarantee. However, the value of this historical obligation would still have to be projected forward to the present time.58 The valuation established by the PODCPF’s actuaries is based on an interpretation of the fund’s enabling l egislation that may not actually be legally valid. In the absence of a court ruling on the matter, the current value of the Government’s unpaid historical obligation simply cannot be definitively determined. 57 Aon Hewitt (March 31, 2016). Public Officers’ Defined Contribution Pension Fund Actuarial Valuation. 58Projecting the value of the unpaid historical obligation forward in time would necessitate agreeing on an appropriate rate of return. This, too, is methodologically problematic as it is unclear whether the rate of return should reflect the Government’s cost of capital (that is, its borrowing costs) or the rates of return the funds would have earned had they been invested by the PODCPF. Regardless, the current value of the unpaid obligation would be substantially less than the value established by the fund’s actuaries. 72 Chapter 4: Public Investment Management 163. Since the mid-2000s, the GoL has been engaged in Public Investment Management (PIM) reforms. The reforms have been designed with intended links to efforts in transforming the broader PFM system. They are anchored in the introduction of the MTEF, the IFMIS, and the creation of an overarching dedicated institutional structure with a formal ‘gatekeeping’ role in project appraisal, approval and execution monitoring. However, improvements have been inadequate. Several key PIM functions have, de facto, been largely missing or exist only in rudimentary form. As the GoL recognizes the urgency for radically transforming the economy, enhancing the PIM’s quality is being prioritized as part of the strategic plan to address the bottlenecks that impede economic growth and development. 164. The focus on efficiency and effectiveness of capital budgeting to enhance growth is essential against the backdrop of high public expenditures. Public spending has surged to over 50 percent of GDP between 2012/13 and 2015/16. Driven by a high wage bill, containing the recurrent budget becomes a real challenge over the short to medium term. With the decline in SACU revenues and heightened fiscal pressures on meeting the growing demand for infrastructure and quality investments to improve the human development indexes, the Government’s plan to shift the composition of resource allocation from recurrent to capital expenditures is the correct policy decision. However, it will need to be accompanied by consistent and well-sequenced PIM reforms. The GoL already allocates a large proportion of government spending to capital investment as compared to the SSA average. In this context, improving the returns to investment will be most important in enhancing growth prospects. 165. This chapter analyzes the institutional and procedural issues in the functioning of the PIM system in Lesotho to shed light on key policy implications for PIM reforms. The chapter is structured as follows. Section I provides an overview of public investment trends. Section II presents an overview of capital budgeting trends over the past 10 years. The chapter also analyzes the pattern of public investment in terms of the level, composition and interrelationship with the dynamics of private investment. Section III presents a brief overview of the institutional setting and the key players in project cycle management (PCM). Section IV covers the main findings in the diagnostic assessment of PIM functionality. Section V concludes with policy implications. I. Trends in Public Investment 166. Hard infrastructure has been receiving a steady, majority portion of capital budget allocation, whereas allocation to the education and health sectors is trending down or remains at relatively low levels. This review covers public investment trends from 2007 as the expenditure side of GDP is only available since 2007 (due to the recent rebasing). Despite the relatively high level of public investment compared to its peers in the region, Lesotho lags in the achievement of major infrastructure and social development outcomes, as will be shown in later sections. This implies an urgent necessity — but also an opportunity, given the high level of spending — to enhance PIM efficiency to effectively convert capital investment into tangible development outcomes. 167. Public investment increased significantly from 2007-2015, whereas private investment remained relatively stable. Figure 35 shows that public investment rose from 4 percent of GDP in 2007 to 12 percent in 2015. During the same period, private investment remained at a relatively stable level, whereas gross capital investment increased.59 This implies that public investment drives the pattern of total 59Lesotho’s national accounts divide investment into central government and all other sectors. The “all other sectors� component may contain investment that is carried out by the SOEs. However, most of the investment is financed through on-lent loans, and the SOEs are the implementing agencies of many projects. Although there might be some discrepancy, the “all other sectors� component can be considered as private investment. 73 investment in the country—contributing to the rise of GFCF as a percentage of GDP? from 21 percent of GDP (2007) to 28 percent (2015), reaching its peak at 32 percent in 2012 (the year when the public investment increased to its highest level of 17 percent of GDP). Figure 35. Lesotho Public, Private, and Gross Capital Formation (as a percentage of GDP), 2007–2015 35% 30% 25% 20% 15% 10% 5% 0% 2007 2008 2009 2010 2011 2012 2013 2014 2015 GFCF (% of GDP) Central government (% of GDP) All other sectors (% of GDP) Sources: Lesotho Bureau of Statistics, national accounts. Note: GFCF= Gross fixed capital formation 168. Public investment in Lesotho is relatively higher than in SACU countries. The level of public investment in Lesotho was at the top of the SACU during 2011/12–2013/14, but was surpassed by Botswana thereafter (figure 36). Public and gross investment in Lesotho were also high as compared to the broader set of 45 SSA countries. Lesotho had an average investment of 20 percent of GDP in 2015, and government investment of 2.7 percent of GDP in 2012. The World Bank (2017) classifies Lesotho in the group of high investment countries (defined as a cluster of countries with public investment of more than 10 percent of GDP). 74 Figure 36. Public Investment in SACU Countries (as a percentage of GDP), 2011/12–2015/16 16.0 14.0 12.0 10.0 8.0 6.0 4.0 2.0 0.0 2011/12 2012/13 2013/14 2014/15 2015/16 Botswana Lesotho Namibia South Africa Swaziland Sources: WDI and Lesotho national accounts. 169. Although public investment in Lesotho has shown an encouraging upward trend, the composition of the capital budget allocation has changed over time. As shown in the table 40, during the period of the NSDP 2012/13-2016/17, the allocation of the capital budget over different sectors fluctuated. However, as indicated in table 37, in FY2016/17, two-thirds of budget was allocated to hard infrastructure, the finance sector and energy. The general trend seems to allocate more resources to some parts of the NSDP, such as public works and transport, than to the key ‘soft’ social sectors such as education and health.60 Chapters 1 and 5 discuss overall spending on the social sectors in greater detail. However, it is important to note that in the social sectors a large portion of spending is allocated to recurrent expenditures.61 In the health sector, for example, 90 percent of the budget is allocated to recurrent expenditures, whereas in the education sector, this figure rises to 95 percent.62 Investment in education was uneven, reaching a level of 5 percent of total public investment (FY2010/11, FY2014/15), but dropping to just less than 2 percent in FY2015/16. Similarly, allocations to the health sector decreased sharply from almost 15 percent (FY2010/11) to 4 percent (FY2016/17) of the total capital budget.63 60 The data used in this chapter is derived mostly from Lesotho’s integrated financial management systems, as discussed in the overall macroeconomic chapter. However, there are several issues with data recording. The IFMIS does not have an economic classification, and the data are only an aggregate at the ministerial level. Also, they might not correspond to the economic classification. 61 The education chapter notes the limitations of school facilities. About 22 percent (as a percentage of reported primary students in ER42 forms) of the primary students are accommodated in community halls and other community and church facilities. Five percent of the primary school students attend school in the open air. There is roughly one latrine per 50 students and 2 percent of these are flush toilets. 62 The figures cited here are the budget allocations in 2016/17; actual spending on recurrent expenditures is probably higher given the higher execution rates of the recurrent budget. 63 The GoL received a five-year compact from the United States Millennium Challenge Corporation in 2008. Through the compact’s health component, 14 outpatient health centers were renovated and a national reference laboratory was built. The decrease in the health allocation is partially attributable to the end of the compact. 75 Table 35: Sector Investment as a Percentage of Revised Total Capital Expenditures in Lesotho (FY2010/11– FY2016/17) Development Expenditure 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 Public works and transport 16.6% 7.9% 14.1% 18.1% 21.3% 23.5% 22.3% Finance and development planning 16.3% 36.0% 5.3% 12.8% 14.0% 19.9% 14.9% Finance 16.3% 36.0% 5.3% 12.2% 13.6% 18.8% 13.1% Development Planning 0.0% 0.0% 0.0% 0.5% 0.5% 1.1% 1.8% Education and Training 5.0% 1.8% 2.3% 3.5% 5.0% 1.8% 2.7% Agriculture and Food Security 1.4% 0.5% 2.8% 3.8% 4.0% 5.7% 4.7% Health 15.1% 0.0% 11.4% 7.8% 2.7% 4.4% 4.0% Energy and Meteorology 24.8% 36.1% 36.8% 28.0% 24.3% 4.4% 7.2% Forestry & Land Reclamation 3.1% 1.5% 1.9% 2.8% 3.1% 3.3% 2.7% Local Government and Chieftainship Affairs 8.0% 7.7% 12.1% 14.1% 8.9% 8.9% 7.6% Home Affairs 0.7% 0.5% 5.3% 1.6% 2.3% 2.5% 3.0% Trade and Industry 1.2% 3.1% 2.0% 2.2% 3.3% 8.7% 2.0% Tourism, Environment and Culture 1.0% 0.5% 0.4% 0.2% 1.1% 1.6% 1.3% Communications, Science and Technology 0.9% 1.9% 1.1% 1.2% 5.2% 4.3% 4.5% Gender, Youth and Sports 4.4% 0.6% 1.0% 1.0% 1.9% 2.0% 0.4% Water 0.0% 0.0% 0.0% 0.0% 0.0% 2.6% 18.5% Other sectors 1.6% 1.8% 3.7% 2.8% 2.7% 6.3% 4.1% Total Investment 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Sources: Lesotho BOOST (World Bank), and authors’ calculations. 170. Although public investment has been trending upward and is much higher than the regional average, Lesotho lags its peers in development outcome attainment in both regional and cross- regional comparisons. Table 38 shows that Lesotho is outperformed by the regional average in all major infrastructure indicators, except for access to water. During 2010–2015, only 23 percent of the population in Lesotho had access to electricity compared with almost 31 percent in SSA. The differences are larger when the comparison is done for the rural population. Similarly, regarding the share of the population with access to the Internet, the country ranks below the SSA average (7.5 percent versus 9 percent, respectively). It should also be noted despite the high level of public investment, the average in all the major infrastructure indicators is either close to the SSA average, or lower than that of other regions or groups of countries. Lesotho performs much better than the SSA average in access to water, implying that the investment in water has indeed translated into better outcomes. However, Lesotho is a water-abundant country compared to SSA. In addition, the MoF and Ministry of Development Planning (MoDP) (2015) stress the critical need to enhance the quality of investment in the human development sectors. The country ranks 160 out of 188 countries on the 2016 United Nations Human Development Index, falling into the category of low human development. Approximately 57.1 percent of the population lives below the national poverty line. An estimated 39 percent of households are also vulnerable to food insecurity. Around 70 percent of the population resides in rural areas and most are subsistence farmers. Despite the availability of water, the amount of irrigated agricultural lands is extremely limited compared to the other lower-middle-income and low-income countries in other regions. 76 Table 36: Comparative Infrastructure Investment Outcomes SSA (HIC Infrastructure indicators (by peers, region and income levels) LSO SWZ ZMB ZWE SSA SAR ECA EAP LAC MENA and UMIC) Access to clean fuels and technologies for cooking (% of population) 30.0 34.2 15.7 30.9 12.1 30.0 86.9 23.4 64.2 49.2 73.7 Agricultural irrigated land (% of total agricultural land) 0.0 3.7 - - 2.1 25.8 7.2 25.8 1.7 2.1 5.9 Mobile cellular subscriptions (per 100 people) 79.1 67.7 64.9 80.3 60.0 70.1 94.8 137.4 116.7 137.4 81.2 Improved water source, urban (% of urban population with access) 94.4 93.4 85.9 97.3 86.8 91.6 96.8 92.2 91.9 68.8 87.1 Improved water source, rural (% of urban population with access) 76.8 67.7 48.6 67.7 59.1 84.0 84.2 75.7 71.7 46.6 71.5 Improved water source (% population with access) 81.3 73.2 63.4 77.4 69.1 86.0 88.6 80.8 82.6 84.5 80.0 Improved sanitation facilities, urban (% of urban population with access) 37.1 63.1 56.0 49.6 40.1 87.8 94.1 78.0 75.6 68.8 88.5 Improved sanitation facilities (% of population with access) 29.5 57.3 43.4 37.3 28.0 53.0 91.5 59.5 98.4 75.7 64.7 Improved sanitation facilities, rural (% of rural population with access) 26.8 55.7 35.0 31.3 20.4 46.5 87.8 51.0 46.6 60.8 54.1 Individuals using the Internet (% of population) 7.5 22.0 15.1 12.5 9.8 13.8 31.0 16.7 19.3 32.2 24.1 Access to electricity, urban (% of urban population) 60.8 100.0 54.1 85.4 59.2 95.2 99.9 85.3 91.5 92.1 87.6 Access to electricity, rural (% of rural population) 9.5 25.2 3.6 11.4 15.3 73.4 99.8 55.3 61.2 75.5 45.4 Access to electricity (% of population) 23.3 57.3 23.9 35.5 31.1 79.2 99.8 62.5 78.2 85.9 69.3 Source: WDI. Note: The regional averages are calculated using the data for lower-middle-income and low-income countries in region. The last column provides the regional average for SSA countries in the high and upper-middle-income categories. EAP = East Asia and the Pacific; ECA = Europe and Central Asia; HIC = High-income country; LAC = Latin America and the Caribbean; MENA = Middle East and North Africa; LSO = Lesotho; SAR = South Asia Region; SSA = Sub-Saharan Africa; SWZ = Swaziland; UMIC = Upper-middle-income country; ZMB =Zambia; and ZWE =Zimbabwe. 171. An analysis of the structure of capital financing by source highlights that a considerable portion of capital investment is funded by external sources. Loan and grants combined already account for almost 40 percent of the total capital investment in FY2015/16 (figure 37). Figure 37 implies that further stretching of the portion of external financing would be a reliable strategy to balance raising the level of capital expenditures and enhancing growth in addition to improving PIM quality and efficiency of the budget financed investment.64 It is worth noting also that the share of grant and loan financing in the total capital budget over the decade (FY2007-FY2017) reached the peak (to the order of 50-54 percent) in the middle of this period, and has been consistently lower thereafter (table 42) Figure 37. Sources of Capital Financing (FY2015/16) Grant funded 13% Loan funded GoL GoL funded 25% funded Loan funded 62% Grant funded 64 Lesotho has very limited access to international capital markets, and most of the external borrowing is limited to concessional financing. Domestic borrowing opportunities are also very limited given the small size of the markets. 77 Source: Lesotho MTEF. 172. The rate of capital budget utilization improved from FY2007/08-FY2016/17. In aggregate (combining all financing sources), the outturn began at just 66 percent of the amount budgeted, but rose sharply and steadily, reaching a level of 80 percent or higher for the remaining years (table 39). However, during the review of the findings, the MoDP’s senior management cautioned that the higher rate of utilization does not necessarily imply more efficient or effective PCM. More specifically, the high outcome in outturn resulted from a last-minute push for disbursements at the end of a fiscal year. This may not have followed the design of the projects. As such, capital resources may have not been spent in accordance with the intended purposes. Table 37: Comparison of Budgets and Outturns for Capital Expenditures (in Maloti, millions) 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 Total capital budget 1,924 2,157 3,450 3,445 4,432 5,423 4,961 5,147 4,888 5,162 GoL funded 1,104 1,112 1,989 1,689 2,072 2,504 2,667 3,130 3,006 3,214 Loan funded 331 355 383 502 770 1,025 963 1,123 930 934 Grant funded 490 691 1,078 1,255 1,591 1,893 1,331 894 952 1,015 Total projected outturn 1,271 1,948 2,954 2,837 4,038 4,137 4,393 3,857 4,536 4,740 GoL funded 757 1,112 2,045 1,879 2,532 1,981 2,794 2,400 2,990 2,830 Loan funded 221 409 274 130 359 649 701 966 595 939 Grant funded 294 427 635 828 1,147 1,507 898 492 952 971 Outturn as% of total budget expenditure 66% 90% 86% 82% 91% 76% 89% 75% 93% 92% GoL funded 69% 100% 103% 111% 122% 79% 105% 77% 99% 88% Loan funded 67% 115% 71% 26% 47% 63% 73% 86% 64% 101% Grant funded 60% 62% 59% 66% 72% 80% 67% 55% 100% 96% Sources: MTFF and author’s calculations. II. Brief Overview of PIM Institutional Mapping The Legal Framework of PIM 173. The Public Financial Management and Accountability Act (PFMA Act), passed in 2011, provides the legal framework for capital budgeting. The Act clarifies the roles and responsibilities of various government agencies and lays the foundation for financial controls, monitoring, and reporting about the use of public resources. The main changes regarding PIM institutional reforms made since the last review in 2012 concern the separation of the MoF and the MoDP, and the establishment of the Public Sector Investment Committee (PSIC) as a replacement for the Project Appraisal Committee (PAC). Key Public Institutions and Stakeholders Related to PIM Functionality 174. The role of the Prime Minister (PM) and the Cabinet of Ministers in PIM. In principle, the PM, the most influential and highest-level decision maker within the executive branch of Government has a strong interest in enhancing the quality of public finance, including the management of the capital budget. While the formal planning, budgeting, and implementation monitoring of public investments are supposed to be managed by the MoDP and the MoF, the Cabinet of Ministers holds decisive authority at all PIM stages. The line ministries (LMs) and other spending agencies sometimes seem to bypass the formal process involving the MoDP/MoF and the PSIC by submitting their project proposals (especially for large-scale projects) directly to the Cabinet of Ministers for approval. Therefore, if PIM reforms are to be successful, they would need to be fully endorsed by the PM and supported by the Cabinet of Ministers. While attaining the utmost support from the top of the executive branch is likely to be challenging, inherently, the PM would have a political interest in gaining popular support on the basis of the improved PIM. 78 175. The role of the Ministry of Public Service in PIM. The Ministry holds the ultimate authority and power in determining the policies regulating the size and composition of the public service. As such, the Ministry has the potential to play a critical role in capacity building for PIM reforms. As noted clearly in the World Bank report (2012), the strategy and plan for staffing and PIM training at the MoDP/MoF and the LMs should be designed and implemented in a way that accommodates requests from the LMs, but without undermining the status and role of the MoDP. 176. The role of the MoF, MoDP, and LMs in PIM. In 2013, the GoL separated the MoF and the MoDP. While the separation may have resulted in both benefits and disadvantages, some potential gains may emerge in the quality of budgeting and development planning from the renewed specialization and focus of the two ministries. However, it could also risk a duality in recurrent and capital budgeting. The MoF is mandated to ensure the effective and efficient utilization and management of public funds to promote macro-fiscal stability. The MoDP deals with development expenditures and is in charge of the formulation of national development plans and programs, as well as monitoring and evaluation, policy advice, mobilization of resources, project management, statistical information management, and national human resource planning and development. The LMs are responsible for and are involved in almost every step of the capital investment project cycle. The LMs conduct the preliminary screening and appraisal of their own project proposals. They then forward them to the MoDP for review, and to the MoF for financing. The sector perspective is to prioritize their project proposals with the primary objective of fitting into the budget envelope and being included in the budget estimates. 177. The role of the Public Sector Investment Committee at the MoDP. In 2013, the Cabinet approved the establishment of the PSIC to replace the PAC (box 4). The PSIC is directly responsible to the MoDP Minister. It is chaired by the permanent secretaries of the MoDP and co-chaired by the permanent secretaries of the MoF. The committee comprises the heads of departments, in particular, the planning units at the LMs, who are not permitted to delegate their responsibilities.65 At a glance, the provisional functions of both the PAC (box 4) and the PSIC appear to be very similar. Both are mandated to serve as gatekeepers for the quality of entry of project proposals to the pool of public-financed projects (specifically listed in the Public Sector Investment Program [PSIP]). However, the PAC has shown its dysfunction due to an unclear mandate, especially its vested authority in relation to the LMs. In an attempt to rectify this problem, the Cabinet has decided to establish the PSIC as a replacement to the PAC, but with upgraded authority. Specifically, the PSIC is now regarded as the highest administrative authority for the appraisal of projects. Only with the approval of the PSICcan proposed projects now be included in the capital estimates and the PSIP. As such, the PSIC is placed front and center as the designated institution in charge of ‘gatekeeping’ for project proposals. The PSIC has also been empowered with a significant increase in staffing. From its inception in September 2013 to mid-2017, the number of PSIC members almost doubled, from 12 to 22. Box 4: History of the PAC The PAC was revived in 2010 after approximately 12 years of inactivity. The mandates of the committee, as approved by by Cabinet, are as follows: • To critically and objectively evaluate all new capital budget project proposals in accordance with the Project Appraisal Guidelines (PAG); • To make recommendations for each project inclusion into the forthcoming budget; 65Each member has a responsibility and authority to bring to the fore any specialist knowledge that is pertinent to the Project Concept Note (PCN) and project proposal; for example, representatives from the Debt Division are well-placed to offer an opinion on the legal aspects and the overall viability of loan-funded projects. The Ministry of Public Works and Transport may pay special attention to the engineering aspects of construction projects. 79 • To approve the final PAC report submitted to inform the Budget Framework Paper (BFP) process; and • To monitor and evaluate ongoing projects. 178. The role of the Department of Project Cycle Management (DPCM) in PIM. The DPCM, also part of the MoDP, plays the role of the Secretariat to the PSIC. In practice, a project should undergo identification and initial screening in an iterative process between the Ministerial Planning Units (MPUs) and the DPCM before a Project Concept Note (PCN) and a project proposal are submitted to the PSIC for formal appraisal. As Secretariat to the PSIC, one of the principal roles of the DPCM is to assist the MPUs in ensuring that their Concept Notes and project proposals are sufficient to ensuring a reasonable probability of approval by the PSIC. Therefore, it is important that the DPCM desk officers are consulted by the planning officers at the LMs from an early stage in the project identification process. The Project Cycle Management and Public Sector Investments Policies and Procedures Manual Report also envisions the role and authority of the DPCM in the financial and physical monitoring of projects, as well as participation in all capital budgeting processes.66 179. The Public Sector Investment Database (PSID) core team. The PSID was launched in 2014. It is designed as the centralized database system containing information on development programs and projects funded from all sources. The objective of the PSID is to introduce and institutionalize transparency and accountability between the GoL, its citizens, and development partners. The PSID core team reports to the Director of DPCM. The team comprises four staff members, including two from the DPCM, one from the Department of Aid Coordination of the MoDP, one from information technologies (IT), and one PSID adviser. At the current stage, the PSID remains confined to donor-funded projects. The MoDP, for its part, continues to face challenges in achieving the buy-in and collaboration by the sectors. III. Assessment of PIM System Performance Assessment Framework and Analytical Methodology 180. The diagnostic assessment of PIM follows the framework as presented in Rajaram and others (2010). This section sheds light on the existing performance and gaps between the current status of the project management cycle and the eight ‘must-have’ features of a well-functioning PIM system. The assessment aims to sketch the trajectory of the PIM reforms over the past few years. It also identifies the current binding constraints that could have direct implications for the design of forthcoming reforms. 181. The gap analysis is organized sequentially across all eight stages of the PIM cycle, and includes: (i) investment guidance, project development, and preliminary screening; (ii) a formal investment appraisal; (iii) an independent review of appraisal; (iv) project selection and budgeting; (v) project implementation; (vi) project adjustment and monitoring during construction; (vii) service delivery; and (viii) project evaluation. The diagnostics are conducted on the basis of a combination of the review of relevant literature and field surveys of officials at the MoF, the MoDP, and other major stakeholders, including selected LMs. During the finalization of this chapter, the authors convened a workshop to present the initial findings and gather the views from stakeholders on the state of PIM, past reform actions, and challenges ahead (annex 4 summarizes the key findings regarding the eight stages). 182. The diagnostics focuses exclusively on the performance of public investments financed by the government budget. As donor-funded projects follow the donor-specific procedures, they are not under the scope of the analysis. However, given the involvement of SOEs in PIM and the critical need to review the 66 The draft manual, version 2016, was developed between the DPCM and European Union for stakeholders’ discussion. 80 performance of the emerging modality of the public-private partnership (PPP) — especially in the health sector — some pertinent assessment is also presented as part of the overall PIM diagnostics. Diagnostic Assessment of PIM Performance in Lesotho Investment Guidance and Preliminary Screening Investment Guidance 183. Strategic guidance for public investment is critical to anchoring the Government’s decisions and guiding sector-level decision makers. In addition, the preliminary screening of projects helps to root out bad projects from the very beginning. This saves time and costs by preventing such projects from proceeding to the detailed appraisal stage. 184. In Lesotho, the process of preliminary screening and evaluating of investment proposals is required according to the Project Appraisal Guidelines (PAG) and the Submission of Project Concept Notes and Proposals Instructions to Ministerial Planning Units and Public Agencies (also later in this chapter, to be referred to as the DPCM Instructions).67 These guidelines were issued by the PSIC and the DPCM respectively in 2013.68 185. In accordance with the PAG, public investment project proposals should be prepared under the guidance of strategic documents, notably the Vision 2020, the NSDP, and/or sector strategies. The Government has established a clear hierarchy in its development strategic plans. At the core is the Vision 2020—developed in 2001. 186. In practice, determining strategic development priorities presents two problems. First, the effective linking of project proposals to strategic documents cannot be properly attained without a credible forecast of the revenue envelope (domestic revenues, SACU, and grants from donors). However, the strategic goals within the scope of the NSDP are broad and aspirational. Sector strategies have been prepared for formal purposes and/or may not have been regularly updated. Hence, the mere connection between a project proposal and such goals can easily be justified. Logic and flow of this paragraph needs work- unclear. Second, it is evident that the link between planning and budgeting under the MTEF is missing, despite the fact that the MTEF is in use for sometime.. As such, policy discussion among the PSIC takes place only when the BFPs have been finalized. However, as the PEFA 2012 shows, a BFP “does not represent a full and integrated MTFF as the Consolidated BFP only includes the recurrent and capital expenditure projections by MDA (Ministries, Departments and Agencies), without the link to overall fiscal aggregates or projections by economic category.� A logical conclusion is that BFPs are scarcely reflected in the final budget. Prescreening 187. Prescreening is essentially an afterthought process. 188. Although there is an established procedure by which the DPCM assigns a desk officer to a proposed project to assist the respective LM in drafting a PCN and a logical framework of the project, such an assignment is deemed ineffective. The PAG and its attachment provide a long checklist of what is to be covered in the PCN/project proposal form — but without detailed guidance as to how such information/analysis is to be attained, or whether the analysis is based on any realistic assumptions. In 67The guidelines also cover the other stages of PIM, including project appraisal and monitoring of project execution. 68The MoDP is concurrently in the process of drafting the Project Cycle Management and Public Sector Investment Policies and Procedures Manual (also referred to in this chapter as the Draft PCM Manual), with the first draft completed in July 2016. 81 addition, the expertise in both sectors (to prepare quality a PCN) and at the DPCM/PSIC (to conduct the review) is lacking. In actuality, projects are prepared mostly on the basis of funding availability (upon the issuance of the Budget Call Circular) and are often subject to a hurried process. Options analysis is neither required, nor ever conducted. Formal Project Appraisal 189. The main purpose of rigorous project appraisal is to avoid financing ill-designed projects and to allow only economically and socially beneficial ones to pass to the next stage of selection and execution. Project Appraisal: The Formality and Practice 190. In Lesotho, appraisal of government-financed projects is the responsibility of the LMs. When conducting their appraisals, sectors are supposed to work closely with and seek assistance as necessary from the PSIC/DPCM. 191. As highlighted, the PSIC and the DPCM have developed the PAG and the instructions, which are supplemented by the manual to guide project appraisal. 192. There is a disconnect between the guidelines and the actual application of the upstream project appraisal. The PAG and other supplemental documents require projects to undergo careful financial and economic analysis as part of appraisal, provided that ‘the project is financially viable and has a positive economic return’. However, neither the PAG itself nor the supplemental instructions would offer any specific guidance on how economic analysis (cost-benefit analysis [CBA], or cost effectiveness analysis) should be conducted. The attachment (accompanying the guidelines) simply provides a long checklist of what should be covered in the PCN/project proposal for submission.69 193. At either the PSIC/DPCM or LMs, there exists no institutional mechanism or discipline to ensure and verify assumptions to ensure that project proposals are within a reasonable range. Consequently, the outputs (if any economic appraisal is ever attempted) would all be wide open to interpretation and negotiation (in the end, the approval would be dependent on budget availability). Chapter 4 (Financial and Economic Analysis), section 4.12 of the PAG, refers briefly to economic viability analysis with the requirement to compute economic cash flow (page 6, PAG) and the use of social discount rate (SDR). This raises a question as to whether an SDR has ever been calculated and applied to economic analysis of all projects.70 194. At the current stage, the process flow in the upstream stages of project prescreening and appraisal is being shaped as a formality. However, no attempt has been made to make it functional. There is no set of specific tools and techniques regarding investment appraisals (in particular, methodologies to establish assumption parameters, and financial and economic cash flows). Further, there are no detailed methodologies on estimating the economic values of inflow and outflow items, and a dedicated SDR that can be applied to all projects uniformly. Without such detailed methodological or technical guidelines or manuals, the other PAG or instructions cannot be implemented simply because they have no basis for application in reality. Even if ministries submit the project document for appraisal, they frequently attempt to increase the likelihood of success by understating the resource implications and/or inflating the projected benefits. In general, project proposals and appraisals do not include an options analysis, reflecting the 69 Even the new draft manual (2016) offers no specific guidance or provisions on applicable appraisal methodology, and/or differentiated approaches to conduct financial and/or economic feasibility analysis of a project proposal. 70 In economic analysis, SDR is a national parameter; once estimated, it is to be applied uniformly to all projects. 82 formality of the entire appraisal and the optimistic bias of the LMs. In fact, no dedicated training in public investment appraisal, either at the PSIC/DPCM or LMs, has ever been conducted. Independent Review of Appraisal 195. In countries where a project appraisal is conducted by spending agencies, it must be subject to independent review to verify its objectivity and integrity. The review is critical to avoid the commonly known problem of ‘optimism bias’ by project promoters/LMs. 196. Lesotho has fully decentralized project appraisals conducted by the LMs, subject to an independent review by the PSIC. 197. Again, this function serves as a formality only. The long-standing problem is attributable to a variety of reasons. First, there is a lack of appraisal tools and techniques, which should be imposed uniformly upon the LMs to apply in their conduct of project appraisals, as well as to the PSIC/DPCM in the review of the appraisals submitted by sectors. Second, without a standardized set of appraisal methodologies, the planning and implementation of training/capacity building (at the PSIC/DPCM and at the LM levels) cannot be undertaken. 198. To date, there are no statistics available to indicate if any project proposal has ever been rejected outright at the PSIC/DPCM review level. In addition, the number of projects being reviewed is limited. For example, during 2013–2016, only 62 projects, mostly small and medium in size, have been evaluated. In addition, such an appraisal review does not have a direct implication for financing: of the total 62 projects approved, only 30 (or more than 48 percent) were financed. 199. Project Selection and Budgeting Appropriate procedures for selection and financing of projects are required to ensure that the projects are implemented and delivered on time, according to their intended purpose, and within budget. 200. Over the past decade, Lesotho has made considerable effort to revive and strengthen the PSIP approach. The latest PSIP aims to provide a framework for and establish a list of ‘pipeline’ projects to be planned, financed, and implemented during 2012/13–2016/17. The five-year PSIP is supposed to set the parameters for the preparation of successive annual capital budgets through the rolling over of three-year PSIPs as part of the annual MTEF process. 201. The effectiveness of the PSIP remains elusive due to a number of reasons. First, the disconnect between planning/MTEF and annual budgeting (including capital budgeting) and the lack of a credible estimate of the midterm revenue envelope have rendered the PSIP a simple wish list of projects. As such, they are not necessarily considered during the annual budget preparation and/or negotiation. Second, no legal binding mechanism exists that requires only those project proposals listed in the PSIP to be selected for annual capital budget allocation. 202. The existing guidelines and templates on project proposal submission and appraisal do not consist of transparent criteria for capital financing prioritization. Therefore, they do not function as a vetting mechanism to control the ‘inflow’ of projects into the PSIP. In actuality, the selection of major projects for government budget financing is not necessarily based on the PSIP list. 203. Regarding budgeting, the first stage is the MoF issues call circular to LMs to prepare BFPs. 83 204. The financing decision about capital projects is effectively the result of a joint give-and-take process during the annual budgeting exercise across the MoF, the MoDP (represented by the PSIC), and the LMs. The selection of major projects for financing (budget-funded projects) is actually based neither on the outcomes of project proposal documents submitted by the spending LMs nor on the PSIC assessment of such proposals submitted by the LMs. Rather, for large-scale projects, sectors may simply bypass the formal process and override the review by the PSIC to obtain direct approval by the Cabinet of Ministers. 205. Even with the MTEF, the DPCM instructions clearly mandate that sectors collect ‘standard unit costs from national and international experience and from technical manuals. Ideally, this information would be updated annually’. This is a vague requirement. There exists no institutional process to enforce the LMs to develop—or the MoDP (or the MoF) to centrally manage—any database of standard unit costs. Despite the MTEF, dual budgeting continues to be the norm. Regarding PIM, there are no accountability mechanisms for the LMs to prepare credible and verifiable recurrent cost implications for capital budgeting. Project Implementation 206. Project construction should only begin when it is ‘shovel ready’,71 that is, when all the financial and human resources are fully mobilized, the construction site has been cleared, the project design approved (with transparent organizational arrangements), and a realistic timetable to complete the project on time and within the budget has been realized. Project Commencement 207. The PFMA Act of 2011 specifies that no expenditures be incurred before the Parliament’s approval, either in the form of a Continuing Appropriation or through a Supplementary Appropriation. There remains, however, a contradiction between the Government’s assessment and actions when dealing with projects considered not fully ready for implementation. In 2012, the MoFDP acknowledged the risk of wasteful spending due to hurried appropriation of the capital budget to such projects. It suggests that funding may not be made available until all preparatory actions have been completed; including, for example, a project site secured, the design completed, all pertinent permits obtained, complementary activities scheduled (such as provision of utilities as applicable), and implementation modalities in place. The reality is the opposite: whenever a project is selected for financing, the pressure to start construction mounts — even when it is clearly not ‘shovel ready’.72 Procurement 208. Long-standing inadequacies in both procurement legislation and practice have made untenable the idea of value for money in the implementation of public investments. The Procurement Regulation of 2007 stipulates the procurement methods and procedures to be followed. Open Competitive Tendering is, however, not defined as the default procurement method. Waivers are often granted on a discretionary basis and tend to undermine the certainty of the public procurement process. 209. The public procurement function is decentralized to all public entities. The Procurement Policy and Advisory Department (PPAD) within the MoF is mandated to play the policy, advisory, and supervisory functions. In practice, however, the PPAD cannot monitor or enforce compliance by sectors due to the 71 Different definitions can be cited for the term ‘shovel ready’. A useful one can be taken from https://en.wikipedia.org/wiki/Shovel_ready, which uses the term to describe ‘construction projects (usually larger-scale infrastructure) where planning and engineering is advanced enough that with sufficient funding, construction can begin within a very short time.’ 72 Pressure on rushing to begin construction is even higher when external financing agencies are involved in financing. Politicians, external financing agencies, and line agencies all push for the project to start as soon as funds have been approved. 84 absence of a system to collect detailed information on procurement from the LMs. Officials from the PPAD voice the following concerns on several aspects of procurement: • The Public Procurement Law has not been enacted. • No standard bidding document has been issued to guide the process. • No monitoring tool has ever been specified. • The procurement manual is not user friendly and may be subject to discretionary interpretations. • Procurement qualifications still need to be accredited. • The PPAD does not have autonomy and authority to enforce compliance. • Capacity building is very limited.73 Project Implementation 210. There are no standardized operational guidelines and/or manuals. As such, neither the MoF nor the MoDP has sufficient tools to enforce uniform procedures and standards in project execution. 211. In interviews, the Government officials shared their view that projects are frequently subject to delays. The delays can lead to higher project costs and less favorable economic returns. They could also endanger the attainment of the project's objectives. 212. Delays in payments of contractors have been persistent and have negatively affected project quality and implementation schedules. Delays in payments to consultants appear to be significant, partially due to a lengthy and complex internal procedure for payment, requiring an authorization of the Principal Secretary as the chief accounting officer, as well as delays in providing comments and approval from the beneficiaries.74 213. Project delays and cost overruns are also major concerns. However, there is no systemic database to capture such problems. Audit 214. The internal and external audit systems are formally well-defined. Under the PFMA Act, each chief accounting officer who is Principal Secretary must establish an internal audit function within their spending unit. Internal audit covers the whole system of internal controls within a ministry, including controls over the performance management system involving the PSIP and the MTEF process. 215. External audit is designed to ensure that the processes of planning and use of public resources conform to rules, regulations, and the specific intended purposes. The Office of the Accountant General is responsible for an audit of government accounts at the end of the financial year and audits specific issues 73By the end of March 2017, there were 12 staff in the PPAD. The PPAD expects to add three more by mid-2017. 74Aide Memoire of Supervision Mission on Public Sector Modernization Project and Public Expenditure Review, March 6 to 17, 2017. 85 within an organization (including performance audits, financial audits, compliance and controls audits, and forensic audits). 216. Some outstanding auditing issues remain a challenge. Physical auditing and coverage of auditing in capital projects are limited. Some audits are not up-to-date due to delays in the submission of financial statements. The Office of the Accountant General considers that at least 50 percent of central government expenditures is covered each year by its detailed audits.75 The audit of the annual statutory accounts lags by two to three years due to delays in receipt of accounts from the Accountant General. Thus, the most recent available audit report is for FY2014/2015. Project Adjustment 217. A capital project typically spans multiple years of construction. Therefore, it is necessary for the PIM system to have some built-in mechanism that would allow for project adjustments to adapt to material changes in project circumstances. Midterm Adjustments 218. With the decision to establish the PSIC, the Cabinet requires that ongoing projects be submitted for reappraisal by the PSIC in the following cases: (i) when conditions fundamental to successful project completion have materially changed; (ii) when there are substantial changes to the project schedule, inputs, activities, organization, or methodology; and/or (iii) when costs are expected to exceed the approved project budget. However, the absence of standard operational guidelines or a manual with specific provisions about both annual and total investment cost controls renders midterm project implementation review and adjustments ad hoc at best. 219. No formal mechanism exists to trigger a review of a project’s continued justification during project implementation. Indeed, projects are frequently treated as if they were perpetual programs and line agencies expand the scope of works, with no limits to escalation of total estimated costs. No progress has been made since the World Bank reviewed arrangements in 2012. Projects are seldom dropped or stopped midway, even when the external macroeconomic environment or the specific contexts of projects under construction have changed, thereby affecting the costs and/or benefits as projected at the design stage. In addition, the MoF does not have the authority to stop projects midway due to mismanagement, which can lead to issues such as cost overrun and/or delays. Monitoring 220. The 2016 draft manual stresses the importance of establishing effective monitoring to ensure that a project is being implemented as planned. 221. Even before the split in 2013, the MoDP put in place specific tools for monitoring, including financial and physical monitoring.76 Currently, monitoring is conducted in an ad hoc manner. The LMs routinely conduct the monthly financial and quarterly physical monitoring. However, the PSIC is neither clearly mandated nor does it have sufficient staff to engage in its own on-the-spot monitoring and checks 75 European Union and ACE International Consultants, PEFA Report. November 5, 2012. 76 Financial monitoring compares actual expenditures against the financial frame for the project. It provides a measure of the utilization of the project's inputs and a basic assessment of the project's progress. Physical monitoring compares the physical progress of the project activities being undertaken against its Plan of Operations. Examples of measures that might be used are: percentage completion rates on a construction project; volume of inputs distributed; and number of immunizations carried out. Physical monitoring of projects is more complex than financial monitoring since it usually relies on several information sources and different types of information and expertise. 86 as needed. Officials from Ministry of Public Works and Transport mentioned that the DPCM has a monitoring tool, but they do not use it. Also, officials from the Ministry of Trade and Industry provided their perspectives on the process, noting that the monitoring system is not computerized. Also, the reporting by the LMs is not automatic. Further, there are various monitoring templates across different LMs. The most serious loophole rendering the entire project monitoring (financial and/or physical) ineffective is that project financial disbursements are disconnected from outcomes of any monitoring findings. Facility Operation 222. Public investment can only be translated into productive assets if it fits into an integrated PIM system that can ensure effectiveness. Since the mid-1990s, the GoL has paid increasing attention to asset management and operations. The NSDP 1 stresses the need to focus on developing a public-sector asset management policy/strategy, as well as a comprehensive asset register. Asset Registry 223. An asset registry was introduced in 1995 as part of an overall attempt to enhance the quality of asset management. Each LM has its own asset register. A central registry—called ‘Contract Management Unit’ under the PPP project—controls all registers. The PPAD also has a formal role to maintain a complete inventory of public assets and provide advice on asset management and disposal. 224. After 20 years, the system remains incomplete, lacks accuracy, and is infrequently updated. The PEFA (draft 2017) presents a thorough assessment of the current state of asset management and gives it a rating of “D� (specifically for the dimension on Nonfinancial Asset Monitoring as a subset of the indicator Public Asset Management (PI-12.2). Post-handover Facility Operation 225. The efficiency of facility operations after the period of asset handover is unclear at best. The current parallel budgeting—whereby the recurrent budget is determined largely on an incremental basis and the capital budget on a residual basis —drives a wedge between annual capital budget allocation and the appropriate size of adequate operations and maintenance planning and budgeting. At the same time, Lesotho has not made any attempt to conduct periodic surveys of the level of use of the newly created assets by the intended project recipients. Project Evaluation 226. In hindsight, ex-post evaluation is essential to control the quality and inadequacies across all stages of PIM, from project design to implementation and facility delivery. Lessons from such evaluations are to be fed into the next round of reviews of overall project appraisal and management procedures, as well as the design of new projects. However, this PIM evaluation function is missing for governmental budget- financed projects in Lesotho. SOEs and Public Investment Management 227. The SOEs, which seek capital budget funds, are formally subject to the same set of guidelines and procedures as applicable to the LMs. These range from project conceptualization to prescreening, appraisal, and implementation. In practice, however, corporate governance in general — and the oversight and monitoring by central financial and planning agencies or by the responsible LMs in particular — show major inadequacies directly affecting the efficiency of PIM (box 5). 87 Box 5: PIM and SOEs in Lesotho: Transparency and Accountability Issues In Lesotho, government ownership of SOEs varies. There are SOEs that are wholly owned by the Government, partially owned by the Government, as well as SOEs within the Lesotho National Development Corporation (LNDC) shareholding structure. The wholly-owned SOEs and partial shareholdings are supervised by their associated LMs, whereas the LNDC shareholdings are supervised by the LNDC (which in turn reports to the Ministry of Trade and Industry). The MoF, with its PSD Unit, is formally tasked with exercising a central monitoring role of SOE finances, as mandated under the PFMA Act. However, with decentralized SOE ownership structures, central monitoring and coordination of the SOE sector faces multiple challenges, including: opportunity for political interference; confusion between ownership and regulatory functions in LMs; fragmented and diffused accountability; insufficient ownership capacity; and lack of adequate oversight of the sector as a whole. Implementation of public investment by SOEs continues to suffer from the lack of transparency and accountability due to deterioration in both auditing and monitoring. Auditing. The Auditor General's Office has responsibility for auditing the SOE accounts. The Public Accounts Committee of the Parliament then receives these audited accounts for review. In practice, due to staff shortages, the Accountant General's Office often subcontracts the SOE audits to private audit firms. Although the PFMA Act empowers the MoF to receive audited financial statements from the SOEs on a timely basis, the PSD Unit in the MoF—responsible for monitoring corporate governance of SOEs —does not receive these statements regularly. The board members of the SOEs also do not report back to the PSD Unit. Therefore, the PSD Unit is not able to monitor or analyze the SOE’s sector performance or its impact on public finances. This includes the net budgetary burden of the sector. Monitoring. Since the winding down of the privatization program during the 1990s to early 2000s, the central oversight function of the SOE sector within the MoF has deteriorated to the extent that there is no effective sector- wide monitoring. The PSD Unit is not involved in the budget process for the SOEs; instead, the SOEs work more closely with the LMs. Therefore, the Unit could not collect the full set of data and financial reports directly from the SOEs, nor could it obtain regular access to data available in other units within the MoF. Source: World Bank (2016a), Lesotho State-Owed Enterprises (Country Policy Note, June 30, 2016). Public-Private Partnerships as an Evolving Modality of PIM in Lesotho 228. A new investment modality, the PPP, has been introduced in the health sector. The PPP becomes an attractive option as Lesotho is faced with mounting problems in public health, particularly the low coverage of essential services and inadequate institutions. The country’s main public hospital, Queen Elizabeth II, has been deteriorating and is in critical need of replacement. In this context, the International Finance Corporation (IFC) played an advisory role in supporting the Government with structuring a new PPP in the health sector. The PPP is being used to finance the construction of the new facility, the QMMH, along with the refurbishment of three feeder clinics and the building of a new one.77 229. Seven years after the conclusion of the bidding process (in 2008), some different views of the outcomes of the health PPP have emerged. IFC (2009), the World Bank (2016b), and the Center for Global Health and Development in Boston University78 considered the PPP a success. It helped the Government to attain better health outcomes, improve the quality of care, and increase accountability for service delivery. The World Bank (2016b) also notes that the success was due to the design of performance-based 77 Tsepong (Proprietary) Ltd., a consortium of Netcare (a private South African health care provider) and several locally owned businesses, won a competitive bidding process in 2008 to partially finance and fully design, build, equip, and operate the new facility. Tsepong also refurbished and managed the three, existing community-based clinics and built a fourth clinic. The new PPP-managed hospital and its four clinics formed an integrated health care network that replaced the government-managed hospital and its three clinics. The renovated clinics opened in May 2010. The PPP-managed hospital and the new clinic opened in October 2011, when the government-managed hospital closed. 78 For Boston University’s evaluation of the project, see https://www.bu.edu/cghd/our-work/projects/evaluation-of-the-queen- mamohato-memorial-hospital-public-private-partnership-ppp-in-lesotho/ (accessed May 22, 2017). 88 contracts and the ongoing Government’s role as a steward of the health sector. However, Oxfam (2014) provides another view of this PPP. It investigated and identified several complaints7980. One concern is that more than half the country’s entire health budget (51 percent) is being spent on payments to the private consortium, led by South-Africa-based Netcare. 230. The Manual for Planning Officers, drafted by the MoDP in 2012, notes that the experience with the PPPs has been mixed and suggests that the Government be cautious, ‘as it is difficult to achieve value-for- money in view of limited internal institutional capacity, [an] inadequate legal framework, and the need to mitigate associated fiscal risks. There is also a risk that ministries use this approach as a way to avoid hard ceilings on the capital budget (since the payments are charged to recurrent expenses).’ An external academic consultant who prepared an unpublished report for the World Bank found that “the project’s financial forecasts suffered from a projection error resulting in a dramatic underestimation of government costs� (Webster 2015). IV. Ranking Priorities in PIM Reforms: The Way Ahead Summary of Findings 231. The GoL has made an effort to build the institutional framework for PIM since 2013; however, key issues across all stages of the PIM cycle have not evolved since the World Bank report (2012). The long-standing challenges render the entire system of capital budgeting and management inefficient. Some highlights can be summarized as follows: (a) The link across policy planning and budgeting under the MTEF is missing due to the absence of credible resource estimates, and the long-standing treatment of capital allocation as a residual in annual budgeting. (b) As dual budgeting persists, including a continuing disconnection between capital financing and allocation of recurrent budget for operations and maintenance. (c) The screening and appraisal of project proposals continues to be done on an ad hoc basis. There is no link between planning and annual capital budgeting. In this regard, the PSIP continues to be an instrument without a credible basis in screening and appraisal; it is regarded as being largely aspirational. The PAG and the DPCM instructions have been published. However, these documents are non-operational because: (i) the guidelines are not supplemented with a manual that would detail the differentiated methodologies for prescreening and appraisal, and that would be compatible to the scale, timing, and technology in project proposals, as well as the institutional capacity at both the LMs and PSIC levels; and (ii) the guidelines are de-linked from the actual capital budget allocation; as such, they are not conducive to generating demand for rigorous project vetting and appraisal. (d) LMs lack staff, technical capacity, and resources for the preparation of quality proposals. (e) The independent review of project appraisals has yet to be done, even as a formality. This function is supposed to be played by the PSIC. (f) Efforts have been made to create the PSID. However, it has been largely confined to donor- funded projects. For government-funded projects, neither the MoF nor the MoDP can obtain 79 Oxfam. A DANGEROUS DIVERSION. Will the IFC’s flagship health PPP bankrupt Lesotho’s Ministry of Health? April 7, 2014. 80 The Guardian Africa Network. Half of Lesotho’s health budget goes to a private consortium for one hospital. April 6, 2014. 89 the most updated and accurate statistics on progress in project implementation — even regarding the basic indicators on project delays and cost overruns. (g) Monitoring, especially physical monitoring during project execution, has been done only as a formality—as it is completely disconnected from onward financing consideration. Policy Implications 232. Given the many issues to be addressed and the politics of PIM, it is critical to prioritize the reform actions in a well-sequenced manner. The reforms are to be built and solidified based on the initial progress made to date, especially the establishment of the PSIC. It also includes ongoing efforts to prepare a full set of policies and procedures in recognition of the long-standing political and institutional challenges to the PIM functioning in key stages. The report recommends the following strategic priorities over the short-to-medium term (one to two years) to longer term (beyond the scope of three years). Empower and Make the PSIC Functional and Effective 233. Empowerment of the PSIC, including a clarification of its mandate and structure, will be necessary. The PSIC has been mandated as the highest authority in upstream reviewing of project appraisals, as well as in making decisions about the selection and inclusion of project proposals in a PSIP. However, the current organizational structure does not have a clear institutional arrangement and mechanism to impart the roles and functions of the MoDP/MoF in managing fiduciary risks during the investment period, as well as in supervising/monitoring project execution, and controlling the quality of construction, timing, and costs. One option is that the Cabinet provide the PSIC with the mandate to serve as the centralized institution for monitoring project execution.81 Some ground for such a mandate and authority of the PSIC does exist. The PSIC has been broadly envisioned to replace and upgrade the functions of the PAC. The latter was already vested with both the upstream project appraisal review and the downstream monitoring of project execution. It is worth emphasizing, however, that the PAC has never been able to implement its specific function of project monitoring. Further policy recommendations on project monitoring are presented in section 2.2. 234. To ensure that the PSIC is fully functional and empowered in playing the role of a ‘gatekeeper,’ the following set of targeted reforms will be required. All reforms are complementary and aimed at instituting the link between the outcomes of the PSIC project appraisal review and the MoF financing decision. Such an objective can only be achieved with the following targeted reforms: (a) Establish a complete, functional structure at the PSIC and its Secretariat (the DPCM) and build appropriate capacity. (b) Review and develop an appropriate set of manuals for project appraisals. (c) Develop a transparent, legally-binding ‘rules of the game’ for creation of a PSIP. (d) Operationalize a unified PIM database system. 235. Once an effective mechanism for the creation of a PSIP is instituted, the program should incorporate public investments from all modalities —including the one financed by donors. This will 81 The MoF would continue to play the lead role in managing certain aspects of project execution (for example, procurement, and financial management). 90 help to ensure a cohesive and effective response to the national development priorities and consistency with the Government’s planning and budgeting. Establish a complete, functional structure for the PSIC and its Secretariat (the DPCM) and build appropriate capacity (Timeline: Years 1–2 and continue training/capacity building) 236. First, the organizational structure of the PSIC and the DPCM need to be upgraded to make it commensurate with the scope of their gatekeeping function.82 This function would cover the entire capital budget planning and proposal process, regardless of the source of financing. The SOEs that bid or receive the budget funds to implement capital projects are also expected to be put on an equal footing with the private sector in following through with the new PIM processes. 237. The upgraded structure of the PSIC and the DPCM will only be functional with full and competent staffing capable of dealing with the various sectors (infrastructure and social). The organizational review will also be used as the basis for detailed analysis of and planning for staffing and training, as follows: (a) Prepare an overall plan for the building and retaining of capacity in project appraisal and review, including staffing needs of the PSIC/DPCM, skills requirements, and training. With the new manual for project appraisal to be published, training should reach the PSIC/DPCM and the LMs. (b) Design and conduct a combination of training of trainers (for example, executive training abroad or in-country training) and in-country training of technical staff on the basis of a capacity-building needs analysis. (c) Engage in knowledge and experience sharing events to learn firsthand the PIM reform lessons from other countries. 238. First, regarding capacity building, comprehensive training for stakeholders involved in PIM will be necessary. The training in project appraisal must not be confined to the staff at the DPCM and the PSIC. It should also include technical staff, primarily, those at the Ministerial Planning Units (MPUs). Line ministries serve the function of project owners, proposers, and implementers. As such, they are the first in line to apply the new methodologies in preparing appraisals to be reviewed by the PSIC/DPCM. Therefore, the functionality of the PSIC would be dependent on the capacity of the LMs to prepare quality financial statements. 239. Second, adopting a long-term perspective will be essential. Training is set to begin immediately in sequence with the timeline for publication of the new PIM Manual. It is important that this training will be sustained over time. To ensure that training is effective, plans should include continuous training, an appropriate human resource policy to motivate and promote trained staff, sufficient budget allocation and, above all, top leadership commitment. Given sufficient resources, this should result in full governmental functionality to conduct project appraisals and reviews. In this context, it should be noted that it took Chile 82The proposed technical assistance to support the PSIC is being led by the World Bank, and aims to facilitate such reform activities. The overall objective of the technical assistance is to: conduct an organizational review of the PSIC; develop the capacity enhancement program in the core technical areas of project appraisal; and propose the restructuring plan. As part of the technical assistance, Dr. Andre Nsabimana has completed the full diagnostic assessment and skills audit (August 2017). The report will be used as the basis and background for further development of a comprehensive capacity enhancement plan, as well as the preparation of the new operational plan, design, and implementation of training programs. 91 more than 15 years of consistent training efforts to succeed in creating and maintaining a critical mass for project appraisals. 240. Third, the forging of partnerships with academic institutions is recommended. Partnership with a university or a college in Lesotho can potentially become a critical element to ensure sustainability in building capacity in various PIM functions. The academic third-party institution could engage in support for the PSIC, notably with: project appraisal techniques and methodologies; training on appraisal, contracting, and project management; and direct participation (on ad hoc basis) in the PSIC review of financial statements for large-scale and/or technically complex projects.83 Within the first two years of reforms, the capacity at such national academic institutions is limited. Therefore, the use of international consultants for training would probably be justified. However, support by international experts will need to be designed in a way that it can be complemented with effective involvement by the national academic institutions to enable them to provide continued training at affordable costs. Review and develop appropriate a set of manuals for project appraisal (Timeline: Year 2 and onward annual review and update). 241. The PSIC should play the lead role in the preparation of a full, simplified, transparent manual for project appraisal. As analyzed, the existing PAG consists of a checklist for spending agencies, but without any reference to specific appraisal methodologies. A comprehensive set of manuals for project appraisal should conform to good international practices. PIM systems keep rely on a complete set of standardized manuals or guidelines, and prevent the inflow of ‘bad’ projects. The detailed appraisal manual would provide clarity with respect to institutional roles and responsibilities in project proposals, appraisals, reviews, and recommendations/decisions about financing. It would also provide for a simplified set of relevant appraisal methodologies applicable to different groups of project proposals (by scale, timing, technology, and sector). 242. The simplest way in which to classify projects into different methodologies is to create investment thresholds. Thresholds will need to be developed with wide stakeholder consultation to ensure a common understanding and buy-in (especially from the LMs). Although CBA and cost-efficiency analysis are the main tools for project appraisal, others (for example, multi-criteria analysis) can also be applied for practical purposes to projects at smaller scales. Lesotho is to consider introducing thresholds and appropriate methodologies for project appraisal. In this context, some lessons from Ireland, with its advanced PIM, can be learned (see box 6). The development of a uniform set of project appraisal methodologies should be piloted and revised for full implementation. Once the set of guidelines and appraisal threshold(s) are established, they need to be formalized through regulations and reflected, at the very minimum, in an inter-ministerial circular. Periodic reviews and revisions are critical to ensuring the currency and practicality of the project appraisal manual even after rollout. Box 6: Formal Project Appraisal in Ireland The Irish Department of Finance develops the Guidelines for the Appraisal and Management of Capital and Expenditure Proposals in the Public Sector (2005). The guidelines specify commensurate methods for appraisal depending largely on the scale of proposed projects. In particular, project proposals are classified into five different groups, with each subject to different appraisal methodologies: • A simple assessment should be carried out for minor projects with an estimated cost below €0.5 million (US$ 0.6 million), such as projects involving minor refurbishment works and so on. • Projects costing between €0.5 million and €5 million (US$ 6 million equivalent) should be subject to a single appraisal incorporating elements of a preliminary and detailed appraisal. 83For an example, as part of a concerted effort to build and maintain capacity in the specific area of project appraisal, Chile has established a partnership with Universidad Catolica and Universidad de Chile. 92 • A multi-criteria analysis should be carried out at a minimum for projects between €5 million and €30 million (US$ 33.4 million equivalent). • A full-fledged CBA should be carried on projects over €30 million (US$ 33.4 equivalent). • A CBA is also applicable to innovative projects costing above €5 million, which (a) involve complex or specialized issues or untried technology, or (b) involve issues that have not been previously investigated in an in-depth manner, or (c) are regarded as pilot projects for which larger programs may be modeled, or (d) would generate additional, substantial and ongoing operating or maintenance costs. Source: The Irish Department of Finance (2005). Develop transparent, legally-binding ‘rules of the game’ for the PSIP (Timeline: Year 2) 243. Learning from past failures, Lesotho needs to develop the PSIP in a comprehensive manner. According to its mandate, the PSIC is the sole central planning agency in charge of the vetting and creation of a PSIP. However, similar to many other developing countries, Lesotho has experienced failed PSIPs in the past. The quality of a PSIP depends on the due application of sufficiently rigorous pre-screening and appraisal procedures for proposed public investment projects. The PSIP should cover investment projects financed from all public sources, including domestic financing, external financing (grants and loans), and PPPs (Allen and Tomasi 2001). (a) First, the PSIP should be credibly created with a proper appraisal and review. (b) Second, the annual budget allocation of public investment should be based on the review of the projects as specifically listed in the PSIP84. Operationalize a unified PIM database (Timeline: Years 2-3 and continued upgrade) 244. The current database will require more work to ensure that it is fully comprehensive. The MoDP has started working toward the creation of PSID. The purpose of the PSID is to help the ministry to track and monitor implementation of all projects. Although progress has been made in tracking the inflow and execution of donor-funded projects, there is still a lack of support and collaboration by the LMs in extending the scope of the PSID to budget-funded public investments. A quick review of this data format indicates that major pieces of information and statistics are missing. This information would help the MoDP, and particularly the PSIC, to monitor the quality of project execution, including basic indicators of cost and time overruns. 245. The following steps and principles need to be considered to make the PSID fully functional: (a) Review the current format of the database and supplement it with key information necessary for project monitoring. (b) The project database template will incorporate the information relevant for decision making at each stage of the entire PIM, ranging from project identification/development to execution and monitoring. It will need to be as comprehensive and user friendly as possible. (c) The new template and format will be used as inputs for the subsequent development of an automated database system. Piloting can be applied to donor-funded projects, 70 percent of 84 As a side note, during fiscal consolidation, it is difficult to protect public investment, although such a policy action would promote efficiency of capital budgeting and thereby growth. A credible PSIP in combination with the discipline of budgeting, as well as effective project monitoring, are among the institutions and policy instruments that would help to lessen the temptation to cut public investments in a blanket approach. 93 which have currently been captured in the existing PSID format. This cluster of projects would lay the foundation for the piloting of a renovated structure. Over the long-term, Lesotho will need to integrate the PSID with the IFMIS system. The existing disconnect between the PSID and IFMIS is detrimental to PIM quality. Specifically, it makes it impossible to effectively monitor the physical execution of projects and their financial performance during any given fiscal year. In particular, the PSID currently has the shell, but its application is confined to the donor- financed projects. In addition, the functional PSID and its integration with the IFMIS would create the platform for the review of not just the flow of new projects, but also for the stock of existing projects. Create a direct link between project monitoring and onward financing (Timeline: From year 3) 246. To ensure the effectiveness and efficacy of project monitoring, outcomes need to be linked directly to the financing decision for on-going projects. Regarding supervision of project execution, two types of monitoring are required: (a) Financial monitoring. Financial monitoring is expected to cover the full spectrum of financial management and fiduciary aspects of project implementation. It reviews and compares the actual execution and relevant rules and regulations as applied to contract management, disbursement (broken down into periodic disbursements as commensurate to the percentage of work being—or to be—completed during the review period) and cost management (both annual and total costs, as projected for the whole project life). (b) Physical monitoring. Effective physical monitoring would look into actual progress in terms of both the quantity and quality of project execution. The information or evidence as obtained during the physical monitoring would be compared directly with the planned project implementation schedule with reference to both annual and total planned physical objectives at the design stage. 247. Financial and physical monitoring are established at the LMs and the MoF/MoDP; however, their effectiveness in managing quality, value-for-money, and timeliness of project execution remains elusive at best. The current practice of project monitoring — without any financing implication — does not exert any real pressure on implementing agencies to perform. It is in complete contrast to the practice in advanced PIM systems, whereby clear roles and responsibilities for project implementation, accounting systems, cost controls, and regular monitoring reports are established and implemented. For example, in South Korea, when monitoring detects the real cost run above 20 percent, the project under construction must be reappraised. Figure 38 depicts the central role of monitoring during project implementation in the case of Ireland. Figure 38. Schematic View of Monitoring at the Center Stage during Project Implementation Project financing decision Contracting Subject to new round of monitoring/supervision Monitoring Project modification (scope/financing) Monitoring outcomes: Issues in project execution? Project review YES Project termination if no longer Project completion cost beneficial Ex-post evaluation 94 Source: Adapted from Ireland Department of Finance (2005). 248. With political will and the necessary institutional adjustments, Lesotho could feasibly make monitoring effective. Some mechanism needs to be established for organizing the MoF/MoDP joint monitoring (both financial and physical) with the LMs, as well as for holding the latter accountable for their own monitoring reports. The PSIC could easily be extended with the mandate and convening authority to collaborate with the MoF and the MoDP. It could be done with support from the Cabinet of Ministers and effective communication with the LMs. They would need to establish an institutionalized mechanism that draws a direct link between monitoring outcomes and onward financing. For this purpose, it is also necessary for the MoDP/MoF to jointly consider reviewing the entire process in management and monitoring of project execution, as well as to prepare a manual/uniform set of guidelines on Standardized Operating Procedures (SOPs). The SOPs will be implemented by the spending agencies and used by the MoDP/MoF to supervise and monitor the quality, costs (annual and total), and timing of project implementation. Based on the SOPs, a report of the failure, either physical or financial or both, may not automatically lead to project cancellation. Rather, it would trigger some sort of inter-ministerial review (to be led by the MoF/MoDP). Based on the outcome of such a review, a decision will be made on whether the project should be stopped or continued — with or without any necessary adjustments in terms of budget, scope, and timing. Harmonize the PPP into a unified PIM framework (Timeline: Year 3+) 249. In theory, PPPs have lots of advantages that can help bridge financing gaps. Using this modality of investment, governments can rely on private sector expertise and efficiency to deliver on infrastructure and/or other public services. However, experience in both developed and developing countries indicates, that PPPs can break the budget if they are not subject to rigorous vetting and monitoring. Without due appraisal, careful design, review, and proper contracting, governments could risk bearing the high cost of contingent liabilities. The PPP projects may fail for a variety of reasons, including: bad deals; poor design (and lopsided risk sharing); procurement problems; and issues in contracting and monitoring. Experience from the successful management of PPPs also suggests that they cannot be separately appraised. Rather, they need to undergo the same rigorous process of project appraisal subject to CBA (depending on the scale and technical complexity) and independent review as applied to ‘traditional’ budget-financed projects. Figure 39 illustrates the decision tree with alternative branches in terms of modality. It shows the comparison between the decision-making process with regard to the PPP financing and budget financing. 95 Figure 39. Decision Tree in Selecting the Public Investment Modality for a Unified PIM Framework Source: Rajaram and others (2014). 250. This system as depicted in figure 39 could be suitable to Lesotho. Its establishment and operationalization would require consistent efforts by both the MoF and the MoDP to initiate and develop the legal framework, necessary institutions, and capacity. As part of an integrated PIM system, the PPP legal framework and procedures need to be built to cover all the 8 stages,85 as typically applicable to a ‘traditional’ PIM system.86 To date, Lesotho has experienced successes and failures in managing the PPPs. The time for trial and error has to come to an end. Ffforts are to be made to harmonize the PPPs into a unified PIM. At the current stage, when the PIM is to be strengthened, the Government should not rush. Rather, it should move with care in the approval of PPP proposals. While the country’s PIM capacity remains limited, Lesotho may use external expertise to help review each PPP proposal in terms of appraisal, risk allocation in contracting, approval, and design of appropriate mechanism for monitoring. 85 Stage 1: Strategic guidance and policy for PPP, Stage 2: Aligned and integrated project appraisal, Stage 3: Independent review of the appraisal, Stage 4: Project selection and budgeting for PPP, Stage 5: PPP project implementation, Stage 6: PPP project adjustment by renegotiation/refinancing, Stage 7: PPP operation and maintenance, Stage 8: PPP project completion and ex-post evaluation. 86 Chapter 7 in Rajaram and others (2014) provides an in-depth discussion of the approaches and country examples in developing a unified approach to managing PIM and PPPs. 96 Annex 1: Composition of Tax Revenue Performance from IFMIS and MTFF Data Sources 2010-11 2011-12 2012-13 2013-14 2014-15 2015-16 Tax Revenue from IFMIS 5,037,801,790 4,324,117,520 4,279,079,720 4,791,141,896 5,502,624,945 6,153,210,775 111 Taxes on Income, Profits, and Capital Gains 2,534,791,317 2,621,078,054 2,367,544,045 2,653,870,045 3,156,417,542 3,714,525,336 1111 Payable by Individuals 2,534,791,706 2,621,079,987 2,367,552,769 2,653,870,045 3,155,309,291 3,714,606,338 1112 Payable by Corporations - - - - - - 1113 Payable by Both Individuals and Corporations (389) (1,933) (8,724) - 1,108,252 (81,002) 112 Taxes on Payroll and Workforce 19,859 1,346,893 78,928 - - - 1120 Other Social Contributions 19,859 1,346,893 78,928 - - - 113 Taxes on Property 1,618,882 300 - - - - 1133 Estate, Inheritance and Gift Taxes - 300 - - - - 1134 Taxes on Financial and Capital Transactions 1,618,882 - - - - - 114 Taxes on Goods and Services 2,100,486,354 1,529,480,182 1,736,889,371 1,936,121,876 2,119,838,588 2,187,483,277 1141 General Taxes on Goods and Services 1,962,160,148 1,436,033,357 1,693,336,715 1,792,088,576 2,020,703,314 2,151,238,000 1142 Excise 131,301,105 75,998,370 35,671,042 131,708,034 91,028,221 18,459,497 1144 Taxes on Specific Services 348,911 697,470 606,547 374,586 1,175,168 1,445,933 1145 Taxes on Use of Goods and Services and on Permission to Use or Perform Activities 6,676,190 16,750,985 7,275,067 11,950,680 6,931,885 16,339,846 115 Taxes on International Trade and Transactions 396,173,198 151,424,219 173,856,002 199,108,938 224,881,079 250,043,540 1152 Taxes on Exports 396,173,198 151,424,219 173,856,002 199,108,938 224,881,079 250,043,540 116 Other Taxes 4,712,180 20,787,871 711,373 2,041,038 1,487,736 1,158,623 1162 Payable by other than Business or Unidentifiable 4,712,180 20,787,871 711,373 2,041,038 1,487,736 1,158,623 Tax Revenue from MTFF 3,881,209,473 4,282,812,500 4,596,302,734 5,147,645,508 5,785,430,664 6,600,236,328 1.1. Taxes on income, profits, and capital gains 1,972,712,769 2,395,012,207 2,348,631,104 2,680,218,262 3,058,588,623 3,643,701,416 1.1.1. Income tax - payable by individuals 1,179,647,339 1,469,101,074 1,445,720,947 1,563,318,970 1,651,905,151 1,701,594,971 1.1.1.1 Government PAYE (LRA) 565,096,125 694,434,314 597,318,615 692,400,000 711,844,668 784,000,000 1.1.1.2 Private Sector PAYE (Residual) 614,551,214 774,666,761 848,402,332 870,918,970 940,060,484 917,594,971 1.1.2. Income tax - payable by corporations and other enterprises 467,450,684 555,773,865 453,121,185 665,261,902 878,079,346 1,121,627,808 1.1.3. Income tax - other 325,614,746 370,137,360 449,789,001 451,637,360 528,604,004 820,478,638 1.3. Taxes on property 105,518,997 125,074,387 149,830,032 127,699,997 139,000,000 - 1.3.2. Recurrent taxes on net wealth (rand monetary compensation) 105,518,997 125,074,387 149,830,032 127,699,997 139,000,000 - 1.3.3. Estate, inheritance, & gift taxes - - - - - - 1.4. Taxes on goods and services 1,402,352,417 1,589,923,706 1,818,772,583 1,886,557,495 2,343,625,244 2,705,332,764 1.4.1. General taxes on goods and services 1,276,386,963 1,369,552,002 1,640,425,171 1,732,300,293 2,116,794,434 2,210,184,082 1.4.1.1. Value-added tax 1,276,386,963 1,369,552,002 1,640,425,171 1,732,300,293 2,116,794,434 2,210,184,082 1.4.1.1.1 Private Sector VAT (MTFF) 991,929,559 1,012,905,798 1,227,486,691 1,268,376,795 1,717,766,837 1,724,240,356 1.4.1.1.2 Government VAT (MTFF) 284,457,404 356,646,204 412,938,480 463,923,498 399,027,596 485,943,726 1.4.1.2. Sales taxes - - - - - - 1.4.2. Excise taxes 112,534,134 200,604,675 163,323,868 132,600,662 206,660,919 469,775,452 1.4.4. Taxes on specific services 7,467,413 3,111,843 7,791,023 9,738,234 8,094,411 9,033,441 1.4.5. Taxes on the use of goods and on permission to use or perform activities 5,963,893 16,655,247 7,232,496 11,918,314 12,075,523 16,339,848 1.4.5.1. Motor vehicle licences 4,960,122 2,606,958 3,156,444 3,458,148 5,971,613 8,376,278 1.4.5.2. Other licences & taxes on use and on permission to use goods or perform activitie 1,003,771 14,048,288 4,076,052 8,460,166 6,103,910 7,963,570 1.4.6. Other taxes on goods & services - - - - - - 1.5. Taxes on international trade and transactions 396,173,187 151,424,225 277,408,203 451,128,815 242,737,762 250,043,533 1.5.1. Customs and other Import duties - - - - - - 1.5.2. Taxes on exports 396,173,187 151,424,225 277,408,203 451,128,815 242,737,762 250,043,533 1.6. Other taxes 4,452,073 21,377,941 1,660,659 2,041,039 1,479,369 1,158,624 1.6.3. Other taxes: stamp taxes 4,452,073 21,377,941 1,660,659 2,041,039 1,479,369 1,158,624 Difference 1,156,592,317 41,305,020 (317,223,015) (356,503,612) (282,805,719) (447,025,553) % difference between IFMIS & MTFF 29.80% 0.96% -6.90% -6.93% -4.89% -6.77% Source: MTFF and BOOST 97 Annex 2: Tax Expenditures: A Primer 1. Tax expenditures are defined as provisions that reduce or postpone revenue for a rather narrow group of taxpayers relative to a benchmark tax. The benchmark tax comprises the rate structure, accounting conventions, deductibility of compulsory payments, provisions to facilitate tax administration, and international fiscal obligations (OECD 2010). Tax expenditures may take a number of different forms, such as allowances, exemptions, preferential tax rates, tax deferrals, and tax credits. 2. Tax expenditures can be measured using one of the following methods: • Initial revenue loss. This method involves calculating the amount by which tax revenue is decreased as a result of the enacted tax expenditure. It does not take into account taxpayers’ behavioral responses. • Final revenue loss. This estimation differs from the initial revenue loss approach by incorporating the change in taxpayers’ behavior and the effects on revenues from other taxes ensuing from the introduced tax expenditure. • Outlay equivalence. This method estimates the direct government expenditure that would have generated the same gain for taxpayers as the considered tax expenditure. This requires measuring the direct spending in pre-tax terms to arrive at the same after-tax effect on taxpayers’ income. Source: Le, Minh Tuan, Leif Jensen, G. P. Shukla, and Nataliya Biletska. 2016. “Assessing Domestic Revenue Mobilization: Analytical Tools and Techniques.� MFM Discussion Paper 15. 98 Annex 3: TADAT - Performance Outcome Areas Source: www.tadat.org 99 Annex 4: Staff Number per Ministry Costing Files Establishment (LA - Lw) 2015/16 2016/17 Actual Vacancy Ministry Approved Approved (15/16) (15/16) MINISTRY OF AGRICULTURE AND FOOD SECURITY 3097 3299 1860 39.94% MINISTRY OF HEALTH 3903 3911 2698 30.87% MINISTRY OF EDUCATION 875 879 15193 N/A MINISTRY OF FINANCE 964 964 738 23.44% MINISTRY OF TRADE 209 221 173 17.22% MINISTRY OF Development Planning 630 629 528 16.19% MINISTRY OF JUSTICE AND 1639 1640 1606 2.01% MINISTRY OF HOME AFFAIRS 693 693 465 32.90% OFFICE OF THE PRIME MINISTER 987 991 635 35.66% MINISTRY OF COMMUNICATIONS, 897 896 855 4.68% MINISTRY OF LAW, CONSTITUTIONAL 445 443 342 23.15% MINISTRY OF FOREIGN AFFAIRS 283 284 187 33.92% MINISTRY OF PUBLIC WORKS 2845 2845 1060 62.74% MINISTRY OF FORESTRY, RANGE AND 800 900 483 39.63% MINISTRY OF ENERGY 202 218 179 11.39% MINISTRY OF LABOUR AND 285 286 189 33.68% MINISTRY OF TOURISM, ENVIRONMENT 279 274 234 16.13% AUDITOR GENERAL'S OFFICE 166 167 142 14.46% HIS MAJESTY'S OFFICE 87 88 59 32.18% PUBLIC SERVICE COMMISSION 53 52 37 30.19% PENSIONS 7620 STATUTORY 5 MINISTRY OF DEFENCE 325 325 4384 N/A NATIONAL ASSEMBLY 208 222 253 -21.63% SENATE 100 100 81 19.00% OFFICE OF OMBUDSMAN 38 38 28 26.32% ELECTORAL OFFICE 257 257 183 28.79% MINISTRY OF LOCAL GOVERNMENT 1600 1601 4153 N/A MINISTRY OF GENDER, YOUTH, 305 298 236 22.62% MINISTRY OF THE PUBLIC SERVICE 266 269 143 46.24% JUDICIARY 1194 1177 1030 13.74% MINISTRY OF SOCIAL DEVELOPMENT 283 283 271 4.24% DIRECTORATE ON CORRUPTION 63 67 59 6.35% MINISTRY OF MINING 160 176 101 36.88% MINISTRY OF POLICE 707 704 4890 N/A MINISTRY OF SMALL BUSINESS 262 262 242 7.63% MINISTRY OF WATER 476 471 325 31.72% Unknown 710 Grand Total 25583 25930 52377 23.30% 100 Annex 5: Composition of Expenditure for Staff Compensation (FY2015/16) Code Description Amount Salaries and Wages 411101 Salaries - Established Posts 4,321,898,267.32 411102 Salaries - New Posts 6,069,040.70 411201 Privy Purse 661,440.00 411203 Chief Justice 559,824.00 411204 Judges 6,250,491.82 411205 Chairman PSC 488,558.00 411207 Members PSC 1,796,688.00 411209 Attorney General 527,657.00 411210 Director Public Prosecutions 477,012.00 411211 Auditor General 472,339.00 411216 Ombudsman 405,036.00 411217 Chairman IEC 429,339.00 411218 Members IEC 772,333.00 411301 Wages - Established Posts 160,398,346.05 411302 Wages - New Posts 134,358,514.66 Sub-total Salaries & wages 4,635,564,886.55 Allowances 411401 Allowances - Non statutory Posts 226,497,366.20 411501 Queens Allowance 529,152.00 411503 King's Allowance 72,000.00 411510 Allowances Ombudsman 65,000.00 411514 Allowances Chairman IEC 76,606.63 411515 Allowances Members IEC 120,000.00 411703 Doctors Housing 4,592,678.85 411704 School and Medical Fees/Foreign Mission 37,485,919.09 Sub-total Allowances 269,438,722.77 Sub-total salaries, wages & allowances 4,905,003,609.32 Social Contributions 412101 GOL Pension Contribution 337,755,021.69 412301 Civil Pensions 293,672,886.79 412302 Gratuities 189,232,140.66 Sub-total social Contributions 820,660,049.14 Total Compensation of Employees 5,725,663,658.46 Source: MoF (IFMIS data). 101 Annex 6: Wage Bill Projection Model 1. Forecasts for the evolution of the wage bill from 2016 to 2022 were developed using a Microsoft Excel-based simulation model. In addition to establishing a baseline projection to inform an understanding of the magnitude and potential trajectory of the wage bill over the medium term, the model was used to analyze how the wage bill might evolve under different assumptions regarding GDP growth and policies governing hiring and compensation (that is, to conduct sensitivity analysis around the baseline projection) and estimate the fiscal impact of potential options for reform. 2. The simulation model reconstructed the wage bill in 2016 using individual employee data and wage tables governing compensation. Employee data for individual workers, including information regarding their grade and point-value (that is, their pay level within their assigned grade) was obtained for the month of October 2016 from the Resource Link. In conjunction with the wage tables governing central government compensation, the Resource Link data was used to create distributions for the number of workers by age and their associated mean wages by age. Once age-specific head counts are multiplied by mean wages and summed, the distributions sourced from the Resource Link accounted for roughly 75 percent of the budget total for 2016/17 for wages and salaries. 3. The workforce was disaggregated into four groups defined primarily by the wage tables that govern their compensation. The disaggregation of the workforce was necessary for two reasons: (a) significant heterogeneity exists across key sectors of the Government in the distribution of their workers and their mean wages by age and (b) sector-specific options for policy reform cannot be simulated unless the workforce is disaggregated in the model. The four groups include Civil Service (which includes some workers in other groups paid using the wage table for civil servants, as well as Chieftainships), Education (teachers and other workers paid using the same wage table), Health (which includes all workers within the MoH), and Police/Defense (which includes workers paid using the wage tables for the police, army and national security, and justice, human rights, and correctional facilities). Table 6.1. Number of Workers and Wages by Employment Group, 2016 Source: IMF staff estimates. 4. Head counts and wages for each employment group were scaled upward to match the budget total for 2016/17 for wages and salaries. The scaling for the number of workers by age and for their associated mean wages was done in equal proportions given the absence of data needed to assess the relative wage expenditures attributable to (a) workers paid outside the Resource Link (which include persons employed under capital projects, local government employees paid through block budget-transfers, and— possibly—‘ghost’ workers) and (b) elements of compensation paid to workers included in the Resource Link but paid outside the payroll (such as allowances and other forms remuneration). The scaled distributions, therefore, capture all wages and salaries paid by the budget, to include local government wages and salaries subsidized by the central government. The Government’s contributions as an employer for social insurance and other levies were estimated separately as a share of total wage expenditure on the basis of the 2016/17 budget to arrive at total wage spending. 5. The distributions of workers by age and their associated mean wages were then projected forward from 2016 to 2022. The distribution of workers by age for each employment group was projected 102 forward on the basis of (a) an input variable representing the number of new hires in each year of the simulation period and (b) assumptions regarding the timing of retirement and the mean age of entry of new entrants into the workforce. The distribution of mean wages by age was projected forward on the basis of (a) an input variable representing the annual rate of increase in wages, historically designed to maintain their real value in the face of inflation and (b) a second input variable intended to capture the value of the ‘notch’ increase (that is, the increase in wages associated with a worker being elevated to higher point- value within his or her grade in the wage table, historically awarded annually). The resulting distributions, along with social levies, were then used to compute total wage spending in each year of the simulation period and to express wage spending relative to nominal GDP using the macroeconomic forecast developed by the World Bank. 6. The structure of the model enabled it to be used to simulate a range of policies governing employment and compensation under varying assumptions for macroeconomic conditions. Collectively, the inputs and assumptions enable the model to simulate the impact on wage spending of policies regarding the extent to which retiring workers are replaced by new recruits, the hiring of workers into newly created positions, the number and remuneration of temporary workers, the renewal of labor agreements for existing contractual workers, changes in the wage tables (including the current policy of providing a wage increase annually for all workers to preserve the inflation-adjusted value of their wages), changes in the progression of workers across the point-values of their pay grade, and changes in the retirement age. The fiscal impact of policy measures that treat some groups of workers disparately—such as measures that treat workers with higher total compensation differently than those with lower total compensation—were analyzed separately before being evaluated, in the aggregate, using the projection model. 7. The model also captured the impact of wage drift on the evolution of the wage bill over the projection period. The term wage drift refers to aspects of public employment policy—such as longevity- related increases in compensation (that is, the ‘notch’ adjustment), promotion patterns and policies, and practices regarding the reclassification of employment positions—that drive wage growth independently of broader decisions regarding the overall size of the workforce or wage levels. Conceptually, wage drift is intended to capture all factors that create upward pressure on wage spending independent of changes in the total number of workers or in their respective wage table. Estimating the impact of wage drift on wage spending relied on the same data from the Resource Link as was used by the simulation model. The analysis was conducted only for civil servants, but the estimates for wage drift are unlikely to change greatly when performed for other employment groups. The Resource Link data included granular-level data on each worker’s grade and point-value. Together with the wage tables, the Resource Link data enabled the impact of wage drift on the wage bill to be estimated and disaggregated into the relative contributions of (a) the ‘notch’ adjustment (whereby all workers are shifted to the next higher point-value of their wage table until they reach the highest point-value for their grade), (b) retirements, and (c) new entrants hired to replace those workers who retired (table 6.2). With the size of the workforce held fixed, the net impact of entrants and retirements is negative because workers always enter the wage grid at the lowest point-value for their grade but typically retire from the highest. It should be noted that the fiscal impact of the notch adjustment is forecasted to gradually decline from 2019 to 2022 because an increasing share of workers across all four employment groupings will have reached the highest point-value for their pay grade. Table 6.2. Impact on Wage Spending (Percentage) 2017/18 2018/19 2018/19 2019/20 Notch 2.78 2.79 2.72 2.67 Entrants 1.87 2.09 1.76 2.79 Retirements (2.11) (2.36) (2.02) (3.24) Net (Drift) 2.55 2.52 2.46 2.23 103 Source: IMF staff estimates. 8. Reform options were evaluated relative to a baseline that reflects assumptions regarding the evolution of the workforce and compensation. Specifically, the baseline projection presumed the authority’s current policies continue to be applied across the projection period. In terms of their fiscal impact, the two most important policies are the following: (a) the provision of annual wage increases to fully preserve the real (that is, inflation-adjusted) value of wages for every point-value of the wage tables over time and (b) the awarding of a notch increase every year for workers who are not yet at the highest point-value of their grade. Wage increases were simulated using inflation and GDP projections sourced from the macroeconomic forecast developed by the World Bank (table 6.3). In line with the authorities’ assumptions used for annual budget preparation, the impact of wage drift was assumed at 2.5 percent for all years of the projection period (that is, mean salaries were adjusted upward by 2.5 percent for everyone in every year irrespective of decisions regarding overall wage increases). Allowances are assumed to grow at the same rate as wages do and changes in compensation policies apply uniformly to all workers in all ministries, agencies, and departments. In the absence of stated policies governing employment, the baseline scenario assumed that the number of central government workers grows in-line with the current rate of growth for the working-age population using demographic projections developed by the United Nations. Table 6.3. Macroeconomic Assumptions 2017/18 2018/19 2019/20 2020/21 Inflation (in Percent) 6.41 7.15 6.61 6.24 Nominal GDP (in Maloti Billions) 37.29 41.33 45.65 50.25 Nominal GDP Growth (in Percent) 9.71 10.85 10.44 10.07 Source: World Bank staff estimates. 104 Annex 7: Summary of Diagnostic of the PIM Performance Stage 1: Investment Guidance and Preliminary Screening Stage of PIM Desirable Features* Actual Functioning in Lesotho Stage 1: • Published development • Strategic documents have been prepared and are supposed to Strategic strategy or vision statement guide the preparation and prescreen project proposals. Guidance and which has unambiguous • Establishment of the PSIC, supported by the DPCM, aims to Preliminary authority. create a gateway to control the quality of project proposals Screening • Centralized approval by the from the conception stage. The PAG and the Submission MoDP or the MoF (or Instruction of Project Concept Notes and Proposals to the delegated) for developing MPUs and public agencies were issued by the PSIC and proposals/explicit ministry DPCM respectively. However, even with the MTEF, level justification with effectively the link between strategic planning, policy, and strategy. annual capital budgeting remains missing. The process of • Clarity of project objectives preparation of a project financial statement is ad hoc, while in terms of outputs and projects are prepared mostly on the basis of funding outcomes. availability and thus render to hurried process. • Consideration of alternative • Both sectors and the PSIC/DPCM exhibit optimistic bias in approaches to objectives. the process of preparation and reviewing of PCN. As such the project concept review and screening are reduced to fulfilling a defined formality. Note: * The desirable institutional features are taken from the diagnostic framework in Rajaram et al. 2010. Stage 2: Formal Project Appraisal Stage of PIM Desirable Features* Current Status in Lesotho Stage 2: Formal • Publicized and • PAG and the instructions for the submission of project concept Project transparent guidance. note and proposals (2013) serve as the formal guidelines for Appraisal • Effective training and project preparation and appraisal. deployment of staff for • The Guidelines are not supplemented with detailed manual on project design and methodologies for project appraisal, and hence they serve simply appraisal. as reference to spending agencies. • Application of • Since no detailed appraisal methodologies are prepared, there has, guidance in project to date, been no real effort to build capacity at PSIC/DPCM and appraisal. LMs, rendering the entire process of appraisal ad hoc and nonuniform across projects and sectors. The PSIC has never rejected any project proposals under its review. Note: * The desirable institutional features are taken from the diagnostic framework in Rajaram et al. 2010. Stage 3: Independent Review of Appraisal Stage of PIM Desirable Features* Current Status in Lesotho Stage 3: • Independent checks to avoid any subjective • The PSIC, supported by the DPCM, is Independent and/or self-serving bias in project appraisal. mandated to review all project proposal Review • Development and maintenance of comprehensive and appraisal. The missing guidance on and effective project database that creates an the detailed appraisal methodologies, inventory of appraised project ranked by priority the lack of capacity at both the for budgetary consideration. PSIC/DPCM and at LMs, and the • Clarity of roles between projects which are minor absence of demand for rigorous and may be dealt with at the departmental level, appraisal indicate the independent and those requiring additional appraisal. review function remains largely missing or ineffective. In formality, the PSIC keeps the list of project reviewed and approved but de facto there is no 105 Stage of PIM Desirable Features* Current Status in Lesotho particular priority ranking within/cross sectors, rendering the review process a formality. Note: * The desirable institutional features are taken from the diagnostic framework in Rajaram et al. 2010. Stage 4: Project Selection and Budgeting Stage of PIM Desirable Features* Current Status in Lesotho Stage 4: • Transparent criteria for • Formally the financing decision of capital projects is the Project selecting projects with joint give-and-take process during annual budgeting across Selection and reference to policy objectives at the MoF, the MoDP (represented by the PSIC), and LMs. Financing ministerial level. The selection of major projects for financing (budget- • Well-structured budget funded projects) is, actually, not based on the outcomes of preparation process with scope project proposal documents submitted by spending LMs or to integrate investment and on the PSIC assessment of such project proposals recurrent implications of submitted by LMs. Instead, sectors may simply bypass the projects. formal process and the review by PSIC to obtain direct • Effective gatekeeping to ensure approval by the Cabinet of Ministers. that only appraised and • The existing guidelines and templates on project proposal approved projects are selected submission and appraisal do not consist of transparent for budget financing. criteria for capital financing prioritization. • Ensuring adequate financing • With the disconnection between the MTEF and the annual for selected projects, including budgeting, parallel budgeting (and thereby the absence of recurrent needs on completion. credible mechanism to integrate capital and recurrent budget) persists. Note: * The desirable institutional features are taken from the diagnostic framework in Rajaram et al. 2010. Stage 5: Project Implementation Stage of PIM Desirable Features* Current Status in Lesotho Stage 5: Project • Published guidelines • A set of standard operational guidelines or manual to Implementation for project enforce uniform procedures and standards in project implementation. execution and fiduciary does not exist. • Cost-effective • Issues in procurement continue to exert negative impact on implementation value for money: The Public Procurement Law has not been through procurement enacted, no standard bidding document, no monitoring tool, and contracting procurement qualifications are not yet accredited, the • Regular monitoring of PPAD does not have autonomy and authority to force financial and compliance, and capacity building is limited. nonfinancial progress • Project delay and cost overrun are major concerns but there • Timely implementation exists no systemic database to capture. reports on major • Delayed payment of contractors has been persistent. projects • Effective budgeting for selected projects Note: * The desirable institutional features are taken from the diagnostic framework in Rajaram et al. 2010. Stage 6: Project Adjustment Stage of PIM Desirable Features* Current Status in Lesotho Stage 6: Project • Active monitoring • No formal mechanisms to trigger a review of a Adjustment project during implementation. 106 • Review of projects’ • Projects are frequently treated as if they were justification during perpetual and the LMs fare aced with no limits to implementation. the cost escalation. Projects are seldom dropped or stopped midway due to changed circumstances. • Monitoring (both physical and financial) is done but its outcomes are not linked to onward financing of projects. Note: * The desirable institutional features are taken from the diagnostic framework in Rajaram et al. 2010. Stage 7: Facility Operation Stage of PIM Desirable Features* Current Status in Lesotho Stage 7: Facility • Asset registry • Asset registry exists at both the central and LMs Operation • Asset registers need to be levels. But asset registers are generally incomplete maintained and asset values and infrequently updated (for example, to track recorded value of assets, net of depreciation). • Facility operation • Efficiency in facility operation is affected by the continued parallel budgeting (capital and recurrent). Note: * The desirable institutional features are taken from the diagnostic framework in Rajaram et al. 2010. Stage 8: Project Evaluation Stage of PIM Desirable Features* Current Status in Lesotho Stage 8: Project • Formal institutional arrangements for ex post • No mechanism for ex post Evaluation evaluation of projects/programs with feedback evaluation applied to budget- into future project designs. funded projects. Note: * The desirable institutional features are taken from the diagnostic framework in Rajaram et al. 2010. 107 Annex 8: Old Age Pension 1. Launched in 2004, Lesotho’s OAP scheme is a social pension scheme for citizens ages 70 and older who are not beneficiaries of a larger civil service pension or social grant.87 The OAP scheme also covers some 432 former members of the LLA, active during the 1970s, and some 1692 former members (and their widows) of the African Pioneer Corps for service during the two world wars. The scheme is noncontributory and budget financed. Eligibility is not contingent on means-testing. 2. The amount of Lesotho’s OAP scheme is not formally indexed to inflation, but has been adjusted annually since 2012. Since the introduction of the scheme in 2004, the real (that is, inflation adjusted) value of its benefit has almost doubled from M 307 to M 580 as of March 2017 (figure 11.1). Comparing the value of the OAP scheme to the value of benefits provided by other countries is methodologically problematic given (a) vast differences in the approaches taken to poverty alleviation worldwide, (b) the limited number of social pension schemes found in comparable lower-middle-income countries (in SSA, for example, comparators include only Cape Verde, Nigeria, and Swaziland), and (c) a lack of current data (Lesotho’s last household budget survey appears to have been conducted in 2010/11). Using that data, the value of the OAP scheme in 2014, expressed relative to three poverty benchmarks, was estimated to be • 1.15 of the international poverty line of $US 1.25 per day adjusted on a purchasing power basis; • 1.70 of Lesotho’s poverty line; and • 3.02 of Lesotho’s food poverty line. Figure 11.1. Value of the OAP Scheme in Constant (2016) Maloti Source: Consultant estimates based on data provided by the authorities. 3. The number of OAP beneficiaries greatly exceeds the population of persons ages 70 and older (figure 11.2). The magnitude of this discrepancy has grown consistently and substantially over time to the 87This section relied heavily on “Lesotho Old Age Pension Diagnostic�—along with its cited references—prepared in February 2015 for the MoF by a World Bank team led by Lucilla Maria Bruni. 108 extent that an estimated 39.9 percent (roughly two in five) of the 87,110 beneficiaries, as of March 2016, were ineligible for benefits based on the rules of the scheme. Figure 11.2. Number of OAP Beneficiaries versus the Population Ages 70 and Above Source: Beneficiary data provided by the authorities; population data sourced from United Nations, Department of Economic and Social Affairs, Population Division (2015). World Population Prospects: The 2015 Revision, DVD Edition. 4. Paying OAP benefits to large numbers of ineligible beneficiaries is expensive . The cost of the OAP scheme—which was initially 1.2 percent of GDP in 2004—has hovered around 1.8 percent of GDP since 2009 (figure11.3) and would have trended higher had GDP grown less rapidly. Figure 11.3. Historical OAP Benefit Costs (Percentage of GDP) Source: Consultant estimates based on data provided by the authorities. 109 5. Eliminating ineligible beneficiaries would reduce the cost of the OAP scheme by an estimated 40 percent (1.1 percent versus 1.8 percent of GDP). Assuming that reforms can be implemented over a period of five years to accomplish this objective, the projected fiscal cost of the scheme should fall rapidly to roughly 1.0 percent of GDP by 2022 and then decline more gradually thereafter (figure 11.4). There are two primary reasons for the projected gradual decline in benefit costs over the coming 25 years: • The population of persons ages 70 and older in Lesotho—which fell from 2008 to 2015—is expected to resume growing but only very slowly (by about 0.5 percentage points, on average, per year); and • The projections presume that GDP will grow in excess of 3.0 percent per year on a real value basis. Even if this assumption proves unduly optimistic, as long as real economic growth exceeds 1.2 percent per year on average, the cost of the OAP scheme will still fall over the projection horizon. Figure 11.4. Projected OAP Scheme Benefit Costs (Percentage of GDP) Source: Consultant estimates based on data provided by the authorities. Note: The solid line shows the cost of the OAP scheme if benefits are only paid to persons eligible to receive them. The dashed line shows the cost of the OAP scheme under reform measures that eliminate ineligible beneficiaries over 2018–2022. 6. The existence of large numbers of ineligible beneficiaries on the OAP scheme’s rolls is attributable primarily to systemic inadequacies in the administration of the scheme. Addressing these inadequacies, however, is frustrated by a handful of contextual challenges, including (a) the use of non- computerized (that is, manual) processes across much of the public sector; (b) a reliance on local chiefs to verify the existence and eligibility of OAP applicants; (c) serious problems with the consistency, completeness, interconnectedness, and integrity of existing public sector databases; (d) nonexistent—or inadequate—mechanisms of accountability and control; (e) limited human resource capacity, particularly with respect to information technology; and (f) difficult topography and the limited reach of the country’s banking system outside the most populated urban areas. 7. Most ineligible OAP beneficiaries fall into one of the following three categories, listed in estimated order of decreasing prevalence: 110 (a) Beneficiaries who have died but have not been dropped from the rolls (that is, ‘zombies’). Eliminating zombies will require better proof-of-life procedures for identifying beneficiaries who have died and purging them from the OAP scheme’s rolls—an objective that is frustrated by the incentive that families have to collect benefits on behalf of deceased relatives, which is then extended to local chiefs who have the authority to certify that deceased beneficiaries are still living. (b) Beneficiaries who are not yet old enough to be eligible or who do not exist (that is, ‘ghosts’). Eliminating these beneficiaries will hinge on the success of audits and other steps to tighten internal controls and the better use of technology to flag falsified beneficiary records. Preventing them in the future will require changes to existing mechanisms for verifying their existence and eligibility—an objective that is frustrated by opportunities for collusion between applicants and local chiefs who have authority to certify eligibility on behalf of applicants. (c) Beneficiaries who are in receipt of a larger public sector pension (that is, ‘double- dippers’). In comparison to the challenge of eliminating zombies and ghosts, eliminating double-dippers should be easier given that public sector retirees are paid through the Resource Link (the payroll and HRM system). 8. In addition to addressing the problem of ineligible beneficiaries, measures are also needed to reduce the cost of delivering OAP benefits. The OPA’s total administrative costs are actually lower than the international benchmark of 10 percent for cash transfer schemes (Grosh et al. 2008). This is remarkable given Lesotho’s contextual challenges. Not surprisingly, however, a very large share—roughly half—of this expense is attributable to the logistical difficulties inherent in making benefit payments—most often in cash—to beneficiaries, some of whom can be reached only by helicopter. In addition, the distribution of large amounts of cash on a regular basis creates serious and intractable opportunities for malfeasance. Table 11.1. Annualized Cost of Administering the OAP Scheme Expense 2017/18 Salaries Paid to Staff (excludes soldiers providing security for benefit payments) 14,095,380 Suppliers and Hotels 8,332,824 Rentals Paid to the Post Office 1,016,976 LDF Air Wing (for helicopters used to deliver benefits to remote areas) 960,000 Computer Business Solutions (contractor) 957,504 Rentals Paid to Lenyora House 442,800 G4S Security (contractor) 388,692 Subsistence Allowances 336,000 Hardship Allowances 62,700 Vehicle Rentals tbd Petrol tbd Total tbd Source: Annualized based on data provided by the MoF. 9. Even without formalized mechanisms for programmatic monitoring and evaluation, the OAP scheme seems highly likely to contribute meaningfully to poverty reduction.88 Whether the cost of the scheme represents the most efficient use of resources for reducing poverty and protecting vulnerable citizens from economic risk, however, is unclear. In 2015, Lesotho spent about 4.5 percent of its GDP on social assistance programs—nearly three times the average spent by countries in SSA and twice that spent 88This assessment is consistent with several (dated) studies and one impact evaluation cited in “Lesotho Old Age Pension Diagnostic� prepared in February 2015 by a World Bank team led by Lucilla Maria Bruni. 111 by other middle-income countries. Roughly half of that spending was attributable to the OAP scheme. Moreover, because benefits are not means-tested, only about half of the OAP spending is believed to have reached beneficiaries who are actually poor (World Bank 2016). When compared against the social pension schemes found in 32 other lower-middle-income countries, the cost of Lesotho’s OAP relative to the country’s GDP is very near the top. Even if the OAP benefits were restricted to eligible beneficiaries, however, the cost of the scheme would still be comparatively very high (figure 11.5). 10. If the budgetary envelope for social spending is fixed, the resources spent on OAP benefits for the elderly necessarily come at the direct expense of spending on programs for other vulnerable members of Basotho society. Figure 11.5. Cost of Social Pension Schemes in Lower-middle-income Countries (Percentage of GDP) Source: Social Pensions Database, May 2017, published by Pension Watch and available online at http://www.pension-watch.net. Note: Values for Lesotho are the consultant’s estimates, where ‘De Jure’ refers to the cost of the OAP scheme under the assumption that ineligible beneficiaries have been identified and removed from the rolls and ‘Actual’ refers to the actual cost of the scheme in 2016. The cost of Georgia’s social pension scheme is not shown because its value (4.8) is more than double that of the rest o f the countries in the sample. 11. Operational reforms to the OAP scheme should first and foremost focus on strengthening its administration with the objective of eliminating ineligible beneficiaries . Doing so successfully would reduce the cost of the scheme by an estimated 0.7 percent of GDP, an amount which exceeds the combined public spending on public assistance, the child grants program, the Orphan Vulnerable Children bursary program, and the post primary bursary program. Only the school feeding program is larger, when measured by the yardstick of total expenditures. The need to strengthen the administration of the OAP scheme and eliminate ineligible beneficiaries is broadly recognized and a number of measures are already being supported by the World Bank under both its PSMP and its Social Assistance Project. 12. Parametric reforms to the OAP scheme (i.e., changes to its eligibility criteria or benefit levels) may also merit consideration but will require further study. The exception to this assertion is the issue of benefit indexation. While indexing the OAP scheme to inflation—or to some acceptable metric of poverty—comes at the cost of fiscal flexibility, it provides essential protection to beneficiaries (by fixing the purchasing power of their benefits) and insulates the budget from political pressure to introduce ad hoc changes that could further escalate the real value of OAP benefits beyond what is needed for poverty 112 alleviation. The case for making other parametric reforms, such as means-testing, however, is less clear. While means-testing is conceptually compelling—providing benefits only to those who need them— means-testing: (a) is not costless (to the extent that its costs can outweigh its savings), (b) can introduce errors of exclusion, (c) carries significant risk of stigma (thereby discouraging some poor from applying), and (d) requires considerable capacity to properly administer. 13. Structural reforms to the OAP scheme should only be considered holistically in the context of a broader conversation surrounding the redesign of Lesotho’s overall social assistance and social insurance strategies. 113 Bibliography Chapter 1: Fainboim Yaker, Israel, and Sailendra Pattanayak. 2011. “Treasury Single Account: An Essential Tool for Government Cash Management.� IMF Technical Note. Honda, Jiro, Manabu Nose, Cesar Sosa-Padilla, Jose L. Torres, Murna Morgan, Fernando. G. Im, and Natalia. A. Koliadina. 2017. “Fiscal Rules: Coping with Revenue Volatility in Lesotho and Swaziland.� IMF Policy Paper. Im, Fernando. 2015. “Some Stylized Facts of Fiscal Policy Making in Southern Africa.� Preliminary draft. Im, Fernando, Cesar Sosa-Padilla, and Jose Torres. 2015. “Volatility of SACU Receipts and Fiscal Rules in Swaziland: A Welfare-Based Approach.� Preliminary draft, World Bank Group, Washington, DC. World Bank. 2016. Implementation Completion and Results Report. 2016. First Growth and Competitiveness Development Policy Grant. Report number: ICR3665. Lesotho National Strategic Development Plan 2012/13–2016/17. Growth and Development Strategic Framework PEFA (Public Expenditure and Financial Accountability). 2017. Public Expenditure and Financial Accountability (PEFA) Assessment 2016. Draft Final Report, February 2017. World Bank. 2016a. Lesotho State-Owed Enterprises. Country Policy Note, June 30, 2016. ———2016b. SACU Receipts and Fiscal Policy in Southern Africa. Report No: AUS13682. ———. 2017. Lesotho Macro-Poverty Outlook, October 2017 Chapter 2: Casanegra de Jantscher, and Milka. 1990. “Problems in Administering a Value-Added Tax in Developing Countries�, edited by Malcolm Gillis, Carl S. Shoup, and Gerardo P. Sicat. Discussion Paper, World Bank, Washington, DC. Chen, Duanjie, and Jack M. Mintz. 2015. “The 2014 Global Tax Competitiveness Report: A Proposed Business Tax Reform Agenda.� The School of Public Policy – University of Calgary. SPP Research Papers. Vol. 8, Issue 4. Cuevas, Alfredo. 2015. “The SACU Transfer Over/Under-Payment Correction and the Volatility of Transfer Payments to the BLNS.� Unpublished manuscript. IMF (International Monetary Fund). 2013. “Lesotho: Issues in Tax Policy and Mining Fiscal Regimes.� Marquez, P.V., A. Gonima, A. Senkal, Zang Sidjou, O. R. 2017. "Tobacco and Alcohol Taxation Policy Reform in Lesotho: Results of Modeling Impact Assessment on Prices, Consumption, and Fiscal Revenues over 2017–2018." Internal Document. Washington, DC: World Bank. 114 Moshoeshoe, Ramaele. 2012. “Consumer Demand for Alcoholic Beverages and Tobacco in Lesotho: A Double-hurdle Approach.� Working Paper 315, Economic Research Southern Africa (ERSA). Schneider F., 2012. “The Shadow Economy and Work in the Shadow: What Do We (Not) Know?� The Institute for the Study of Labor (IZA). March 2012. WHO African Region. 2015 Report on the Status of Major Health Risk Factors for Noncommunicable Diseases. Chapter 3: OECD (Organisation for Economic Co-operation and Development). 2016. OECD Pensions Outlook 2016. OECD Publishing. Chapter 4: Allen, Richard., and Daniel Tomasi. 2001. Managing Public Expenditure: A Reference Book for Transition Countries. Paris: Organization for Economic Co-operation and Development (OECD). Rajaram, Anand, Tuan Minh Le, Nataliya Biletska, and James Brumby. 2010. “A Diagnostic Framework for Assessing Public Investment Management.� Working Paper 5397, World Bank, Washington, DC. Glenday, Graham, Kai Kaiser, and Tuan Minh Le. 2014. “Approaches to Better Project Appraisal.� Chapter 4. In The Power of Public Investment Management Edited by Anand Rajaram, Le, Kaiser, Kim, and Frank. World Bank. IFC (International Finance Corporation). 2009. Breaking New Ground: Lesotho Hospital Public-Private Partnership—A Model for Integrated Health Services Delivery. Ministry of Finance and MoDP. 2015. Draft Budget Strategy Paper. September 2015. Ministry of Finance and Development Planning. 2012. Manual for Planning Officer. Draft November 1, 2012. Oxfam. 2014. A Dangerous Diversion: Will the IFC’s flagship Health PPP Bankrupt Lesotho’s Ministry of Health? PEFA (Public Expenditure and Financial Accountability). 2017. Public Expenditure and Financial Accountability (PEFA) Assessment 2016. Draft Final Report. Rajaram, Anand, Tuan Minh Le, Kai Kaiser, Jay-Hyung Kim, and J. Frank. 2014. The Power of Public Investment Management. World Bank. Webster, Paul C. 2015. “The Lesotho's Controversial Public-Private Partnership Project.� The Lancet, 386 (10007):1929–1931. World Bank. 2012. Lesotho Public Investment Management Efficiency Review. Report No: 65694-LS. ———. 2016a. “Lesotho State-Owed Enterprises.� Country Policy Note, June 30. ———. 2016b. Lesotho Health Network Public-Private Partnership (PPP). February 19. ———. 2017. Africa’s Pulse. Volume 15, April 2017. 115 Annex 8: Margaret Grosh, Carlo del Ninno, Emil Tesliuc, and Azedine Ouerghi. 2008. “For Protection and Promotion: The Design and Implementation of Effective Safety Nets.� World Bank, Washington, DC. World Bank. 2016. “Project Appraisal Document for a Social Assistance Project.� World Bank, Washington, DC. 116