Kingdom of Lesotho Agriculture Public Expenditure Review 2019 Kingdom of Lesotho: Agriculture Public Expenditure Review Standard Disclaimer This volume is a product of the staff of the International Bank for Reconstruction and Development/The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Copyright Statement The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www .copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Table of Contents Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi Abbreviations and Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix CHAPTER 1: Introduction: Why an Agriculture Public Expenditure Review for Lesotho? . . . . . . . . . . . . . . . . . . 1 The Central Role of Lesotho’s Agriculture Sector in the Economy, for Poverty and Food CHAPTER 2:  Security, and Challenges Moving Forward . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 The Data and Methodology Used for Conducting This Agriculture Public Expenditure CHAPTER 3:  Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Description of the data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 CHAPTER 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? . . . . . . . . . . . 15 Budget process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Level—trends, fluctuations, and outliers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Composition—trends, fluctuations, and outliers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Assessing PEA to nutrition-sensitive agriculture and climate-smart agriculture . . . . . . . . . . . . . . . . . . . . . . . . . 46 CHAPTER 5: Conclusions and Recommendations: How to Improve the Quality of PEA in Lesotho . . . . . . . 54 Implications of the review’s findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 ANNEX 1. Detailed Methodology and Data Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 ANNEX 2: Policy Coherence Analysis Summary Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 PEA per Ministry, Program, and Project, Total 2010/11 to 2015/16, ANNEX 3:  in Million Constant 2010 Maloti . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 ANNEX 4: Local Government Expenditures, by Project, Share of Total Expenditures 2010/11 to 2015/16 . 78 ANNEX 5: Rate of Subsidy per Fertilizer Type, Year and Region, 2012/13 to 2016/17 . . . . . . . . . . . . . . . . . . . 79 Input Procurement and Sale Value, per Input Category, 2012/13 to 2016/17, ANNEX 6:  in Million Constant 2010 Maloti . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 ANNEX 7: Table of Donor Projects, Constant and Current Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 BOXES Box 1:  Lesotho’s climate and international climate commitments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Box 2:  Nutrition in Lesotho . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Box 3:  Lesotho’s input subsidy program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Box 4:  Agricultural Productivity Programme for Southern Africa (APPSA) . . . . . . . . . . . . . . . . . . . . . . . . 38 Box 5:  Lesotho’s “Fato Fato” program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Box 6:  PEA under the El Niño drought response plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Box 7:  What agricultural activities can be considered nutrition-sensitive? . . . . . . . . . . . . . . . . . . . . . . . 47 iii iv  n Agriculture Public Expenditure Review Box 8: Overview of the climate finance tracking methodology of the multilateral development banks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Box 9:  Linking PEA to agricultural policy objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Box 10: Malawi’s Joint Sector Review: Bringing together stakeholders for a broad rural development agenda . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 FIGURES Figure E1:  Cost center breakdown of MAFS PEA, in million constant 2010 Maloti . . . . . . . . . . . . . . . . xiii Figure E2:  Cost center breakdown of MFRSC PEA, in million constant 2010 Maloti . . . . . . . . . . . . . . . xiii Figure E3:  Share of agriculture subsectors in AgBOOST and in AgGDP, in percent . . . . . . . . . . . . . . . xiv Figure 1:  Percent growth in GDP and agricultural GDP in Lesotho, 1988–2017 . . . . . . . . . . . . . . . . . . . . 2 Figure 2:  Lesotho’s four agro-ecological zones and ten districts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Figure 3:  Lesotho’s agrofood exports and imports, million US$ (2014–2016 average) . . . . . . . . . . . . . .9 Figure 4:  Lesotho’s agrofood exports by commodity, share of US$ (2014–2016 average) . . . . . . . . . . 9 Figure 5:  Lesotho’s agrofood Imports by commodity, share of US$ (2014–2016 average) . . . . . . . . . 10 Figure 6:  Functional and subfunctional perimeters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Figure 7: Final PEA and share in total public expenditures (%), PEA growth rate, and public expenditures growth rate (%) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Figure 8: Administrative composition of PEA, in million constant 2010 Maloti, actual expenditures . . . 21 Figure 9:  Cost center breakdown of MAFS PEA, in million constant 2010 Maloti . . . . . . . . . . . . . . . . . 22 Figure 10:  MFRSC capital expenditures, per output, average share for 2011/12, 2014/15, and 2015/16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Figure 11:  Cost center breakdown of MFRSC PEA, in million constant 2010 Maloti . . . . . . . . . . . . . . . . 22 Figure 12:  District-level PEA, MAFS (left) and MFRSC (right), in million constant 2010 Maloti . . . . . . . 24 Figure 13: Budgeted and final PEA (constant 2010 million Maloti); and PEA and total public expenditure execution rates (%) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Figure 14: Execution rates for MAFS at central and district departments and for operating costs at the district level; and execution rates for MAFS operating costs and personal emoluments at central and district levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Figure 15:  Execution rates for MFRSC at central and district departments and for operating costs at the district level; and execution rates for MFRSC operating costs and personal emoluments at central and district levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Figure 16:  Detailed economic structure of personal emoluments, by district and central level, MAFS, in total million constant 2010 Maloti . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Figure 17: Detailed economic structure of personal emoluments, by district and central levels, MFRSC, in total million constant 2010 Maloti . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Figure 18:  Share of graded employees per grade in the MAFS and the MFRSC . . . . . . . . . . . . . . . . . . 32 Figure 19:  Share of graded employee salary bill per grade in the MAFS and the MFRSC . . . . . . . . . . 32 Figure 20:  Breakdown of PEA by subsector, in percent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Figure 21:  Share of subsectors in AgBOOST and in AgGDP, in percent . . . . . . . . . . . . . . . . . . . . . . . . . 34 Figure B1:  Total procurement value 2012/13 to 2016/17, in million constant 2010 Maloti . . . . . . . . . . . 36 Figure 22:  Share of agricultural research salaries, operating and program costs, and capital investment, ASTI and AgBOOST, 2011–2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Figure 23:  Trend of PEA in support of aggregate functional categories, in million constant 2010 Maloti, 2010/11 to 2015/16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Figure B2:  Breakdown of El Niño response plan expenditures, in million Maloti . . . . . . . . . . . . . . . . . 40 Figure 24:  Assessing the appropriate level of government involvement in the agriculture sector . . 42 Figure 25:  PEA, by private sector and public sector functions, MAFS and MFRSC, in million constant 2010 Maloti . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Figure B3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Figure 26:  Nutrition-sensitive agriculture as share of total public expenditures to agriculture, 2015/16, in constant 2010 Maloti . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Table of Contents n v Figure B4:  Policy and expenditure cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Figure 27:  Distribution of budget lines with unlabeled program names in AgBOOST . . . . . . . . . . . . . 67 TABLES Table E1:  Main PEA indicators (million constant 2010 Maloti) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii Table E2:  Impacts, cost, and timeframe for proposed actions and reforms . . . . . . . . . . . . . . . . . . . . . . xvi Table 1:  Lesotho’s crop production (2014–2016 average) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Table 2:  Main PEA indicators, in million constant 2010 Maloti . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Table 3:  AgPER and ReSAKSS PEA, in million constant 2010 US$ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Table 4: PEA (million constant 2010 US$) and share of total public expenditures (%), 2010–2016 averages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Table 5:  Rural road expenditures by local governments and PEA, in million constant 2010 Maloti . . 23 Table 6:  District office PEA heat table, in million constant 2010 Maloti . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Table 7: Share of district-level PEA for MAFS (average 2010/11 to 2015/16) and share of fields planted (2009/10 census) per district, in percent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Table 8: Donor PEA in AgBOOST, OECD’S CRS, with SACU, SADC, and Africa average, in million constant 2010 US$ (BOOST) and million constant 2016 US$ (CRS) . . . . . . . . . . . . 26 Table 9:  Donor aid to Lesotho’s agriculture sector, OECD’S CRS and World Bank, in million constant 2016 US$ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Table 10:  Budgeted and actual PEA, recurrent and development, operating costs and personal emoluments with economic classification breakdown, in million constant 2010 Maloti . . . 29 Table 11: Recurrent and development expenditures and economic groupings, share of AgBOOST PEA and BOOST public expenditures, in percent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Table 12:  Farmers per extension worker, selected countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Table 13:  Heat table of the share of agricultural functions in PEA, 2010/11 to 2015/16 . . . . . . . . . . . . . 35 Table 14:  AgBOOST and ASTI PEA for research, 2010/11 to 2015/16, in million constant 2010 US$ . . 37 Table 15:  Functional count of agricultural interventions in the NSDP, count and percent of total . . . 43 Table 16:  Deviation between functional share of NSDP intervention count and PEA, in percentage points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Table 17:  Prioritization of sectors in the NSDP and in PEA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Table 18:  Functional count of agricultural interventions of the BFPs, count and percent of total . . . . 45 Table 19:  Deviation between functional share of BFPs intervention count and PEA, in percentage points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Table 20:  Prioritization of sectors in the BFPs and in PEA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Table 21:  Overview of NSA activities per district . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Table 22:  Overview of key projects that supported adoption of CSA practices (deflated to constant 2010 US$) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Table 23:  Impacts, cost, and timeframe for proposed actions and reforms . . . . . . . . . . . . . . . . . . . . . . 62 Table 24:  Administrative, economic, functional, and source of funding variables of the BOOST PEA database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Table 25:  Program codes in the BOOST database for MAFS and MFRSC . . . . . . . . . . . . . . . . . . . . . . . 68 Table 26:  Data limitations, implications, and recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Table 27:  AgPER subfunctional perimeter, based on selected categories from FAO’s MAFAP methodology (comprises COFOG categories 70421, 70422, 70423, and 7084) . . . . . . . . . . 71 Table 28:  Inclusion of AUGN-enhanced COFOG functions in the AgPER perimeter . . . . . . . . . . . . . . .72 Table 29:  Keywords master list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Table 30:  Administrative perimeter of the AgPER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Table 31:  Donor expenditures in current and constant million 2010 US$ . . . . . . . . . . . . . . . . . . . . . . . . 81 Table 32:  Donor PEA in AgBOOST, OECD’S CRS, with SACU, SADC, and Africa average in million constant 2010 US$ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 Table 33:  Donor aid to Lesotho’s agriculture sector, in million constant 2010 US$ . . . . . . . . . . . . . . . . 82 Acknowledgments This report was prepared by a World Bank team led by Åsa Giertz (Senior Agricultural Economist), consisting of Alban Mas Aparisi (Agricultural Policy Specialist, Consultant, Lead Author), Christine Heumesser (Economist), Yuxuan Zhao (Consultant), and Seitebatso Tsemane (Team Assistant). Amy Gautam edited the report. ­ The team gratefully acknowledges the collaboration with the Government of Lesotho on this work led by the Lesotho Ministry of Agriculture and Food Security, in particular the Hon. Lits’oane Lits’oane, Minister, Ministry of Agricul- ture and Food Security; Former Minister Mahala Molapo, Ministry of Agriculture and Food Security; Dr. Nthabiseng Makoae Deputy Minister, Ministry of Agriculture and Food Security; Mr. Malefetsane Nchaka, Principal Secretary, Ministry of Agriculture and Food Security; Mme Mathoriso Molumeli, former Director Planning and Policy Analysis, Ministry of Agriculture and Food Security; Mme Mabolaoana Phakisi, Acting Director Planning and Policy Analysis, Ministry of Agriculture and Food Security; Dr. Lefulesele Nteletsana Le­ besa, Director, Department of Agricultural Research, Ministry of Agriculture and Food Security; Mr. Ntitiea Tuoane, Director, and Department of Field Services, Ministry of Agriculture and Food Security; Mr. Selebalo Moeketsi, Director of Crops, Ministry of Agriculture and Food Security; and Mme Monica Lephole, Chief Research Officer, Ministry of Agriculture and Food Security. The team is grateful for the contributions to this work from the Ministry of Forestry, Range and Soil Conservation, in particular the Hon. Leshoboro Mohlajoa Minister, Ministry of Forestry, Range and Soil Conservation; Mr. Maseithati Mabeleng, Principal Secretary, Ministry of Forestry, Range and Soil Conservation; Mr. Sekoati Sekaleli, Director Forestry, Ministry of Forestry, Range and Soil Conservation; and Mr. Refuoe Boose, Director Conversation, Minis- try of Forestry, Range and Soil Conservation, as well as by the Disaster Risk Management Authority team led by Mr. Haretsebe Mahosi, Chief Executive Officer, Disaster Management Authority. The team gratefully acknowledges the coordinating role of the Ministry of Development and Planning, in particular Mr. Tlohelang Aumane Minister, Ministry of Development Planning; Ms. Nthoateng Lebona, Principal Secretary, Ministry of Development Planning; and Mrs. Malisebo ‘Mokela, Deputy Principal Secretary, Ministry of Development Planning. Finally the team grate- fully acknowledges the support from the Ministry of Finance, in particular the Hon. Dr. Moeketsi Majoro, Minister, Ministry of Finance; Ms. Motena Tsolo, Principal Secretary, Ministry of Finance; Mrs. ‘Mamakopoi Letsie, Deputy Principal Secretary, Ministry of Finance; and Maleshoane Lekomola, Budget Controller. Without their intense and trustful cooperation, this Agriculture Public Expenditure Review would not have been possible. The team would also like to thank World Bank colleagues Joachim Boko (Senior Social Protection Specialist), Ademola Braimoh (Senior Natural Resources Management Specialist), Holger Kray (Practice Manager), Gert Van ­ Der Linde (Lead Financial Management Specialist), Tahira Syed (Senior Rural Development Specialist), Meeta Sehgal (Senior Agricultural Specialist), and Bobojon Yatimov (Senior Agricultural Specialist) for sharing their insights. Critical guidance and support in the work for this report were provided by Paul Noumba Um (Former Country Director), Emmanuel Noubissie Ngankam (Country Program Coordinator), Janet K. Entwistle (Country Representa- tive), Mark E. Cackler (Former Practice Manager), Erwin De Nys (Program Leader), and Sebastien Dessus (Former Program Leader). The report greatly benefited from peer reviewers Ulrich Schmitt (Lead Agriculture Economist), Irina Schuman (Senior Economist), Rachel Lemay Ort (Public Sector Specialist), and Leopold Ghins (Agriculture Policy and Public Expenditure Specialist, External Peer Reviewer). vi Abbreviations and Acronyms AgBOOST Agricultural BOOST AgPER Agriculture Public Expenditure Review APPSA Agricultural Productivity Programme for Southern Africa ASTI Agricultural Science and Technology Indicators AU African Union AUGN African Union Guidance Note BFP Budget Framework Paper CAADP Comprehensive Africa Agriculture Development Programme COFOG Classification of the Functions of Government CPI Consumer Price Index CRS Creditor Reporting System CSA Climate-Smart Agriculture CSAIP Climate Smart Agriculture Investment Plan DAO District Agricultural Office DAR Department of Agricultural Research DO District Office ENSO El Niño Southern Oscillation FAO Food and Agriculture Organization of the United Nations FY Fiscal Year GDP Gross Domestic Product GEF Global Environmental Facility GHG Greenhouse Gas GoL Government of Lesotho HQ Headquarters ICP Intensive Crop Production Programme IFAD International Fund for Agricultural Development IFMIS Integrated Financial Management Information System IFPRI International Food Policy Research Institute IPCC Intergovernmental Panel on Climate Change  JSR Joint Sector Review Kg Kilogram MAFAP Monitoring and Analysing Food and Agricultural Policies MAFS Ministry of Agriculture and Food Security MCC Millennium Challenge Corporation MDB Multilateral Development Bank MDP Ministry of Development and Planning vii viii  n Agriculture Public Expenditure Review MF Ministry of Finance MFD Mobilizing Finance for Development MFRSC Ministry of Forestry, Range and Soil Conservation MHA Ministry of Home Affairs MPS Ministry of Public Service MSBDCM Ministry of Small Business Development, Cooperatives, and Marketing MTEF Medium Term Expenditure Framework MTI Ministry of Trade and Industry NAIP National Agriculture Investment Plan NATO Nutrition Area Technical Officer NSA Nutrition-Sensitive Agriculture NSDP National Strategic Development Plan OECD Organisation for Economic Co-operation and Development PCA Policy Coherence Analysis PEA Public Expenditures to Agriculture PEFA Public Expenditure and Financial Accountability PER Public Expenditure Review PSMP Public Sector Modernisation Project R&D Research and Development ReSAKSS Regional Strategic Analysis and Knowledge Support System RISDP Regional Indicative Strategic Development Plan SACU Southern African Customs Union SADC Southern African Development Community  SADP Smallholder Agriculture Development Project SDG Sustainable Development Goal UNDP United Nations Development Programme WHO World Health Organization WMP Watershed Management Programme EXCHANGE RATES 2010 2011 2012 2013 2014 2015 2016 2017 Maloti/US$* 7.3198 7.2599 8.212 9.6492 10.8789 13.9865 14.7101 13.4 *Throughout the report, unless otherwise indicated, 2010 constant values are being used for Maloti and US$. Source: Central Bank of Lesotho, 2019. Executive Summary Why an Agriculture Public Expenditure Review? Almost 60 percent of Lesotho’s population live in rural areas and depend directly or indirectly on agriculture for their livelihood, and growth in the sector is crucial for poverty reduction. Nevertheless, agricultural produc- tivity is low and value addition is limited. Poverty is high in Lesotho: about 49.7 percent of the population live below the national poverty line and 24.1 percent fall below the extreme poverty line, with expenditures below minimum food requirements; 75.6 percent of those below the poverty line are engaged in agrciculture compared with 65.4 per- cent of non-poor (World Bank 2019, forthcoming). As most of those engaged in agriculture are smallholders with less than one hectare per family, growth in the sector will be central for poverty reduction in Lesotho, but productivity and cereal yields are low (on average below 1,000 kg per hectare, compared with the Southern African Development Community (SADC) target of at least 2,000 kg per hectare, and land productivity averages about US$70 per hectare annually compared to the regional average of about US$120 per hectare (2008–2013 figure). Value addition is low outside the wool and mohair industry, and most agroprocessing takes place in South Africa. Low investment in agriculture has rendered the sector vulnerable to weather impacts and poorly equipped to adapt to climate change. The sector’s low productivity is largely a result of low investment in infrastructure, including irrigation, low uptake of new technologies and inputs, poor quality extension and advisory services, and limited access to credit. Environmental challenges such as land and natural resource degradation, erosion, and low soil fertility exacerbate the situation, leaving the sector vulnerable to risks. High intraseasonal and interannual rainfall variability with frequent droughts, including the El Niño Southern Oscillation (ENSO), have resulted in poor crop harvests and large livestock losses for rural farmers. According to the Lesotho Vulnerability Assessment Com- mittee, Lesotho experienced an almost 62 percent decline in crop production during the unprecedent El Niño event ix x  n Agriculture Public Expenditure Review in 2015/2016 (United Nations Office of the Resident Coordinator 2016). The Intergovernmental Panel on Climate Change (IPCC) categorizes Lesotho as one of the countries highly vulnerable to the impacts of climate change, with low capacity among the population to adapt to changing weather conditions (Dejene et al. 2011). Low agricultural productivity and poverty are closely linked with food insecurity and malnutrition in Leso- tho. About one-third of all children under five are stunted in Lesotho, and malnutrition is closely linked to poverty: almost one-half of all children under five are stunted in the lowest income quintile compared to 10 percent of children in the highest. The burden of malnutrition doesn’t just fall on the individual but also affects the economy as a whole: it’s estimated that over 7 percent of gross domestic product (GDP) is foregone annually due to lower education outcomes, decreased productivity, poor health, and high child mortality, and premature deaths as a result of malnutrition. Margins are small for Lesotho’s poor, as 41 percent of all families spend more than one-half of their income on food, and recent extreme weather events in Lesotho have exposed the population’s vulnerability to food insecurity. Adverse weather impacts on agricultural production left around 709,000 people food-insecure in 2015/16; over 200,000 people were in need of humanitarian assistance in 2017; and over 485,000 were reported being at risk of food insecurity in May 2019 (Lesotho Vulnerability Committee 2017; Lesotho Disaster Risk Manage- ment Authority 2019). Recognizing the agriculture sector’s importance for “job creation and inclusive economic growth under a new growth path led by the private sector,” the Government of Lesotho (GoL) selected agriculture as one of four productive sectors1 central to its new National Strategic Development Plan (NSDP) II for 2018/19–2022/23. To strategically support agriculture’s role in climate change mitigation and adaptation, and in nutrition and food security, the GoL is in the process of developing an Integrated Program for Agriculture and Food System Develop- ment (a National Agriculture Investment Plan, NAIP), a Multi-Sectoral Nutrition Strategy, and with the support of the World Bank, a Climate Smart Agriculture Investment Plan (CSAIP)—all of which are expected to be ready for adoption in 2019. These are supported by significant World Bank–financed investments approved in FY19 under the Second Phase Smallholder Agriculture Development Project (SADP-2, US$52 million), and the Agricultural Pro- ductivity Programme for Southern Africa (APPSA, US$20 million). In the pipeline for preparation is also a project to support Health and Nutrition Services. To strengthen the alignment between strategic objectives and the allocation of public expenditures to agri- culture, the GoL asked the World Bank for support to conduct an Agriculture Public Expenditure Review (AgPER) as part of a larger GoL effort to strengthen Public Financial Management. Starting in FY17, the World Bank together with the GoL conducted the Public Expenditure Review (PER): Improving Expenditure Efficiency for Inclusive Development and Growth (World Bank 2018) with the overarching objectives to: identify inefficiencies in government spending and revenue collection; identify opportunities for creating fiscal space in the context of a significant decline in Southern African Customs Union (SACU) and domestic revenues; analyze the performance of critical areas and/or sectors; and improve fiscal planning and management for better service delivery. The GoL expressed early on a wish for World Bank support in conducting an AgPER to better understand how expenditures can be aligned with objectives for the sector, and guidance for evaluating the impacts of public support. As a signa- tory to the Maputo Declaration, Lesotho has committed to allocate about 10 percent of its public spending to agri- culture. The update to the Maputo Declaration is the Malabo Declaration of 2014, which also emphasizes agriculture public expenditure quality. Other sector-specific PERs that have been conducted are the Education Public Expendi- ture Review (World Bank 2019) and the Public Health Sector Expenditure Review (World Bank 2017). In response to the GoL’s request, the development objective of this AgPER is to identify measures to improve the quality of public expenditures in agriculture. The hypothesis of this AgPER is that despite strategic objectives, policies, and international commitments that seek to address important key challenges for Lesotho such as climate 1 The other three productive sectors included in the draft NSDP II are Manufacturing, Tourism and Creative Industries, and Technology and Innovation. Executive Summary n xi change impacts, nutrition, and private sector development, public expenditures to agriculture (PEA) are not well aligned with these policies and priorities. The main audience of the report is policy makers and technical staff work- ing on agricultural budgets and expenditures in the Ministry of Agriculture and Food Security (MAFS), the Ministry of Forestry, Range and Soil Conservation (MFRSC), and the Ministry of Finance (MF). In addition, this report can be seen as an important document for the Ministry of Development and Planning (MDP) and for multilateral and bilateral partners engaged in the sector. This AgPER was conducted in close collaboration with the GoL, particularly the MAFS, the MFRSC, and the MF. In response to a request from the MAFS Policy Department to use this AgPER as an opportunity to transfer knowledge to the GoL, a one-day training on “Assessing the Expenditures in Agriculture” was organized in November 2018. In addi- tion to the MAFS, technical staff from the MFRSC, the MDP, the Ministry of Small Business Development, Cooperatives, and Marketing (MSBDCM), and the Ministry of Trade and Industry (MTI) participated in the training. The results of the AgPER were presented to stakeholders in a workshop in Maseru in March 2019. Participants came from the public and private sectors, and their feedback on the findings and their recommendations were incorporated in this report. Data and Methodology For the purpose of this AgPER, the MF’s BOOST database was converted into an Agricultural BOOST data- base (AgBOOST) containing public expenditures in support of agriculture (PEA). The Lesotho BOOST data- base,2 which reproduces the data from the Integrated Financial Management Information System (IFMIS), is used by the MF and line ministries to plan, execute, and monitor the budget. Although the IFMIS and BOOST have certain limitations, these are the most comprehensive data sources available to analyze Lesotho’s official budget. AgBOOST covers approved and revised estimates, as well as warranted release and final expenditures, for the period 2010/11 to 2015/16. The data were also converted into Maloti and US$ expressed in 2010 constant values. Official exchange rate and Consumer Price Inflation (CPI) data from the Lesotho Central Bank were used. Categorical variables in the AgBOOST database allow for analysis of the budgetary data, but they lack depth on aspects that are of importance for the review. Other variables were thus created using various sources of qual- itative and quantitative information. Four types of variables can be distinguished: administrative, economic, func- tional, and source of funding. However, the GoL currently uses only one level of functional coding, which makes it difficult to use AgBOOST for functional analysis, and District Office (DO) spending is only broken down by economic attributes, not by function. To strengthen the analysis and allow for more details, qualitative and quantitative infor- mation from the MAFS and the MFRSC were used to create functional, sectoral, and geographical variables. In line with the African Union Guidance Note (AUGN), extra budgetary funds were generally not included in this AgPER. Agriculture Public Expenditures: How much is allocated and where does it go? Between 2010/11 and 2015/16, 322.7 million Maloti, or 1,513.5 Maloti per rural capita, were allocated to PEA, accounting for less than 3 percent of total public expenditures in Lesotho. This is in the context of the sector contributing 10 percent to GDP and over half of the population depending on agriculture for their live- lihoods. PEA are below what Lesotho has committed under the Comprehensive Africa Agriculture Development Programme (CAADP) (10 percent of total public expenditures) and fall below allocations of other countries in the region. Lesotho stands almost 3 percentage points below the SADC average (excluding the outlier service econo- mies of Seychelles, South Africa, and Mauritius), including Madagascar (7.4 percent of public expenditures), Malawi (16 percent), Zambia (8.1 percent), and Zimbabwe (8.3 percent); although it is worth noting that, Malawi aside, none of these countries reach the Maputo target. Nevertheless, actual PEA are about twice the amount that Lesotho reports to ReSAKSS (Regional Strategic Analysis and Knowledge Support System). Note that while public expendi- tures increased 4.9 percent annually from 2010/11 to 2015/16, PEA only increased 0.9 percent on average in the same period (Table E1). 2 The BOOST database is produced by the MF of Lesotho with support from the World Bank. xii  n Agriculture Public Expenditure Review TABLE E1:  Main PEA indicators (million constant 2010 Maloti) 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 Average Budgeted PEA 389.9 417.1 313.9 428.5 443.5 475.2 411.3 Final PEA 298.3 318.7 282.1 379.7 348.1 309.0 322.7 PEA per capita (in Maloti) 1,069.9 1,130.3 988.0 1,312.6 1,187.5 1,040.1 1,121.4 PEA per rural capita (in Maloti) 1,422.7 1,513.3 1,329.8 1,776.2 1,615.7 1,423.1 1,513.5 Budgeted PE 10,787.2 12,497.8 12,321.0 12,435.3 13,067.3 12,437.7 12,257.7 Final PE 8,418.1 9,789.1 8,885.5 9,882.8 11,245.3 11,219.6 9,906.7 Agricultural GDP (AgGDP) 895.2 942.3 910.7 1,132.9 1,173.3 1,094.2 1,024.8 GDP 17,286.5 18,479.5 19,588.0 19,949.1 20,571.9 21,141.0 19,502.7 Share AgGDP in total GDP 5.2% 5.1% 4.6% 5.7% 5.7% 5.2% 5.3% Share budgeted PEA/PE 3.6% 3.3% 2.5% 3.4% 3.4% 3.8% 3.4% Share final PEA/PE 3.5% 3.3% 3.2% 3.8% 3.1% 2.8% 3.3% Share final PEA/AgGDP 33.3% 33.8% 31.0% 33.5% 29.7% 28.2% 31.6% Share final PEA/GDP 1.7% 1.7% 1.4% 1.9% 1.7% 1.5% 1.7% PEA growth — 6.4% –13.0% 25.7% –9.1% –12.7% 0.6% Public expenditures growth — 16.3% –9.2% 11.2% 13.8% –0.2% 4.9% Source: Authors and ReSAKSS 2019. PEA are unpredictable and affect the GoL’s capacity to support agriculture sector development. The volatility of PEA partly reflects that of the general budget,3 with strong jumps and falls; the largest was a 13 percent drop in 2012/13, followed by a 25.7 percent growth spurt. Nevertheless, PEA exhibit their own pattern and appear even more unpredictable than the general budget. They seesawed through the entire period of analysis, except for two consecutive years of negative growth in 2014/15 and 2015/16. Budget execution for both the MAFS and the MFRSC is volatile, with execution rates both below and above budgeted amounts depending on circumstances. In certain years, this is due to late payments that shift execution from one fiscal year to the next. Breaking down execution rates between ministry headquarters (HQ) and DOs shows that transfers from the MAFS’s DOs have supported the Intensive Crop Production Programme (ICP) at HQ. Execution rates for personal emoluments are largely in line with budgeted amounts, while operating costs are reduced when cuts have to be made. As discussed in the report, the low operating costs reportedly affect program implementation. Almost all PEA are allocated through the MAFS (64 percent) and the MFRSC (33 percent); the highest shares go to the former’s Crops Department and the latter’s Conservation Department. Only 3 percent of PEA go through other ministries: the MF, the MTI, and the Ministry of Home Affairs (MHA). In the MAFS, the Crops Depart- ment has the highest share of spending (37 percent of total spending); in the MFRSC, 75 percent of total expendi- tures go to the Conservation Department. The MAFS’s DOs receive about 30 percent of MAFS’s spending and the MFRSC’s DOs about 15 percent. Geographically, PEA are allocated largely according to the spatial distribution of agricultural fields across the country. In the MAFS, the Crops Department spends more than four times the amount of the next unit and more than the 10 DOs combined (Figure E1). This is because the Crops Department is in charge of the large input subsidy Summer Cropping Programme, also referred to as the ICP. The rest of the MAFS’s expenditures are spread between Lesotho Agricultural College (119 million Maloti), the Extension (71 million Maloti) and Livestock Departments 3 PEA are expressed in constant terms and for actual expenditures. The growth rate variation thus captures inflation and budget execution issues. Executive Summary n xiii FIGURE E1:  Cost center breakdown of MAFS PEA, in million constant 2010 Maloti 0104 Crops 0101 Administration 0106 Lesotho Agricultural College 0107 Extension 0103 Livestock 0108 DAO Maseru 0113 DAO Mohales Hoek 0105 Research 0111 DAO Berea 0110 DAO Leribe 0109 DAO Butha Buthe 0112 DAO Mafeteng 0116 DAO Thaba Tseka 0114 DAO Quthing 0115 DAO Qachas’ Nek 0117 DAO Mokhotlong 0102 Planning 0 50 100 150 200 250 300 350 400 450 500 Million constant 2010 Maloti (61 million Maloti), the Research Department (39 million Maloti), and the Planning Department (18 million Maloti). MAFS DOs spent 369 million Maloti in total, or 30 percent of MAFS expenditures. The MFRSC’s Conservation Department registers three-quarters of the ministry’s total expenditures, while other departments and DOs have smaller allocations (Figure E2). The MFRSC’s flagship program is the Water- shed Management Programme (WMP), also referred to as the Integrated Catchment Management Programme or the Poverty Alleviation Programme, but most commonly known as Fato Fato. This program largely consists of build- ing anti–soil erosion infrastructure, in particular gully structures, stone lines, and diversion furrows. Interviews with GoL staff and the literature review point to Fato Fato’s social objective (“cash for work”) on top of its conser- vation focus. The program is therefore heavily reliant on unskilled labor. The Rangeland Management Department activities are similar but marginal, at 2 percent of MFRSC expenditures. The Forestry Department (2 percent of MFRSC PEA) focuses on subsidizing households with free fruit trees—the trees are bought from producers that are FIGURE E2:  Cost center breakdown of MFRSC PEA, in million constant 2010 Maloti 1403 Conservation 1405 Districts 1401 Administration 1402 Forestry 1404 Range Resources Management 0 50 100 150 200 250 300 350 400 450 500 Million constant 2010 Maloti xiv  n Agriculture Public Expenditure Review supported with extension services—and planting forest trees (afforestation). MFRSC DOs (15 percent of MFRSC PEA) are all managed under one department, though they do have a subcost center status and appear in the IFMIS/ BOOST. The MFRSC does not have a research department. Even though Lesotho has a relatively high wage bill in general, the share of PEA going to wages (56.1 per- cent) is higher than that of total public expenditures (37.7 percent). This is problematic for agriculture sector programs, as operating costs are reportedly too low to enable staff to implement them properly. In the MAFS DOs, the salary bill is as high as 91 percent of total spending, leaving little resources to finance operating costs (e.g., for extension workers). Similarly, the Livestock Department, which generally has high expenses for vaccines and other operating costs, spends 71 percent of its budget on staff. Even the MFRSC’s capital-intensive Fato Fato absorbs 70 percent of the MFRSC wage bill. This goes to the many temporary laborers employed under Fato Fato as part of its dual objectives of environmental restoration and cash for work. The wage structure differs for the MAFS and the MFRSC, but for both ministries, the number of staff is relatively high at low grade levels, such as support staff. Relative PEA to individual subsectors are not aligned with their contributions to agricultural GDP. Most sector-specific expenditures go to the crop sector (83 percent), followed by livestock (14 percent) and forestry ­ (3 percent). Consultations with the ministries confirm that multi-sectoral expenditures largely follow the same pattern. This composition is not in line with the subsectors’ relative contribution to agricultural GDP, for which the livestock sector accounts for 52 percent, crops 28 percent, and forestry 20 percent (Figure E3). The ICP accounts for 69 percent of crop expenditures. FIGURE E3:  Share of agriculture subsectors in AgBOOST and in AgGDP, in percent 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Livestock Crops Forestry Share in AgGDP Share in AgBOOST Source: Authors and Central Bank of Lesotho (AgGDP data). Note: Shares in AgGDP from the Central Bank data are 50 percent for livestock, 25 percent for crops, 19 percent for forestry, and 6 percent for services. Services were removed to compute relative shares of the three sectors. PEA going to recurrent expenditures and expenditures of private nature are high, while development expenditures are low. Of PEA under the MAFS, about 40 percent are spent on goods and services that are of a private nature; for the MFRSC, 97 percent of expenditures go to public goods and services. The high level of PEA spent on private goods by the MAFS is largely because of the ICP. In addition, these subsidies go to inputs rather than capital. Public investment in infrastructure is very low (0.5 percent), as are PEA that support agricultural pro- cessing and marketing activities (0.2 percent), although donor-funded projects not reported in BOOST complement this allocation. Public financing to agricultural research and development (R&D) is 0.8 percent. While development expenditures account for about two-thirds of total PEA, they are overestimated as they include expenditures that are recurrent in nature, especially the ICP and salaries under Fato Fato. Executive Summary n xv Likely as a result of the current budget process, PEA are not particularly well aligned with the main strate- gic documents for the sector—the NSDP and the Budget Framework Papers (BFPs). Budgeting of PEA is done by the individual departments in each ministry and while timely, it is not an integral part of the implementation of key strategic documents for the sector. Further, no institutionalized coordination of investment in the sector occurs between the GoL and the donor community, and donor funding was not mainstreamed into the budget in the review period. As a result, an assessment of the policy coherence of PEA shows that input subsidies and extension and advi- sory services are overfunded compared to what is envisioned in strategic documents, while agricultural processing and marketing, irrigation, and R&D are severely underfunded. Allocations to nutrition-sensitive agriculture (NSA) and to climate-smart agriculture (CSA) are low given the GoL’s commitments to address the challenges related to malnutrition and to climate change. Recommendations Perhaps the most significant finding of this AgPER is how poorly PEA are aligned with policy development objectives for the sector. Instead, PEA are largely spent on recurrent expenditures, which are unlikely to contrib- ute to growth and longer term development. Going forward, the link between policy and allocations of PEA must be strengthened across ministries. It is also recommended that the MAFS—in consultation with the MFRSC, and as part of the NSDP II—adopt an evidence-based agricultural policy strategy with a NAIP to guide allocations of PEA.4 In the implementation of the NSDP II and a sector strategy/NAIP, it will be important to establish a feedback loop structure with rigorous results indicators for evaluating the impacts of the different programs, especially those that absorb a large share of PEA. Having such structure in place would help ministries better evaluate the effectiveness of different expenditures and better target resources where they have the most impact. Setting up such a structure would include establishing a Joint Sector Review (JSR)—in line with CAADP directives—to coordinate investment in the agriculture sector across ministries, other government agencies, donors, and nongovernmental actors. Continuing to update AgBOOST and refine classification will help analysts from the MAFS, the MFRSC, the MF, and other bodies evaluate PEA’s allocation and impacts. However, technical staff in the MAFS and the MFRSC Policy Department/Unit will need training on how to use AgBOOST in their day-to-day work. Also, making reporting of PEA more consistent across different ministries and including donor financing in the budget will ensure that anal- yses are comprehensive. Including functional classifications in BOOST will further allow for more detailed analysis. Where possible, it is recommended that the GoL provides public rather than private goods and services, and strives for a better balance between development and recurrent spending. It is not clear that the GoL has found its optimal role in the agriculture sector at this point, and it would be good to better assess current justifications for large programs of private nature, such as the ICP. Rebalancing the budget more toward development expenditures includes allocating more PEA to capital investments, which are currently low. Similarly, increased allocations to R&D would help the GoL meet its aspirations for a competitive agriculture sector that is a key source of employment. It is recommended that MAFS and MFRSC cost centers be better balanced, especially given the sector’s struc- ture. For example, the MAFS could consider balancing out the budget more toward the Livestock, Research, and Planning Departments, for which funding is currently low given the relative contribution of sectors and the empha- sis in the NSDP. Similarly, balancing personnel costs and capital investments against the necessary operating and maintenance costs will be important to effectively implement certain programs in the sector and to ensure sustain- ability of, e.g., infrastructure investment. Table E2 provides an overview of expected impacts, assessed relative costs, and a proposed implementation time- frame for the recommendations proposed in this report. 4 It can be noted that work to develop a NAIP is currently on-going with the support from FAO and in coordination with the African Union/CAADP. The report is, at the time of writing, in the approval process. It has at the time of writing this report not been appraised for Malabo Declaration (2014) compliance. xvi  n Agriculture Public Expenditure Review TABLE E2:  Impacts, cost, and timeframe for proposed actions and reforms Impact on expenditure quality Timeframe Cost Medium (<5y) Long (>5y) Short (<1y) Moderate Medium High High Low Low Responsible Priorities and measures agency Priority 1: Adjusting the level and composition of Public Expenditures to Agriculture to address challenges and achieve objectives for the sector (i) Optimizing the GoL’s support to private goods and MAFS High Medium Low services • Evaluating the impacts of the Intensive Crop Production High Short Low Programme (ICP), including on private input providers and implementing steps to minimize interference (ii) Rebalancing the budget more toward development rather MAFS & High Medium Moderate than recurrent allocations MFRSC (iii) Increasing the share of operating costs so that they MAFS & High Medium Moderate allow for implementation of, e.g., extension and livestock MFRSC services, and for operation and maintenance of capital investment High Medium Low (iv) Rebalancing allocations between cost centers so that MAFS & they are more in line with the contribution to Agriculture MFRSC GDP, priority topics, or other socioeconomic objectives, including Climate-Smart Agriculture, Nutrition-Smart Agriculture, and private sector led job growth Priority 2: Establishing policy-based expenditure allocations (i) Aligning the MAFS’s and the MFRSC’s budgets with MAFS & High Short Moderate relevant policies, especially the National Strategic MFRSC Development Plan (NSDP) II MAFS in High Medium Moderate (ii) Adopting an evidence-based agricultural policy consultation strategy with a National Agriculture Investment Plan for with MFRSC implementing the policy Priority 3: Better tracking Public Expenditures to Agriculture (i) Updating AgBOOST MAFS & Medium Medium Low MFRSC (ii) Including functional classifications in AgBOOST MAFS, Medium Medium Moderate MFRSC, MF (iii) Including donor funded activities in Department and MAFS, Ministry budget planning with the objective of including MFRSC, MF Medium Short Low donor funding in the Integrated Financial Management Information System (IFMIS), pending compliance with donors’ financial management requirements MAFS & High Short Low (iv) Training relevant staff on how to use AgBOOST in budget MFRSC planning and policy evaluation Priority 4: Better evaluating the impacts of Public Expenditures to Agriculture (i) Establishing a feedback loop with results indicators to MAFS & High Short Moderate evaluate impacts of large programs MFRSC (ii) Establishing a Joint Sector Review under the new National MAFS & Medium Medium Low Agriculture Investment Plan MFRSC CHAPTER 1: Introduction: Why an Agriculture Public Expenditure Review for Lesotho? Almost 60 percent of Lesotho’s population live in rural areas and depend directly or indirectly on agricul- ture for their livelihood, and growth in the sector is crucial for poverty reduction. As a share of the total econ- omy, agriculture is a relatively small sector, contributing less than 10 percent of Lesotho’s gross domestic product (GDP), but the majority of people in rural areas are engaged in agriculture, making it the main sector for livelihoods and food security. Poverty is high in Lesotho: about 49.7 percent of the population live below the national poverty line and 24.1 percent fall below the extreme poverty line, with expenditures below minimum food requirements; 75.6 percent of those falling below the poverty line are engaged in agriculture compared with 65.4 percent of non- poor (World Bank 2019, forthcoming). As most of those engaged in agriculture are smallholders with less than one hectare per family, growth in the sector will be central for poverty reduction in Lesotho. Despite the importance of growth in agriculture, productivity remains low and value addition activities in the sector are limited. The sector consists mainly of rain-fed crop production (30 percent, of which cereals are grown on 85 percent of arable land) and livestock production (70 percent). Cereal yields are below 1,000 kilograms (kg) per hectare on average, compared with the Southern African Development Community (SADC) Regional Indic- ative Strategic Development Plan (RISDP) target of at least 2,000 kg per hectare; land productivity averages about US$70 per hectare annually, compared to the regional average of about US$120 per hectare (2008–2013 figure). Value addition is low outside the wool and mohair industry, and most agroprocessing takes place in South Africa. Low investment in agriculture has left the sector vulnerable to adverse weather impacts and poorly equipped to adapt to climate change. The sector’s low productivity is largely a result of low investment in infrastructure 1 2  n Agriculture Public Expenditure Review including irrigation, low uptake of new technologies and inputs, poor quality extension and advisory services, and limited access to credit. Environmental challenges such as land and natural resource degradation, erosion, and low soil fertility exacerbate the situation, rendering the sector vulnerable to risks. On interannual timescales, the El Niño Southern Oscillation (ENSO) phenomenon particularly affects climate variation in Lesotho (World Bank 2016a). High intraseasonal and interannual rainfall variability with frequent droughts have resulted in poor crop harvests and large livestock losses to rural farmers. As a result, growth in the sector is highly volatile and frequently below the Comprehensive Africa Agriculture Development Programme’s (CAADP) 6 percent growth objective (Fig- ure 1). Looking forward, the Intergovernmental Panel on Climate Change (IPCC) categorizes Lesotho as one of the countries highly vulnerable to the impacts of climate change, with low capacity among the population to adapt to changing weather conditions (Dejene et al. 2011). FIGURE 1:  Percent growth in GDP and agricultural GDP in Lesotho, 1988–2017 40 20 0 –20 –40 88 90 92 94 96 98 00 02 04 06 08 10 12 14 16 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 GDP growth (annual %) Agriculture, value added (annual % growth) Source: WDI database 2018. Low agricultural output and productivity, poverty, and heavy reliance on food imports are closely inter- twined with food and nutrition insecurity in Lesotho, which impose high costs on the country. An esti- mated 80 percent of the food consumed in Lesotho is imported. This applies to staples (e.g., 60 percent of maize is imported) as well as highly nutritious foods such as fruits and vegetables. Forty-one percent of all families spend more than one-half of their income on food, yet diets are still poor and malnutrition is high in Lesotho. One-third of children under five in Lesotho are stunted, with almost one-half of all children under five being stunted in the lowest income quintile. The challenges of malnutrition continue into adulthood; chronic malnutrition costs Lesotho an estimated 7.3 percent of its GDP, and 19.5 percent of child mortality is due to malnutrition (9,272 child deaths in the period 2008–2014). Recent extreme weather events in Lesotho have exposed the population’s vulnerability to food insecurity: the drought during the 2015/16 growing season was the most severe on record, leaving around 709,000 people food-insecure in terms of access to adequate energy intake (Lesotho Vulnerability Assessment Committee 2016). In 2017, over 200,000 people were in need of humanitarian assistance due to weather risks. Countrywide below aver- age rainfalls from September to December 2018 resulted in decreased availability of crop and livestock products in markets around Lesotho. A recent Rapid Assessment estimates that 487,587 people are currently food insecure in Lesotho (Lesotho Disaster Risk Management Authority 2019). Recognizing the agriculture sector’s importance for “job creation and inclusive economic growth under a new growth path led by the private sector,” the Government of Lesotho (GoL) selected agriculture as one of four productive sectors5 central in its new National Strategic Development Plan (NSDP) II for 2018/19–2022/23. 5 The other three productive sectors included in the draft NSDP II are Manufacturing, Tourism and Creative Industries, and Technology and Innovation. Chapter 1: Introduction: Why an Agriculture Public Expenditure Review for Lesotho? n 3 Further, to strategically support agriculture’s role in climate change mitigation and adaptation, and in nutrition and food security, the GoL is in the process of developing an Integrated Program for Agriculture and Food System Devel- opment (a National Agriculture Investment Plan, NAIP), a Multi-Sector Nutrition Strategy, and with the support of the World Bank, a Climate Smart Agriculture Investment Plan (CSAIP)—all of which are expected to be ready for adoption in 2019. These are supported by significant World Bank-financed investments approved in FY19 under the Second Phase Smallholder Agriculture Development Project (SADP-2, US$52 million), the Agricultural Productivity Programme for Southern Africa (APPSA, US$20 million). A project to support Health and Nutrition services is in the pipeline for preparation. To strengthen the alignment between objectives and the allocation of public expenditures to the sector, the GoL asked the Bank for support to review Public Expenditures in Agriculture (AgPER) as part of a larger GoL effort to strengthen Public Financial Management. Starting in FY17, the World Bank together with the GoL conducted the Public Expenditure Review (PER): Improving Expenditure Efficiency for Inclusive Development and Growth (World Bank 2018) with the overarching objectives to: identify inefficiencies in government spending and revenue collection; identify opportunities for creating fiscal space in the context of a significant decline in Southern African Customs Union (SACU) and domestic revenues; analyze the performance of critical areas and/ or sectors; and improve fiscal planning and management for better service delivery. The PER focused on assessing the detailed fiscal situation and identifying fiscal space that looked at overall revenue performance, taxation, public sector employment, and the wage bill, and public sector investment management. In addition to this work, the GoL asked for the World Bank’s support to conduct sector-specific PERs and expressed early on a wish for World Bank (i) support in conducting an AgPER to better understand how expenditures can be aligned with objectives for the sector, and (ii) guidance for evaluating the impacts of public support to the agriculture sector. As a signatory to the Maputo Declaration, Lesotho has committed to allocate about 10 percent of its public spending to agriculture. The update to the Maputo Declaration is the Malabo Declaration of 2014, which also emphasizes agriculture public expenditure quality. Other sector-specific PERs that have been conducted are the Education Public Expenditure Review (World Bank 2019) and the Public Health Sector Expenditure Review (World Bank 2017). In response to the GoL’s request, the development objective of this AgPER is to identify measures to improve the quality of public expenditures in agriculture. The hypothesis of this AgPER is that despite strategic objectives, policies, and international commitments that seek to address important key challenges for Lesotho, such as climate change impacts, nutrition, and private sector development, public expenditures to agriculture (PEA) are not well aligned with these policies and priorities. The main audience of the report is policy makers and technical staff work- ing on agricultural budgets and expenditures in the Ministry of Agriculture and Food Security (MAFS), the Ministry of Forestry, Range and Soil Conservation (MFRSC),6 and the Ministry of Finance (MF). In addition, this report can be seen as an important document for the Ministry of Development and Planning (MDP) and for multilateral and bilateral partners engaged in the sector. To achieve its objective, this report seeks to answer the following questions: 1. What share of public expenditures is allocated to agriculture? Is it aligned with the size of the sector in the economy and with international commitments under CAADP? 2. What is the composition of spending in agriculture? What is the balance of spending between development and recurrent expenditures? 3. Are public expenditures in agriculture prioritized and sustainable, or are public expenditures to the sector heavily reliant on external funding from donors and other sources? 4. Are expenditures consistent with agricultural policies and sector development goals? 5. Are the GoL’s commitments to climate-smart agriculture (CSA) and nutrition reflected in public expenditures to the sector? 6 Former Ministry of Forestry and Land Reclamation (MFLR). 4  n Agriculture Public Expenditure Review This AgPER was conducted in close collaboration with the GoL, in particular the MAFS, the MFRSC, and the MF. In response to a request from the MAFS Department of Policy and Planning to use this AgPER as an opportunity to transfer knowledge to the GoL, a one-day training in Assessing the Expenditures in Agriculture was organized in November 2018. The training covered technical components, tools, and methodological principals for PEA analysis, sources and classification methods, key analytical indicators, and approaches for analyzing the special topics cov- ered in this report (Climate Smart Agriculture/CSA, Nutrition-Sensitive Agriculture/NSA, and Mobilizing Finance for Development/MFD in agriculture). In addition to staff and Deputy Directors from the MAFS, technical staff from the MFRSC, the MDP, the Ministry of Small Business Development, Cooperatives, and Marketing (MSBDCM), and the Ministry of Trade and Industry (MTI) participated in the training. Further, the results of the AgPER were presented to stakeholders in a workshop in Maseru in March 2019. Participants came from the public and private sectors, and their feedback on the findings and their recommendations were incorporated in this report. The report is structured as follows: Chapter 2 provides an overview of the agriculture sector in Lesotho, out- lines some of its main challenges, and provides an overview of the main policy documents that guide investment in the sector. Chapter 3 describes the data and methodology used for the analysis of this AgPER and is particularly intended for technical staff who work with agricultural budgeting and expenditures. The detailed description of the methodology in this chapter can be used as guidelines for relevant staff in their future work on agricultural budgets and public expenditures. Chapter 4 is the main analytical chapter. It describes the budget cycle for alloca- tion and reporting PEA; looks at discrepancies between budgeted and executed amounts; assesses the nature of PEA between 2011 and 2017; discusses how Lesotho’s PEA align with its own policy objectives for the sector as well as with international commitments, especially under CAADP; and seeks to understand if the public support to the sector is of a nature that may crowd out private investment. Finally, Chapter 5 summarizes the findings of the analysis and provides recommendations for how to strengthen effectiveness and quality of public expenditures to the agriculture sector. CHAPTER 2: The Central Role of Lesotho’s Agriculture Sector in the Economy, for Poverty and Food Security, and Challenges Moving Forward This chapter provides an overview of Lesotho’s agriculture sector, its size in the economy, its contribution to employ- ment, and its role in poverty reduction and food security. This chapter also describes the main strategies and policy documents that guide the GoL’s investment in the sector. Finally, the chapter gives detailed insights into two main chal- lenges that currently face Lesotho and in which agriculture has a key role: climate change and nutrition. 5 6  n Agriculture Public Expenditure Review Key Points in This Chapter • Agriculture contributes less than 10 percent of Lesotho’s GDP (5.2 percent in 2016, at US$160 million). However, over half of the country’s population lived in rural areas and depended, directly or indirectly, on agriculture for their livelihood. Most peo- ple in the rural population are engaged in subsistence production, with land lots below 1 hectare. • Lesotho’s agriculture sector is dominated by livestock production (52 percent), followed by crops (28 percent) and forestry (20 percent) (contribution to agricultural GDP, 2014 figure). Maize, sorghum, and wheat are planted on 85 percent of the country’s arable land, which comprises about 10 percent of Lesotho’s total land area. Sheep and goats dominate the livestock sector, reared mainly for wool and mohair. • Crop yields are low, with average cereal yields below 1,000 kg per hectare, failing to meet the SADC RISDP target of achieving at least 2,000 kg per hectare (Nhemachena, Matchaya, and Nhlengethwa 2016). For the period 2008–2013, land productivity averaged about US$70 per hectare per year compared to the regional average of about US$120. • Constraints to improved agricultural productivity include outdated farm technologies and farm management practices, limited technical expertise, low and suboptimal use of inputs, low access to irrigation, land degradation, inadequate rural infrastructure, and limited access to credit. Going forward, limited R&D in the sector along with rudimentary extension and advisory services impede the technological shift necessary for agricultural transformation. • Climate change is a key challenge for the agriculture sector; weather risks such as droughts, floods, frosts, snow, hailstorms, and strong winds and high intraseasonal and interannual rainfall variability, with frequent droughts, have already resulted in poor crop harvests and large livestock losses. • The vulnerability of Lesotho’s agriculture sector—characterized by the IPCC as highly vulnerable to climate change impacts— has led to almost one-half of a million people currently facing food insecurity in Lesotho as a result of poor rainfalls in 2018. • As a result of the low productivity and environmental challenges, the marketable surplus of agricultural products is low; Lesotho imports about 10 times more agrofood products than it exports. Exported agrofood products largely consist of cereals, milled flower, and wool, while imported agrofood products to a larger extent are processed, like breakfast cereals, meat, and dairy. • The agrofood system does not deliver for nutrition and food security in Lesotho. Weather-related risks regularly put a large share of the agricultural population at risk of food insecurity, and malnutrition is prevalent. Lesotho experiences the double burden of high rates of both child stunting and adult overweight/obesity (stunting in children under five years is 33 percent; adult overweight and obesity are 31 percent, and 45 percent among women). Dietary diversity is low at all income levels, and availability of fruits and vegetables in Lesotho is below the WHO’s recommended intake. • Recognizing the significant role of agriculture in Lesotho’s overall economic development, poverty reduction, and food secu- rity and nutrition, the GoL is currently undertaking critical measures to develop the sector. This includes putting agriculture at the center in its new NSDP II for 2018/19–2022/23; developing an Integrated Program for Agriculture and Food System Development (a National Agriculture Investment Plan, NAIP), and a Climate Smart Agriculture Investment Plan (CSAIP) and a Multi-Sector Nutrition Strategy; conducting the Lesotho Zero Hunger Strategic Review (LZHSR); and being an Early Adopter of the World Bank’s Human Capital Project (HCP), launched in 2018. • The GoL is also increasing its investment in agriculture and nutrition through three key projects: the Second Smallholder Agriculture Development Project (SADP-2), the Lesotho Agriculture Productivity Programme for Southern Africa (APPSA), and a project to support health and nutrition services (pipeline). Lesotho is a middle-income country, but inequality is high and over one-half of the population live in poverty—most of them in rural areas. The Kingdom of Lesotho is a small, mountainous, landlocked country in ­ Southern Africa with a population of 2.23 million (2017 figure). Lesotho is classified as a lower-middle-income country with a per capita gross domestic product (GDP) of US$1,181 (2017 figure). The service sector is the biggest contributor to GDP, followed by manufacturing and mining and then agriculture (WDI, 2019). Unemployment is high and averaged 26 percent between 2013 and 2017. With a Gini coefficient of 44.6 percent, Lesotho is one of the most unequal countries in the world (ILOSTAT database). An estimated 49.7 percent of the population live below the national poverty line and 24.1 percent below the extreme poverty line. 75.6 percent of those falling below the poverty line are engaged in agriculture compared with 65.4 percent of non-poor (World Bank 2019, forthcoming). Even though the agriculture sector is relatively small in the total economy, the sector plays a significant role in Lesotho’s economy and for the livelihood of a majority of the population. Agriculture contributes less than 10 percent of Lesotho’s GDP (5.2 percent in 2016 with US$160 million). However, over half of the country’s Chapter 2: The Central Role of Lesotho’s Agriculture Sector in the Economy n 7 population live in rural areas and depend, directly or indirectly, on agriculture for livelihood, which also accounts for 10 percent of formal employment. As most of those engaged in agriculture are smallholders with less than one hectare per family, broadly distributed growth in the sector could significantly increase food security, reduce rural poverty, and generate both on- and off-farm employment. Lesotho’s agriculture sector is dominated by livestock production (52 percent), followed by crops (28 per- cent) and forestry (20 percent) (contribution to agricultural GDP, 2014 figure; Central Bank of Lesotho 2015). Based on rainfall patterns, altitude, soil- and water-holding capacity, growing seasons, and physiological features, Lesotho has four agro-ecological zones: the lowlands (17 percent), the foothills (15 percent), the Senqu River Valley (9 per- cent), and the highlands (59 percent) (Figure 2; FAO AQUASTAT 2015; Moeletsi and Walker 2013). About 10 percent of Lesotho’s total land area (30,355 square kilometers) is arable. The main crops—maize, sorghum, and wheat—are planted as monocrops on 85 percent of the country’s arable land, or almost 130,000 hectares. Production of other crops is limited, with fruit and vegetable production occupying less than 7,000 hectares and potato production 7,400 hectares (FAO 2014–2016 figure average). It can be noted that commercial vegetable production currently ranges from 100–600 hectares (depending upon rainfall), while commercial potato production accounts for less than 500 hectares. Sheep and goats dominate the livestock sector, reared mainly for wool and mohair through extensive grazing. FIGURE 2:  Lesotho’s four agro-ecological zones and ten districts Source: Moeletsi and Walker 2013. Lesotho’s agriculture sector is characterized by low productivity and low value addition. Most of the rural population is engaged in subsistence, rain-fed, undiversified farming (primarily cereal production) and extensive livestock grazing. Land plots are small—the average landholding is about 1.0 hectares per family—and yields are low: average cereal yields are below 1,000 kg per hectare, failing to meet the SADC RISDP target of at least 2,000 kg per hectare (Table 1) (Nhemachena, Matchaya, and Nhlengethwa 2016). For the period 2008–2013, land produc- tivity averaged about US$70 per hectare per year compared to the regional average of about US$120. Constraints to improved agricultural productivity include outdated farm technologies and farm management practices, limited technical expertise, low and suboptimal use of inputs, low access to irrigation, inadequate rural infrastructure, and limited access to credit. Massive soil erosion and degradation of agricultural land further contribute to the low pro- ductivity.7 Going forward, limited R&D in the sector along with rudimentary extension and advisory services will impede the technological shift necessary for agricultural transformation. 7 The estimated annual cost of land degradation in Lesotho is US$57 million, equivalent to 3.6 percent of the country’s GDP (https://www.unccd .int/sites/default/files/inline-files/Lesotho.pdf). 8  n Agriculture Public Expenditure Review TABLE 1:  Lesotho’s crop production (2014–2016 average) Beans Maize Sorghum Wheat (dry) Potatoes Vegetables Fruit Yield (kg/ha) 648 230 872 181 16,837 9,001 4,102 Area harvested (ha) 94,690 22,259 10,326 17,011 7,408 3,522 3,323 Gross production value (constant 2004–2006 8,777 786 1,422 1,835 21,057 5,973 4,759 1,000 I$) Source: FAOSTAT 2019. Note: I$ denotes international dollars. Climate change is a notable challenge facing Lesotho’s agriculture sector, and the IPCC categorizes Lesotho as highly vulnerable to climate change impacts. Lesotho already experiences weather hazards, such as droughts, floods, frosts, hailstorms, and strong winds, and high intraseasonal and interannual rainfall variability. Specifically, droughts have resulted in poor crop harvests and negatively affected livestock sector, contributing to degradation of rangeland in the past, a situation expected to continue unless adaptation and resilience-building measures are pro- moted (World Bank 2019b). Below-average rainfalls in the last quarter of 2018 had significant impacts on agricultural production, and it is estimated that almost one-half of a million people are currently food insecure in Lesotho (Lesotho Disaster Risk Management Authority 2019). The 2015/16 El Niño resulted in one of the worst droughts experienced in the last three decades and left around 709,000 people food-insecure (Lesotho Vulnerability Assessment Committee 2016; CIAT and World Bank 2018; World Bank 2019a, forthcoming). Agricultural risks are expected to intensify with climate change and to cause significant damage unless adaptation and resilience-building measures are promoted. Due to high population pressure on the available land and natural resources, substantially degraded soils, the high level of food insecurity and poverty, and lack of infrastructure, the local population’s capacity to adapt to severe climate change, variability, and hazards is relatively low (Dejene et al. 2011). Box 1 gives more details on Lesotho’s climate BOX 1:  Lesotho’s climate and international climate commitments Over the period 1981 to 2012, Lesotho observed an increase in temperature and a decrease in precipitation amounts. Climate projections suggest that temperatures are likely to increase by an average of 2°C by 2050 and up to 2.4°C by 2070. In contrast to the past trend of reduced rainfall, projections indicate the possibility of a very slight increase in the country’s rainfall (up to 1.6 percent) by 2070. However, the projected changes in rainfall are not uniform throughout the year or across regions. For instance, winter rainfall is projected to decrease strongly, summer and autumn rainfall to experience no significant change, and spring rainfall to gradually increase. All production systems are expected to be at least somewhat adversely affected by climate change, although specific impacts depend on the production system; wheat, potatoes, and vegetables will face the most significant impacts (CIAT and World Bank 2018). An aggravating factor is the interannual ENSO phenomenon, which is expected to affect climate variations (World Bank 2016a). Lesotho’s Nationally Determined Contributions (NDC 2017) include both adaption and mitigation goals. To guide the adaptation process, seven sectoral vulnerability and adaptation assessments were completed to determine appropriate adap- tation measures. Within the agriculture sector, key activities, conditional and unconditional, were identified, including the need to diversify livestock and improve rangeland management, build adaptation capacities in climate-resilient agronomic practices, and promote drought- and heat-tolerant crop varieties as well as improved land use practices, to name a few. The mitigation tar- gets were set in energy, industrial processes, agriculture (crop and livestock), land use, land use change and forestry, and waste. Lesotho is committed to reduce unconditionally 10 percent of its greenhouse gas (GHG) emissions compared to a business-as- usual (BAU) scenario by 2030. The conditional target refers to an additional 25 percent by 2030, estimated to require an overall investment of US$590 million. The country developed a National Climate Change Policy (NCCP) and National Climate Change Policy Implementation Strategy in 2017. The NCCP specifically mentions the need for climate-smart practices, and marks a significant step in inte- grating CSA into the country’s policies and programs. It aims at developing climate-resilient agriculture and food systems, and strengthening (i) the needed human resource capacity for climate-resilient agriculture, (ii) the enabling environment for the participation of vulnerable groups, and (iii) agricultural research. Chapter 2: The Central Role of Lesotho’s Agriculture Sector in the Economy n 9 and international climate commitments. While the GoL acknowledges the need for climate adaptation and resilience building, the question remains the extent to which the GoL supports these measures through its PEA. As a result of the low productivity and environmental challenges, the marketable surplus of agricultural products is low, and Lesotho imports about 10 times more agrofood products than it exports. Lesotho’s agro- food export averaged US$36.2 million in the period 2014–2016, while agrofood imports were US$386.5 million (Figure 3). An estimated 80 percent of the food consumed in Lesotho is imported. FIGURE 3:  Lesotho’s agrofood exports and imports, million US$ (2014–2016 average) 450 400 350 300 Million US$ 250 200 150 100 50 0 Export Import Source: FAOSTAT 2019. Export products mainly consist of primary agriculture and milled products, while imported agrofood prod- ucts are a mix of primary and processed goods. Wheat flour stands out as Lesotho’s most important agrofood export product according to FAOSTAT, contributing 42 percent of total agrofood exports and averaging US$12.5 mil- lion. Wool is the second most important—it averaged US$9.8 million from 2014 to 2016, or 27 percent of total agri- cultural exports. Lesotho’s exports of horticultural products are negligible. Other export commodities include maize products and cotton (Figure 4; FAOSTAT 2019). On the import side, maize and maize flour are by far the most import- ant commodities and make up more than one-half of all agrofood imports. Imported maize accounts for 60 percent of the maize consumed in Lesotho. Otherwise, the main imported agrofood products include l ­ ivestock-sourced food such as chicken meat and processed dairy products, as well as wheat, wheat flour, and breakfast cereals (Figure 5; FAOSTAT 2019). FIGURE 4:  Lesotho’s agrofood exports by commodity, share of US$ (2014–2016 average) 3% 7% 13% 43% Wheat flour Wool Maize Maize flour 34% Fruit (dried) Cotton (carded, combed) Source: FAOSTAT 2019. 10  n Agriculture Public Expenditure Review FIGURE 5:  Lesotho’s agrofood imports by commodity, share of US$ (2014–2016 average) 10% Maize flour Maize 10% Chicken meat Wheat 42% Wheat flour 8% Cereals, breakfast Milk, skimmed cow Beverages, non-alcoholic 4% Pig meat and sausages Cotton lint 4% Sugar raw centrifugal 4% Rice—total (rice milled equivalent) 3% Sunflower oil 2%2% 3% 2% 3% 3% Other Source: FAOSTAT 2019. The agrofood system does not deliver for nutrition and food security in Lesotho. As described earlier in this chapter, weather-related risks regularly put a large share of the agricultural population at risk for food insecurity and in need of food aid and other safety nets. Malnutrition is prevalent, and Lesotho is challenged with the double burden of high rates of both child stunting and adult overweight/obesity. Income is associated with malnutrition in Lesotho: 41 percent of all families spend more than one-half of their income on food, and almost one-half of all children under five are stunted in the lowest income quintile compared to 10 percent of children in the highest. However, dietary diversity is low across income quintiles, likely due to inadequate availability of certain food groups and lack of nutrition knowledge. Box 2 provides more details of the scope and costs of malnutrition in Lesotho. BOX 2:  Nutrition in Lesotho High malnutrition exists in Lesotho. Stunting among children under the age of five is 33 percent, and micronutrient deficiency (e.g., anemia and vitamin A deficiency) is high among children and adults alike. At the same time, Lesotho is undergoing a dietary shift that has driven overweight and obesity around the country. An estimated 31 percent of the adult population are overweight or obese; among women, this figure reaches 45 percent. Stunting, micronutrient deficiency, and overweight and obesity imply a high burden for the individual but are also costly for society as a whole. In 2008–2014, 9,272 child deaths (19.5 percent of all child mortalities) were a result of undernutrition. Lesotho loses an estimated 7.13 percent of its GDP due to chronic malnutrition.8 Dietary diversity is low in Lesotho and is likely a result of a malfunctioning food system. While stunting and overweight and obesity are two widely different forms of malnutrition, they are related in that stunted children have a higher probability of becoming obese as adults. Both these forms are also a result of a food system that does not avail an adequate diet. In Lesotho, only 20 percent of households with acceptable levels of food consumption have high dietary diversity and as little as 11 percent of children aged 6–59 months have sufficient dietary diversity. Fruit and vegetable availability in Lesotho is estimated to be 128 grams per person per day, compared with the World Health Organization’s (WHO) recommended intake of 400 grams per person per day. Low dietary diversity among households with acceptable levels of food consumption indicates that low dietary diversity is not just a result of low incomes, but likely also due to inadequate availability of certain food groups and nutrition knowledge— both areas that indicate failures of the food system. Strengthening Lesotho’s domestic food system would be a win-win outcome. Agriculture has been identified as a key sector for reducing stunting in Lesotho, and given that the agrofood sector is productive, nutrition-sensitive investments in the food sys- tem would be a win-win. Such investments would decrease malnutrition and associated costs to individuals and society, as well as improve livelihoods, generate incomes, and create jobs for the 60 percent of the population that depend on the agrofood sector. 8 It is not clear if this number includes the cost of overweight, obesity, and diet-related noncommunicable diseases. Chapter 2: The Central Role of Lesotho’s Agriculture Sector in the Economy n 11 Over the past decade multiple strategic documents have laid out the GoL’s aspirations for the agriculture sector, including the National Vision 2020, the NSDP 2012/13–2016/17, the National Action Plan for Food Secu- rity (NAPFS) 2007–2017, the National Forestry Policy (2008), the National Soil and Water Conservation Policy (NSWCP) 2013, the National Range Resources Management Policy, and the Lesotho Food and Nutrition Policy (LFNP) 2016–2025. Importantly, to achieve the National Vision 2020 aspirations, the GoL’s medium-term implementation ­ strategy, namely the NSDP 2012/13–2016/17, set the policy framework for implementation of the CAADP process and identified four priority areas for the agriculture development agenda: (i) Reducing vulnerability and managing risk; (ii) Promoting commercialization and diversification in agriculture; (iii) Strengthening capacity of farmers and institutions for increased and optimal agricultural performance; and (iv) Sustainable natural resource management. Recognizing agriculture’s significant role in Lesotho’s overall economic development, poverty reduction, and food security and nutrition, the GoL is currently undertaking critical measures to develop the sector as laid out in its various strategies and programs. The GoL selected agriculture as one of four productive sec- tors for “job creation and inclusive economic growth under a new growth path led by the private sector” in its new NSDP II for 2018/19–2022/23. Further, to strategically support agriculture’s role in climate change mitigation and adaptation, and in nutrition and food security, the GoL is developing an Integrated Program for Agriculture and Food System Development (a NAIP), a Multi-Sector Nutrition Strategy, and with the support of the World Bank, a CSAIP—all of which are expected to be ready for adoption in 2019. The Lesotho Zero Hunger Strategic Review (LZHSR) was recently commissioned to address the challenges that prevent Lesotho from achieving zero hunger by 2030. Developed within the framework of Sustainable Development Goal (SDG) number 2, the LZHSR pillars include access to adequate food and healthy diets all year round, an end to malnutrition and all its forms, sustain- ability of food systems, a doubling of smallholder productivity and income, and elimination of food loss and waste. Finally, Lesotho confirmed its commitment to ending malnutrition by being an Early Adopter of the World Bank’s HCP, launched in 2019. The GoL is also increasing its investment in resilient, productive agriculture and nutrition through three key donor-financed projects over the next 5+ years: SADP-2 (US$52 million), APPSA (US$20 million), and a project to support health and nutrition services (pipeline). However, ensuring the quality in the overall PEA will be important to ensure that strategic objectives and commitments are being met for the sector. The GoL’s regular bud- get will play an important role in achieving the objectives set, provided that it is allocated efficiently and according to sector strategies. CHAPTER 3: The Data and Methodology Used for Conducting This Agriculture Public Expenditure Review9 This chapter provides a brief summary of the data and methodology applied for this AgPER and discusses the develop- ment and use of different databases in conducting the analysis. Importantly, this chapter describes which expenditures qualify as PEA and how these expenditures were identified. The methodology and data used for the review are covered more extensively in Annex 1, which can also be used as a methodological guideline for technical staff who work on PEA in Lesotho. 9 This section draws largely from Mas Aparisi (2018), an unpublished methodological note also used in the unpublished Agricultural Sector Review for Kenya (World Bank undated). 12 Chapter 3: The Data and Methodology Used for Conducting This Agriculture Public Expenditure Review n 13 Key Points in This Chapter • This AgPER relies on data from the Lesotho BOOST database, which reproduces, in Excel format, the data from the Integrated Financial Management Information System (IFMIS) used by the MF and line ministries to plan, execute, and monitor the budget. The IFMIS and BOOST have limitations but are the most comprehensive data sources available to analyze Lesotho’s official budget. • For the purpose of this AgPER, the MF’s BOOST database was converted into an Agricultural BOOST database (AgBOOST) containing public expenditures in support of agriculture (PEA). • AgBOOST covers approved and revised estimates, as well as warranted release and final expenditures, for the period 2010/11 to 2015/16. AgBOOST includes very minimal donor-funded expenditures—2 percent. • The data were converted into million Maloti and million US$ expressed in 2010 constant values. Official exchange rate and CPI data from the Lesotho Central Bank were used. • Categorical variables in the AgBOOST database allow for analysis of the budgetary data. Four types of variables can be distin- guished: administrative, economic, functional, and source of funding. • The functional variables that exist in Lesotho’s IFMIS/BOOST do not allow for a detailed functional and sectoral analysis. The data were therefore recoded with functional and sectoral categories, using a conversion table of keywords to function. • Annex 1 outlines the full methodological approach for conducting this AgPER and can be used as a guideline for technical staff in relevant ministries going forward. DESCRIPTION OF THE DATA This AgPER relies on data from the Lesotho BOOST database,10 which reproduces, in Excel format, the data from the Integrated Financial Management Information System (IFMIS) used by the MF and line ministries to plan, execute, and monitor the budget. Although the IFMIS and BOOST are not without limitations, they are the most comprehensive data sources available to analyze Lesotho’s official budget. For the purpose of this AgPER, the MF’s BOOST database was converted into an Agricultural BOOST database (AgBOOST) containing public expenditures in support of agriculture (PEA). AgBOOST covers approved and revised estimates, as well as warranted release and final expenditures, for the period 2010/11 to 2015/16. It includes recurrent and development expenditures for all relevant ministries (heads), cost centers, and subcost centers. AgBOOST includes very minimal donor-funded expenditures—2 percent. The data were converted into million Maloti and million US$ expressed in 2010 constant values. Official exchange rate and CPI data from the Lesotho Central Bank were used. Categorical variables in the AgBOOST database allow for analysis of the budgetary data. Four types of variables can be distinguished: administrative, economic, functional, and source of funding. In addition, the following variables, all categorical, were created for the purpose of the analysis: • Functional category: This variable was created by coding budget lines using information available in the administrative, economic, functional, and source of funding variables. • Sector: This variable was created by coding budget lines using information available in the administrative, economic, functional, and source of funding variables. • Region: This variable was created to identify transfers to District Offices (DOs) by coding budget lines using information available in the cost center labels of the MAFS and the MFRSC. • Economic grouping (this variable was created by MF and World Bank BOOST analysts and reused for this exer- cise): This variable aggregates economic categories under two groups: operating costs and personal emolu- ments. For instance, “41 Compensation of Employees” falls under personal emoluments. The limitations in the data are outlined in detail in Annex 1. 10 The BOOST database is produced by the MF of Lesotho with support from the World Bank. 14  n Agriculture Public Expenditure Review METHODOLOGY This section lays out the approach followed to define the scope of agriculture expenditures and create the functional and sectoral variables. Functional scope To conduct this analysis, a functional perimeter of expenditures considered “agriculture expenditures” (PEA) was delineated. In line with the African Union Guidance Note (AUGN) on tracking and measuring the levels and quality of government expenditures for agriculture, the functional perimeter includes: “agriculture (including livestock)” (70421); “forestry” (70422); “fishing and hunting” (70423); and “agricultural research and development” (7084) (African Union 2015). A more fine-tuned functional perimeter was then established drawing on the MAFAP meth- odology (Figure 6). This is further detailed in Annex 1. All government units spending on the functions outlined above were included. Nongovernmental organizations and the private sector were excluded to solely capture public spending, in line with the AUGN recommendations. Using this perimeter as a starting point, BOOST was trans- formed into AgBOOST by using a multistage keyword filtering technique (outlined in Annex 1). FIGURE 6:  Functional and subfunctional perimeters A. Subsidies A1. Input A2. Capital E. Inspection and quality control F. Infrastructure B. Research F1. F2. F3. Feeder Irrigation Other roads C. Extension and advisory G. Storage services D. Training H. Markets and marketing Note: Administrative costs are not represented on this diagram but form another category of the perimeter. COFOG: agriculture (70421); fisheries/hunting (70422); forestry (70423); and agricultural R&D (7084). Functional and sectoral classification The functional variables that exist in Lesotho’s IFMIS/BOOST do not allow for a detailed functional and sectoral analysis. The data were thus recoded with functional and sectoral categories, using a conversion table of key- words to function. The approach followed is explained in Annex 1. CHAPTER 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? This chapter provides the analysis of this report. It describes the budget process, how much of total public expendi- tures are devoted to agriculture, and how the share changes over time. Further, this chapter looks at how PEA are allocated between central and local levels, and how expenditures are allocated geographically. The second part of this chapter analyzes if and to what extent PEA align with the GoL’s stated policy objectives for the sector. This includes more in-depth discussions on three key topics: private sector investment, nutrition-sensitive agriculture (NSA), and climate-smart agriculture (CSA). 15 16  n Agriculture Public Expenditure Review Key Points in This Chapter • Between 2010/11 and 2015/16, 322.7 million Maloti, or 1,513.5 Maloti per rural capita, were allocated to PEA, accounting for less than 3 percent of total public expenditures in Lesotho. This is in the context of the agriculture sector contributing close to 6 percent to GDP and almost 60 percent of the population depending on agriculture for their livelihood. While public expendi- tures increased 4.9 percent annually from 2010/11 to 2015/16, PEA increased only 0.9 percent on average in the same period. • PEA are below what Lesotho has committed under CAADP (10 percent of total public expenditure) and fall below allocations of other countries in the region. Nevertheless, actual PEA are about twice the amount Lesotho reports to ReSAKSS. • Almost all PEA are allocated by the MAFS (64 percent) and the MFRSC (33 percent). Only 3 percent of PEA go through other ministries: the MF, the MTI, and the Ministry of Home Affairs (MHA). • Budgeting PEA is done by the individual departments in each ministry and while timely, it is not an integral part of the imple- mentation of key strategic documents for the sector. Further, no institutionalized coordination of investment occurs in the sec- tor between the GoL and the donor community, and donor funding was not mainstreamed into the budget in the review period. • In the MAFS, the Crops Department has the highest level of spending (37 percent of MAFS’s total spending); in the MFRSC, 75 percent of total expenditures go to the Conservation Department. • MAFS DOs receive about 30 percent of its spending and MFRSC DOs about 15 percent. Geographically, PEA are allocated largely according to the spatial distribution of farm households across the country. • Donor funding is generally not recorded in the budget, and only accounts for 2 percent of total PEA in the BOOST (about 15 million Maloti). The OECD’s Creditor Reporting System (CRS) reports about 511 million Maloti, or 12 percent of PEA + CRS combined, which is an underestimate of the actual donor funding in this period as not all donors have reported their funding. The World Bank, International Fund for Agricultural Development (IFAD), and the Global Environmental Facility (GEF) make up 94 percent of total donor funding to agriculture in this period. • Budget execution for both the MAFS and the MFRSC is volatile, with execution rates both below and above the budgeted amounts depending on circumstances in certain years due to late payments that shift execution from one fiscal year to the next. Breaking down execution rates reveals that transfers from MAFS DOs were made to support the ICP at MAFS Headquar- ters. Execution rates for personal emoluments are largely in line with budgeted amounts, while operating costs are reduced when cuts have to be made. • Development expenditures account for about two-thirds of total PEA, but are overestimated as they include recurrent expen- ditures, especially the ICP and salaries under the WMP. • Even though Lesotho has a relatively high wage bill in general, the share of PEA going to wages (56.1 percent) is higher than that of total public expenditures (37.7 percent). This is problematic for agriculture sector programs, as operating costs are too low to enable staff to implement them properly. In MAFS DOs, the salary bill is as high as 91 percent of total spending, leaving little resources to finance operating costs (e.g., for extension workers). Similarly, the Livestock Department, which generally has high expenses for vaccines and other operating costs, spends 71 percent of its budget on staff. Even the capital-intensive WMP program absorbs 70 percent of the MFRSC wage bill. However, the MFRSC employs temporary staff rather than perma- nent, as some of its programs have a dual objective of environmental restoration and providing cash for work. Thus a large share of its wage bill goes to temporary employment. • Most sector-specific expenditures go to the crop sector (83 percent), followed by livestock (14 percent) and forestry (3 per- cent). Consultations with the ministries confirm that multi-sectoral expenditures largely follow the same pattern. This com- position is not in line with their relative contribution to agricultural GDP, to which the livestock sector contributes 52 percent, crops 28 percent, and forestry 20 percent. The ICP comprises 69 percent of crop expenditures. • Subsidies to the sector go to inputs, especially to the ICP, rather than capital. Public investment in infrastructure is very low (0.5 percent), as is PEA that support agricultural processing and marketing activities (0.2 percent), although donor-funded projects not reported in BOOST complement this allocation. Public financing to agricultural R&D is 0.8 percent. • Of PEA under the MAFS, about 40 percent are spent on goods and services of private nature, largely because of the ICP; for the MFRSC, 97 percent of expenditures go to public goods and services. • PEA are not particularly well-aligned with the main strategic documents for the sector—the NSDP and the Budget Frame- work Papers (BFPs). Compared to what has been outlined in these strategic documents, input subsidies and extension and advisory services are overfunded, while processing and marketing, irrigation, and R&D are severely underfunded. • Allocations to nutrition-sensitive agriculture (NSA) and CSA are low given the GoL’s commitments to address the challenges related to malnutrition and to climate change. Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 17 BUDGET PROCESS The budget preparation process in Lesotho is well structured according to comprehensive guidelines for budget preparation and conducted in a timely manner. The budget preparation process lasts from approxi- mately September to February. It starts in September with a call for each line ministry to prepare a Budget Frame- work Paper (BFP) according to an MTEF (Medium Term Expenditure Framework) budget approach, followed by detailed estimates that are submitted to the MF in December. In both the MAFS and the MFRSC, budgets are devel- oped by individual departments, then consolidated centrally in the respective ministry before being sent to the MF. Donor financing is generally not included in these budgets.11 Following the submission of ministry budgets to the MF, discussions are held with the line ministries. The final steps are to present the budget to the Cabinet Budget Committee in January, and then to the Cabinet a few weeks later, before it is presented in the Parliament in February. This gives the Parliament time to process and approve the budget ahead of the new fiscal year, which starts in April (PEFA 2017). Budget ceilings are frequently not respected. Even though the MF gives specific budget ceilings for the individ- ual ministries, the Public Expenditure and Financial Accountability (PEFA) 2017 Performance Assessment Report notes that these are often not respected by the line ministries, which submit budgets that exceed these ceilings, often based on the previous year’s budget and expenditures. Instead, the MF adjusts these budgets in the final budget (PEFA 2017). To address this issue, the new Central Budget Management System (CBMS) implemented by the IFMIS Project under the Public Financial Management Reform Project has automated hard controls so that submitted budgets do not exceed assigned budget ceilings and ministries cannot spend more than what is allocated (and legally entrenched in the Appropriation Act). According to the Constitution, upward adjustments to the budget should be approved ex ante by the Parliament; however, the GoL submits one supplementary bill per year ex post that contains multiple amendments to the budget. This practice has been accepted by the Parliament. Other restric- tions to budget adjustments are for transfers between programs of more than 10 percent of capital expenditures or 20 percent of recurrent expenditures under a program, which needs approval from the MF.12 PEA are not guided by a NAIP or a formal Agriculture Policy in Lesotho. Instead, the strategic objectives for the sector are outlined in the NSDP. Further, as discussed later in the chapter, significant disparities arise between agriculture and nutrition objectives and policies, and the allocations of PEA. This is not exclusive to PEA, but applies to the line ministries in general, according to the PEFA 2017 Performance Assessment Report (PEFA 2017). Budgets for DOs are done centrally and not according to demand from individual DOs. The Maseru DAO (Dis- trict Agricultural Office) resembles a mini department for Extension Services, and it budgets and allocates expendi- tures to other DAOs. From the perspective of BOOST, this makes the Maseru DAO look like it has a very large budget, while expenditures for the other DAOs look smaller than they are in reality. Reporting of PEA is limited, making it difficult for the GoL to evaluate the impacts of spending and if policy objectives are being met. Each department consolidates reports with an overview of activities and outputs over the year. While the reports are a good first step, they are output-based and thus insufficient for impact monitoring. These reports also do not seem to be conducted systematically and it is not clear if they report back to the MDP, which is currently responsible for the main agriculture policy objectives in the NSDP. While Lesotho has signed the Malabo Declaration and is in the process of compiling its first report on PEA, Lesotho does not have a Joint Sector 11 Donor-funded projects are off-budget and ring-fenced in separate systems as a result of historical wariness about fiduciary compliance and systems. The IFMIS system has been upgraded under the World Bank–financed PFM Reform Support Project, and the upgraded version went live in April 2019. The upgraded IFMIS can be used to (i) incorporate donor-funded projects into the Consolidated Fund and budget and support the financial management of such projects through the IFMIS making full use of government processes; or (ii) support the project financial man- agement requirements separately from the Consolidated Fund. Future use of the system with regards to donor financed projects will depend on the results from the upgraded IFMIS system confirming more reliable financial management, accounting, and financial reporting and agreed between the GoL and individual donors. 12 For more information about the general budget process in Lesotho, see PEFA (2017). 18  n Agriculture Public Expenditure Review Review (JSR). A JSR is a requirement under CAADP and ensures that governments and donors mobilize resources to agriculture toward the same objectives and report on the same indicators. LEVEL—TRENDS, FLUCTUATIONS, AND OUTLIERS Between 2010/11 and 2015/16, 322.7 million Maloti, or 1,513.5 Maloti per rural capita, were allocated to PEA, accounting for less than 3 percent of total public expenditures in Lesotho. Lesotho’s PEA have fallen behind the increase in general public expenditures, and the allocation to rural capital has remained virtually unchanged since 2010. In real terms, PEA have grown by 0.6 percent per year while general public expenditures have expanded by 4.9 percent. In particular, they decreased by 9.1 percent and 12.7 percent in 2014/15 and 2015/16, respectively, against growth rates of 13.8 percent and –0.2 percent for general public expenditures. This demonstrates a stagna- tion of PEA, and even a contraction in 2012/13, 2014/15, and 2015/16, in a context of tightening fiscal resources for the GoL. The level of PEA per rural capita in 2015/16 almost reverted below its 2010/11 level (Table 2). TABLE 2:  Main PEA indicators, in million constant 2010 Maloti 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 Average Budgeted PEA 389.9 417.1 313.9 428.5 443.5 475.2 411.3 Final PEA 298.3 318.7 282.1 379.7 348.1 309.0 322.7 PEA per capita (in Maloti) 1,069.9 1,130.3 988.0 1,312.6 1,187.5 1,040.1 1,121.4 PEA per rural capita (in Maloti) 1,422.7 1,513.3 1,329.8 1,776.2 1,615.7 1,423.1 1,513.5 Budgeted PE 10,787.2 12,497.8 12,321.0 12,435.3 13,067.3 12,437.7 12,257.7 Final PE 8,418.1 9,789.1 8,885.5 9,882.8 11,245.3 11,219.6 9,906.7 Agricultural GDP (AgGDP) 895.2 942.3 910.7 1,132.9 1,173.3 1,094.2 1,024.8 GDP 17,286.5 18,479.5 19,588.0 19,949.1 20,571.9 21,141.0 19,502.7 Share AgGDP in total GDP 5.2% 5.1% 4.6% 5.7% 5.7% 5.2% 5.3% Share budgeted PEA/PE 3.6% 3.3% 2.5% 3.4% 3.4% 3.8% 3.4% Share final PEA/PE 3.5% 3.3% 3.2% 3.8% 3.1% 2.8% 3.3% Share final PEA/AgGDP 33.3% 33.8% 31.0% 33.5% 29.7% 28.2% 31.6% Share final PEA/GDP 1.7% 1.7% 1.4% 1.9% 1.7% 1.5% 1.7% PEA growth — 6.4% –13.0% 25.7% –9.1% –12.7% 0.6% Public expenditures growth — 16.3% –9.2% 11.2% 13.8% –0.2% 4.9% Source: Authors and ReSAKSS 2019. Unpredictable PEA affects the GoL’s capacity to support agriculture sector development. It prevents medium- and long-run policy planning and undercuts operations on the ground when the budget drops. The volatility of PEA partly reflects that of the general budget,13 with strong jumps and falls. The largest was a 13 percent drop in 2012/13, followed by a 25.7 percent growth spurt (Figure 7). Nevertheless, PEA exhibit their own pattern and appear even more unpredictable than the general budget. They seesawed through the entire period of analysis, except for two consecutive years of negative growth in 2014/15 and 2015/16. PEA as a share of total public expenditures fall below Lesotho’s commitments under the African Union’s (AU) CAADP and do not align with the share of the population engaged in agriculture. The share of PEA in 13PEA are expressed in constant terms and for actual expenditures. The growth rate variation thus captures inflation and issues in budget execution. Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 19 FIGURE 7:  Final PEA and share in total public expenditures (%), PEA growth rate, and public expenditures growth rate (%) 400 4.5% 30% 350 4.0% 25% 300 3.5% 20% 3.0% 15% 250 2.5% 10% 200 2.0% 5% 150 1.5% 0% 100 1.0% –5% 50 0.5% –10% 0 0.0% –15% 1 2 3 4 5 6 ge 1 2 3 4 5 6 ge /1 /1 /1 /1 /1 /1 /1 /1 /1 /1 /1 /1 ra ra 10 11 12 13 14 15 10 11 12 13 14 15 e e 20 20 20 20 20 20 20 20 20 20 20 20 Av Av Final PEAS Share final PEAS/PE PEA growth rate PE growth rate total public expenditures has ranged between 2.8 percent and 3.8 percent, well below the 10 percent target agreed by the GoL in the Maputo (2003) and Malabo (2014) Declarations under CAADP (African Union 2003, 2014). This is broadly in line with the share of agriculture in the country’s GDP—an average 5 percent—but very low considering that 73 percent of the population live in rural areas and derive their livelihood from the low-productivity agriculture sector. Despite the already low levels of PEA, this AgPER finds that the PEA of the Maputo target tracking indicators of the AU’s ReSAKSS initiative are significantly underreported—by an average of 22.3 million constant US$ (156 million Maloti) per year14 (Table 3). This is around 48 percent of the average annual PEA reported in this AgPER. One likely explanation is that ReSAKSS’s measurement leaves out expenditures by the MFRSC (ReSAKSS 2019). If the AgPER’s PEA are corrected to remove MFRSC spending, the average difference is narrowed down to 7.2 million Maloti per year, or 50 million Maloti, which is 15 percent of annual PEA. The mismatched ratio of AgPER and ReSAKSS PEA is systematic for the first three years of the review period, but doubles in 2013/14 and 2014/15 before dropping below 1 in 2015/16. This illustrates that the data presented in this review (and any other assessment, such as that of ReSAKSS) are subject to a margin of error depending on the sources used, timing of data TABLE 3:  AgPER and ReSAKSS PEA, in million constant 2010 US$ 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 Average AgPER 42.6 45.5 40.3 54.2 49.7 44.1 46.1 AgPER—no Forestry 25.6 26.3 29.8 37.3 34.2 32.7 31.0 ReSAKSS 18.8 21.2 20.8 16.0 13.7 52.3 23.8 AgPER/ReSAKSS ratio 2.3 2.1 1.9 3.4 3.6 0.8 2.4 AgPER—no Forestry/ReSAKSS 1.4 1.2 1.4 2.3 2.5 0.6 1.6 AgPER “Maputo target” 3.7% 3.4% 3.3% 4.0% 3.2% 2.9% 3.4% ReSAKSS “Maputo target” 1.8% 2.0% 1.7% 1.5% 4.0% 3.8% 2.5% Source: Authors and ReSAKSS 2019. 14 ReSAKSS reports agriculture public expenditure data in constant 2010 million US$—the Lesotho public expenditure data were thus converted into the same currency to allow comparison. 20  n Agriculture Public Expenditure Review collection, and assumptions made to delineate agriculture expenditure. Nonetheless, a good degree of consistency exists between ReSAKSS and the AgPER data: the average “Maputo target” figure for the former stands at 3.4 per- cent, and at 2.5 percent for the latter. The share of public expenditures that Lesotho allocates to agriculture is below that of other African coun- tries, especially those with large rural economies. Lesotho’s share of public expenditures going to agriculture is very slightly above the SACU, SADC, and Africa average (0.75 percentage points). However, Lesotho stands almost 3 percentage points below the SADC average when excluding the outlier service economies of Seychelles, South Africa, and Mauritius (Table 4). The SADC average is pulled upward by four rural economies: Madagascar (7.4 per- cent), Malawi (16.0 percent), Zambia (8.1 percent), and Zimbabwe (8.3 percent). It is worth noting that, Malawi aside, none of these countries reach the Maputo target. TABLE 4:  PEA (million constant 2010 US$) and share of total public expenditures (%), 2010–2016 averages PEA 2010–2016 average Share of public expenditures Lesotho  499 3.4 SACU  672 2.9 SACU—no South Africa  109 2.6 SADC  672 2.1 SADC—no South Africa  181 5.4 Africa 1,019 2.9 Source: Authors and ReSAKSS 2019. COMPOSITION—TRENDS, FLUCTUATIONS, AND OUTLIERS15 The composition branch of the analysis evaluates the quality of the total budget spent in support of the agriculture sector in the country over the period. In particular, it looks at the breakdown of the budget following four analytical lenses: administrative, source of funding, economic, and functional.  Administrative analysis The administrative breakdown analysis assesses the repartition of PEA between central and local governments on one hand, and within central and local governments’ administrative units on the other hand. The aim is to find out whether the repartition of PEA between such units matches national policy objectives and efficiency criteria. Ministerial breakdown of expenditures Agriculture public expenditures in Lesotho are concentrated in two ministries: the MAFS (64 percent) and the MFRSC (33 percent) (Figure 8). The MAFS essentially spends to increase agricultural production and produc- tivity, while MFRSC expenditures support sustainable production, with two-thirds allocated to fighting soil erosion in rural areas. Three other ministries contribute to 3 percent of PEA: the MF, the MTI, and the MHA (Figure 8). The MF con- tribution (2 percent of total PEA) comes through its funding for the agricultural census in 2010/11 and 2011/12 as well as its participation in the Rural Financial Intermediation Program, aimed at increasing access to financial ser- vices in rural areas, in particular for actors of the agrofood sector (Annex 3). The MTI is involved through its Depart- ment of Agricultural Marketing. This department was formerly under the MAFS with the aim to improve marketing 15The material below draws largely from Mas Aparisi (2018), an unpublished methodological note also used in the unpublished Agricultural Sector Review for Kenya (World Bank, undated). Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 21 FIGURE 8:  Administrative composition of PEA, in million constant 2010 Maloti, actual expenditures16 400 Million Maloti (constant 2010) 350 300 250 200 150 100 50 0 2010–11 2011–12 2012–13 2013–14 2014–15 2015–16 01 Agriculture and food security 04 Finance 05 Trade and industry 08 Home affairs 14 Forestry and land reclamation of domestic agricultural produce. It was transferred to the MTI in 1999 and then moved, in 2015, to the new Minis- try of Small Business Development, Cooperatives and Marketing (MSBDCM). The department is involved in control and regulation of import substitution agricultural commodities (import/export licenses; restrictions; price controls and subsidies); market support (horticulture, wool) under the Enhanced Integrated Framework supported by IFAD; market research (value chain analysis, market information system); and market access (construction of market infrastructure). However, the only spending entry recorded for the MTI/MSBDCM in BOOST was in 2015/16; it accounted for 1.55 million Maloti, or 0.5 percent of the year’s PEA. It can be assumed that similar expenditures apply for previous years. The MHA contributes to PEA with its involvement in livestock marking and registration. Cost center analysis In the MAFS, the Crops Department is, by far, the highest spending cost center, with spending accounting for 27 percent of MAFS total expenditures, and exceeding that of the 10 DAOs combined. It spent a total of 462 million Maloti, more than four times the amount of the next unit, the Administration Department (Figure 9). This is because the Crops Department is in charge of the hefty input subsidy Summer Cropping Programme, also referred to as the ICP. The program is further analyzed below. The rest of MAFS expenditures are spread between Lesotho Agricultural College (119 million Maloti), the Extension (71 million Maloti) and Livestock Departments (61 million Maloti), the Research Department (39 million Maloti), and the Planning Department (18 million Maloti). MAFS DOs spent, in total, 369 million Maloti, or 30 percent of MAFS expenditures. In the MFRSC, the Conservation Department registers three-quarters of total expenditures—495 million Maloti over the period (Figure 11). It corresponds to the MFRSC’s flagship program, the Watershed Management Programme (WMP), also referred to as the Integrated Catchment Management Programme or the Poverty Allevia- tion Programme, but more commonly known as Fato Fato (Box 5). This program largely consists of building anti– soil erosion infrastructure, in particular gully structures, stone lines, and diversion furrows as highlighted in the MFRSC capital expenditures breakdown (Figure 10). Interviews with GoL staff and the literature review (FAO 2016; Kardan, O’Brien, and Masasa 2017) point to the WMP’s social objective (“cash for work”) on top of its conservation focus. It is therefore heavily reliant on unskilled labor. The MFRSC’s other departments and DOs have smaller allocations. The Rangeland Management Department activities are similar but are marginal, at 2 percent of MFRSC expenditures. The Forestry Department focuses on subsidizing households with free fruit trees—the trees are bought from producers that are supported with extension services—and planting forest trees (afforestation). This is a relevant expenditure for agriculture sector 16 Unless stated otherwise, all figures in the report are actual expenditures. 22  n Agriculture Public Expenditure Review FIGURE 9:  Cost center breakdown of MAFS PEA, in million constant 2010 Maloti 0104 Crops 0101 Administration 0106 Lesotho Agricultural College 0107 Extension 0103 Livestock 0108 DAO Maseru 0113 DAO Mohales Hoek 0105 Research 0111 DAO Berea 0110 DAO Leribe 0109 DAO Butha Buthe 0112 DAO Mafeteng 0116 DAO Thaba Tseka 0114 DAO Quthing 0115 DAO Qachas’ Nek 0117 DAO Mokhotlong 0102 Planning 0 50 100 150 200 250 300 350 400 450 500 Million constant 2010 Maloti FIGURE 10:  MFRSC capital expenditures, per output, average share for 2011/12, 2014/15, and 2015/16 40% 35% 30% 25% 20% 15% 10% 5% 0% s g r es s s n s t ng en bo re m w llie tin tio in di rro da tu pm La an uc gu el ee c fu on od ru pl d lo W of an n ve st St e pr io g e de ly s in s Tr er nd ul er ap dd s G iv Po nk Sh Fo D Ta FIGURE 11:  Cost center breakdown of MFRSC PEA, in million constant 2010 Maloti 1403 Conservation 1405 Districts 1401 Administration 1402 Forestry 1404 Range Resources Management 0 50 100 150 200 250 300 350 400 450 500 Million constant 2010 Maloti Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 23 development: fruit trees and agroforestry or planting trees on farms are common practices that provide income-­ generating opportunities for farmers and environmental benefits in the form of reduced land degradation and enhanced soil quality, which can enhance agricultural productivity. In addition, agroforestry is considered one of the most effective agricultural practices for reducing greenhouse gas (GHG) emissions and enhancing carbon seques- tration through biomass growth. The Forestry Department (2 percent of MFRSC PEA) also contributes to watershed and catchment area management, which ultimately can have a positive effect on soil quality and agricultural pro- ductivity. DOs (15 percent of MFRSC PEA) are all managed under one department, though they do have a subcost center status and appear in the IFMIS/BOOST. The MFRSC does not have a research department. Central and regional expenditures Local governments receive budgetary transfers from the central government but these cannot be attributed to the agriculture sector. Decentralization is enshrined in the Lesotho National Constitution of 1993 and is legally established by the Local Government Act of 1997 (Government of Lesotho 2014). According to the information avail- able in the IFMIS, budgetary transfers for the period under review are spread across seven cost centers: Administra- tion, District Administration, Decentralization, Lands, Housing, Chieftainship, and Engineering and Infrastructure (Annex 4). The cost centers’ attributions do not include agriculture but are rather focused on waste management and transport as well as urban and rural infrastructure. Local governments are involved in rural road construction and maintenance, and although these roads sup- port the agriculture sector, they contribute to multiple rural development objectives and are not consid- ered within this review’s perimeter (see Chapter 3). For the sake of comparison, local government expenditures on rural roads are presented below. They represent an average 35 percent of PEA, with the share reaching up to 51 percent in 2011/12 and 53 percent in 2012/13 (Table 5). This illustrates the substantial indirect support pro- vided to the agriculture sector by local governments through this item of expenditure. TABLE 5:  Rural road expenditures by local governments and PEA, in million constant 2010 Maloti 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 Total Rural roads, local government 87 161 150 121 77 75 672 PEA 298 319 282 380 348 309 298 Rural roads/PEA 29% 51% 53% 32% 22% 24% 35% Transfers to DOs within the central administration stand at 30 percent and 15 percent of MAFS and MFRSC expenditures, respectively. The former share is sizeable and in line with the MAFS’s objective of providing local-level extension services, but the latter can be considered a low figure (Table 6). DOs are not fiscally decentralized but are, by and large, responsible for providing extension and advisory services to farmers; however, the MAFS Maseru DOs account for 4 percent and the MFRSC Maseru DOs for 2 percent of the respective minis- tries’ allocations, despite the fact that they are arguably located in areas without significant activities. The reason is because the Maseru DOs allocate supplementary funds to DOs around the country. Thus apart from administrative costs, these funds do not remain with the Maseru DOs. Expenditures for MAFS DOs are skewed toward Maseru and the Mohale’s Hoek district (Figure 12). This appears appropriate, as a good share of Mohale’s Hoek’s surface is along the Senqu River Valley agro-ecological zone, considered to have good potential for growing maize, the national staple (Moeletsi and Walker 2013). MFRSC regional spending is highest in Leribe and Berea. These districts are situated in Lesotho’s western lowlands and are major agricultural production regions, also strongly affected by soil erosion (Lageat 1999; Bisangwa 2013). It is thus reasonable to see high MFRSC transfers toward these regions. Interviews with MAFS staff suggest that the Maseru DOs act as parallel Extension Departments, redistributing funds to other DOs. This would explain their high 24  n Agriculture Public Expenditure Review TABLE 6:  District office PEA heat table (red—low; green—high), in million constant 2010 Maloti Region Expenditures, total Expenditures, MAFS Expenditures, MFRSC Berea 49.5 39.3 10.3 Butha-Buthe 47.2 38.6 8.6 Leribe 51.0 38.7 12.3 Mafeteng 46.2 36.3 9.9 Maseru 61.4 47.5 13.8 Mohale’s Hoek 54.4 46.4 8.1 Mokhotlong 42.4 33.6 8.8 Qacha’s Nek 42.8 34.4 8.4 Quthing 44.7 35.5 9.1 Thaba-Tseka 45.1 36.0 9.1 Total 484.7 386.3 98.4 FIGURE 12:  District-level PEA, MAFS (left) and MFRSC (right), in million constant 2010 Maloti level of PEA although, from a budget efficiency point of view, it could be appropriate to assess the rationale for hav- ing seemingly duplicate administrative units (Extension Department and Maseru DOs). No strong discrepancies exist between budgetary allocations and agricultural field concentration per dis- trict. Table 7 shows the comparison between the share of district-level PEA for the MAFS and the share of total agri- cultural fields planted per district. The district of Qacha’s Nek is the most “overfunded,” with a share of ­district-level PEA that is 5 percentage points above its share of fields planted, while the district of Leribe is the most “under- funded,” with a negative deviation of 6 points. Note that the number of fields planted is not equivalent to the number of agricultural households in a dis- trict, as one household may own multiple fields. However, it can be assumed that less bias occurs toward large landowners than by comparing PEA district allocation with share of total area planted, because large landowners tend to own a small number of large fields rather than dozens of small fields. Thus whenever there are many fields planted in a district, this can be assumed to reflect the fact that there are many agricultural households, though some will own multiple fields. To understand the equity of allocations from a household perspective, the exact number of agricultural households should be used for this comparison. However, such data could not be obtained for this AgPER. Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 25 TABLE 7:  Share of district-level PEA for MAFS (average 2010/11 to 2015/16) and share of fields planted (2009/10 census) per district, in percent Share of district-level Share of total number Deviation (percentage PEA MAFS of fields planted points) Qacha’s Nek  9%  4%  5 Butha Buthe 10%  6%  4 Quthing  9%  5%  4 Mohale’s Hoek 12% 10%  2 Thaba Tseka  9%  9%  0 Mafeteng  9% 10%  0 Berea 10% 13% –3 Maseru 12% 15% –3 Mokhotlong  9% 12% –3 Leribe 10% 16% –6 Source: Authors, with Bureau of Statistics 2014. Sources of funding The analysis by source of funding highlights the share of donor public resources in total PEA, thereby indicating the level of reliance of the sector’s budget on external funding. This section also illuminates donor priorities in the sector and evaluates whether discrepancies arise with areas of intervention prioritized by the GoL. Both of these indicators offer valuable insights on the GoL’s policy space in the agriculture sector. Donor expenditures on the agriculture sector Donor financing for Lesotho’s agriculture sector is generally not recorded on the budget, making it diffi- cult to capture the full scope of allocations to the sector. Only two donor-funded projects with expenditures are recorded in the AgBOOST database: the Sustainable Agriculture and Natural Resources Management Project (2.3 million Maloti financed by MCC), and the Rural Financial Intermediation Programme (12.5 million Maloti financed by IFAD). This is 2 percent of total PEA and obviously a strong underestimation of total donor funding for the agriculture sector in Lesotho. The data from the OECD’s CRS show higher levels of donor spending but still lower than actual donor financing to the sector. BOOST data can be compared with the Creditor Reporting System (CRS), which records aid transfers submitted to the OECD by bilateral and multilateral donors (OECD 2019). The OECD data report that US$37 million17 (around 511 million Maloti) were transferred in the form of agricultural aid over the period, 12 per- cent of total GoL AgBOOST plus CRS PEA. In absolute amounts, this is still low by OECD aid standards: SACU and SADC countries received about two and nine times more agricultural aid, on average (Table 8). However, given the size of Lesotho’s population—2.25 million people—it is likely equivalent to what other countries receive. Four donor organizations make up 94 percent of agricultural aid in the country: IFAD, the World Bank, FAO, and the GEF (Table 9). In the review period IFAD was the largest funding agency for agriculture (US$23.5 million), followed by the World Bank (US$11.4 million) and the GEF (US$4.5 million, all in constant 2016 US$). This strong concentration of donor support means that Lesotho’s agricultural budget is highly dependent on the four organiza- tions’ own financial cycles. This “grant-dependent” pattern poses a problem for budget planning and efficient bud- get execution, since it is harder to absorb aid and smooth public expenditures over time. On the positive side, donor 17 These are constant 2016 million US$, the measure used by OECD’S CRS. 26  n Agriculture Public Expenditure Review TABLE 8:  Donor PEA in AgBOOST, OECD’S CRS, with SACU, SADC, and Africa average 2011 2012 2013 2014 2015 2016 Total Average AgBOOST  0  1  0  1  0  0   2 0.3 CRS Lesotho 18  1  0 16  1  0  37 6 CRS SACU 26  7  7  7 10  9  66 11 CRS SADC 54 59 56 38 63 58 328 55 CRS Africa 57 67 74 68 80 86 433 72 Source: Authors and OECD 2019. Note: AgBOOST data are reported in million constant 2010 US$; CRS data are reported in million constant 2016 US$. Annex 7 provides all values in constant 2010 US$. TABLE 9:  Donor aid to Lesotho’s agriculture sector, OECD’S CRS and World Bank, in million constant 2016 US$   2011 2012 2013 2014 2015 2016 Total Share IFAD 7.1 — — 16.4 — — 23.5 50% World Bank 4.9 1.3 0.3 1.3 1.1 2.9 11.9 25% GEF 3.9 0.5 0 — — — 4.4 9% Ireland 1.3 0.1 0.1 — — — 1.5 3% United Kingdom — — — — 0.9 — 0.9 2% Japan 0.1 0.1 0.1 0.1 0.1 0.1 0.6 1% Canada 0.4 — — — — — 0.4 1% Korea 0.1 0.1 — 0 0 0 0.2 0.4% FAO — — 0.1 — — 3.8 3.9 8% Total 17.8 2.1 0.6 17.8 2.1 6.8 47.3 100% Source: OECD 2019, World Bank 2019b, FAO Project Documents (see pp. 18–20 for more detail). Note: OECD’S CRS depends on donors’ voluntary reporting to the OECD, thus this table is likely to be incomplete in terms of actual donor financing/aid in the period. Figures for the World Bank, FAO, and IFAD were obtained separately; annual disbursements could not be obtained for all donors, and yearly allocations may therefore represent that of multiple years. Annex 7 provides all values in constant 2010 US$. concentration reduces transaction costs by limiting administrative back-and-forth with a multitude of donors and thus offers a significant windfall in recurrent budget efficiency. Donor expenditures concentrate on support to productivity improvements and market-oriented agricul- ture, many of which also support climate-smart agriculture practices. For example, the Lesotho Smallholder Agriculture Development Project (SADP 2011–2020, IDA Credit of US$20 million + IFAD US$10 million, nominal US$) provides technical and investment support to increase marketed output among smallholder farmers through sup- port for improved productivity and processing technologies and practices. Many of the grant-financed subprojects have been in high-value, highly nutritious foods for commercial purposes. The IFAD-financed Wool and Mohair Promotion Project (WAMPP, 2015–21, US$38.93 million, nominal US$) is aimed at boosting the economic and climate resilience of poor, smallholder wool and mohair producers and provides support to farmers to improve livestock nutrition and health and marketing of fiber. Both projects support Lesotho’s climate adaptation and mitigation goals in the agriculture sector, as described later in this chapter. Economic analysis At the core of the economic analysis lies the concept of efficiency. Efficiency is “how economically resources/inputs (funds, expertise, time, etc.) are converted to results” (OECD 2010). The focus of this section is therefore on budget execution—whether what was budgeted was actually spent—and repartition between recurrent and development Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 27 expenditures. It should be noted that budget execution is a proxy for efficiency under the assumption that all things being equal, each marginal actual outlay out of the budgeted amount (i.e., the input) will be converted into addi- tional output. This may not be the case and a true efficiency measure should look at delivery of results in the form of concrete outputs (e.g., physical infrastructure) or, even better, developmental outcomes. Assessing the relationship between budgetary outlays and concrete outputs can be done with methods such as a public expenditure tracking survey, but this resource-intensive exercise goes beyond the scope of this review. The same applies to the investiga- tion of causality between public expenditures and development outcomes. Budget execution The average execution rates of PEA (79 percent) and general public expenditures (81 percent) are over- all similar but do not follow the same pattern. The latter was especially the case in 2012/13 and 2015/16 (Figure 13). The poor execution rate in 2015/16 (65 percent) comes from a combination of low execution for the ICP—an 18 million Maloti shortfall—and a budgeted transfer to an extra budgetary unit that did not materialize (56 million Maloti). FIGURE 13:  Budgeted and final PEA (constant 2010 million Maloti) (left); and PEA and total public expenditure execution rates (%) (right) 500 100 450 90 400 80 350 70 (constant 2010) Million Maloti 300 Percent 60 250 50 200 40 150 30 100 20 50 10 0 0 1 2 3 4 5 6 1 2 3 4 5 6 e /1 /1 /1 /1 /1 /1 /1 /1 /1 /1 /1 /1 ag 10 11 12 13 14 15 10 11 12 13 14 15 er 20 20 20 20 20 20 20 20 20 20 20 20 Av Budgeted PEA Final PEA Execution rate PEA Execution rate PE Execution rates have been volatile at the MAFS, mainly reflecting issues in the ministry’s ability to ade- quately plan activities—especially input subsidies—and to procure inputs on time; nonetheless, allocations to personnel are relatively stable. In 2012/13, when the MAFS launched the ICP, it overspent on operating costs at the central level (128 percent) and cut operating costs significantly at the DAO level (79 percent), implying a resource transfer from DAOs to headquarters (HQ) (Figure 14). The operating cost underspending at the DAO level continued in 2013/14 but was reversed in 2014/15 and more importantly in 2015/16, when execution rates were 100 percent versus 44 percent for MAFS HQ. Whether at the HQ or DAO level, execution rates are better and also more stable for personal emoluments than operating costs on average, though the distribution is different as out- lined above. Operating costs act as the adjustment variable for budget unpredictability. According to MAFS staff, due to delays in input procurement, the budget for the ICP is sometimes not fully spent during the fiscal year it is budgeted for. This results in displacement of expenditures to the next fiscal year, thus creating an underspending/ overspending pattern. The pattern of the PEA data, however, goes in the opposite direction: in 2012/13, when the ICP was launched, the MAFS spent more than it budgeted for, while in 2013/14, it spent less than budgeted for. For the MFRSC, execution rates have been good on average both at HQ and district levels but they do exhibit instability. In 2012/13 they plummeted down to 61 percent at the central level, where MFRSC resources are con- centrated, before rocketing to 116 percent the next year (Figure 15). This reflects volatility in expenditures under the MFRSC’s flagship Fato Fato. It should also be noted that personal emoluments at the district level remained 28  n Agriculture Public Expenditure Review FIGURE 14:  Execution rates for MAFS at central and district departments and for operating costs at the district level, in percent (left); and execution rates for MAFS operating costs and personal emoluments at central and district levels, in percent (right) 140 180 160 120 140 100 120 Percent Percent 80 100 60 80 60 40 40 20 20 0 0 1 2 3 4 5 6 e 1 2 3 4 5 6 e /1 /1 /1 /1 /1 /1 ag /1 /1 /1 /1 /1 /1 ag 10 11 12 13 14 15 10 11 12 13 14 15 er er 20 20 20 20 20 20 20 20 20 20 20 20 Av Av Central departments District departments Op.costs central Emoluments central Total Op. costs district Op. costs district Emoluments district FIGURE 15:  Execution rates for MFRSC at central and district departments and for operating costs at the district level, in percent (left); and execution rates for MFRSC operating costs and personal emoluments at central and district levels, in percent (right) 140 160 120 140 100 120 100 Percent Percent 80 80 60 60 40 40 20 20 0 0 1 2 3 4 5 6 e 1 2 3 4 5 6 e /1 /1 /1 /1 /1 /1 ag /1 /1 /1 /1 /1 /1 ag 10 11 12 13 14 15 10 11 12 13 14 15 er er 20 20 20 20 20 20 20 20 20 20 20 20 Av Av Central departments District departments Op.costs central Emoluments central Total Op. costs district Emoluments district stable and close to 100 percent throughout the entire period. The MFRSC districts could thus be pockets (1 percent of total PEA) of budgetary resilience, but the figure could also be the outcome of an accounting artifact due to mis- reporting of actual salary expenditures by the districts. Execution rates have been much higher for recurrent than development expenditures (96 percent versus 66 percent) and for personal emoluments than operating costs (92 percent versus 66 percent) (Table 10). As could be expected, recurrent salaries and operating costs have been stable, with 96 percent and 95 percent execution rates (travel and transport at 101 percent), while development operating costs have borne the brunt of budgetary inefficiency, with a 58 percent execution rate. This is either the reflection of overbudgeting or large neg- ative budget reallocations for the MAFS’s ICP and the MFRSC’s Fato Fato. Development emoluments—largely the MFRSC’s Fato Fato salaries—have been more consistent, at 86 percent execution, although they are less predictable than the recurrent emoluments. Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 29 TABLE 10:  Budgeted and actual PEA, recurrent and development, operating costs and personal emoluments with economic classification breakdown, in million constant 2010 Maloti Economic Economic Recurrent Development Total grouping classification Budg. Exec. Rate Budg. Exec. Rate Budg. Exec. Rate Operating costs Travel and 131 132 101% 44 27 61% 175 159 91% transport Operating costs 109 94 86% 576 377 65% 685 471 69% Transfers 50 49 97% 80 0 0% 130 49 37% Other expenses 0 1 — 27 2 8% 27 3 9% Acquisition of 2 2 85% 307 197 64% 309 199 64% nonfinancial assets Total 292 278 95% 1,033 602 58% 1,326 880 66% Personal Compensation 752 721 96% 390 334 86% 1,142 1,056 92% emoluments of employees Total 1,045 999 96% 1,423 937 66% 2,468 1,936 78% Recurrent and development expenditures The share of development versus recurrent expenditures using BOOST data can be misleading, as this classification does not fully reflect the socioeconomic purposes of certain expenditures. Importantly, the ICP accounts for 32.2 percent of development expenditures even though it is questionable whether input subsidy expenditures recurring on an annual basis to boost productivity for that given year should be considered develop- ment expenditures. Furthermore, the program existed before 2012/13 but does not appear in the IFMIS data in 2010/11 and 2011/12 because, according to interviews with MAFS finance staff, it was included in recurrent costs before 2012/13. It is not clear why the ICP was transferred to development expenditures. Additionally, 9 percent of recurrent expenditures comprise operating costs, which could be classified as development expenditures due to the nature of these activities. The MFRSC’s Fato Fato occupies another 32.8 percent of development expenditures, although it should be noted that 67 percent of the project’s expenditures are reported under the “Compensation of Employees” economic category, and are thus for salaries. This echoes interviews with some GoL staff, who portray Fato Fato as a “cash-for-work” program. It is unclear whether “Compensation of Employees” reflects salaries for MFRSC staff or for the labor building the anti-erosion infrastructure. In both cases, it casts doubt as to the development nature of the program. It also raises accounting issues. Indeed, a global analysis of the BOOST database reveals that it is an uncommon practice to report expenditures of the “Compensation of Employees” category under “Development” (Table 11). Similarly, the share of PEA going to personal emoluments exceeds that of general public expenditures (56.1 percent versus 37.7 percent). An important way to look at the economic structure of PEA is to investigate the balance between operating costs and personal emoluments, which straddle the recurrent and development categories. The relatively high share of PEA allocated to staff costs is not an issue per se, since the mix between oper- ating costs and personal emoluments is highly sector dependent. For instance, personal emoluments account for 71.8 percent of the Ministry of Education and Training’s expenditures, while they total 14 percent for the Ministry of Public Works and Transport. The 56 percent share could reflect agriculture’s balanced needs for “soft” (training, extension, research) and hard infrastructure. However, PEA’s composition of high personal emoluments and low operating costs appears problematic as it leaves little room to implement programs with fuels, vehicles, material, and other items. First, DAOs 30  n Agriculture Public Expenditure Review TABLE 11:  Recurrent and development expenditures and economic groupings, share of AgBOOST PEA and BOOST public expenditures, in percent Recurrent/development Economic grouping Share of AgBOOST PEA (%) Share of BOOST PE (%) Recurrent Operating costs 16.4 41.9 Personal emoluments 38.8 35.9 Total 55.2 77.8 Development Operating costs 27.4 20.3 Personal emoluments 17.2 1.8 Total 44.6 22.1 Total Operating costs 43.9 62.2 Personal emoluments 56.1 37.7 account for 56 percent of the MAFS’s salary bill (Figure 16). This is reflected in DAOs’ own economic structure, with salaries making up 91 percent of their expenditures. Although DAOs are primarily geared toward staff-intensive extension and advisory services, they do require a strong operating budget to cover the ground: vehicles and fuel, appropriate offices, and farming technologies. Interviews with the Extension Services Department suggest that DAOs indeed strongly lack this operating budget. International comparisons show that Lesotho has a low number of farmers per extension worker compared to other countries and hence a high number of extension workers per farmer. A crude estimate of the number of farmers per extension worker is 828, close to that of South Africa, which has a much more significant agricultural budget (Table 12). The Livestock Department’s personal emoluments make up close to 70 percent of its budget when it should be more of an operating cost-intensive department (vaccines and artificial insemination material, for instance). This resonates with evidence from interviews held with the MAFS Livestock Department—staff claim that the Department’s budget to carry out basic disease control and animal productivity enhancement activities on the ground is very limited. In the meantime, the salary-intensive Department of Agricultural Research (DAR) has a FIGURE 16:  Detailed economic structure of personal emoluments, by district and central level, MAFS, in total million constant 2010 Maloti 4114 Allowances in cash—non-statutory posts 9 District 4111 Non-statutory salaries in cash 279 4123 Unfunded pensions and gratuities 4 4114 Allowances in cash—non-statutory posts 4 Central 4113 Wages in cash 6 4111 Non-statutory salaries in cash 211 0 50 100 150 200 250 300 Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 31 TABLE 12:  Farmers per extension worker, selected countries Farmers per Country extension worker Lesotho  828 South Africa  878 Vietnam 1,331 Kenya 1,200 Rwanda 1,500 Note: The number of extension workers in Lesotho was calculated as a proxy by multiplying the total number of employees at the MAFS in 2015/16 by the share of the MAFS salary bill going to DAOs (including the Maseru DAO but not including the 0107 Extension cost center, i.e., the MAFS Extension Department). The number of farmers per extension worker was obtained by dividing the number of agricultural households (Bureau of Statistics 2014) by the number of extension workers and multiplying by 4 to account for the number of farmers per household. lower budget than even some DAOs. Nevertheless, it should be added that interviews conducted with MAFS finan- cial staff suggest that capital expenditures made by DAOs are underreported in the IFMIS. For the MFRSC, 70 percent of personal emoluments are through Fato Fato. On the one hand, the program itself should be heavier on operating costs due to its focus on environmental infrastructure works. On the other hand, much of its personal emoluments are thought to reflect temporary labor under the program’s cash-for-work objec- tive. This means the salary bill for MFRSC civil servants accounts for only 30 percent of the ministry’s total. Fato Fato also makes the MFRSC top-heavy: 83 percent of personal emoluments are at central level, leaving DOs short of funds (Figure 17). The nature of personnel costs differs between the MAFS and the MFRSC: 72 percent of the MFRSC’s salary bill (85 percent at the central level) comprises cash wages, by and large under Fato Fato, as opposed to 95 percent of salaries as posts for the MAFS (Figure 16 and Figure 17). This reflects once again the focus of the MFRSC’s activities on Fato Fato, a program heavily reliant on temporary workers. FIGURE 17:  Detailed economic structure of personal emoluments, by district and central levels, MFRSC, in total million constant 2010 Maloti 4114 Allowances in cash—non-statutory posts 2 District 4113 Wages in cash 1 4111 Non-statutory salaries in cash 55 4114 Allowances in cash—non-statutory posts 1 Central 4113 Wages in cash 272 4111 Non-statutory salaries in cash 48 0 50 100 150 200 250 300 32  n Agriculture Public Expenditure Review In spite of already low levels, 47.4 percent of operating costs at the MAFS level goes to the ICP. As outlined above, this can be seen as problematic, given that input subsidy programs are meant to be temporary support mea- sures rather than investments that expand the country’s agricultural production frontier or improve the enabling environment for private sector growth. The distribution of employees is very bottom-heavy in the MAFS and the MFRSC, more than in other minis- tries. The two ministries are thus likely to be overreliant on support staff. Forty-eight percent of graded employees at the MAFS are in grades A to C (37 percent on grade A alone), the lowest levels—support staff. An equal share of employees is concentrated in mid-level grades D to F (48 percent)—mid-managers and technical ­specialists—while high-level grades G–L total a low 4 percent of staff (Figure 18). In the MFRSC, grade A–C employees account for 53 percent of the total number (33 percent in grade A), while 32 percent are in grades D–F and 15 percent are in the higher grades G–L. In comparison, other ministries employ 43 percent of their graded employees in grades A–C (21 percent grade A), 39 percent in grades D–F, and 18 percent in grades F–L. FIGURE 18:  Share of graded employees per grade in the MAFS (left) and the MFRSC (right) L L K K J J I I MFRSC grade MFRSC grade H H G G F F E E D D C C B B A A 0% 10% 20% 30% 40% 0% 10% 20% 30% 40% The distribution of wages is middle-heavy in the MAFS, in stark contrast with the MFRSC and other minis- tries, where it tends to be skewed on the top. Whereas the MFRSC and other ministries concentrate their wage bill (excluding ungraded workers) to high grades G–L—38 percent and 36 percent, respectively—the MAFS wage bill stands at 18 percent for higher grades and is instead skewed toward middle grades, at 62 percent (Figure 19). In addition, although the MAFS and the MFRSC have many more employees in lower grades than other ministries, the share of these employees’ wages in the total graded wage bill is similar across all ministries: 21 percent for the MAFS, 22 percent for the MFRSC, and 23 percent for other ministries. This suggests that A–C grade employees are less well paid at the MAFS and the MFRSC, thus compensating financially for their number but potentially with a negative effect on their productivity. FIGURE 19:  Share of graded employee salary bill per grade in the MAFS (left) and the MFRSC (right) L L K K J J I I MFRSC grade MFRSC grade H H G G F F E E D D C C B B A A 0% 5% 10% 15% 20% 25% 30% 0% 5% 10% 15% 20% 25% 30% Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 33 Functional analysis The functional analysis looks at the allocation of PEA across agricultural function, such as input subsidies, research, and irrigation. This functional breakdown is compared to government priorities for the sector (policy coherence analysis) and used to understand whether the government is using public resources on public or private goods and on areas of expenditure conducive to private sector investment.  Subsectoral indicators Assessing the allocation of PEA between sectors reveals that 70 percent are multisectoral in nature (Fig- ure 20). That is because the MFRSC’s Conservation and Rangeland Departments’ expenditures are considered multi-sectoral, as well the MAFS’s Administration recurrent costs and expenditures for the DOs, Lesotho Agricul- tural College, and the DAR. However, for 2010/11 and 2011/12, the ICP is blended into recurrent costs and can- not be distinguished, except for some sporadic occurrences of “input subsidies” in the MAFS budget. Based on a gross calculation—average recurrent costs 2012/13 to 2015/16 minus average recurrent costs from 2010/11 to 2011/1218—the crop-specific ICP represented an estimated additional 47.5 million Maloti in 2010/11 and 2011/12. This would mean the actual share of multi-sectoral PEA is 68 percent, with crops at 27 percent. FIGURE 20:  Breakdown of PEA by subsector, in percent 4% 1% 25% 70% Multi-sectoral Crops Livestock Forestry Crops make up 83 percent of sector-specific expenditures, with livestock and forestry accounting for 14 per- cent and 3 percent, respectively. A major cause for this is that the MAFS’s ICP aims at improving crop productivity. In addition, based on evidence from interviews conducted with the MAFS and the MFRSC, it is safe to assume that the internal orientation of multi-sectoral expenditures follows the same pattern. For instance, all of the DAR’s sig- nificant active projects are about crops. The breakdown of sector-specific expenditures shows that allocations are not in line with the relative com- position of agricultural GDP. Livestock and forestry account for 70 percent of agricultural GDP but receive 17 per- cent of PEA (Figure 21). In addition, a systematic bias tends to arise against livestock and forestry in the agricultural GDP of African countries, because economic transactions are harder to record and measure due to the higher 18 The ICP is not labeled in 2010/11 and 2011/12, but anecdotal evidence suggests it is contained in the MAFS’s recurrent costs for these years. Average recurrent costs in 2010/11 and 2011/12 stand at 147.25 million Maloti, while they drop to 123.5 million Maloti for the period 2012/13 to 2015/16, with a very narrow standard deviation (3.7 million). The year 2012/13 coincides with the appearance of the ICP in the development budget. It is thus assumed the average differential in recurrent costs between the two periods is a gross estimate of the true value of the ICP in 2010/11 and 2011/12. 34  n Agriculture Public Expenditure Review FIGURE 21:  Share of subsectors in AgBOOST and in AgGDP, in percent 90 80 70 60 Percent 50 40 30 20 10 0 Livestock Crops Forestry Share in AgGDP Share in AgBOOST Source: Authors and Central Bank of Lesotho (AgGDP data). Note: Shares in AgGDP from the Central Bank data are 50 percent for livestock, 25 percent for crops, 19 percent for forestry, and 6 percent for services. Services were removed to compute relative shares of the three sectors. intensity of informal and/or illegal trade, and to nonmonetized value such as ecosystem services or self-collected wood fuel (Muthami and Behnke 2011; Behnke and Nakirya 2013; Lebedys and Li 2014). The imbalance could thus be higher than it appears but, on the other hand, the AgBOOST data do not include IFAD’s Wool and Mohair Promo- tion Project (WAMPP), a very substantial contribution to PEA (US$20.3 million, or 298 million Maloti) in support of livestock. There need not be a perfect match between the subsectoral composition of PEA and agricultural GDP: some sectors are more intensive in public goods than others, while some are productive and competitive enough to thrive with limited direct government support. Nevertheless, the large disproportion between the contribution of the agriculture subsectors to the economy and the level of support they receive warrants further scrutiny. In particular, in the case of Lesotho, the livestock industry contributes not only to GDP but also to the country’s trade balance, given its significant wool and mohair exports. The IFMIS/BOOST data are not granular enough to allow commodity-specific analysis. This could be improved for future agriculture public expenditures by coding MAFS and MFRSC budget lines by the commodity they sup- port. However, 69 percent of crop expenditures go to the ICP—which mainly supports staple cereal production: maize, wheat and sorghum. Therefore, around 57 percent of sector-specific and 17 percent of total PEA (at least) single-handedly benefit these staple cereals in the form of input subsidies. The real level of public support to the cereals is likely to be much higher, with a good share of district-level activities (52 percent of multi-sectoral expen- ditures) certainly geared toward them. Functional indicators PEA are balanced between the three functional poles of (i) subsidies (25 percent), (ii) knowledge produc- tion and dissemination (35 percent), and (iii) infrastructure expenditures (26 percent) (Table 13). It should be stressed that the input subsidy figures for 2010/11 and 2011/12 are underestimated due to their reporting as recurrent costs. As stated above, the subsidy could be underestimated by 47.5 million Maloti for 2010/11 and 2011/12. This means the true share of input subsidies in PEA would be closer to 30 percent for the period. In addi- tion, a large share of the Fato Fato program is classified under “other infrastructure” due to its objective of building environmental infrastructure, but it could also be considered as a social transfer and thus a form of subsidy. Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 35 TABLE 13:  Heat table of the share of agricultural functions in PEA (red—low share; green—high share), 2010/11 to 2015/16 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 Average Subsidies 10% 13% 35% 32% 30% 28% 25% Capital subsidies 5% 5% 1% 7% 5% 5% 5% Input subsidies 3% 8% 33% 25% 25% 23% 20% Services subsidies 2% 1% 0% 0% 0% 0% 0% Knowledge 41% 38% 33% 30% 33% 38% 35% Research 4% 3% 2% 2% 2% 2% 2% Extension and advisory services 32% 30% 27% 23% 26% 29% 28% Training 5% 5% 5% 4% 5% 6% 5% Infrastructure 31% 34% 18% 26% 26% 22% 26% Irrigation 1% 0% 0% 0% 0% 1% 0% Other infrastructure 29% 33% 18% 25% 25% 20% 25% Processing and marketing 0% 0% 0% 1% 0% 1% 0% Administrative 17% 16% 14% 12% 12% 13% 14% Nonclassified 1% 0% 0% 0% 0% 0% 0% Subsidies therefore account for a very sizeable share of PEA, at close to 30 percent, although this is roughly comparable to what is observed in a number of other African countries. For instance, Jayne et al. (2018) found that the average share of PEA allocated to input subsidies in ten African countries in 2014 was 19 percent; Ghins, Mas Aparisi, and Balié (2017) found a share of 35 percent for a different sample of nine countries for the period 2006–2013. The composition of subsidies is strongly oriented toward inputs rather than capital. Yet input subsidies are less likely to increase productivity in the long run, because the productivity boost effect of the fertilizers and improved seeds will dissipate after one year, as opposed to that of farm equipment, cattle, or other forms of capital. Although input subsidies are meant to be a temporary support measure, they have not abated since 2010/11. In the current context of narrow fiscal space, the input subsidy program blocks one-third of the GoL’s budget every year—more when considering the salaries of civil servants involved in administrative tasks related to the ICP. This warrants close scrutiny on the rationale for the subsidy program and the results obtained (Box 3). The current structure of knowledge and infrastructure PEA is unlikely to address structural constraints to productivity increase and agricultural transformation. Although expenditures on extension and advisory services are high, they may not be efficient due to a concentration on wages over operating costs. And although low levels of irrigation in Lesotho are a key constraint, allocations to irrigation infrastructure are limited. Further, to support a path to agricultural transformation, the GoL has a key role in financing R&D for increased productivity; yet spending on agricultural research is dismal, at an average 8 million Maloti per year (Fan, Gulati, and Thorat 2008; Mogues, Fan, and Benin 2015) The share of PEA allocated to agricultural research in AgBOOST is low, equivalent to 0.8 percent of agricul- tural GDP over the period. This is below the 1 percent commitment enshrined in the 2006 Khartoum Decision 36  n Agriculture Public Expenditure Review BOX 3:  Lesotho’s input subsidy program The current input subsidy program was launched by the GoL in 2010/11. Initially labelled the Summer Cropping Programme, it was renamed the Intensive Crop Production Programme (ICP) due to the fact that inputs are subsidized during both winter and summer cropping seasons. Through the ICP, the GoL buys fertilizer, seed, and pesticide packages (Figure B1) in South Africa and sells them at discounted prices to private traders. The traders then sell the inputs to producers at a subsidized price, factoring in their margin. The inputs are sold by the MAFS to the traders against cash. Traders sell exclusively to beneficiaries that appear on a list prepared at the district level, but no transparent or official criteria exists for drawing the list. Since there are no explicit targeting criteria, it can be assumed that the program is regressive, given the unequal structure of land ownership in Lesotho. Indeed, 1.2 percent of households own land above 6 hectares, with some of these landowners holding dozens of hectares, while 45 percent of households own land of less than 1 hectare (Bureau of Statistics 2014). FIGURE B1:  Total procurement value 2012/13 to 2016/17, in million constant 2010 Maloti 15 60 210 Fertilizer Seeds Chemicals Source: Authors, from MAFS 2018. The average subsidy expenditures over the period were 64 million Maloti, and 87 million Maloti since 2012/13. The average GoL fertilizer price subsidy rate—that is, the ratio of trader sale price over procurement price—was 62 percent (Annex 5). The rate was higher for highlands (63 percent) than lowlands (56 percent) due to costlier production (MAFS 2018). However, 93 percent of the total input sale value for the period comes from lowland sales (MAFS 2018). Indeed, this is where agricultural production is concentrated. An important point is that since the GoL does not give away the inputs procured, but sells them, there is a recoup on the subsidy. The average annual recoup value is estimated at 24 percent of procurement value (Annex 6). Thus, the true cost for the GoL of the ICP is about one-­ quarter less than what appears in the expenditure side of the budget. Furthermore, evidence from the Capital Progress Reports communicated by the MAFS implies that ICP expenditures have not been allocated 100 per- cent to input procurement. An estimated 6 percent went to capital subsidies (equipment procurement), 4 percent to irrigation, 2 percent to research (Geographic Information System), and 2 percent to inspection/quality control (plant protection). However, this breakdown is not reported in the official budget. Finally, it is worth assessing the ICP project with the SMART criteria. The literature refers to a “SMART” benchmark—Specific tar- geting, Measurable impacts, Achievable objectives, Result-oriented, Time-bound—to estimate whether input subsidy programs (ISP) are likely to be efficient (Minde et al. 2008; Dorward 2009; Wanzala-Mlobela, Fuentes, and Mkumbwa 2013). Wanzala-­ Mlobela, Fuentes, and Mkumbwa (2013) summarize these objectives as such: smart ISPs (i) target a specific socioeconomic group, (ii) contribute to a competitive and open market, and (iii) come with an exit strategy. The ICP does not appear to match any of the criterion, since there is no clear targeting and no evidence of a private sector crowding-in and/or phasing out strategy given the persistence of the program in PEA over the years. In addition, there is no indication of increased yield or production other than bumper harvests due to favorable rainfall (FAO 2018). for Agriculture and Technology of the Executive Council of the African Union (African Union 2006).19 For the sake of accuracy, the AgBOOST figures are compared to the figures published by the International Food Policy Research Institute’s (IFPRI) Agricultural Science and Technology Indicators (ASTI) initiative. Interestingly, ASTI figures are superior by a factor of three on total for the period (Table 14). This may reflect the donor funding that is absent in 19 The exact commitment is to allocate a budget equivalent to 1 percent of GDP to science and technology. For the agriculture sector, this would mean 1 percent of AgGDP. Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 37 TABLE 14:  AgBOOST and ASTI PEA for research, 2010/11 to 2015/16, in million constant 2010 US$ 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 Total AgBOOST 1.0 0.9 0.9 1.1 0.8 1.0  5.6 ASTI 2.5 2.7 2.8 2.2 4.7 2.8 17.7 Source: Authors and ILO 2019. AgBOOST and purportedly included in the ASTI dataset—although donor funding accounts for a mere 6.9 percent of the ASTI total, not enough to explain the differential. Despite the need to invest in and maintain advanced equipment to conduct the research necessary for a competitive agriculture sector, the AgBOOST and ASTI datasets both portray a strong direction of the research budget toward salaries: 93 percent and 79 percent, respectively. Research is a salary-intensive activ- ity, but it does require high enough operating costs and investment in equipment. Anecdotal evidence from inter- views conducted with DAR triangulates observations from the AgBOOST and ASTI data, pointing to the fact that the department runs on excessively low operating costs and capital investment, preventing it from conducting normal research activities (Figure 22). The ASTI data do record a large capital investment in 2015, however. This may cor- respond to an internal lump-sum transfer from the ICP that was mentioned during interviews with DAR staff but does not appear on the IFMIS. The World Bank–financed APPSA project will largely fill this gap (Box 4), and it is important that the project is reported on the budget to adequately capture these important expenditures. FIGURE 22:  Share of agricultural research salaries, operating and program costs, and capital investment, ASTI and AgBOOST, 2011–2015 100 90 80 70 60 Percent 50 40 30 20 10 0 ASTI ASTI ASTI ASTI ASTI AgBOOST AgBOOST AgBOOST AgBOOST AgBOOST 2011 2012 2013 2014 2015 Salaries Operating and program costs Capital investment Source: Authors, with ASTI 2019. The PEA on infrastructure represent, at 99 percent, the MFRSC’s Fato Fato and its related soil and range- land conservation measures. Fato Fato indeed has a strong infrastructure component, with the building of stone lines, diversion furrows, and gully and water harvesting structures (Box 5). Although it is outside the scope of this review to assess the appropriateness and results of the Fato Fato engineering measures, the line of expenditure appears to be going in the right direction by providing an essential public good in support of agricultural produc- tion. However, as outlined above, 67 percent of Fato Fato expenditures are reported under the “Compensation of 38  n Agriculture Public Expenditure Review BOX 4:  Agricultural Productivity Programme for Southern Africa (APPSA) In an effort to increase public financing to agricultural research and to improve productivity in the sector, the GoL will participate in the World Bank–financed Agricultural Productivity Programme for Southern Africa (APPSA). While agricultural productivity has increased in Southern Africa, Lesotho’s average yields are below the regional average. APPSA is expected to help narrow these gaps with investment in technology adaptation and dissemination. The total financing is US$20 million; the first compo- nent, which is entirely focused on support for agricultural R&D, is US$6.8 million. Activities financed under APPSA include Innovative Research and Development (R&D) technology generation and dissemina- tion activities associated with the commodity groups or technology themes being targeted by participating countries, as well as strengthening of the institutional and enabling environment for technology adaptation in these countries. Malawi, Mozambique, and Zambia have participated in APPSA for several years while Lesotho and Angola just joined. The project was approved by the World Bank Board of Executive Directors on December 19, 2018. BOX 5:  Lesotho’s “Fato Fato” program The Watershed Management Program is a public works program formally started by the MFRSC in 2013. It is commonly known as the Fato Fato program, meaning “to dig” in Sesotho. Fato Fato public work programs have been conducted since at least the 1990s in Lesotho (Pule 1999). In continuation with previous programs, the current Fato Fato serves the dual objectives of environmental conservation and social safety net with a cash-for-work approach. Through Fato Fato, households are offered a maximum one month of work per year to participate in conservation activities, mainly small infrastructure such as stone lines, diversion furrows, and gully and water har- vesting structures (FAO 2016). Also, only one household member at a time may participate (FAO 2016; Kardan, O’Brien, and Masasa 2017). One study found that in 2013 Fato Fato participants were paid 48 Maloti per day for up to 20 days and that ­ 58,000–115,000 individuals participated (World Bank 2013). It can be noted that despite its poverty objectives, the Fato Fato program has no target- ing criteria; beneficiaries are selected on a “first come, first served” basis (FAO 2016; Kardan, O’Brien, and Masasa 2017). The experience of the Fato Fato program has shown challenges in achieving the program’s objectives, both in availing a viable safety net for the poorest segments of the rural population and in implementing sustained land conservation. To address these challenges, WFP Lesotho is supporting the government and the MFRSC since 2015 in reforming the Fato Fato program through the following: • Defining of a long-term vision for the program, along with a multiyear, results oriented programming for greater impact; • Establishing a community-based process for the identification of the desired outcome: promoting ownership; • Identifying and prioritizing assets around outcomes vs. transfers; • Application and use of planning tools that promote coordination, integration, and partnership; • Improving target and selection mechanism of program beneficiaries; • Improving quality assurance: technical and technological appropriateness with focus on quality and not only quantity; • Ensuring maintenance, utilization, and management of assets created; and • Improving the program Monitoring and Evaluation system. WFP Lesotho supported the development of new guidelines on effective implementation of public works programs, in line with integrated catchment management principles, and is currently supporting the implementation of the other proposed reforms through pilot initiatives in Maseru, Botha-Bothe, and Leribe. An evaluation of the pilot is currently underway, and lessons learnt are intended to support wider reforms of the program. WFP also supported, in collaboration with the World Bank, knowledge exchange with the Ethiopia Productive Safety Nets Public Works from which lessons could also be learnt to improve the Fato Fato (source: communication with WFP Lesotho). Employees” economic category—reflecting the cash-for-work nature of the program. In addition, Fato Fato expen- ditures dipped in 2012/13 for a reason that is unclear; this is reflected in total PEA spending on infrastructure (Figure 23). Given that Fato Fato appropriates such a high share of the MFRSC’s budget and of Lesotho’s agri- culture-related infrastructural expenditure, it would be appropriate to review its efficiency and effectiveness in delivering on its dual objectives. Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 39 FIGURE 23:  Trend of PEA in support of aggregate functional categories, in million constant 2010 Maloti, 2010/11 to 2015/16 140 120 100 80 60 40 20 0 2010–11 2011–12 2012–13 2013–14 2014–15 2015–16 Administrative costs Infrastructure Knowledge Marketing Subsidies MAFS expenditures, on the other hand, are almost devoid of infrastructure support, although these are vital to improve the sector’s enabling environment. Irrigation in particular has received only 0.5 percent of PEA, although rainfall dependency and risk of drought represent a major impediment to the sector’s growth. This was illustrated in 2015/16, when a severe episode of drought caused by the El Niño phenomenon left around 709,000 people food-insecure (Lesotho Vulnerability Assessment Committee 2016). The gravity of this episode led to the implementation of a 155 million Maloti for a response plan (Box 6). PEA in support of processing and marketing activities have been extremely minimal, at 0.2 percent. They mostly reflect the contribution of the marketing department of the MSBDCM. The department provides a range of interesting market-oriented services to the agriculture sector (see Administrative analysis section) but these account for a fraction of the total PEA budget. In addition, a number of marketing department activities involve trade and price-distortive measures, such as import restrictions, recommended producer prices, and the handling of consumer price subsidies, as was the case in 2016/17 when a 162 million Maloti subsidy was provided via the MSBDCM to maize flour millers (DMA 2016; MSBDCM 2018). Note that actual processing and marketing PEA, as well as infrastructure PEA, are likely higher than their AgBOOST level, because they benefit from off-budget IFAD and World Bank support. Indeed, the AgBOOST functional structure of PEA reveals GoL spending priorities, given that donor expenditures are barely included. It is manifest that the GoL is focusing expenditures on production-oriented, labor-heavy, and private good areas (input subsidies, extension services, WMP), while donors fund market-oriented, infrastructure, and public good areas of expenditure. This discrepancy is a reflection of aid fungibility: with donor agencies taking up the aforementioned functions, the GoL divests the domestic budget toward functions that are left unsupported. Yet with donor sup- port concentrated in the World Bank/IFAD duo since the disengagement of several other donors over the period20 (World Bank 2016), the GoL is very exposed to shifts in funding and policy priorities of the two organizations. It may be a good strategy for the GoL to balance domestic support more evenly across agricultural functions. For instance, agricultural research has been left underfunded for years in the absence of substantial donor support. 20 The World Bank NSDP Review report highlights that over the 2005–2013 period, Lesotho experienced the suspension of the second compact of the Millennium Challenge Corporation (MCC), cancellation of budget support of €26.85 million by the European Union under its EDF 10 assis- tance program, as well as phasing out of resident development partners such as Ireland and the United Kingdom (World Bank 2016). 40  n Agriculture Public Expenditure Review BOX 6:  PEA under the El Niño drought response plan Following the catastrophic drought of 2015/16, the GoL, with support from donors, put together an emergency response plan. The plan was costed at 584 million Maloti, and 155 million Maloti were raised. The plan was coordinated by the Disaster Man- agement Authority (DMA) of the Prime Minister’s Office (PMO). Evidence communicated by the DMA suggests that the MAFS received 6 percent of the 155 million Maloti, equivalent to 5 percent of the MAFS’s budget for 2015/16 (Figure B2). This amount was used to increase the input subsidy in a bid to help production recover in time for the 2016/17 winter cropping, lest Lesotho should face a second food crisis. It should also be noted that these expenditures may not capture the entire drought response budget. FIGURE B2:  Breakdown of El Niño response plan expenditures, in million Maloti 10 13 132 Ministry of Health Ministry of Water MAFS Source: DMA 2019. From an agriculture sector perspective, a more sustainable way of mitigating drought effects may be to invest in irrigation infrastructure, research on drought-resistant varieties, and support for farmers to access appropriate financial and insurance instruments. Source: FAO undated. Public and private sector functions21 An angle of functional analysis is to examine whether the GoL has been allocating public resources on the agricultural functions it is best placed to take on. The economic literature offers a useful reference to define what such functions are for agriculture (López 2005; López and Galinato 2007; Fan, Gulati, and Thorat 2008; Mogues, Fan, and Benin 2015). Most of it relies on the distinction between “public goods” and “private goods.” In short, public goods are nonrival and nonexcludable: their use by an economic agent does not reduce that of another one, and an economic agent cannot exclude another one from using them. Private goods, on the contrary, are rival and excludable. Although many goods do not fall neatly into one or the other category, governments are generally recognized as better placed to provide goods that come the closest to the “public good” definition. Indeed, the mar- ket will tend to underprovide them because of free rider issues, among other factors: those who do not pay market price for the good may still be able to enjoy it, creating a disincentive to pay. On the contrary, from a resource effi- ciency point of view, the market is best placed to provide private goods, because it will do so at a lower social cost than government. When there are “market failures”—that is, when the market is not able to provide private goods in a socially optimal quantity and at an optimal price (lack of information, asymmetries of information between buyer and seller, lack of access to a market due to difficult geography, etc.)—then the government can compensate by subsidizing optimal private good provision. However, government provision of private goods risks crowding out the private sector, which means the provision of the good will unsustainably weigh on public resources. In addition, 21The introductory paragraph below draws largely from Mas Aparisi (2018), an unpublished methodological note also used in the unpublished Agricultural Sector Review for Kenya (World Bank undated). Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 41 when government provides private goods, it effectively transfers public resources (belonging to everyone) toward private owners. This is not without political economy implications.  The optimal role for government in the agriculture sector depends on the country context. The World Bank there- fore developed a framework for assessing how well the market provides different functions versus where the government can temporarily play a role in the provision of private goods to the agriculture sector (Fig- ure 24). In the absence of such market failures, government is best suited to provide public goods and services, address information asymmetries, and/or minimize negative externalities in the sector. While it is beyond the scope of this analysis to determine where exactly Lesotho can be placed in the MFD (Maximiz- ing Finance for Development) ladder, it is possible to provide an overview of the allocation of PEA to “private sector” and “public sector” functions (Figure 25). Private sector functions include all subsidies, whereas public sector functions include knowledge-related expenditures as well as public infrastructure. Administrative costs were attributed evenly to each category. The MAFS has focused on private good provision (40 percent) through the ICP. This raises the issue of sus- tainability of the GoL’s agricultural budget. Government transfers for private goods tend to prevent private sector crowding-in and thus weigh on the budget, thereby reducing availability of resources to support public good func- tions underprovided by the market, such as better extension services, higher quality agricultural research, and expanded irrigation infrastructure. The MFRSC has targeted public goods with Fato Fato and the DOs’ extension services (97 percent). Neverthe- less, a more fine-tuned analysis of Fato Fato’s items of expenditure would certainly reveal private good transfers. Indeed, it is likely that cash-for-work transfers under Fato Fato contain a significant subsidy proportion—i.e., the share of the transfers that is above the opportunity cost of labor to build the infrastructure. In addition, not all forms of conservation infrastructure are public. For instance, stone lines, if built on privately owned land, may end up belonging to the owner and thus would be a private good. Policy coherence analysis The National Strategic Development Plan (NSDP) 2012/13–2016/17 and the Budget Framework Papers (BFPs) are used as policy benchmarks for the policy coherence analysis (PCA). The PCA looks at the coherence between the functional structure of PEA and the GoL’s policy objectives for the agriculture sector over the period. The NSDP is the GoL’s main strategic document and sets out its policy objectives for every sector, including agricul- ture. It is the reference agricultural policy document for the period under review (except 2010/11 and 2011/12). The BFPs are produced by the MF and set out the government budget priorities for each sector, with a triennial budget projection for line ministries. Three batches of available BFPs were used for the PCA: one for 2011/12, one for 2012/13 to 2014/15, and one for 2015/16 to 2017/18. The formal linkage between policy strategy documents and policy implementation in Lesotho remains very much a work in progress (World Bank 2016), but it is valuable to assess whether stated objectives and budget allocations were congruent even in the absence of a strong public financial management system. NSDP policy coherence analysis The NSDP was the reference policy orientation document for the agriculture sector for the period under review; yet it was not sector-specific, nor was it costed. No agricultural policy strategy was articulated to the NSDP I between 2012/13 and 2016/17. In addition, although agriculture-specific priority interventions are listed in the NSDP, they do not come with an implementation plan or a budget. This is a considerable policy shortcoming, reducing budget predictability, transparency, and accountability. Furthermore, although a second NSDP is currently under preparation with more detailed agricultural strategic objectives, no evidence exists of a stand-alone, bud- geted agricultural strategy in the pipeline. FIGURE 24:  Assessing the appropriate level of government involvement in the agriculture sector 42  Spectrum of potential actions to promote responsible food and agriculture investments ■ Strengthen country capacity to assess and mitigate/regulate environmental and social risks Is the private sector Yes ■ Promote private sector alignment with the principles of responsible investment doing it? ■ Support inclusive business models to improve linkages among smallholders and firms of all sizes No Spectrum of potential actions to increase space for private sector investments Is this because of Yes ■ Support competition and associated policy reform, including of state owned enterprises limited space for ■ Strengthen investment policy and dialogue to open space for global investment private sector activity? ■ Reduce government intervention in agricultural financial markets to open space for private financial service providers No Spectrum of potential actions to improve the policy and regulatory environment for private sector investments and to reduce the distortionary effects of public spending ■ Reduce distortionary effects of public spending policies Improve incentives and reduce transaction costs Reduce private sector investment risk n Agriculture Public Expenditure Review Is this because of policy Yes ■ Lower trade costs ■ Ensure macroeconomic and political stability and regulatory gaps or ■ Improve policies and regulatory regimes of input ■ Improve the stability and predictability of policies weaknesses? markets ■ Improve land tenure security and access to land ■ Improve policy and regulatory environment for ■ Shift public policies from farm production support toward agri-finance improving access for farms and agribusiness to risk ■ Strengthen food safety systems management instruments that can increase lending No Spectrum of potential public investments to reduce private sector transaction costs and risk Improve incentives and reduce transaction costs Reduce private sector investment risk ■ Invest in public infrastructure based on clear ■ Support political risk insurance for financial institutions and Can public investment Yes private sector needs private investors help crowd-in private ■ Invest in public inspections and quality assurance ■ Consider use of market pull incentive mechanisms investment? ■ Improve coordination to reduce transaction costs ■ Provide direct financing to value chain actors ■ Consider public-private partnerships No Use public resources to invest in public or quasi-public goods and services Where there is no viable private sector return: Pursue purely public ■ Invest agricultural public spending in public goods and services (e.g., human capital, agricultural research) financing ■ Support complementary public investment in other sectors (such as rural roads, energy) to enable the commercialization and competitiveness of national agricultural production, processing, and marketing Source: Townsend et al. 2018. Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 43 FIGURE 25:  PEA, by private sector and public sector functions, MAFS and MFRSC, in million constant 2010 Maloti 350 300 Million Maloti (constant 2010) 250 200 150 100 50 0 MAFS MFLR MAFS MFLR MAFS MFLR MAFS MFLR MAFS MFLR MAFS MFLR 2010–11 2011–12 2012–13 2013–14 2014–15 2015–16 Avg Public sector function Private sector function TABLE 15:  Functional count of agricultural interventions in the NSDP, count and percent of total Functional category Count of interventions Share Administrative costs 26 39% Processing and marketing  8 12% Extension and advisory services  7 10% Training  6  9% Research  5  7% Irrigation  4  6% Other infrastructure  4  6% Capital subsidies  3  4% Inspection/quality control  3  4% Services subsidies  1  1% Input subsidies  0  0% The NSDP puts forward three overarching strategic objectives for the agriculture sector, which can be bro- ken down into 15 actions and 67 interventions. The 67 interventions were classified according to their functional objective and subsector. The number of interventions per functional category and subsector was then counted and ranked by order of magnitude (summarized in Table 15). Even though the NSDP does not propose development of a specific agricultural policy strategy for the sector, “Administrative costs” is by far the category with the highest number of interventions mentioned in the doc- ument. As the average number of interventions per function is 6, the outstanding category is administrative costs, with 26 interventions. Many of these interventions have to do with policy and institutional design (“develop an inte- grated land and water management policy and framework;” “establish interministerial coordination mechanism) or policy studies (“undertake regular public expenditure reviews;” “identify options for agricultural insurance prod- ucts”). This is paradoxical, because these interventions do not include the design and costing of an agricultural policy strategy, which could have been a major objective in itself. 44  n Agriculture Public Expenditure Review In the absence of a costed strategy that can be compared with actual expenditures, only a crude estimate of the GoL’s agricultural policy priorities can be provided. The functional composition of PEA is expressed in monetary value, which cannot be directly compared with the count of interventions in the NSDP. However, the pri- oritization of agricultural functions in the NSDP and PEA can be compared by looking at their respective rankings (Table 16). This approach comes with caveats, because the count of interventions in support of a function is not the same as its budgetary allocation. One single intervention may imply a large sum of expenditures, while a dozen small interventions may amount to a small budgetary value. TABLE 16:  Deviation between functional share of NSDP intervention count and PEA, in percentage points Functional category NSDP intervention count share PEA share Deviation Input subsidies  0 20 20 Other infrastructure  6 25 19 Extension and advisory services 10 28 17 Capital subsidies  4  5 0 Services subsidies  1  0 –1 Training  9  5 –4 Inspection/quality control  4  0 –4 Research  7  2 –5 Irrigation  6  0 –6 Processing and marketing 12  0 –12 Administrative costs 39 14 –25 Source: Authors, with Government of Lesotho 2012. Agricultural administrative costs and processing/marketing were much less prioritized in PEA than they were in the NSDP, while subsidies and extension services were overprioritized in PEA compared with their NSDP rank. This adds an interesting angle on the functional analysis above, because it shows that the functions underfunded by the GoL were also those prioritized in the NSDP, while those actually prioritized are not mirrored in the NSDP objectives. For instance, the document makes no reference to input subsidies. The overprioritization of “other infrastructure” (i.e., the MFRSC’s Fato Fato) is an outcome of the absence of conservation policy priorities in the agriculture part of the NSDP document. The lack of prioritization of processing/marketing can be explained by the fact that the SADP, which is not accounted for in this review, mainly covered this function. For other sectors, the ranking of NSDP interventions is not very far off the sectoral composition of PEA. Livestock (10 percent of interventions) do not receive the same attention as crops (31 percent), while the majority of interventions are multi-sectoral in nature (Table 17). TABLE 17:  Prioritization of sectors in the NSDP and in PEA Count of Deviation Sector interventions Share in NSDP Share in PEA (percentage points) Multi-sectoral 39 58% 70% 18 Crops 21 31% 25% –6 Livestock  7 10%  4% –6 Forestry  0  1%  1% 0 Source: Authors, with Government of Lesotho 2012. Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 45 BFP policy coherence analysis Like the NSDP, the 2011/12 and 2012/13–2014/15 BFPs do not come with a costing of policy objectives but are rather in the form of a policy orientation document for individual ministries. They are limited to trien- nial budgetary projections at the ministry level. However, in the 2015/16 to 2017/18 BFP, an improvement can be noted: ministry budget projections are broken down at the program level, bringing this BFP closer to a policy ori- entation document. For instance, for the MAFS, the following programs are budgeted: General Administration, Live- stock Services, Crop Services, Agriculture Research Service, Agricultural Training, Planning & Policy Analysis, and Extension Services Management. Yet these programs actually correspond to MAFS cost centers and are a repackag- ing of administrative layers into programmatic categories. In addition, the program names are descriptive in nature and differ from the MAFS’s policy priority areas stated (but not budgeted) in the BFPs. Policy priority areas in BFPs do not come with a budget; instead, performance indicators were used as proxies of policy priorities. The intervention functional count method was used to compare prioritization of agri- cultural functions in the documents with actual PEA. The priority areas of both the MAFS and the MFRSC were extracted and coded with functional and subsectoral categories. In total, 56 interventions were identified—26 in the 2011/12 BFP, 19 in the 2012/13 to 2014/15 BFP, and 11 in the 2015/16 to 2017/18 BFP. Interventions from the 2012/13 to 2014/15 BFP were duplicated for every year of the period covered by the BFP. The final number of interventions analyzed is thus 94. Finally, the 2011/12 BFP does not include priority policy areas, only performance indicators for each ministry. Most interventions of the BFPs, as is the case for the NSDP, are administrative in nature (23 percent) and involve policy planning, design, and analysis (Table 18). Since the BFP PCA allows to capture MFRSC priorities, the importance of “other infrastructure” (i.e., conservation infrastructure) is better reflected than in the NSDP PCA, and this function has the second highest number of occurrences (20 percent). Interestingly, a great deal of attention is placed on farmer training in the BFPs, with 17 interventions (18 percent of the total). The policy incoherence areas are broadly the same as with the NSDP benchmark, with administrative costs, processing, and marketing ranked lower in PEA than in the BFPs, while input subsidies and extension ser- vices are ranked much higher (Table 19). A notable difference is that the BFPs give more priority to inspection TABLE 18:  Functional count of agricultural interventions of the BFPs, count and percent of total Functional category Count of interventions Share Administrative costs 22 23% Other infrastructure 19 20% Training 17 18% Capital subsidies  9 10% Inspection/quality control  8  9% Extension and advisory services  6  6% Processing and marketing  4  4% Input subsidies  3  3% Research  3  3% Irrigation  2  2% Services subsidies  1  1% Source: Authors, with Ministry of Finance, Planning and Economic Development 2012, 2014, and 2017. 46  n Agriculture Public Expenditure Review TABLE 19:  Deviation between functional share of BFPs intervention count and PEA, in percentage points Functional category BFP count share PEA share Deviation Extension and advisory services  6 28 21 Input subsidies  3 20 17 Other infrastructure 20 25 5 Services subsidies  1  0 –1 Research  3  2 –1 Irrigation  2  0 –2 Processing and marketing  4  0 –4 Capital subsidies 10  5 –5 Inspection/quality control  9  0 –9 Administrative costs 23 14 –10 Training 18  5 –13 Source: Authors, with Ministry of Finance, Planning and Economic Development 2012, 2014, and 2017. and quality control and training than the NSDP—and both were very underfunded over the period. Also, the BFPs give less priority to extension services than the NSDP. Finally, it is worth highlighting that the MFRSC’s PEA appear quite consistent with BFP priority areas—that is, they have a large focus on building conservation infrastructure. The sectoral allocation of PEA is also in line with the sectoral share of BFP interventions, except for a small bias against forestry (Table 20). This again confirms findings from the NSDP PCA—the GoL has favored the subsectors it said it would. TABLE 20:  Prioritization of sectors in the BFPs and in PEA Sector Count of interventions Share in BFPs Share in PEA Deviation (percentage points) Multi-sectoral 57 61% 70% 9 Crops 23 24% 25% 1 Livestock  7  7%  4% –3 Forestry  7  7%  1% –6 Source: Authors, with Ministry of Finance, Planning and Economic Development 2012, 2014, and 2017. ASSESSING PEA TO NUTRITION-SENSITIVE AGRICULTURE AND CLIMATE-SMART AGRICULTURE Allocations to nutrition-sensitive agriculture in PEA The GoL has made multiple national and international commitments related to food security and nutrition. These include: • SDG 2.2: by 2030, end all forms of malnutrition, including achieving, by 2025, the internationally agreed tar- gets on stunting and wasting in children under five years of age, and address the nutritional needs of adoles- cent girls, pregnant and lactating women, and older persons (Zero Hunger Strategic Review Study) • Being an early adopter of the Human Capital Project • Under the NSDP I: malnutrition prevalence target of 30 percent (actual from 39 in 2009 to 33 in 2014) Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 47 • Under the NSDP II: (i) Strengthen Micronutrient Supplementation, Fortification and Diet Diversification, and (ii) Strengthen Nutrition Governance and Capacity Development • Developing a National Multi-Sectoral Nutrition Strategy and Action Plan • Developing the Integrated Program for Agriculture and Food System Development (a NAIP) It is recognized that agriculture plays an important role in achieving these objectives and implementing strategies, not the least in rural communities. Internationally recognized pathways for agricultural interventions toward nutri- tion outcomes—commonly referred to as NSA—are described in Box 7. Activities that fall under the MAFS include technical assistance for establishing kitchen gardens, food preservation demonstrations, and other diet-related activities. Further, envisioned scaling up of biofortified crop production would be the responsibility of the MAFS. The MFRL is in turn responsible for enhancing access to fruits through its fruit tree planting program. PEA to Nutrition Allocations to NSA come from the MAFS and the MFRSC. The MAFS has a separate Nutrition Unit that implements NSA activities through so-called Nutrition Clubs at community levels. These activities include advising communities in setting up keyhole (kitchen) gardens, cooking, and food preservation demonstrations. The unit also coordinates the NSA work in other departments and has, for example, an R&D Desk Officer to coordinate biofortified research. The MAFS employs a total of 54 nutrition staff: 6 centrally in the Nutrition Unit, 10 District Nutrition Officers, and 38 Nutrition Area Technical Officers (NATOs). The MAFS reports that 29 NATO positions are yet to be filled. Other nutrition-sensitive expenditures under the MAFS would include research into biofortified crops and sanitary activ- ities, but such activities were not ongoing in the time period analyzed. BOX 7:  What agricultural activities can be considered nutrition-sensitive? FIGURE B3:  Impact pathway for nutrition-sensitive agriculture Nutritional status IMPACTS Diet Health Food access Care practices Health and sanitation environment On-farm Food Income Women’s Nutrition Natural availability, environment empowerment knowledge resource OUTCOMES diversity, and in markets (time, labor, assets, and norms management safety of food income control) practices Intervention Public sector interventions in agriculture that are considered nutrition-sensitive include those that: (i) increase the on-farm availability, diversity, and safety of food; (ii) support a nutritious food environment in the markets; (iii) increase incomes for those involved in the sector; (iv) support women’s empowerment; (v) enhance nutrition knowledge and norms; and (vi) pro- mote and invest in natural resource management practices (Figure B3). Important is that the purpose of a given intervention is to achieve an intended nutrition outcome. Source: Herforth and Ballard 2016. 48  n Agriculture Public Expenditure Review In the MFRSC, the fruit tree planting program is classified as nutrition-sensitive, as one of its objectives is to increase access to fruits for food and nutrition security purposes.22 The MFRSC’s program for fruit tree plan- tation is considered to be part of the GoL’s NSA activities, as one objective of this program is to enhance the food environment. The annual target for the program is, according to the MFRSC, to plant some 500,000 fruit trees and bushes under a cash-for-work model.23 The fruit tree planting program plants peaches, apples, and vines/grapes, but farmers lack the equipment to raise these trees properly, so their survival rate is about 60–70 percent after the first year when the MFRSC stops managing the new trees. About 2.3 percent of total agriculture expenditures go to NSA activities, but the MFRSC contributes almost twice as much as the MAFS (Figure 26).24 Of total expenditures to NSA, 37 percent comes from the MAFS and 63 percent from the MFRSC. Under the MAFS, only 1.2 percent, or 2.6 million Maloti of its total budget of 215 million Maloti, goes to the Nutrition Unit and related staff. This can be compared with the 64–87 million Maloti that the MAFS spent annually between 2012/13 and 2016/17 on cereal production under the ICP (Box 3). Total operational expenditures for the Nutrition Unit are 0.7 million Maloti, while the total cost for nutrition staff is estimated at 1.9 million Maloti. (Note that staff cost may be slightly underestimated as these are based on the MAFS’s average staff cost, and the nutrition positions are likely to be paid higher wages.) For the MAFS, this does not include R&D to biofortified crops, as it is estimated to be marginal at this point. NSA research will, however, be more important going forward given the scope of APPSA. For the MFRSC, 3 percent, or 4.5 million Maloti of the MFRSC’s total budget of 118.5 million Maloti, goes to the fruit tree planting program (2.6 million Maloti for procurement of fruit trees and 1.8 million Maloti for actual planting of trees). FIGURE 26:  Nutrition-sensitive agriculture as share of total public expenditures to agriculture, 2015/16, in constant 2010 Maloti MAFS MFLR 0.85% 1.45% Other PEA 97.7% Source: Lesotho BOOST 2018; MAFS 2019; MFRSC 2019.25 22 At the same time, the fruit tree planting program can also be considered climate-smart, as trees—fruit trees and agroforestry trees—are known to reduce land degradation and soil erosion and thus enhance agro-ecosystem resilience, and to sequester carbon through biomass growth. In some instances, the establishment of certain agroforestry trees and trees on farms also support agricultural productivity growth. Thus, trees support the achievement of the CSA triple-win objectives (see the next section). 23 Consultations with the MFRSC Department of Forestry. 24 All figures are for budget year 2015/16, 2010 constant values. 25 Several posts of NSA expenditures fall outside the budget. Note that these figures do not include the Ministry of Education’s Farm to School Program, under which produce for school meals is procured in the local community. Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 49 Geographically, the number of NATOs (filled and unfilled) are not allocated according to malnutrition chal- lenges in the districts. For every district, six or seven NATOs have been planned for, with the exception of Thaba Tseka, where eight positions have been planned for, and Berea, where five positions have been planned for. The number of Nutrition Clubs is also not established according to the share or number of stunted children in the dis- trict. Comparing NATOs and Nutrition Clubs with allocations of general PEA or the number of fields planted as per Table 21 of this report does not reveal any specific pattern either. It is clear that the districts with the highest number or share of stunted children and those with the highest numbers of Nutrition Clubs do not overlap. This finding is likely to also apply to other forms of malnutrition. This means that there may be a choice to be made on where the levels of efficiency from spending are the highest. TABLE 21:  Overview of NSA activities per district Number District Nutrition NATO positions of Nutrition District Ranking—stunting Officer # of NATOs to be filled Clubs/Groups Maseru 1 no stunted ✓ 4 3 19 Qacha`s Nek ✓ 0 6  8 Thaba Tseka 2 share stunted ✓ 6 2 — Berea ✓ 5 — 18 Mokhotlong 1 share stunted ✓ 2 4  6 Butha- Buthe 3 share stunted ✓ 4 3 21 Mohale`s Hoek 3 no stunted ✓ 4 3 32 Mafeteng ✓ 5 2  8 Leribe 2 no stunted 0 5 2 — Quthing 3 4 Share of stunted District children (mean) Butha-Buthe 39.4% Leribe 31.1% Berea 31.4% Maseru 31.2% Mafeteng 24.6% Mohale’s Hoek 38.2% Quthing 35.6% Qacha’s Nek 32.1% Mokhotlong 48.4% Thaba Tseka 39.8% Source: World Bank 2018, unpublished. 50  n Agriculture Public Expenditure Review Allocations to climate-smart agriculture through donor-funded programs Climate-smart agriculture (CSA) reflects an ambition to improve the integration of agricultural develop- ment and climate responsiveness. CSA initiatives aim at addressing three pillars: sustainably increasing agricul- tural productivity to support equitable increases in incomes, food security, and development; adapting and building resilience to climate change; and reducing or removing GHG emissions from agriculture. Thereby, practitioners are expected to be cognizant about tradeoffs and synergies between these three pillars. Many CSA practices already exist worldwide and are used by farmers to cope with various production risks. Promising CSA practices already in use or promoted in Lesotho include: conservation agriculture practices, including minimum soil disturbance, crop rotation, permanent soil cover, and improved varieties; mulching; grassland restoration, conservation, and rota- tional grazing; adopting improved varieties, crop rotation, and intercropping; improved application of manure and fertilizer; and efficient irrigation systems and water management practices (CIAT and World Bank 2018). Recent studies demonstrate the benefit of CSA to farmers and its potential to support Lesotho’s climate com- mitments. The Lesotho’s Climate-Smart Agriculture Investment Plan (World Bank 2019b) shows the positive impact of adopting CSA on farmers’ profitability compared to conventional practices. The annual net income of climate-smart crop and livestock production for a representative farming household is about three to five times higher than that earned using conventional practices. CSA practices have notable public good benefits and were found to reduce soil erosion and generate carbon sequestration. For crops, the adoption of CSA practices reduced greenhouse gas (GHG) emissions per ton of produce between 1.1 tCO2eq and 20.4 tCO2eq; switching to climate-smart livestock practices led to a decline in livestock emission intensity, ranging from 1 percent for sheep and goats to 37 percent for poultry. These findings show the large potential of CSA to contribute to climate adaptation and mitigation commitments. However, the adoption of key CSA practices is rather low, often due to lack of access to finance and knowl- edge. Among significant obstacles to higher rates of adoption and retention are that CSA practices tend to be labo- rious, and farmers have poor access to critical labor-saving equipment such as jab planters, direct seeders, and rippers, as well as limited knowledge and capacity. As CSA practices, particularly those regarding soil health, rarely provide immediate benefits, there is a risk that farmers adopt and then quickly abandon them (CIAT and World Bank 2018). World Bank (2019b) suggests a range of investments that could promote CSA adoption in the future, including the promotion of on- and off-farm investment in hydraulic infrastructure (e.g., water harvesting, mod- ernization of infrastructure, implementation of sustainable water management practices); or the development of market hubs and agricultural clusters to strengthen market access. Given the potential benefits of CSA, the question arises of the extent to which agriculture expenditures were dedicated to promoting CSA in the past years. An understanding of the extent of CSA and its potential benefits will support the GoL to enhance resource allocation to CSA. Public financing of the CSA agenda is crucial as it sets the stage for the necessary shifts in investments and practices in the sector. In Lesotho, CSA-related public expendi- tures are not tagged or monitored through BOOST. Lesotho’s public expenditures do not account for expenditures for climate adaptation or mitigation finance, as is the case in other countries (e.g., Philippines, Bangladesh). Con- versations with government officials made it evident that most CSA expenditures are linked to donor programs. The assessment thus focused on development partners’ recent expenditures, which are outside of BOOST. However, within government expenditures, one notable public expenditure on CSA practices is support for fruit trees and agroforestry, as outlined in the previous section. The assessment of CSA-related public expenditures relied on expert consultation, a literature review, and the Multilateral Development Banks (MDB) Climate Finance Tracking Methodology. Discussions with stakeholders identified relevant donor projects aimed at strengthening climate adaptation, mitigation, and resilience building. The literature research identified typical CSA practices relevant for Lesotho that would be counted as CSA-related expenditure. The CIAT and World Bank (2018) Climate-Smart Agriculture Country Profile served as a key resource. The share of CSA-related expenditures in these projects was determined using the project costing and expenditures files. Based on the financing for CSA, a share of financing for project management was attributed to supporting the Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 51 BOX 8:  Overview of the climate finance tracking methodology of the multilateral development banks In the scope of the multilateral development banks (MDB) methodology, climate finance refers to the amounts committed by MDBs to finance climate change mitigation and adaptation activities in development projects. The methodology provides guid- ance to track adaptation finance and mitigation finance. In agriculture lending projects, both adaptation and mitigation finance are often related to the promotion of CSA practices that explicitly aim at enhancing productivity and resilience and reducing/ removing GHG emissions. Climate Adaptation Finance: This type of finance aims at lowering the current and expected risk or vulnerabilities posed by climate change. For activities to count toward adaptation finance, they must fulfil the following: (i) the project must set out the vulnerability context of the target area and beneficiaries; (ii) make an explicit statement of the intent to address climate vulner- abilities through the specified activity; and (iii) articulate a clear and direct link between the climate vulnerability context and the specific project activity. Climate Mitigation Finance: Climate change mitigation promotes efforts to reduce, limit, or sequester GHG emissions to reduce the risk of climate change. But not all activities that reduce GHGs are eligible to be counted toward MDB mitigation finance. Miti- gation finance is based on a list of activities compatible with low-emission pathways. For the agriculture sector this list includes, for instance: reduction in energy use in traction (such as efficient tillage), irrigation, and other agricultural processes; agricul- tural projects that improve existing carbon pools (such as rangeland management, collection and use of bagasse, rice husks, or other agricultural waste, reduced tillage techniques that increase carbon contents of soil, rehabilitation of degraded lands, peatland restoration, and so on); reduction of non-CO2 GHG emissions from agricultural practices and technologies; and live- stock projects that reduce methane or other GHG emissions (for example, manure management with biodigesters, and improved feeding practices to reduce methane emissions). Source: AfDB et al. 2016. promotion of CSA, and prorated. This process builds on, but is less rigorous than, the 2016 Climate Finance Tracking Methodology of the Multilateral Development Banks (MDBs) (AfDB et al. 2016) (Box 8). In contrast to the MDB meth- odology, CSA-related financing—which may serve climate adaptation and/or mitigation in agriculture—is accounted for whenever the term “climate-smart agriculture” or improved practices that have features of CSA are mentioned, irrespective of the project’s vulnerability context or whether a clear link between climate vulnerability and the activ- ity is articulated, or whether the specific wording to count as mitigation finance is used. This AgPER identified four development partners that promoted CSA practices and resilience-building measures through one or a series of projects in the timeframe covered (Table 22). All values are reported in constant 2010 USD values. (For nominal values, see Annex 7.) These are: • The World Bank and IFAD-funded SADP, approved in 2011, with a total allocation of US$19 million. The proj- ect was expected to close in 2018; due to its success the closing date was extended to 2020 and the project received additional financing (AF) of US$7 million from the World Bank;26 • Several grant projects supported by FAO: • Emergency response to the El Niño induced drought in Lesotho from November 2016 to June 2018, US$0.81 million;27 • Increase in resilience of Lesotho food and nutrition security through upscale of Climate-Smart Agricultural and Functional Land Resources Database from 2014 to 2016, US$0.5 million;28 26 The project is cofinanced by the World Bank, IFAD, as well as beneficiary and government contributions; at project approval the resource envelope for the project was US$20 million, including US$10 million in IDA financing and US$10 million in IFAD financing. The IDA financing is provided in the form of a credit; 50 percent of the IFAD financing is provided in the form of a grant and 50 percent in the form of a loan. The remaining US$4.46 million are beneficiary and government contributions. Values in current 2011 USD. The Additional Finance is supported by World Bank with US$10 million and by IFAD with US$4.3 million in current 2017 USD. 27 This project was supported by the GoL through the World Bank, which contributed US$1.1 million to the FAO project UTF/LES/054/LES, entitled “Emergency response to the El Niño-induced drought in Lesotho.” In current US$ 2016 values. 28 Financial contributions came from the European Commission Directorate General Humanitarian Aid and Civil Protection (ECHO), and the Swiss Development Cooperation (SDC) contributed US$210,000 (current 2016 values). 52  n Agriculture Public Expenditure Review TABLE 22:  Overview of key projects that supported adoption of CSA practices (deflated to constant 2010 US$) Share of Total budget/ CSA-related Absolute CSA-related Project disbursed amount expenditures expenditures IFAD: US$30.7 million (allocated) US$9.2 million (allocated) Wool and Mohair Promotion US$3.6 million (disbursed, 12/2018) US$1.1 million (disbursed) 30% Project (2015–2021) World Bank/IFAD: SADP (2011–2020) US$26 million (allocated) 26% US$7.4 million (allocated) US$13.9 million (disbursed) US$3.8 million (disbursed) FAO: US$0.8 million (reallocated from US$2.1 million (allocated and SADP) disbursed) Emergency Response (two in US$0.7 million 2016) 74% Rapid Response to Drought (2016) US$0.8 million Resilience Building US$0.5million (2014–2016) US$61 million (allocated) Sum (allocated): US$19 million Average: 43% US$20 million (disbursed) Sum (disbursed): US$7 million Note: The names of FAO projects are abbreviated; all values are reported in constant 2010 US$ and are in line with the remainder of the report. Disbursements of the IFAD Wool and Mohair Promotion Project (WAPP) and World Bank/IFAD SADP are accounted for until December 2018 and March 2019, respectively. The amount dedicated to CSA was calculated by using the share of CSA finance for the entire project, applied to the disbursed funds. • FAO emergency response to El Niño drought in Lesotho (2016), US$0.7 million;29 • Lesotho rapid response to drought, which took place in 2016, US$0.83 million.30 • IFAD’s Wool and Mohair Promotion Project, with total financing of US$30.7 million31 between 2015 and 2021; and • GEF-UNDP’s project “Reducing Vulnerability from Climate Change in the Foothills, Lowlands, and the Lower Senqu River Basin.”32 FAO’s programs responded to an emergency situation in 2016, but within each program promoted resilience-­ building and CSA measures that are expected to support agriculture sector development in the longer term. The section below analyzes the share of funding dedicated to promotion of CSA and resilience-building practices. The received allocated and disbursed values do not correspond with the OECD’s CRS database. According to the CRS database (2019), total donor funding for 2011 to 2016 was 36.6 million Maloti. The values received for the analysis below do not match those in the CRS database, as international and multilateral organizations often do not report rigorously to OECD. The absolute values should thus be interpreted with caution, with a focus instead on the share of CSA-related expenditures. The World Bank and IFAD SADP project supports CSA-related expenditures with 26 percent of total financ- ing, thus promoting CSA practices with approximately US$7.4 million. The World Bank and IFAD-financed SADP, 29 While the proposal is funded by ECHO with €1 million, FAO contributed €36,548 from its Special Fund for Emergency Rehabilitation and Activities (SFERA) (current 2016 values). 30 The funding came from United Nations Central Emergency Response Fund (CERF). 31 The financing will stem from different sources, with 14.9 percent from an IFAD grant as well as an IFAD loan; 12 percent from OPEC’s Fund for International Development; 18 percent from the Adaptation for Smallholder Agriculture Programme Trust Fund, which is managed by IFAD; 4 percent from the Lesotho Wool and Mohair Growers Association; 10.2 percent from the GoL; and the remaining 7.3 percent from other financiers. 32 The team was in contact with UNDP, but the actual figures were not available at the time of writing the report. Chapter 4: Agriculture Public Expenditures: How Much Is Allocated and Where Does It Go? n 53 approved for the period 2011–2020, has a total allocation of US$26 million constant 2010 US$ (including additional financing). The project aims to increase marketed output among project beneficiaries in Lesotho’s smallholder agriculture sector. The project has three components: Component 1, Increasing Agricultural Market Opportunities; Component 2, Increasing Market-oriented Smallholder Production, wherein subcomponent 2.1 deals with Prepa- ration and Implementation of Agricultural Investment Plans and subcomponent 2.2 with distribution of improved Technology Packages for Smallholders; and Component 3, Project Management. CSA practices are promoted in sub- component 2.1.; for a successful promotion of CSA activities, project management (Component 3) is crucial and the financing therefore prorated. The share of CSA-related expenditures in total financing is thus 26 percent. FAO supported CSA-related expenditures with 74 percent of total funding, resulting in US$2.1million. Three out of four projects responded to the El Niño occurrence in 2015/16, which highlighted the prevailing vul- nerabilities in Lesotho. While the projects were part of an emergency response, they promoted climate-smart and ­ resilience-building strategies that are relevant for long-term agriculture sector development and therefore consid- ered in this analysis. Three out of four projects aimed at enhancing agricultural productive assets through support of CSA practices and inputs. The remaining funding was targeted to social protection, dietary, and nutrition measures. For instance, the Resilience of Lesotho Food and Nutrition Security aimed at upscaling climate-smart agricultural and land use information by expanding the capacity development efforts initiated by FAO and the GoL during the Emergency and Resilience Programme (ERP) Response to the 2012 drought. Several public and private sector stake- holders were trained on improved agricultural techniques that mitigate the impact of climate change. FAO designed a Drought Response Plan (2016–2017) that laid the foundation for upscaling CSA in Lesotho. The overall share of CSA-related expenditures in total FAO expenditures was assessed at 74 percent. These programs were initiated as response to El Niño—while financing long-term agricultural development strategies. Thus, US$2.1 million of CSA- related donor financing, or 30 percent of the total disbursed amount, were releases as a response to an extreme event. IFAD’s Wool and Mohair Promotion Project’s share of CSA-related expenditures is 30 percent, expected to result in US$9.2 million for CSA promotion at the end of the project. Out of this amount, the project has disbursed US$1.6 million for CSA to date. The project aims to boost the economic and climate resilience of poor, smallholder wool and mohair producers to adverse effects of climate change in the mountain and foothill regions of Lesotho. It has three components. Component A aims at Climate-Smart Rangeland Management, with subcomponents on Effective Information on Climate-Smart Rangeland Management and Climate-Smart Participatory Rangeland Man- agement. Component B aims at Improved Livestock Production and Management; its three subcomponents aim at Improving Livestock Nutrition, Improving Livestock Breeding, and Improving Livestock Health. The subcompo- nent on Livestock Nutrition supports improved fodder production, intercropping, promotion of crop residues, and research on improved species under various climate conditions. These activities thus qualify as CSA-related activ- ities. Component C aims at Wool and Mohair Fibre Handling and Marketing. A last component deals with project coordination. The assessment shows that out of US$30.7 million, 30 percent of funding is dedicated to CSA-related activities in the project timeframe 2015–2021. On average, the share of funding toward CSA-related activities can be considered modest, at 43 percent, or up to US$19 million at project completion, while the potential returns on such investments can be substantial. Investing in CSA to build resilience in the sector and mitigate losses in the sector can potentially have high returns. Table 22 provides a summary of projects, the initial allocation, and disbursements to date. In conclusion, donor projects support the promotion of CSA with 43 percent of their funding on average, resulting in about US$7 million disbursed for CSA in the last years, and expected to result in US$19 million when all funds are disbursed at the end of the respective projects. These expenditures are directly contributing toward meeting Lesotho NDC’s commit- ments, which explicitly mentions the introduction of climate-smart agriculture practices (e.g., conservation agricul- ture, agroforestry, drought and heat-tolerant varieties) as promoted in reviewed projects as promising. However, the assessment has caveats. For example, CSA-related expenditures are accounted for when improved management practices that aim at increasing productivity and production are mentioned, as long as climate change and hazards are identified in project documents. It is expected that more recent project documents provide a clearer narrative on climate change, adaptation, mitigation, and CSA and may demonstrate higher levels of CSA-related finance. CHAPTER 5: Conclusions and Recommendations: How to Improve the Quality of PEA in Lesotho This chapter discusses the findings of the analysis in Chapter 4 and provides recommendations for the GoL. As per the objective of this report, these recommendations are intended to improve the quality of PEA based on the GoL’s priorities for the sector in the context of a constrained fiscal space. At the end of the chapter, a table lists the recommendations with an indication of their respective impacts, cost, and timeframe for proposed actions and reforms. 54 Chapter 5: Conclusions and Recommendations: How to Improve the Quality of PEA in Lesotho n 55 Key Points in This Chapter • Better linking PEA with strategic objectives for the sector under the NSDP II will be crucial to achieve the objectives set for the sector, as the current allocation of expenditures is not balanced in a way that helps address its challenges. It is also rec- ommended that the MAFS—in consultation with the MFRSC, and as part of the NSDP II—continue to develop and adopt an evidence-based agricultural policy strategy with a NAIP for implementing the policy. • Continuing to update AgBOOST and refine classification would help analysts from the MAFS, the MFRSC, the MF, and other bodies evaluate PEA’s allocation and impacts. However, technical staff in the MAFS and the MFRSC Policy Department/Unit would need training on how to use AgBOOST in their day-to-day work. • Making reporting of PEA more consistent across different ministries would ensure that the analyses are comprehensive. Including functional classifications in BOOST would further allow for more detailed analysis. Including donor financing in the budget process would provide a more complete overview of actual funding to the sector and alignment of PEA against set objectives. • Establishing a feedback loop system with rigorous results indicators would be important going forward so that the impacts of the different programs can be evaluated, and especially those that absorb a large share of resources. This would include establishing a Joint Sector Review (JSR) to coordinate activities in the agriculture sector across ministries, other government agencies, donors, and nongovernmental actors. • Where possible, it is recommended that the GoL provide public rather than private goods and services and strive for a better balance between development and recurrent spending. It is not clear that the GoL has found its optimal role in the agriculture sector at this point, and it would be good to better assess current justifications for large programs of private nature, such as the ICP. Rebalancing the budget toward more development expenditures includes allocating more PEA to capital investments, which currently are low. Similarly, allocations to R&D need to be increased if the sector is to meet the GoL’s aspirations for a competitive sector that is a key source of employment. • It is recommended that MAFS and MFRSC cost centers be better balanced, especially given the structure of the sector. For example, the MAFS could consider balancing the budget more toward the Livestock, Research, and Planning Departments, for which funding is currently low given the relative contribution of the sectors and their emphasis in the NSDP. • Similarly, balancing personnel costs and capital investments against the necessary operating costs would be important to effectively implement certain programs in the sector and to ensure sustainability of, e.g., infrastructure investment. • Assessing and justifying allocations to CSA and NSA would be important given that current allocations to both areas are seem- ingly low as a share of total PEA, given the scope of climate change impacts and of malnutrition in the country. Objectives and targets for these programs should be established and resources allocated accordingly. IMPLICATIONS OF THE REVIEW’S FINDINGS Looking back at the original research questions for this AgPER, it is clear that while room may not exist to increase the share of PEA in the current fiscal environment, the quality of PEA for Lesotho can be improved. The share of PEA in terms of total public expenditures—less than 3 percent over the review period—is well below the 10 percent target that the GoL committed to under the CAADP. Similarly, it is below the contribution of the sector to GDP and arguably miniscule compared to the share of Lesotho’s population that depends on agriculture. Given the current fiscal constraints, it is unlikely that government allocations will increase significantly; however, ample room remains to improve the quality of PEA. Perhaps the most significant finding of this AgPER is how poorly PEA are aligned with policy objectives for the sector. As seen in Chapter 4, PEA in Lesotho are not aligned with agricultural objectives in the NSDP; instead, PEA are largely spent on recurrent expenditures that are unlikely to contribute to growth and longer term develop- ment. Public expenditures are perhaps the most important instrument that a government has available for imple- menting policies for the sector, but if expenditures are not aligned with sectoral objectives, they will not contribute to achieving set results. This is especially important in a resource-constrained environment. Box 9 provides insights into how agricultural policies, budgeting, and monitoring and evaluation should be linked to achieve intended results in the sector. 56  n Agriculture Public Expenditure Review BOX 9:  Linking PEA to agricultural policy objectives Any budget process should be predictable and transparent FIGURE B4:  Policy and expenditure cycle in terms of allocations to priorities, and reporting on spend- ing, outcomes, and other indicators against set objectives should feed back into the process. Public expenditures are arguably the most important policy instrument a government has, and the budget process should be an integral part of agri- cultural policy making. Results-based budgeting requires a feed- Evaluation of Sector needs back loop from sector needs assessments, strategic priorities, results assessment allocations, implementations, and results evaluation (Figure B4). The basis for this type of budgeting is clearly defined priorities for the sector according to which expenditures are allocated. Fur- ther, monitoring and evaluation of activities and results should be transparent for stakeholders. Preferably, the process should Prioritized programs Implementation and investments have mechanisms for stakeholder feedback and consultations. In line with this, the African Union and CAADP recommend mem- bers to (i) develop a multiyear Agriculture Investment Plan with a results framework, to guide allocations of agriculture expen- Allocations of ditures; and (ii) ensure an annual Joint Sector Review of imple- expenditures mentation and results for stakeholders. The composition and balance in PEA are not in line with the support needed for the sector, given its chal- lenges. Agricultural productivity is low in Lesotho due to a number of factors that are not adequately addressed with current PEA allocations. As outlined in Chapter 2, these include limited access to irrigation, low uptake of new technologies and inputs, poor-quality extension and advisory services, and limited access to credit. Environmental challenges such as land and natural resource degradation, erosion, and low soil fertility further contribute to the low productivity, rendering the sector vulnerable to climate change impacts and other production risks, with high losses for producers and the need for a costly emergency response from the public sector. Malnutrition in the form of high levels of stunting among children, micronutrient deficiency across the population, and high overweight and obesity especially among adult women, have disclosed a malfunctioning food system, whereby Lesotho’s agricul- ture sector is currently unable to address the existing deficit of highly nutritious food. However, PEA have been skewed toward recurrent rather than development expenditures, with high levels of spending on goods of private nature, and with low investment in infrastructure and R&D. A large share of PEA have gone to wages, including for extension agents, but to the extent that the wage bill has crowded out necessary operating costs to provide relevant services. Activities to address some of the key challenges, such as climate change and nutrition, have seen limited funding. While donors, including the World Bank, over the next years are providing significant funding to many of these areas, donor funding is not included in the budget and is not reported on, nor do plans exist for allocating PEA to these areas over time as donors phase out. To improve the quality of PEA and to ensure that PEA are prioritized and sustainable in the medium and long term, there are several steps that the GoL can take. Chapter 5: Conclusions and Recommendations: How to Improve the Quality of PEA in Lesotho n 57 RECOMMENDATIONS Adjusting PEA level and composition to address challenges and achieve set objectives for the sector General forestry and agriculture For MAFS PEA should focus on public rather than private goods and services to avoid crowding out the private sector. Although it varies from year to year, around 30 percent of MAFS expenditures go to private goods, largely the ICP. Evaluating the ICP to understand the impacts on the sector and if government involvement crowds out private actors in this field would be important, especially given how much the program absorbs of total PEA. If impacts are limited and do not address the supposed market failure for input use in the agriculture sector, it would be important to develop an exit plan for this program. Other ways to strengthen the program and avoid crowding out the private sector are through clear and transparent targeting of farming segments in the sector where market failures exist and introducing an (e-)voucher system for distribution of the subsidy. It would also be useful for both government- and donor-financed allocations to agriculture to assess where exactly Lesotho is placed on the MFD ladder to guide public support to the sector. For MAFS and MFRSC A rebalancing of development versus recurrent funding would be useful. For the MFRSC, this entails evaluat- ing the balance between spending on the Fato Fato program—which has a strong social safety net dimension—and other more cost-effective conservation infrastructure expenditures. For the MAFS it would be important to assess the rationale behind the functional balance of the budget that is heavily skewed toward input subsidies against infrastructure and knowledge production. Even though APPSA will allocate significant funding to R&D over the next five years, it would be important to start scaling up allocations to agricultural research over time as APPSA approaches its closing date in 2024. Similarly, funding for resilience-building activities is much needed, such as irrigation infrastructure and CSA. For the next years these will receive funding under SADP-2, but after the project closes, such investment needs to be better financed by domestic PEA. Further, as part of an agricultural policy strat- egy under the NSDP II, the GoL should reassess the rationale behind the ICP and evaluate its impact on agricultural production. Instead, the MF may consider increasing allocations to MSBDCM to support agricultural marketing and processing activities to foster demand-driven growth in the sector. Any decision to make investments in private goods should be based on an assessment of existing market failures according to the MFD ladder. For MAFS and MFRSC While it is recommended that the share going to capital investments be increased as a share of the budget, it is important to allow for sufficient allocations for operations and maintenance. The low levels of operating costs are problematic not only for extension services and the livestock department, but also for operating and main- taining capital investments in the sector. Importantly, several donor-financed programs finance such investments, but there is limited allocation in the budget for operations and maintenance. As these investment projects approach their closing date, it is especially important to make sure that budget lines are in place and provisions made for these costs are in the budget. For MAFS and MFRSC A more balanced allocation of staff versus operating costs under extension services and other departments would enable programs to support the sector effectively. This AgPER found that as much as 91 percent of MAFS DOs expenditures went to personal emoluments, which reportedly prevents staff from effectively carrying out their work and implementing the program. Given that Lesotho has a relatively high extension worker/farmer, room may remain to decrease the number of extension workers over time as people retire or leave for other employment. At the same time, the number of support staff in the MAFS and MFRSC is relatively high compared with technical levels, and there may be room for rebalancing the staffing in the ministries for more effective implementation of programs according to the GoL’s objectives. A review of government staffing, functions, and grades has been conducted under 58  n Agriculture Public Expenditure Review the Public Sector Modernisation Project (PSMP) spearheaded by the Ministry of the Public Service (MPS), and this can be built on with a specific functional review for MAFS and MFRSC to better balance staff costs with other needs. Another option is to shift resources over time from the ICP to increase operating costs. For MFRSC It is recommended that the very strong concentration of MFRSC expenditures in the conservation cost cen- ter is assessed against the objectives for the sector. The MFRSC places an emphasis on the conservation cost center through Fato Fato; an evaluation of the program could be useful to determine its impacts and the rationale for this imbalance in allocations under the ministry. It is also recommended that the tree planting program be eval- uated to measure the extent to which it achieves its objectives, and a plan developed to ensure higher survival of trees. Further, properly coordinating the program with the MAFS’s Nutrition Unit would ensure that the food secu- rity objectives of the fruit tree planting program are being achieved. For MAFS In a similar fashion, it is recommended that the allocation of expenditures between MAFS cost centers is assessed against the objectives for the agricultural sector. The MAFS could consider balancing the budget more toward the livestock, research, and planning departments. All three departments are important to support the development of Lesotho’s agriculture sector. The livestock department appears currently underfunded in regard to the sector’s contribution to AgGDP, while the research department has a crucial role to play in improving agricul- tural productivity in the long run. It can also be noted that even with high allocations to extension services, advice to the sector will not be efficient unless support to extension services are balanced with investments in agricultural R&D so that advice provided to farmers reflects new technology. The planning department could support general improvements in budget efficiency, which is all the more important given the constrained fiscal environment that Lesotho and the MAFS face. Nutrition and CSA sectors For MAFS and MFRSC The responsibilities and expected results of the MAFS’s and MFRSC’s NSA activities would benefit from being better defined and budget allocations made accordingly, while at the same time aligning allocations geographically according to levels of malnutrition. PEA to NSA activities under the MAFS and MFRSC are low and seemingly not in line with what is required of the agrofood sector to meet the GoL’s nutrition objectives and international commitments. Given the needs to improve nutrition in Lesotho and the potential returns from such investments on the overall economy, reallocating funds from ICP, that largely support cereal production, toward pro- duction of more nutritious foods, could be considered. However, the exact mandate of the Nutrition Unit and how it relates to other nutrition services, e.g., under the Ministry of Health, are not clear, making it difficult to evaluate the unit’s needs. Similarly, neither the expected nor the actual impacts from the MFRSC’s fruit tree investment program, are identified in any detail, and the program does not seem to be linked to other nutrition programs in the country. Better allocating resources according to the nature and levels of malnutrition around the country, and better coor- dinating activities with those of other agencies, would nevertheless be important to strengthen the quality of these expenditures. Note that the MAFS will receive funding for NSA activities under SADP-2 in the period 2019–2025. For MAFS and MFRSC Scaled-up allocations to CSA would benefit the sector. About 30 percent of disbursed CSA-related finance stems from programs initiated in response to the 2015/2016 El Niño drought. Given the scope of the expected benefits of adopting CSA under critical climate conditions as well as regular conditions, the funding of CSA—an average 43 per- cent of donor expenditures, which may result in US$19 million (in constant 2010 US$) upon project ­ completion— seems modest. It is therefore recommended that a larger share of funds be allocated to CSA-related activities, and that CSA-related expenditures are tracked in the regular government budget. This would support achievement of the ambitious agriculture sector policy goals for CSA, climate adaptation, and mitigation. Under SADP-2, an addi- tional US$26 million (or US$17 million in constant 2010 USD) will be allocated between 2019 and 2025 to activities that will support CSA in the sector. To fully assess the scope of what the GoL spends on risk management in terms of Chapter 5: Conclusions and Recommendations: How to Improve the Quality of PEA in Lesotho n 59 impact mitigating investments as well as emergency response and other coping mechanisms after agricultural risk events occur, and how those compare with the actual impacts of these risks, GoL may also consider conducting an analysis of the Fiscal Implications of Agricultural Risks according to the World Bank methodology. Better planning PEA: Establishing policy-based expenditure allocations All sectors For MAFS and MFRSC Going forward, strengthened links across ministries between policy and allocations of PEA will be critical. Better linking PEA with strategic objectives for the sector under the NSDP II and other strategic documents will be crucial to achieve the GoL’s objectives for the sector. For MAFS It is recommended that the MAFS—in consultation with the MFRSC, and as part of the NSDP II—continues to develop and adopt an evidence-based agricultural policy strategy with a NAIP for implementing the pol- icy. Such an evidence-based agricultural policy strategy would include analyses of priority sectors for agricultural development and their needs for public investment. Note that during the work for this AgPER, a series of value chain analyses have been conducted with the support of FAO for the purpose of developing a NAIP, and a draft NAIP has been produced, although it is not yet validated for compliance with the Malabo Declaration (2014). It is recommended that this work continue. Another step would be to further improve the BFPs by introducing more fine-tuned, program-based budgeting. Better tracking PEA All sectors For MAFS and MFRSC Continuing to update AgBOOST and refining classification would help analysts from the GoL and other bod- ies evaluate the allocation and impacts of PEA. It is recommended that AgBOOST become a permanent tool for the MAFS and the MFRSC Planning and Policy Department/Unit. While the cost of doing this is low, especially compared with the scope of the budget, relevant staff would need training to be able to maintain and use AgBOOST. Such training would include strategic budgeting, results monitoring, and budget analysis. It is also recommended that the MAFS’s financial controller, who is highly familiar with the budget and expenditure data, support the policy department more closely and participate actively in the training of analysts. For MAFS, MFRSC, and MF Including functional classifications in AgBOOST would help track expenditures and analyze impacts. This would enable tracking, e.g., expenditures for adaptation, CSA, resilience building, and climate mitigation. The MAFS and the MFRSC could provide a more accurate overview of the share of recurrent versus development support to the sector by classifying ICP and WMP expenditures as “recurrent.” For the purpose of assessing allocations and ultimately impacts, it would also be important to clearly identify input subsidies, and where applicable, report them as separate items in the budget. Finally, the MF and the MAFS could consider revising the budget classification for years before 2012/13 so that input subsidies appear more clearly in the IFMIS. For MF and all line ministries Ensuring that financial reporting is consistent under the different ministries would help in analyzing and reporting PEA. The assessment shows that inconsistencies arise in how the relevant ministries report PEA, which makes it difficult to compare allocations. An example is the reporting for DOs under the MAFS and the MFRSC: DOs are reported under different cost centers for the former but under one department for the latter. Going forward, applying uniform classification of expenditures across all line ministries can also help in getting a more compre- hensive overview of expenditures of a cross-sectoral nature, such as expenditures allocated to nutrition and agri- cultural risk management, etc. Addressing inconsistencies in how costs are reported is therefore recommended. 60  n Agriculture Public Expenditure Review For MAFS, MFRSC, and MF Reporting donor funding both in the budget and in actual expenditures would better capture the full scope of PEA. Doing so will more accurately reflect the size and nature of PEA in Lesotho, as well as improve predictabil- ity, accountability, and transparency in the allocation of resources. Further, it will allow for longer term planning and help the GoL with linking investments to operations and maintenance. Currently, most lines for donor funding (although not all) exist in AgBOOST, but they do not report their budget value. Interviews with MFRSC staff suggest that there are plans to report donor financing in the IFMIS from FY19/20. However, the GoL’s financial management is still not complying with the fiduciary requirements of some donors, and all funds can therefore not be reported on the GoL systems. Until GoL’s budget systems comply with donors’ fiduciary requirements, MAFS and MFRSC may consider including donor funds in their internal planning and budget preparation, to get a better overview of the overall budget for their programs. Better evaluating PEA’s impact All sectors For MAFS and MFRSC Establishing a feedback loop system with rigorous results indicators for evaluating the impacts of the dif- ferent programs would ensure more optimal use of resources. This is especially needed for those that absorb a large share of resources, like the extension services, the tree planting program, Fato Fato, and the ICP. Consultations with ministry staff indicate that despite high levels of funding for these programs, they face challenges in imple- mentation that may impact results. For example, even though afforestation is needed to prevent costly soil erosion around Lesotho, and significant resources are spent on the tree planting program, interviews indicate that most trees do not survive after the first year. Similarly, Lesotho has a high number of extension workers per farmer, but operating costs are too low to effectively reach out to all communities on a regular basis. The same applies to the Nutrition Unit, where NATOs work mainly with those communities that are in proximity to the DOs, as operating costs do not allow them to travel to more distant communities. And as discussed in Chapter 4, any impacts under the ICP are temporary and do not seem to have increased demand for inputs in any meaningful manner, while cereal yields in Lesotho remain below regional averages and the figures set in the GoL’s objectives. At the same time, the size of the program prevents allocation of public resources to investments with higher returns, or to operating costs to extension services, for example. Establishing a better feedback loop from budgeting through program implemen- tation to evaluation and into the next budgeting cycle will ensure that these scarce resources are used optimally. For MAFS and MFRSC CAADP requirements for a JSR could be met by incorporating this in the new NAIP. Lesotho has signed on to the Malabo Declaration and is in the process of compiling its first report on PEA. However, Lesotho has not set up a JSR, even though it is a requirement under CAADP. A JSR is an institutional structure to coordinate and monitor financing and activities in agriculture across ministries and donors. Such a structure consists of sector and techni- cal committees that meet regularly around the year to coordinate. A policy forum is organized annually to present investments and progress under the NAIP, in which private sector and academia participate along with government agencies and donors. Note that this was recommended by one of the participants in the AgPER workshop. Box 10 describes how the JSR functions in Malawi. Table 23 provides an overview of expected impacts, assessed relative costs, and a proposed implementation time- frame for the recommendations proposed in this report. Chapter 5: Conclusions and Recommendations: How to Improve the Quality of PEA in Lesotho n 61 BOX 10:  Malawi’s Joint Sector Review: Bringing together stakeholders for a broad rural development agenda The Joint Sector Review (JSR) framework has been in place in Malawi since the signing of the CAADP Compact in 2010.33 The Ministry of Agriculture, Irrigation and Water Development (MoAIWD) is the convener of JSR meetings, and invites other gov- ernment ministries, private sector actors, civil society organizations, leading farmer organizations, development partners, aca- demic and research institutions, NGOs, and other stakeholders to partake in the biannual JSR meetings. The other government ministries that attend include the Ministries of Finance, Economic Planning and Development (MoFEPD); Industry and Trade (MoIT); Local Government and Rural Development (MoLGRD); Lands, Housing and Urban Development (MoLHUD); Transport and Public Infrastructure (MoTPI); and Natural Resources, Energy and Mining (MoNREM).34 In addition, the Office of the Pres- ident and Cabinet (OPC), specifically the principal secretary for nutrition and HIV/AIDS, is invited to sector review meetings. Some prominent development partners that regularly attend the JSR include: the UK Department for International Development (DfID), the US Agency for International Development (USAID), the European Union (EU), the World Bank, and FAO. The annual JSR meeting takes place in November and documents the performance of the agriculture sector based on input, output, and outcome targets set in the government’s Agriculture Investment Program: The Agriculture Sector-Wide Approach Program (ASWAp). The mid-year JSR meeting reviews progress on commitments made during the annual JSR meeting.35 The ASWAp tries to align the key and strategic policies of the Malawi Growth and Development Strategy (MGDS), the CAADP Compact for Malawi, and the Malawi Development Assistance Strategy (DAS). After each JSR meeting, a report is produced that documents key decisions, commitments, and issues that need to be addressed. Weaknesses identified with regard to Malawi’s JSR reports are that they mainly focus on MoAIWD activities while activities by other stakeholders in the agriculture sector are underreported, and capacity at the district level for monitoring and evaluation is weak. Source: ReSAKSS 2014. 33 http://ebrary.ifpri.org/utils/getfile/collection/p15738coll2/id/130346/filename/130557. pdf (p. 13). 34 http://ebrary.ifpri.org/utils/getfile/collection/p15738coll2/id/130346/filename/130557. pdf (p. 7). 35 Ibid. 62  n Agriculture Public Expenditure Review TABLE 23:  Impacts, cost, and timeframe for proposed actions and reforms Impact on expenditure quality Timeframe Cost Medium (<5y) Long (>5y) Short (<1y) Moderate Medium High High Low Low Responsible Priorities and measures agency Priority 1: Adjusting the level and composition of Public Expenditures to Agriculture to address challenges and achieve objectives for the sector (i) Optimizing the GoL’s support to private goods and MAFS High Medium Low services • Evaluating the impacts of the Intensive Crop Production High Short Low Programme (ICP), including on private input providers and implementing steps to minimize interference (ii) Rebalancing the budget more toward development rather MAFS & High Medium Moderate than recurrent allocations MFRSC (iii) Increasing the share of operating costs so that they MAFS & High Medium Moderate allow for implementation of, e.g., extension and livestock MFRSC services, and for operation and maintenance of capital investment High Medium Low (iv) Rebalancing allocations between cost centers so that MAFS & they are more in line with the contribution to Agriculture MFRSC GDP, priority topics, or other socioeconomic objectives, including Climate-Smart Agriculture, Nutrition-Smart Agriculture, and private sector led job growth Priority 2: Establishing policy-based expenditure allocations (i) Aligning the MAFS’s and the MFRSC’s budgets with MAFS & High Short Moderate relevant policies, especially the National Strategic MFRSC Development Plan (NSDP) II MAFS in High Medium Moderate (ii) Adopting an evidence-based agricultural policy consultation strategy with a National Agriculture Investment Plan for with MFRSC implementing the policy Priority 3: Better tracking Public Expenditures to Agriculture (i) Updating AgBOOST MAFS & Medium Medium Low MFRSC (ii) Including functional classifications in AgBOOST MAFS, Medium Medium Moderate MFRSC, MF (iii) Including donor funded activities in Department and MAFS, Ministry budget planning with the objective of including MFRSC, MF Medium Short Low donor funding in the Integrated Financial Management Information System (IFMIS), pending compliance with donors’ financial management requirements MAFS & High Short Low (iv) Training relevant staff on how to use AgBOOST in budget MFRSC planning and policy evaluation Priority 4: Better evaluating the impacts of Public Expenditures to Agriculture (i) Establishing a feedback loop with results indicators to MAFS & High Short Moderate evaluate impacts of large programs MFRSC (ii) Establishing a Joint Sector Review under the new National MAFS & Medium Medium Low Agriculture Investment Plan MFRSC References AfDB, ADB, EBRD, EIB, IDBG, and WBG. 2016. 2016 Joint Report on Multilateral Development Banks’ Climate Finance. Available at: www.ebrd.com/2016-joint-report-on-mdbs-climate-finance.pdf African Union (AU). 2003. Assembly of the African Union, Second Ordinary Session, 10–12 July 2003, Maputo, Mozambique. Declarations. _____. 2006. 8th Ordinary Session of the Executive Council (16–21 January 2006), Khartoum, Sudan. 277. _____. 2014. Assembly of the Union. Twenty-Third Ordinary Session. 26–27 June 2014. Malabo, Equatorial Guinea. Decisions, Declarations and Resolutions. _____. 2015. The AU Guidance Note on Tracking and Measuring the Levels and Quality of Government Expenditures for Agriculture. African Union, Addis-Ababa, Ethiopia. Behnke, R., and M. Nakirya. 2013. “The Contribution of Livestock to the Ugandan Economy.” IGAD Livestock Policy Initiative Working Paper, 2(12): 1–37. Bisangwa, E. 2013. “The Influence of Conservation Agriculture Adoption on Input Demand and Maize Production in Butha, Lesotho.” University of Tennessee, Master’s Thesis. Tennessee, United States. Bureau of Statistics. 2014. 2013/2014 Agricultural Production Survey Crops. Maseru, Lesotho. Central Bank of Lesotho. 2015. Macroeconomics Statistics Database. Maseru, Lesotho. https://www.centralbank .org.ls/index.php/statistics/macroeconomics-statistics. Accessed on 4 June, 2019. CIAT and World Bank. 2018. Climate-Smart Agriculture in Lesotho. Country Profile. Dejene, A., S. Midgley, M. V. Marake, and S. Ramasamy. 2011. Climate Change Adaptation in Agriculture: Experience and Lessons from Lesotho. Food and Agriculture Organization of the United Nations, Rome. http://www.fao.org/ docrep/014/i2228e/i2228e00.pdf DMA. 2016. El Niño Disaster Response Plan Budget. Maseru, Lesotho. _____. 2019. Personal communication with Mr. Haretsebe Mahosi, Chief Executive Officer, Disaster Management Authority. Maseru, Lesotho. Dorward, A. 2009. “Rethinking Agricultural Input Subsidy Programmes in a Changing World.” In Non-Distorting Farm Support to Enhance Global Food Production, ed. A. Elbehri, and A. Sarris, pp. 311–374. Rome: FAO. Fan, S., A. Gulati, and S. Thorat. 2008. “Investment, Subsidies, and Pro-Poor Growth in Rural India.” Agricultural Eco- nomics, 39(2): 163–170. doi:10.1111/j.1574-0862.2008.00328.x FAO. 2016. Strengthening Coherence between Agriculture and Social Protection. Lesotho Country Case Study Report. FAO, Rome. _____. 2018. AQUASTAT Database. doi:http://www.fao.org/nr/water/aquastat/data/query/index.html?lang=en _____. 2018. FAOSTAT Database. doi:http://www.fao.org/waicent/portal/statistics_en.asp _____. 2019. FAOSTAT Database. doi:http://www.fao.org/waicent/portal/statistics_en.asp Ghins, L., A. Mas Aparisi, and J. Balié. 2017. “Myths and Realities About Input Subsidies in Sub-Saharan Africa.” Devel- opment Policy Review Vol. 35: O214–O233. doi:10.1111/dpr.12258 Government of Lesotho. 2012. National Strategic Development Plan 2012/13–2016/17. Maseru, Lesotho. _____. 2014. National Decentralisation Policy. doi:10.1016/S0735-1097(98)00194-6 LK. Maseru, Lesotho. 63 64  n Agriculture Public Expenditure Review Herforth, A., and T. Ballard. 2016. “Nutrition indicators in agriculture projects: current measurements, priorities and gaps.” Global Food Security. Available at: www.sciencedirect.com/science/article/pii/S22119123415300109 (accessed 25.08.2016). ILO. 2018. ILOSTAT Database. doi:https://www.ilo.org/ilostat/ _____. 2019. ASTI Database. https://www.asti.cgiar.org/data. Accessed on 4 June 2019. International Monetary Fund (IMF). 2014. Government Finance Statistics Manual 2014. Washington, DC: IMF. Jayne, T. S., et al. 2018. “Review: Taking Stock of Africa’s Second-Generation Agricultural Input Subsidy Programs.” Food Policy. Elsevier Vol. 75: 1–14. doi:10.1016/j.foodpol.2018.01.003 Kardan, A., C. O’Brien, and M. Masasa. 2017. Shock-Responsive Social Protection Systems Research. Oxford Policy Man- agement. United Kingdom: Oxford. Lageat, Y. 1999. “L’erosion des Sols au Lesotho.” Cahiers D’outre-Mer 52(205): 3–22. Lebedys, A., and Y. Li. 2014. “Contribution of the Forestry Sector to National Economies 1990–2011.” Available at: http://www.fao.org/3/a-i4248e.pdf Lesotho BOOST. 2018. World Bank and Government of Lesotho. Lesotho Disaster Risk Management Authority, May 2019. Maseru, Lesotho. Lesotho Vulnerability Assessment Committee (LVAV). 2016. “Market Assessment Report.” Lesotho Vulnerabil- ity Assessment Committee, Maseru. https://reliefweb.int/report/lesotho/lesotho-vulnerabilityassessment- committee-market-assessment-report-2016. Lesotho Vulnerability Assessment Committee (LVAC). 2017. “Vulnerability Assessment and Analysis Report.” Mas- eru, Lesotho. López, R. 2005. “Why Governments Should Stop Non-Social Subsidies: Measuring Their Consequences for Rural Latin America.” Policy Research Working Paper No. 3609. World Bank, Washington, D.C. López, R., and G. I. Galinato. 2007. “Should Governments Stop Subsidies to Private Goods? Evidence from Rural Latin America.” Journal of Public Economics 91(5): 1071–1094. https://doi.org/10.1016/j.jpubeco.2006.10.004. MAFAP. 2015. “Methodology Working Paper: Volume II—Analysis of Public Expenditures on Food Security and Agri- culture.” Maseru, Lesotho. Mas Aparisi, A. 2018. “World Bank Agricultural Sector Expenditure Reviews: A Methodological Note.” Unpublished note. Minde, I., et al. 2008. “Promoting Fertilizer Use in Africa : Current Issues and Empirical Evidence from Malawi, Zam- bia, and Kenya.” Available at: http://www.aec.msu.edu/fs2/aamp/seminar_2/seminar_1b_jayne.pdf Ministry of Agriculture and Food Security (MAFS). 2018. ICP Reconciliation Report 2012–18. Maseru, Lesotho. _____. 2019. Maseru, Lesotho. Ministry of Forestry, Range and Soil Conservation (MFRSC). 2019. Maseru, Lesotho. Ministry of Finance, Planning and Economic Development (MoFPED). 2012. Budget Framework Paper for Financial Year 2011–12. Maseru, Lesotho. _____. 2014. Consolidated Budget Framework Paper—FY 2012/14–2014/15. Maseru, Lesotho. _____. 2017. National Budget Framework Paper 2015/16–2017/18. Maseru, Lesotho. Ministry of Small Business Development, Cooperatives, and Marketing (MSBDCM). 2018. “Department of Marketing—PowerPoint Presentation.” Maseru, Lesotho. ­ Moeletsi, M. E., and S. Walker. 2013. “Agroclimatological Suitability Mapping for Dryland Maize Production in Leso- tho.” Theoretical and Applied Climatology Vol. 114(1–2): 227–236. doi:10.1007/s00704-012-0829-1 References n 65 Mogues, T., S. Fan, and S. Benin. 2015. “Public Investments in and for Agriculture.” The European Journal of Develop- ment Research Vol. 27(3): 337–352. doi:10.1057/ejdr.2015.40 Muthami, D., and R. Behnke, R. 2011. “The Contribution of Livestock to the Kenyan Economy.” IGAD Livestock Policy Initiative Working Paper, 3–11: 1–62. Nhemachena, C., G. Matchaya, and S. Nhlengethwa. 2016. “Agricultural Growth Trends and Outlook in Lesotho.” Organisation for Economic Co-operation and Development (OECD). 2010. Glossary of Key Terms in Evaluation and Results Based Management. OECD, Paris. _____. 2019. Creditor Reporting System. Available at: https://stats.oecd.org/Index.aspx?DataSetCode=crs1 (Accessed: 17 February 2019). Public Expenditure and Financial Accountability (PEFA). 2017. PEFA 2017 Performance Assessment Report. Avail- able at: https://pefa.org/assessments/lesotho-2017 (Accessed March 23, 2019). Pule, Neville. 1999. “Power Struggles in the BCP.” Lesotho Social Science Review 5(1): 1–30. ISSN 1028-0790. Faculty of Social Science, Maseru, Lesotho. Regional Strategic Analysis and Knowledge Support System (ReSAKSS). 2014. Malawi Joint Sector Review Assess- ment: Advancing Mutual Accountability through Comprehensive, Inclusive, and Technically Robust Review and Dialogue. _____. 2019. ReSAKSS Indicators. Available at: https://www.resakss.org (Accessed: 19 February 2018). Showers, K. B. 2005. Imperial Gullies: Soil Erosion and Conservation in Lesotho. Ohio University Press (Ecology and History Series), Ohio, USA. Townsend, R., L. Ronchi, C. Brett, and G. Moses. 2018. “Future of Food: Maximizing Finance for Development in Agri- cultural Value Chains.” World Bank, Washington, DC. UNICEF and World Bank. 2018. “Lesotho Public Health Sector Expenditure Review 2017.” doi:10.1596/29344 Wanzala-Mlobela, M., P. Fuentes, and S. Mkumbwa. 2013. “Practices and Policy Options for the Improved Design and Implementation of Fertilizer Subsidy Programs in Sub-Saharan Africa.” NEPAD Agency Policy Study. IFDC, Muscle Shoals, Alabama, United States. World Bank. 2013. “Lesotho. A Safety Net to End Extreme Poverty.” World Bank, Washington, DC. _____. 2016. “Final Report NSDP Review and NSDP 2 Scoping.” World Bank, Washington, DC. _____. 2016a. “Lesotho Water Security and Climate Change Assessment.” World Bank, Washington, DC. _____. 2017. “Kingdom of Lesotho Public Health Sector Expenditure Review. World Bank, Washington, DC. _____. 2018. “Kingdom of Lesotho Public Expenditure Review: Improving Expenditure Efficiency for Inclusive Devel- opment and Growth. World Bank, Washington, DC. _____. 2019. “Kingdom of Lesotho Education Public Expenditure Review”. World Bank, Washington, DC. _____. 2019 forthcoming. “Kingdom of Lesotho Poverty Assessment”. World Bank, Washington DC. _____. 2019a, forthcoming. “Agricultural Sector Risk Assessment and Risk Financing Diagnostics.” World Bank, Wash- ington, DC. _____. 2019b. “Lesotho Climate Smart Agriculture Investment Plan.” World Bank, Washington, DC. _____. 2019c. “World Bank Operations Portal.” World Bank, Washington, DC. _____. Undated. “Kenya Agricultural Sector Expenditure Review.” World Bank, Washington, DC. “World Development Indicators, The World Bank” accessed in 2019, at https://databank.worldbank.org/source/ world-development-indicators ANNEX 1. Detailed Methodology and Data Section DETAILED DESCRIPTION OF VARIABLES Variables under administrative, economic, functional, and source of funding are levelled, summarized, and exem- plified in Table 24. TABLE 24:  Administrative, economic, functional, and source of funding variables of the BOOST PEA database Type Variable—L1 Variable—L2 Variable—L3 Variable—L4 Administrative Head Cost center Subcost center — Example 01—Agriculture and 0104—Crops 010401—Crops admin food security Economic Econ1 Econ2 Econ3 Econ4 Example 41—Compensation of 411—Wages and 4111—Nonstatutory 411101—Salaries, employees salaries salaries in cash established posts Functional Program Subprogram — — Example 0103—Crops services 010301—Crops services Source of funding Source of funding Donor label Project — Example 2—Development 119—Millennium 0336—Smallholder partner, grants Challenge Corporation agricultural (MCC) development DATA LIMITATIONS The MF’s IFMIS system is relatively new and not comprehensive. The IFMIS system was introduced in FY09/10 as part of the GoL’s efforts to improve Public Financial Management (PFM), and has gradually strengthened under the PFM Reform Action Plan 2012/13 to 2017/18 (World Bank 2018). The 2018 World Bank Lesotho Public Expen- diture Review: Improving Expenditure Efficiency for Inclusive Development and Growth (World Bank, 2018) stresses that “there are considerable differences between IFMIS data and the Medium-Term Expenditure Frame- work (MTEF) both in revenues and expenditures.” In addition, the review notes 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” (World Bank 2018). The BOOST database, which draws from IFMIS data, is an even more recent exer- cise. It was initiated in support of the aforementioned PER. As a consequence, the BOOST has several limitations: • First and foremost, the BOOST database has no module for the agriculture sector. It is a transcription of IFMIS data into a more accessible and groomed Excel dataset. AgBOOST was created for this review by extracting the relevant data for the sector. To support future analysis of agriculture sector expenditures, it is recommended that the MAFS update AgBOOST on a regular basis. • Data for the year 2016/17, in the BOOST database used for this review, are not complete and were not used for the analysis. 66 Annex 1. Detailed Methodology and Data Section n 67 • Donor expenditures are almost entirely missing from the IFMIS and BOOST. Over the period 2010/11 to 2015/16, total donor expenditures recorded in BOOST accounted for 2 percent of total PEA. • Functional coding of the IFMIS/BOOST data is incomplete and minimal. Out of US$276 million total PEA in the AgBOOST over the period, US$74.3 million were not coded with an explicit program name (27 percent). The distribution of budget lines with unlabeled program names is, however, clearly skewed toward the early years of IFMIS’s adoption (2010/11 in particular) and exhibits a strong downward trend (Figure 27). By 2014/15, no budget lines have unlabeled program values. This is a sign of improvement in the use of func- tional classification. • There are eight labelled functional codes for expenditures under the MAFS (Table 25). While these are very valuable, two of them are sectoral rather than functional (livestock services; crop services) and two are over- lapping in subject matter and are thus difficult to distinguish (extension services; extension services manage- ment).36 In addition, there is only one level of functional coding—even though a “subprogram” variable exists, it only repeats the values of the program variable. For the MFRSC, only two program codes are labelled: gen- eral administration and management, and land management services. This minimal, binary contrast under- mines the analytical power of the functional classification. • To understand the function of development expenditures beyond its program label (if any), the most dis- aggregated level of information available in AgBOOST is a project’s name, located either under the subcost center (administrative level 3) or project (source of funding level 3). Of 644 development budget lines, 631 are labelled under a project name, but this is not enough information to allow for an in-depth functional analysis of the projects. For instance, the large Summer Cropping Program, also known as the Intensive Crop Production Programme (ICP), is not broken down into its different components (input subsidy, irrigation, machinery . . .). In addition, this major item of expenditures was blended into recurrent costs before 2012/13 and could not be singled out using BOOST. FIGURE 27:  Distribution of budget lines with unlabeled program names in AgBOOST 600 550 500 450 400 Budget line count 350 300 250 200 150 100 50 0 2010–11 2011–12 2012–13 2013–14 2014–15 2015–16 Year 36 Note that these two functions do not overlap in terms of amounts, so no risk of double counting expenditures arises. 68  n Agriculture Public Expenditure Review TABLE 25:  Program codes in the BOOST database for MAFS and MFRSC Program code Note Ministry of Agriculture and Food Security 0100 Not labeled 0101—General administration and management Functional 0102—Livestock services Sectoral 0103—Crops services Sectoral 0104—Agricultural research services Functional 0105—Extension services Overlaps with 0108 0106—Agricultural training Functional 0107—Planning and policy analysis Functional 0108—Extension service management Overlaps with 0105 Ministry of Forestry, Range and Soil Conservation 1400 Not labeled 1401—General administration Functional 1402—Land management service Functional • The level of granularity of the data does not allow for a strong functional analysis of MFRSC expenditures, for which there is only one project—Watershed Management. The subcost centers correspond to the MFRSC’s three departments plus its administration and DOs: Forestry; Soil and Water Conservation; Range Resources Management; Administration; and one subcost center per district. • Expenditures for DOs are “black boxes.” Even though 25 percent of AgBOOST expenditures go to districts (Forestry—15 percent; Agriculture—30 percent), they are only broken down by their economic attributes ­ (wages, travel and transport, operating costs . . .). It is thus hard to analyze district-level expenditures. • No commodity coding of expenditures occurs in the BOOST database. It is thus not possible to map expendi- ture flows to agricultural commodities, and the analysis stops at a sectoral level (livestock, crops, forestry). While this is not a standard budgetary reporting exercise, it could be valuable to institutionalize coding of budget lines by commodity in AgBOOST. • No geotracking of expenditures occurs in the BOOST database. It is therefore not possible to analyze PEA allo- cation by agro-ecological zone (e.g., highlands versus lowlands). Table 26 provides an overview of existing data limitations, the implications these data limitations have for the anal- ysis, and recommendations for how the GoL can address existing limitation going forward. Annex 1. Detailed Methodology and Data Section n 69 TABLE 26:  Data limitations, implications, and recommendations Data limitation Implication Recommendation No agriculture sector public Analysts created an agricultural Update the AgBOOST created for the expenditure database BOOST database for this review. review on a regular basis (every second year, for example). Incomplete data for 2016/17 2016/17 is not part of the period of Update BOOST to incorporate 2016/17. analysis. Donor expenditures almost entirely The figures underestimate total PEA Incorporate donor funding in the IFMIS. missing from the IFMIS and BOOST in Lesotho. This recommendation seems to have already been taken into account— discussions with MFRSC staff suggest that donor funding will be integrated in the IFMIS in FY19/20, although it still will not go through the Treasury. Incomplete functional coding The IFMIS/BOOST lacks analytical Ensure that all budget lines are power for a sectoral PER. systematically and clearly coded in the IFMIS/BOOST and AgBOOST. Minimal functional coding The IFMIS/BOOST lacks analytical At least for AgBOOST, code all budget power for a sectoral PER. lines according to their socioeconomic function in every update. Particular care should be given to the MFRSC’s budget, currently poorly coded under the “programme,” “subprogramme,” and “project” variables. Black box DO expenditures The IFMIS/BOOST lacks Improve reporting of district-level analytical power for district-level expenditures in the IFMIS—go beyond expenditures. economic classification to include functional and project-coded data. No commodity-specific variable Analysts created their own Add a commodity variable in BOOST or categorizing budget lines commodity variable. in the IFMIS and update systematically. No geotracking of national-level It is impossible to know how much Add geotracking variable in BOOST or in expenditures goes to which agro-ecological zone the IFMIS and update systematically. on aggregate (national plus county- level spending). Analysts created a regional variable to analyze central transfers to MAFS and MFRSC DOs. METHODOLOGY This section lays out the approach followed to define the scope of agriculture expenditures and create the func- tional, sectoral, and commodity variables. Scope To conduct this analysis, a perimeter of expenditures considered “agriculture expenditures” (PEA) was delineated. The perimeter has four characteristics: time, functions, administrative units, and economic categories. Time perimeter The analysis focuses on the period for which data are available in BOOST: 2010/11 to 2015/16. This timeframe corresponds to the standard PER period (around five years)—for instance, the recent Lesotho Public Health Sector Expenditure Review spans fiscal years 2011/12 to 2015/16 (World Bank 2017). 70  n Agriculture Public Expenditure Review Functional and subfunctional perimeters According to the International Monetary Fund (IMF)’s Government Finance Statistics (GFS) manual, a functional classification of expense “provides information on the purpose for which an expense was incurred” (IMF 2014). The functional core perimeter of agriculture for the AgPER thus defines what is considered an expenditure whose function is to directly support agriculture. Expenditures that support agriculture indirectly (through positive exter- nalities or spillover effects) are not included in the functional perimeter. The AgPER functional perimeter is based on the Classification of the Functions of Government (COFOG). The COFOG is a functional accounting system developed by the OECD and the United Nations Statistics Division. It classifies public expenditures by “functions or socioeconomic objectives that general government units aim to achieve” (IMF 2014). The AUGN on tracking and measuring the levels and quality of government expenditures for agricul- ture recommends using the following COFOG categories to define “agricultural public expenditures”: “agriculture (including livestock)” (70421); “forestry” (70422); “fishing and hunting” (70423); and “agricultural research and development” (7084) (African Union 2015). For this AgPER, this perimeter was further disaggregated into a subfunctional perimeter, using the FAO’s (Food and Agriculture Organization of the United Nations) methodology for analysis of public expenditures on food and agri- culture. The methodology developed by FAO’s Monitoring and Analysing Food and Agricultural Policies (MAFAP) program offers subfunctional categories labelled “COFOG-compatible”; that is, they can fall within the COFOG perimeter (MAFAP 2015). Some MAFAP categories, however, go beyond that perimeter to include rural expendi- tures and expenditures in support of consumers. These were removed to obtain the AgPER subfunctional perimeter presented in Table 27. These categories are all in line with the COFOG perimeter described above. Although this perimeter corresponds to the AUGN’s COFOG perimeter, it does not quite match its “enhanced COFOG” perimeter. The latter perimeter includes several additional categories deemed too broad for the purpose of this analysis. Table 28 indicates enhanced COFOG categories that are retained or not in the AgPER perimeter and pro- vides a justification why. Data were extracted from the BOOST database to produce an AgBOOST database that would contain only expendi- tures that fall within the agricultural functional and subfunctional perimeter. For this, a two-pronged approach was used: administrative and keyword-based. Administrative approach: All expenditures under the MAFS and the MFRSC were automatically considered, in line with the AUGN on sectors that constitute agriculture. Keyword-based approach: To capture agriculture expenditures that fell under ministries other than the MAFS and the MFRSC, a list of agriculture-relevant keywords was established and applied as a filter on selected categorical variables of the BOOST database. Variables filtered: • Administrative variables: cost center • Functional variables: program, subprogram • Source of funding variables: project Annex 1. Detailed Methodology and Data Section n 71 TABLE 27:  AgPER subfunctional perimeter, based on selected categories from FAO’s MAFAP methodology (comprises COFOG categories 70421, 70422, 70423, and 7084) Functional category Description A. Subsidies Transfers to private agents in the agriculture sector in the form of variable inputs or assets. A1. Input subsidies Transfers to private agents in the agriculture sector in the form of partial or total payment of seeds, fertilizers, pesticides, fuel, electricity, credit, and other of the like. A2. Capital subsidies Transfers to agents in the agriculture sector in the form of partial or total payment of agricultural equipment, machinery, on-farm infrastructure, livestock, and other of the like. B. Research Transfers to public or private agents in the form of partial or total payment of agricultural research activities. C. Extension and advisory services Transfers to public or private agents in the form of partial or total payment of agricultural extension and advisory services. D. Training Transfers to public or private agents in the form of partial or total payment of training of agricultural private agents. E. Inspection and quality control Transfers to public or private agents in the form of partial or total payment of inspection and quality control activities: livestock vaccination campaigns, inspection of produce quality for marketing, and other of the like. F. Agricultural infrastructure Transfers to public or private agents in the form of partial or total payment of the construction of infrastructure that directly supports the agriculture sector. F1. Feeder roads Transfers to public or private agents in the form of partial or total payment of the construction of roads that connect production areas to the market. F2. Irrigation Transfers to public or private agents in the form of partial or total payment of the construction of irrigation for agricultural production that is not on- farm (for on-farm, see A2): irrigation dams, canals, and other of the like. F3. Other infrastructure Transfers to public or private agents in the form of partial or total payment of the construction of other agricultural infrastructure that are not on-farm (wells for livestock, for instance). G. Storage Transfers to public or private agents in the form of partial or total payment of the construction of public or collectively owned storage infrastructure (for private ownership, see A2). H. Processing and marketing Transfers to public or private agents in the form of partial or total payment of the construction of collectively owned market and processing infrastructure (for private ownership, see A2), marketing or processing training, and other of the like. Other—Administrative costs Transfers to public agents in the form of partial or total payment of the costs of maintaining efficient administrative functions in support of agriculture. This essentially includes salaries of staff in agricultural government agencies involved in managerial and secretarial functions and the maintenance running costs of these agencies. Source: Authors, based on MAFAP 2015. 72  n Agriculture Public Expenditure Review TABLE 28:  Inclusion of AUGN-enhanced COFOG functions in the AgPER perimeter Inclusion in the AUGN-enhanced COFOG function perimeter Justification Food and Nutrition Security (FNS) No No standard methodology exists to identify FNS expenditures, which can involve a far-reaching perimeter encompassing health and environment, food aid and humanitarian expenditures, and education, among others. Their inclusion risks heavily skewing the indicators. Rural/feeder roads No/Yes Only identifiable feeder roads are retained. Rural roads entail massive public expenditure items that aim to support all rural economy sectors—health, education, resource extraction. . . . Their inclusion risks heavily skewing the indicators. Rural land administration No Rural land titling and administration have multiple institutional, social, and economic objectives and support agriculture only indirectly. Sustainable Natural Resource Yes/No Expenditures in direct support of the agriculture sector that Management (NRM) include sustainable NRM are included—including agro-ecology, agroforestry, and soil and rangeland conservation measures. Other environmental expenditures are not included as they support agriculture only indirectly and their inclusion risks skewing the indicators. Multi-sectoral/multipurpose projects Yes Whenever they include an agricultural component, these projects are considered. Mandated functions of state-owned Yes Whenever semi-autonomous government agencies (including enterprises state-owned enterprises) receive public funds to spend on agriculture, the public funds are included. Agricultural marketing Yes See category H of the perimeter in Table 27. Capacity development for agriculture Yes See category D of the perimeter in Table 27. development Rural electrification for agriculture No Rural electrification entails massive public expenditure items that aim to support all rural economy sectors—health, education, resource extraction. . . . Their inclusion risks heavily skewing the indicators. Information and communication Yes ICT for agriculture is included through different categories of the technology (ICT) for agriculture perimeter in Table 27 (B, C, D, E, H, mainly). Subnational expenditures Yes Annex 1. Detailed Methodology and Data Section n 73 TABLE 29:  Keywords master list *agricultur* *agro* *irrigat* *farm* *seed* *fertili* *pesticid* *crop* *animal* *veteri* *forest* *pastor* *tree* *fish* *aquaculture* *pisci* *green* *lake* *wood* *livestock* *leather* *feeder* *food* *rural* *extensio* Budgetary lines matching with the keyword filter were automatically included in the perimeter. After combing through the results, a second filter was applied to weed out nonagricultural matches (for instance, “street lighting” obtained with the keyword search “tree”). In particular, the keyword filter reaped a large number of budget lines (779 out of 4,882) associated with rural development but not the agriculture sector sensu stricto: rural electricity, rural roads, and rural water. Although these are supportive of agriculture sector development, they contribute to a breadth of other sectors and were not retained in the perimeter. Nevertheless, they are presented in the report as a complement to the PEA analysis. Although in an AgPER the vast majority of budgetary items can be easily classified as within or outside the func- tional perimeter, there are typically “grey” lines of expenditures that require thorough examination: (i) multipur- pose development projects (COFOG category: 70474) that include an agricultural component; and (ii) “on-the-fence” projects that are either considered wholly in the perimeter or wholly outside of it. An assumption had to be made for MFRSC expenditures. Indeed, it could be argued that some of them are in direct support neither of forestry nor agriculture, but rather for environmental preservation. Because of insufficient detail in the BOOST data, it was dif- ficult to untangle purely environmental preservation expenditures from PEA. The biggest expenditure item in the MFRSC—70 percent—is for the WMP under the Conservation Department. Qualitative information gathered on the 74  n Agriculture Public Expenditure Review project through interviews with conservation department staff suggests that the program’s focus is on soil protec- tion (stone lines and terraces against erosion, farrows and diversions against floods . . .). This was triangulated with budgetary data shared by the conservation department. Given Lesotho’s endemic soil erosion problem and close relationship with agricultural production (Showers 2005; Bisangwa 2013), it was decided to include the WMP and other MFRSC soil and rangeland conservation measures. Administrative perimeter The administrative perimeter follows from the functional perimeter, as outlined in Table 30. TABLE 30:  Administrative perimeter of the AgPER Administrative unit Included in the perimeter Ministry of Agriculture and Food Security Yes, all lines of expenditure Ministry of Forestry, Range and Soil Yes, all lines of expenditures (see section above for clarification). Conservation Other ministries: Yes, but only the lines of expenditures that are in direct support of Ministry of Finance agriculture. The Disaster Management Authority (DMA) of the Prime Ministry of Small Business Development, Minister’s Office was also interviewed to pin down its contribution to PEA. Cooperatives, and Marketing It is not included in the perimeter because no direct contribution to the Ministry of Home Affairs sector over the period was identified, but rather transfers to the MAFS. In addition, BOOST data do not allow to disentangle PEA from other DMA expenditure. Semi-autonomous government agencies SAGAs are not in the IFMIS and BOOST database. Interviews were held (SAGAs): with Lesotho Highland Development Authority and Lesotho National None Development Corporation to consider their role in the agriculture sector, but they are not in the perimeter. Donors: Yes, but only lines of expenditures that are in direct support of agriculture MCC and that are on the budget. This corresponds to 2 percent of the total IFAD AgBOOST. Nongovernmental organizations No Private sector No Economic perimeter The economic perimeter for this review includes the following: approved and revised estimates, as well as war- ranted release and final expenditure. Debt service and revenue forgone are not considered in the perimeter, in line with AUGN recommendations. Functional and sectoral classification Functional analysis is an important part of the AgPER. It entails understanding the socioeconomic objective of pub- lic expenditures. However, as mentioned in previous sections, the functional variables that exist in Lesotho’s IFMIS/ BOOST do not allow for a detailed functional analysis. The data were thus recoded with the categories presented in (Table 27). AgBOOST contains 3,954 budget lines and it is impossible to manually attribute a function and sector to each line. To do so, a conversion table of keywords to function was created. The conversion table was used to auto- matically assign a functional category value to budget lines based on their values for the following variables: proj- ect; subprogram; subcost center, and economic level 4—see Table 24 for a description of the variables. For instance, the function “research” was assigned to all budget lines that contained the word “research” in the columns project, subprogram, subcost center, or economic level 4. The functional data were then inspected and checked for incon- sistencies using data visualization, especially for functions that represented a sizeable share of public expenditures, such as input subsidies and extension services. The sectoral classification (crops, livestock, forestry) was done using the same method, with a keyword-to-sector conversion table. ANNEX 2: Policy Coherence Analysis Summary Table Indicator Observed Target Source of target Gap Recommendation The table below summarizes all dimensions of PEA policy coherence. Level 3.3% 10% Malabo (2014) Declarations, 6.7% 2010 CAADP Compact Technical efficiency Execution rates 79% No target, high value desirable Economic efficiency Share of development spending 44.6% No target, high in PEA value desirable Share of operating costs in PEA 43.9% No target, high value desirable Functional efficiency Allocation of PEA across sectors Multi-sectoral 70% 58% NSDP –12 Crop 25% 31% NSDP +6 Livestock 4% 10% NSDP +6 Forestry 1% 6% NSDP 0 Allocation of PEA across subfunctions Input subsidies 0% 20% NSDP 20 Other infrastructure 6% 25% NSDP 19 Extension and advisory services 10% 28% NSDP 17 Capital subsidies 4% 5% NSDP 0 Services subsidies 1% 0% NSDP –1 Training 9% 5% NSDP –4 Inspection/quality control 4% 0% NSDP –4 Research 7% 2% NSDP –5 Irrigation 6% 0% NSDP –6 Processing and marketing 12% 0% NSDP –12 Administrative costs 39% 14% NSDP –25 75 Share of private sector-related 42% No target, low World Bank (2011) functions in PEA value desirable ANNEX 3: PEA per Ministry, Program, and Project, Total 2010/11 to 2015/16, in Million Constant 2010 Maloti Ministry Program Project PEA 01 Agriculture and 0100 0000 141.1 Food Security 0002 Sustainable Agriculture and Natural Resources 4.6 Management 0003 Procurement of Agricultural Machinery 5.4 0004 Recapitalisation of Lesotho Agricultural College 3.6 0005 Irrigated Crop Production 3.6 0006 Special Program for Food Security 1.2 0122 Rehabilitation of Farmers Training Centre 6.8 0307 0.0 0101 General 0000 Administration and 100.2 Management 0102 Livestock Services 0000 38.4 0006 Special Program for Food Security 0.7 0315 Wool and Mohair Prod Development 2.9 0363 Integrated Poultry Production 0.0 0397 Wool Sheds 4.2 0478 Climate Resilient Wool and Mohair Product 0.0 0483 0.7 0484 0.0 0103 Crops Services 0000 77.9 0003 Procurement of Agricultural Machinery 15.6 0005 Irrigated Crop Production 5.2 0424 Summer Cropping Programme 333.5 0104 Agricultural Research 0000 32.4 Services 0105 Extension Services 0000 308.5 0106 Agricultural Training 0000 54.6 0004 Recapitalisation of Lesotho Agricultural College 27.1 0107 Planning and Policy 0000 14.0 Analysis 0108 Extension Services 0000 21.6 Management 0002 Sustainable Agriculture and Natural Resources 6.5 Management 0122 Rehabilitation of Farmers Training Centre 11.7 0336 Smallholder Agriculture Development 14.6 76 Annex 3: PEA per Ministry, Program, and Project, Total 2010/11 to 2015/16 n 77 Ministry Program Project PEA 04 Finance 0400 0048 Rural Financial Intermediation Programme 18.1 0127 2009/2010 Agricultural Census 8.5 0139 Rural Water Supply and Sanitation 0.0 0403 Macroeconomic 0048 Rural Financial Intermediation Programme 12.3 Policy Mgt 05 Trade and Industry 0503 Industrialisation, 0474 Agricultural Production and Trade Development Commercialisation and (EIF) Tier II 1.5 Market Diversification 08 Home Affairs 0802 Livestock Marking 0000 23.8 and Registration 0451 Livestock Registration 0.0 14 Forestry and Land 1400 0000 105.0 Reclamation 0077 Watershed Management 221.9 0354 0.4 1401 General 0000 Administration and 16.8 Management 1402 Land Management 0000 64.9 Services 0077 Watershed Management 225.6 0354 0.2 1404 0354 0.0 ANNEX 4: Local Government Expenditures, by Project, Share of Total Expenditures 2010/11 to 2015/16 Share Cost center Sub cost center Project of total 4201 Administration 420101 Administration 0000 8% 0105 Assistance to Decentralisation 0% 0108 Solid Waste Management (Land Fill) Ts’oeneng 1% 0110 Designs of Urban Roads 0% 0111 Local Government Infrastructure 0% 0112 Removal of Hazardous Sludge 0% 0114 Solid Waste Management—Maseru 0% 0144 Construction of Pitso Houses 0% 4202 District 420201 Maseru 0000 5% Administration 0104 Development Fund for Councils 4% 420202 Mafeteng 0000 4% 420203 Berea 0000 4% 420204 Leribe 0000 5% 420205 Botha Bothe 0000 3% 420206 Mokhotlong 0000 3% 420207 Qacha’s Nek 0000 3% 420208 Thaba Tseka 0000 3% 420209 Quthing 0000 3% 420210 Mohale’s Hoek 0000 4% 4203 Decentralisation 420301 Decentralisation 0000 1% 0111 Local Government Infrastructure 0% 0317 Lesotho Local Development Programme 0% 0391 Solid Waste Management Urban Councils 0% 0392 Construction of Bus Terminals 0% 4204 Lands 420401 Lands 0000 3% 0106 0% 4205 Housing 420501 Housing 0000 0% 0143 0% 0389 Construction of Low Income Houses 0% 4206 Chieftainship 420601 Chieftainship 0000 0% 0113 Principal Chiefs Offices 0% 4207 Engineering and 420701 Engineering and 0000 0% Infrastructure Infrastructure 0107 Upgrading of Urban Roads 26% 0109 Development of Rural Community Roads 18% 0110 Designs of Urban Roads 0% 78 ANNEX 5: Rate of Subsidy per Fertilizer Type, Year and Region, 2012/13 to 2016/17* Fertilizer type Year 2:3:2 (22) 3:2:1 (25) 6:2:1 (31) Lan Urea Average 2012/13 Rate of subsidy highlands 63% 73% — — 68% Rate of subsidy lowlands 66% 75% 71% 70% 75% 71% 2013/14 Rate of subsidy highlands 64% — 64% — — 64% Rate of subsidy lowlands 61% 51% 62% — — 58% 2014/15 Rate of subsidy highlands 0% — — — — 0% Rate of subsidy lowlands 61% 0% 54% 38% 2015/16 Rate of subsidy highlands — — — — — — Rate of subsidy lowlands 63% 63% 62% 65% 63% 2016/17 Rate of subsidy highlands — — — — — — Rate of subsidy lowlands 60% 62% 60% 61% Average Rate of subsidy highlands 42% NA 69% NA NA 56% Rate of subsidy lowlands 62% 50% 62% 67% 75% 63% Rate of subsidy national 55% 50% 64% 67% 75% 62% Source: Authors, from MAFS 2019. Note: *Data unavailable for seeds and pesticides. 79 ANNEX 6: Input Procurement and Sale Value, per Input Category, 2012/13 to 2016/17, in Million Constant 2010 Maloti Input category Year Chemicals Fertilizer Seeds Total 2012/13 Total procurement value 4.79 55.28 25.93 86.00 Total sale value 0.10 8.64 2.40 11.14 Sale recoup on subsidy 2% 16% 9% 13% 2013/14 Total procurement value 2.64 58.52 14.54 75.70 Total sale value 0.41 11.25 3.15 14.82 Sale recoup on subsidy 16% 19% 22% 20% 2014/15 Total procurement value 4.60 54.10 12.65 71.35 Total sale value 0.38 18.21 3.22 21.80 Sale recoup on subsidy 8% 34% 25% 31% 2015/16 Total procurement value 3.23 42.13 7.02 52.38 Total sale value 0.04 6.25 0.53 6.82 Sale recoup on subsidy 1% 15% 8% 13% 2016/17 Total procurement value 1.51 53.64 10.10 65.26 Total sale value 0.29 26.91 1.75 28.95 Sale recoup on subsidy 19% 50% 17% 44% Total Total procurement value 16.77 263.68 70.25 350.69 Total sale value 1.22 71.25 11.05 83.52 Sale recoup on subsidy 7% 27% 16% 24% 80 ANNEX 7: Table of Donor Projects, Constant and Current Values Since the AgBOOST is presented in constant 2010 US$, most donor expenditures in section 4.4.2 are presented in constant 2010 US$ values as well. The OECD CRS database presented in section 4.3.2 presented donor expendi- ture sin constant 2016 US$. Table 31 provides expenditures in current values as well as constant 2010 US$ values for completed, ongoing and planned (not an exhaustive list) donor projects for easier comparison and reference. Table 32 provides Donor PEA in AgBOOST, OECD’S CRS in constant 2010 US$; and Table 33 provides Donor aid to Lesotho’s agriculture sector, in constant 2010 US$ values. TABLE 31:  Donor expenditures in current and constant million 2010 US$ Nominal values, in US$ Values in constant 2010 US$ Donor-funded projects (completed and under implementation) IFAD: Wool and Mohair Promotion 38.93 30.7 Project (2015–2021) World Bank/IFAD: SADP (2011–2020) 30 26 IFAD/GEF: Adaptation of Small-scale 4.3 3.01 Agriculture (LASAP, 2017–2020) FAO: Emergency Response (two 1.137 0.8 projects in 2016) 0.97 0.7 Rapid Response to Drought (2016) 1.13 0.8 Resilience Building (2014–2016) 0.61 0.5 Donor-funded projects (planned) World Bank: Agricultural Productivity 20 13.2 Programme for Southern Africa, APPSA (2018–2025) Second Phase Smallholder Agriculture 52 34.3 Development Project (SADP-2,) (2019–2026) TABLE 32:  Donor PEA in AgBOOST, OECD’S CRS, with SACU, SADC, and Africa average in million constant 2010 US$ 2011 2012 2013 2014 2015 2016 Total Average AgBOOST 0 1 0 1 0 0 2 0.3 CRS 13 1 0 12 1 0 27 4 Lesotho CRS SACU 19 5 5 5 7 7 49 8 CRS SADC 40 44 41 28 47 43 242 40 CRS Africa 42 49 55 50 59 64 319 53 Source: Authors and OECD 2019. 37 Project financed by SADP. 81 82  n Agriculture Public Expenditure Review TABLE 33:  Donor aid to Lesotho’s agriculture sector, in million constant 2010 US$ 2011 2012 2013 2014 2015 2016 Total Share IFAD 5.2 — — 12.1 — — 17.4 50% World Bank 3.6 1.0 0.2 1.0 0.8 2.1 8.7 25% GEF 2.9 0.4 0.0 — — — 3.2 9% Ireland 1.0 0.1 0.1 — — — 1.1 3% United — — — — 0.7 — 0.7 2% Kingdom Japan 0.1 0.1 0.1 0.1 0.1 0.1 0.5 1% Canada 0.3 — — — — — 0.3 1% Korea 0.1 0.1 — 0.0 0.0 0.0 0.1 0.40% FAO — — 0.1 — — 2.8 2.9 8% Total 13.1 1.6 0.4 13.2 1.6 5.0 34.9 100% Source: Authors and OECD 2019.