101232 Trajectories for Sustainable Development Goals Framework and Country Applications Susanna Gable, Hans Lofgren, and Israel Osorio Rodarte Foreword by Mahmoud Mohieldin Trajectories for Sustainable Development Goals Framework and Country Applications Susanna Gable, Hans Lofgren, and Israel Osorio Rodarte Foreword by Mahmoud Mohieldin © 2015 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 18 17 16 15 This work is a product of the staff of The World Bank with external contributions. The findings, interpretations,  and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, 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. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to following copy, distribute, transmit, and adapt this work, including for commercial purposes, under the ­ conditions: Attribution—Please cite the work as follows: Gable, Susanna, Hans Lofgren, and Israel Osorio Rodarte. 2015. “Trajectories for Sustainable Development Goals: Framework and Country Applications.” Washington, DC: World Bank. License: Creative Commons Attribution CC BY 3.0 IGO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official content or error in this translation. World Bank translation. The World Bank shall not be liable for any ­ Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party- owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to the Publishing and Knowledge Division, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@ worldbank.org. Cover photo: © World Bank. Further permission required for reuse. Cover design: Bill Pragluski / Critical Stages LLC. Contents Foreword. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1 Framework for Country Development Diagnostics Post-2015 . . . . . . . . . . . . . . . . . 5 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2. Step One: Benchmarking SDG Progress. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 3. Step Two: SDG Business-as-Usual Projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 4. Step Three: Benchmarking Determinants and Identifying Spending Priorities. . . . . . . . . . . . . 10 5. Step Four: Identifying Fiscal Space. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 6. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Country Briefs 2 Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 2. SDG Indicators: History and Projections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3. Fiscal Space. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 4. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Annex 2A: Data Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 3 Jamaica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 2. SDG Indicators: History and Projections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 3. Fiscal Space. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 4. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Annex 3A: Data Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 4 The Kyrgyz Republic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 2. SDG Indicators: History and Projections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 3. Fiscal Space. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 4. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Annex 4A: Data Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 5 Liberia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 2. SDG Indicators: History and Projections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 iii 3. Fiscal Space. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 4. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Annex 5A: Data Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 6 Nigeria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 2. SDG Indicators: History and Projections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 3. Fiscal Space. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 4. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 Annex 6A: Data Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 7 Pakistan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 2. SDG Indicators: History and Projections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 3. Fiscal Space. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 4. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 Annex 7A: Data Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 8 Peru . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 2. SDG Indicators: History and Projections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 3. Fiscal Space. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 4. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Annex 8A: Data Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 9 The Philippines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 2. SDG Indicators: History and Projections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 3. Fiscal Space. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 4. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Annex 9A: Data Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 10 Senegal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 2. SDG Indicators: History and Projections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 3. Fiscal Space. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 4. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Annex 10A: Data Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186 11 Uganda  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 2. SDG Indicators: History and Projections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 iv Contents 3. Fiscal Space. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197 4. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Annex 11A: Data Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209 Boxes 1.1 Using GNI per Capita for SDG Benchmarking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.2 Projecting GDP and GNI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.3 SDG Business-as-Usual Projections for 2030. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.4 Measures of Government Effectiveness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Figures I.1 Recent and Projected Levels of Access to Electricity (% of Population), and Real GNI per Capita Growth, by Country. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.1 Uganda—Primary School Net Enrollment and GNI per Capita (Left); Primary School Completion and GNI per Capita (Right) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.2 Uganda—Secondary School Gross Enrollment and GNI per Capita (Left); Secondary School Completion and GNI per Capita (Right) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.3 Uganda—Historical Data and Projections for Real GDP per Capita (2011=100). . . . . . . . . . . 10 1.4 Uganda—Expenditure per Primary Student and GNI per Capita (Left); Expenditure per Secondary Student and GNI per Capita (Right) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.5 Uganda—Primary Pupil-Teacher Ratio and GNI per Capita (Left); Secondary Pupil-Teacher Ratio, Secondary and GNI per Capita (Right) . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.6 Uganda—Tax Revenues 1990–2011 (% of GDP) (Left); Tax Revenues (% of GDP) versus GNI per Capita (Right) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 1.7 Uganda—ODA (% of GNI) versus GNI per Capita (Left); ODA (per Capita) versus GNI per Capita (Right). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.1 Ethiopia—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.2 Ethiopia—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012. . . . . . . . . . 29 2.3 Ethiopia—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting. . . . . . . . .31 3.1 Jamaica—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 3.2 Jamaica—Percentile Cross-Country Ranking for SDG Indicators since 2000. . . . . . . . . . . . . . 48 3.3 Jamaica—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting. . . . . . . . . 50 4.1 The Kyrgyz Republic—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 4.2 The Kyrgyz Republic—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 4.3 The Kyrgyz Republic—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 5.1 Liberia—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Contents v 5.2 Liberia—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012. . . . . . . . . . . 84 5.3 Liberia—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting. . . . . . . . . . 86 6.1 Nigeria—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 6.2 Nigeria—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012. . . . . . . . . . 102 6.3 Nigeria—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting . . . . . . . . 104 7.1 Pakistan—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 7.2 Pakistan—Percentile Cross-Country Ranking for SDG Indicators since 2000. . . . . . . . . . . . 120 7.3 Pakistan—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting . . . . . . . 122 8.1 Peru—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 8.2 Peru—Percentile Cross-Country Ranking for SDG Indicators since 2000. . . . . . . . . . . . . . . . 139 8.3 Peru—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting. . . . . . . . . . . 140 9.1 The Philippines—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 9.2 The Philippines—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012. . . .158 9.3 The Philippines—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 10.1 Senegal—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 10.2 Senegal—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012 . . . . . . . . . 176 10.3 Senegal—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting . . . . . . . . 178 11.1 Uganda—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191 11.2 Uganda—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012 . . . . . . . . . 196 11.3 Uganda—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting . . . . . . . . 199 Map I.1 The Ten Country Briefs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Tables 1.1 Uganda—Historical and Projected Growth from Various Sources. . . . . . . . . . . . . . . . . . . . . . . 11 1.2 Uganda—SDG Projections for 2030. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.3 Uganda—Policy-Relevant SDG Determinants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1.4 Government Fiscal Space—Recent Indicators and Future Directions of Change. . . . . . . . . . . 16 2.1 Ethiopia—SDG Indicators: Evolution since 2000 and Projections to 2030 . . . . . . . . . . . . . . . . 24 2.2 Ethiopia—Fiscal Space: Revenue, Spending, and Government Efficiency. . . . . . . . . . . . . . . . . 30 2.3 Ethiopia—Summary Results for SDG Indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 3.1 Jamaica—SDG Indicators: Evolution since 2000 and Projections to 2030. . . . . . . . . . . . . . . . . 42 3.2 Jamaica—Fiscal Space: Revenue, Spending, and Government Efficiency. . . . . . . . . . . . . . . . . . 49 3.3 Jamaica—Summary Results for SDG Indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .54 4.1 The Kyrgyz Republic—SDG Indicators: Evolution since 2000 and Projections to 2030. . . . . . 60 4.2 The Kyrgyz Republic—Fiscal Space: Revenue, Spending, and Government Efficiency . . . . . . 68 4.3 The Kyrgyz Republic—Summary Results for SDG Indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . 72 vi Contents 5.1 Liberia—SDG Indicators: Evolution since 2000 and Projections to 2030. . . . . . . . . . . . . . . . . . 83 5.2 Liberia—Fiscal Space: Revenue, Spending, and Government Efficiency. . . . . . . . . . . . . . . . . . 85 5.3 Liberia—Summary Results for SDG Indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 6.1 Nigeria—SDG Indicators: Evolution since 2000 and Projections to 2030 . . . . . . . . . . . . . . . . . 96 6.2 Nigeria—Fiscal Space: Revenue, Spending, and Government Efficiency. . . . . . . . . . . . . . . . . 104 6.3 Nigeria—Summary Results for SDG Indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 7.1 Pakistan—SDG Indicators: Evolution since 2000 and Projections to 2030 . . . . . . . . . . . . . . . 114 7.2 Pakistan—Fiscal Space: Revenue, Spending, and Government Efficiency. . . . . . . . . . . . . . . . 121 7.3 Pakistan—Summary of Results for SDG Indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 8.1 Peru—SDG Indicators: Evolution since 2000 and Projections to 2030. . . . . . . . . . . . . . . . . . . 132 8.2 Peru—Fiscal Space: Revenue, Spending, and Government Efficiency . . . . . . . . . . . . . . . . . . . 140 8.3 Peru—Summary Results for SDG Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 9.1 The Philippines—SDG Indicators: Evolution since 2000 and Projections to 2030. . . . . . . . . 152 9.2 The Philippines—Fiscal Space: Revenue, Spending, and Government Efficiency. . . . . . . . . . 159 9.3 The Philippines—Summary of Results for SDG Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 10.1 Senegal—SDG Indicators: Evolution since 2000 and Projections to 2030. . . . . . . . . . . . . . . . 170 10.2 Senegal—Fiscal Space: Revenue, Spending, and Government Efficiency. . . . . . . . . . . . . . . . . 177 10.3 Senegal—Summary of Results for SDG Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 11.1 Uganda—SDG Indicators: Evolution since 2000 and Projections to 2030. . . . . . . . . . . . . . . . 190 11.2 Uganda—Fiscal Space: Revenue, Spending, and Government Efficiency. . . . . . . . . . . . . . . . . 198 11.3 Uganda—Summary Results for SDG Indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Contents vii Foreword The experience with the Millennium Development Meanwhile, a team of prominent experts Goals (MDGs) has provided many lessons which at  the World Bank was working on various could usefully be applied to the Sustainable approaches and models to analyze MDG prog- Development Goals (SDGs). The SDGs continue ress. They were asked to assist with the develop- to tackle the issues that the MDGs attempted to ment of a framework through which the ability solve, but they go further, highlighting a global of countries to achieve the new goals could be partnership to end extreme poverty and protect assessed. After deliberations and technical dis- the planet for the next generations, while leaving cussions, the decision was made to proceed with no-one behind. A major issue confronted during a practical approach to the questions at hand the MDG era was managing financing and ser- using case studies of a representative group of vice delivery gaps at the country level. Thus iden- countries. Their selection was based on a variety tifying specific service delivery and financing of criteria which was not limited to per capita solutions to close attainment gaps lies at the crux income and initial conditions, but also included of meeting our development goals. access to natural resources, size, and geography. In 2013, discussions around the Post-2015 The framework applied in this publication Agenda were taking shape. A High-Level Panel consists of four steps, first illustrated through a of Eminent Persons (HLP) was established by the pilot study on Uganda: UN Secretary General on this agenda and, in • Benchmarking current level of progress for addition to producing a report on how to eradi- each SDG for the country being analyzed rel- cate poverty and transform economies through ative to other countries, given Gross National sustainable development, occasioned multiple Income (GNI) per capita engagements on how to best tackle the world’s • Projecting the country’s values for SDGs by most pressing challenges. In this context, I was 2030, given projected business-as-usual involved in several meetings with leaders and development of GNI per capita members of the HLP. Two of these meetings were • Turning to determinants of SDG outcomes of direct relevance and significance to this work: and seeking to identify ways of achieving out- the first was with H.E Ellen Johnson Sirleaf, comes that are more ambitious than those of President of Liberia, and Ms. Amina Mohammed, business-as-usual. This includes discussion of Special Adviser to the UN Secretary-General on potential changes in policies and spending in Post-2015 development planning; the second priority areas was with H.E Ngozi Okonjo-Iweala, then- • Discussing ways to expand fiscal space for Coordinating Minister for the Economy and the priority SDG spending, including additional Minister of Finance of Nigeria. The emphasis of domestic or foreign financing and efficiency these meetings was the importance of overcom- gains. ing the financing and service delivery challenges that the SDGs pose. Later discussions with H.E This framework may be used to analyze the Maria Kiwanuka, then-Minister of Finance of likely progress in SDGs and their determinants Uganda, confirmed this priority. Conversations and discuss policy and financing options to with these distinguished leaders, policymakers, accelerate progress. This publication includes and other influential individuals highlighted the ten examples of geographically dispersed low-­ demand for a tool to address these challenges income, middle-income, landlocked, fragile, which featured so prominently in countries’ ­ ountries. natural resource-rich, and small island c experience with the MDGs. The countries included are: Ethiopia, Jamaica, ix Kyrgyzstan, Liberia, Nigeria, Pakistan, Peru, this framework improves policymakers’ ability Philippines, Senegal and Uganda. to identify and address opportunities and chal- While I believe that there is value in this lenges for achievement of the SDGs. publication for government officials and devel- ­ opment practitioners working on these ten countries, I hope additional value comes from sharing and applying the relevant lessons Mahmoud Mohieldin learned  to other countries. Ultimately, I trust Corporate Secretary and President’s Special Envoy x Foreword Acknowledgements This book was written by Susanna Gable, Hans Sultanova (the Kyrgyz Republic); Inguna Lofgren, and Israel Osorio Rodarte under the Dobraja, Errol Graham, Kristen Himelein and leadership of Mahmoud Mohieldin, Corporate Douglas Sumerfield (Liberia), Olatunde Secretary and President’s Special Envoy, with Adetoyese Adekola, Irajen Appasamy, Kebede joint guidance from Marilou Uy and Jos Verbeek. Feda and Khwima Nthara (Nigeria), Mehwish We thank Hans Timmer, Elena Ianchovichina, Ashraf, Uzma Basim, Saadia Refaqat, Richard and Punam Chuhan for their valuable sugges- Spencer, Muhammad Waheed, and Inaam tions as peer reviewers. We are also grateful for Haq  (Pakistan), Alberto Rodriguez, Pedro comments from Farida Aboulmagd, Lily Chu, Rodriguez, Jamele Rigolini, Maria Eugenia Anton Dobronogov, Eric Feyen, Marcelo Giugale, Genoni, Luciana Velarde, and  Ekaterina Julius Gwyer, Gloria Grandolini, Raj Nallari, Vostroknutova (Peru), Rogier van den Brink, Alberto Portugal, Sajjad Shah, Marco Scuriatti, Caryn Bredenkamp and Lynnette Perez (the Chris Thomas, and Debrework Zewdie. Philippines), Philip English (Senegal) and For the country briefs we again thank Jos Anton Dobronogov (Uganda). Verbeek for the overall guidance and the coun- The findings, interpretations, and conclu- try teams for their crucial feedback. In particu- sions expressed in this book are entirely those lar we are grateful for input from Lars Moller of  the authors. They do not necessarily rep- and Thanh Thi Mai (Ethiopia); Adam Fuchs, resent  the views of the International Bank for Todd Johnson, Kathy Lalazarian, Jonna Reconstruction and Development/World Bank Lundwall, Harriet Nannyonjo, Suzette Samms- and its affiliated organizations, or those of the Lindsay, Sona Varma and Davide Zucchini Executive Directors of the World Bank or the (Jamaica); Jean-Michel Happi and Gulmira governments they represent. xi Introduction In anticipation of the December 2015 expiration of income per capita. Subject to data ­ constraints, of the Millennium Development Goals (MDGs), the framework strives to address the aspects of the UN General Assembly adopted in September the SDG agenda that matter to ­ country-level 2015 an ambitious agenda for sustainable devel- decision making. opment up to 2030. The agenda is based on The country briefs cover countries with diverse Sustainable Development Goals (SDGs) for 2030 prospects for realizing the ambitions of the global that cover economic, social, and environmental SDG agenda by 2030: For example, they include dimensions of development.1 It is made up of 17 countries that face adverse growth prospects and goals and 169 targets that together offer a com- that, compared to other countries with similar prehensive view of what is needed for sustainable per  capita incomes, are underperforming for at human well-being with a special emphasis on least some of the SDGs, two considerations that ending poverty, environmental sustainability, together make it difficult for them to reach the and inclusiveness. Individual countries face the 2030 targets for these SDGs. They also include challenge of translating these global ambitions countries that expect to grow rapidly and, for at into feasible strategies with clear targets and spe- least some SDGs, are currently performing better cific policies based on country circumstances than expected for their income per capita level; and initial conditions, and linked to country however, because of low initial outcomes, they are priorities. still projected to find it difficult to reach the global This book presents the Country Development 2030 targets. Other cases offer yet another exam- Diagnostics Post-2015 framework, developed by ple: because of low growth projections, they are the World Bank Group to assess the country-level not expected to realize global ambitions in spite of implications of the post-2015 global agenda, as strong initial SDG outcomes. well as brief, “at-a-glance” applications of the Figure I.1 exemplifies the diversity across our framework to ten countries: Ethiopia, Jamaica, country sample for one SDG indicator, access to the Kyrgyz Republic, Liberia, Nigeria, Pakistan, electricity, for which it shows current and pro- Peru, the Philippines, Senegal, and Uganda—see jected outcomes; it also shows the growth projec- map I.1. This set of countries is highly diverse in tions for GNI per capita that affect differences terms of growth prospects, per capita incomes, between countries in the rates of improvements outcomes for SDG target indicators, natural in electricity access.2 While some of the countries resource endowments, and physical access to may reach or get close to the global target of uni- international trade (landlocked or not). For each versal access, it is also clear that this target is very country, the analysis is designed to offer a start- distant for others. This exemplifies the fact that it ing point for a discussion of how policies and will be very difficult to realize global ambitions in financing should be designed to speed up devel- all countries, that to even get close to such out- opment outcomes by (a) benchmarking recent comes, countries should look for policies that outcomes for SDG target indicators and the fac- permit them to progress beyond what is typical tors (including policies) that influence them; (b) given expected levels of incomes in 2030. For projecting 2030 outcomes for those indicators; some countries it may be more productive to set and (c) assessing options to increase financial their own targets, striking a more appropriate resources for spending in support of the SDG balance between ambition and realism. agenda. The analysis is done from a cross-­country The country briefs also illustrate differences perspective, with the outcomes assessed relative in  SDG progress via provision of more and to what is typical for countries at the same level higher-quality public services. For example, given ­ Introduction 1 Map I.1  The Ten Country Briefs Kyrgyz Republic (Resource rich; landlocked) Pakistan Jamaica Republic of Philippines (Small island) Senegal Yemen Nigeria (FCS) Ethiopia (Resource Liberia rich) Uganda (Resource rich; FCS) (Landlocked) Peru (Resource rich) Low income Lower-middle income Upper-middle income This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank GSDPM Map Design Unit Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. Figure I.1  Recent and Projected Levels of Access to Electricity (% of Population), and Real GNI per Capita Growth, by Country 100 6 90 80 5 70 4 60 50 3 40 30 2 20 1 10 0 0 n lic es a u a ia ia a l ga ta i c r nd op er er ub Pe n ai is ne pi ga b ig m hi ep k Li ilip N Pa Ja Et Se U R Ph z r gy Ky Access to electricity, recent Access to electricity, 2030 GNIpc growth 2013–30 Source: Authors’ calculations. Note: Recent data is the most recent country observation during 2010–13. 2 Introduction current levels of taxes and other revenue sources, In outline, the contents of this book are some countries may find it difficult to raise o ­ rganized as follows: chapter 1 presents the resources for SDG-targeted policies without the Country Development Diagnostics Post-2015 additional government resources that a growth framework in detail, including examples from acceleration would make available. In addition, for Uganda. Chapters 2–11 present the ten country most SDGs, accelerated inclusive growth would briefs, country by country. also promote the SDGs by improving the living conditions of disadvantaged groups and others. In contrast, other countries may find it very difficult to foster growth beyond what is already projected; Notes however, it may be easier to increase resources 1. Other key documents on the 2030 SDG agenda through other means, such as higher taxes or include the report of the Open Working Group of improvements in government efficiency. the General Assembly on Sustainable Development Goals (UN 2014) and the High-Level Panel of Each country brief provides a starting point Eminent Persons on the Post-2015 Development for an in-depth analysis of the SDG agenda that Agenda (UN 2013). Our analysis covers the SDG could draw more fully on the framework and indicators for which data are available in multi-­ cover additional indicators. In general, the country databases. framework and available cross-country data offer 2. For a specific country, the higher rate of GNI per analysts in developing countries and the broader capita growth, the stronger improvement in access to electricity. However, comparing across coun- international community a useful starting point tries, the impact from economic growth is smaller for analyses that translate the global ambitions of the smaller initial GNI per capita (the increase in the agenda into country-specific targets and pol- resources are smaller) and the higher initial access icies that help accelerating progress. to electricity in the country (diminishing returns). Introduction 3 Chapter 1 Framework for Country Development Diagnostics Post-2015 1. Introduction (a)  poverty reduction and shared prosperity, (b)  infrastructure (water, sanitation, electricity, In setting the post-2015 Sustainable Development roads, and information and communications Goals (SDGs), the global community will need to technology, or ICT), (c) education, (d)  health, take cognizance of challenges to implementation and (e) climate change. Several indicators are and financing at the country level. This will used to measure progress of goals in each of necessitate integrated discussion of the develop- these areas, limited by what is available in cross-­ ment goals and the associated financing frame- country data sets. The aim of this chapter is to work. Financing in particular will have to be concisely present the analytical framework, structured in a way that taps into and leverages a using data for Uganda for illustration, it is more variety of financing sources beyond aid, and the selective in terms of both SDGs and the indica- policy framework will have to ensure ­ private sec- tors used. tor efficiency and improved public sector pro- The questions that the framework helps to ductivity. The ability to leverage diverse financing address include: For any country, what would be will differ from country to country, typically with a set of feasible development targets for 2030 if less ability for low-income and/or conflict-­ the country were to develop with the current affected countries. Given vastly d ­ ifferent capabil- income projections? What policy areas should ities, histories, starting points, and circumstances, the country’s government consider in order to the SDG agenda adopted by the UN General accelerate progress? How could it create the fiscal Assembly in September 2015 states that each resources needed to achieve more ambitious government should choose the appropriate level development outcomes? of ambition for each target, since every country More concretely, the framework benchmarks cannot be expected to reach the same absolute country performance in SDGs, policies, and target.1 other determinants (factors that influence SDGs). This chapter presents the Post-2015 Country It makes projections for SDGs to the year 2030, Development Diagnostics framework, devel- analyzes spending adjustments in ­ priority areas, oped by the World Bank Group with the aim of and discusses sources of fiscal space. Cross- providing a starting point for policy makers and country regressions of SDGs and their determi- researchers who are analyzing the implications nants on GNI per capita play a c­ entral role in the of the challenges of achieving the SDG agenda in analysis. The advantages and ­ disadvantages of different countries. The framework is designed (typically more elaborate) cross-country regres- for application in countries with a wide variety of sions have been discussed extensively in the liter- characteristics, including differences in initial ature.3 Our use of this tool is simple and conditions and access to financing, and provides transparent, drawing on the observation that a starting point for more detailed analysis.2 It many development indicators, including SDGs benchmarks a country’s achievements, provides and their determinants, are highly correlated projections up to 2030, and helps policy makers with GNI per capita. For such indicators, we ask questions about SDG targets and policy view GNI per capita as a summary indicator of options. It covers the following SDG areas: the  basic capacity of a country to bring about Framework for Country Development Diagnostics Post-2015 5 outcomes, both for SDGs and their determinants. • Step Four addresses challenges related to This does not translate into GNI being a direct or expanding fiscal space. In this context, the single determinant of outcomes—it is merely a analysis considers Uganda’s options for creat- benchmark and starting point for how a country ing fiscal space (through additional financ- performs relative to others at its income level. It ing  and government efficiency gains), again is noteworthy also that certain indicators, such as by looking at Uganda’s current situation com- the income share of the bottom 40 percent (a key pared to what is expected for a typical country measure of shared prosperity), are largely unre- at its GNI per capita. These findings for fiscal lated to GNI per capita. This points to the fact space are then compared with the assessment that purposeful measures are crucial to change of spending priorities identified in Step Three. for many development outcomes: in this case, The chapter concludes with a summary of growth does not, in any regular fashion, directly ­ findings  for Uganda and a discussion of how or indirectly, stimulate processes that bring forth this  framework may be applied to a variety of shared prosperity. countries. Underpinning the analysis is a database that covers all low- and middle-income countries, designed to include available indicators relevant to the post-2015 agenda, including SDGs, their 2. Step One: Benchmarking SDG determinants, and indicators related to financing Progress options. Subject to data availability, the database In this step, cross-country regressions are used to covers key aspects of the post-2015 agenda that assess the performance of the case study country can be meaningfully analyzed in a framework of in terms of SDGs, relative to its level of GNI per the type developed here. capita (box 1.1 provides the rationale). The purpose of this chapter is to illustrate our Here we will exemplify the SDG benchmark- framework, drawing on examples from Uganda. ing approach by analyzing primary and second- The infographic on the next page presents a sim- ary education in Uganda.4 Figure 1.1 shows two plified, visual overview of the framework. The scatter plots with each observation representing analysis is made up of four steps: a country’s position relative to its GNI per • Step One benchmarks Uganda’s current SDG capita and the SDG, the latter represented by outcomes against those of other countries, ­ primary school enrollment on the left and given the levels of GNI per capita. primary completion on  the right. The fitted, ­ • Step Two projects Uganda’s business-as-usual straight line represents expected school enroll- (BAU) levels for the SDGs in year 2030, draw- ment or completion levels for countries at ing on GNI per capita projections. different levels of GNI per capita. Countries ­ • Step Three tries to assess how to achieve more outside the shaded area are significantly over- ambitious targets than those suggested by the ­ or underperforming relative to their GNI per BAU projections. To this end, it b ­ enchmarks capita. Hence, for Uganda, net enrollment in the current levels of the determinants of the primary is significantly higher than expected, various SDGs for Uganda and compares them while primary completion rates are significantly to those of other countries in order to assess lower than expected. Figure 1.2 shows similar spending priorities. Determinants for  which information for secondary education in Uganda is ­ significantly lagging behind other Uganda: gross enrollment rates are significantly countries with a similar level of GNI per capita lower than expected but completion rates are as are singled out for special consideration. expected.5 6 Framework for Country Development Diagnostics Post-2015 BOX 1.1  Using GNI per Capita for SDG Benchmarking GNI per capita plays a central role in the analysis. Its level is highly correlated with most SDG indicators for several reasons, perhaps most importantly due to the fact that GNI per capita is highly correlated with determinants of SDGs, including (a) per capita household incomes, parts of which is spent on items that contribute to SDGs (for example, on health, education, and electricity); and (b) tax revenue and govern- ment capacity, which contributes to the fiscal space for government spending in areas that, directly or ­ indirectly, contribute to SDGs (most importantly, government services and infrastructure). Causality may also go in the opposite direction: the levels for different SDGs (for example, those related to health and education) may influence GNI per capita. Cross-country, constant-elasticity regressions are first used to benchmark current SDG outcomes—that is, to assess whether a country is over- or underperforming for an SDG relative to its GNI per capita.a Hence, for individual countries, deviations from predicted SDG values may be viewed as an indication of how well a country does relative to its capacity to achieve outcomes and provide inputs (determinants). Instead of GDP per capita (a production measure), GNI per capita, an income measure, is used since it conceptually is more closely related to a country’s capacity to achieve SDGs. a. These simplified regressions are useful for current purposes (benchmarking and projections). However, they do not claim to sort out interactions between different indicators, a difficult task given high degrees of correlation, lagged effects, complex time- and ­space-specific relationships, and data limitations. Figure 1.1  Uganda—Primary School Net Enrollment and GNI per Capita (Left); Primary School Completion and GNI per Capita (Right) 5.0 5.0 School enrollment, primary (% net) total (% of relevant age group) Primary completion rate, UGA 4.5 4.5 4.0 4.0 UGA 3.5 3.5 150 400 1,100 2,980 8,100 150 400 1,100 2,980 8,100 GNI per capita (constant 2005 US$) GNI per capita (constant 2005 US$) ln(DET) = 3.924*** + .073*** ln(INC); R 2 = .198 ln(DET) = 3.315*** + .153*** ln(INC); R 2 = .421 Sources: WDI, EdStats. 8 Framework for Country Development Diagnostics Post-2015 Figure 1.2  Uganda—Secondary School Gross Enrollment and GNI per Capita (Left); Secondary School Completion and GNI per Capita (Right) 5 6 School enrollment, secondary Secondary completion rate 4 4 (% gross) 2 UGA 3 0 UGA –2 2 150 400 1,100 2,980 8,100 150 400 1,100 2,980 8,100 GNI per capita (constant 2005 US$) GNI per capita (constant 2005 US$) ln(DET) = 2*** + .297*** ln(INC); R 2 = .55 ln(DET) = −.348 + .48 ln(INC); R 2 = .072 Sources: WDI, EdStats. 3. Step Two: SDG Business-as- be expected given a country’s initial conditions, projected growth in GNI per capita, typical rates Usual Projections of progress according to cross-country patterns, If the relationship between GNI per capita and and gradual convergence to close gaps between an SDG is considered tight enough, then the GNI observed and expected values.7 For any SDG, data for the country in question are used, not projections are presented only if the fit between only to benchmark the initial SDG outcome but GNI per capita and the SDG is considered suffi- also to project BAU SDG outcomes for 2030. For ciently tight (box 1.3). this, we need projections of GNI per capita. Table 1.2 presents recent values and BAU Box 1.2 discusses alternative sources for GDP projections to 2030 for Uganda for a set of ­ and GNI projections, which are available for most SDG  indicators, including those shown in countries. Figure 1.3 uses three of these sources figures 1.1 and 1.2, using a 2030 GNI per capita ­ to show Uganda’s projected (indexed) levels of of US$817. As explained under Step 1, Uganda GDP per capita up to 2030 (and, for comparison, is currently overperforming in its primary the historical development since 1990), while school net enrollment rate (indicated by green table 1.1 presents growth rates. We opted for the text in table  1.2); however, the cross-country CEPII’s EconMAP projection, which for Uganda relationship is not tight  enough to make a has a growth rate for GNI per capita of 4.0 percent relevant BAU ­ ­ projection for  2030. For the per year (at constant 2005 US$), translating to an primary school completion rate,  Uganda is ­ increase from US$378 in 2011 to US$817 in 2030 under­ performing (indicated by red  text). The (both at constant 2005 prices), a level similar to projected BAU value in 2030 is 66.1 percent, an the current levels of countries such as Vietnam, increase due mainly to GNI per capita growth India, and Senegal.6 Considering the range of but influenced also by the convergence effect. alternative projections, an annual per capita Substantial progress is recorded for other indi- growth rate of 4  percent seems realistic, if per- cators, but without realizing global ambitions: haps erring on the moderately optimistic side. for example, the extreme poverty rate declines The levels of selected SDGs are projected to very strongly. 2030. These BAU projections reflect what can Framework for Country Development Diagnostics Post-2015 9 BOX 1.2  Projecting GDP and GNI Aggregate growth projections covering most countries are produced by various international organiza- tions, including the World Bank, IMF, CEPII, OECD, and IIASA, but also by most governments and other sources, such as Hausmann et al. (2011). From the projections, it is difficult to determine which source is most reliable. Moreover, given the fact that available sources project only GDP while our framework uses GNI data, we have to assume, for most countries quite reasonably, that projected GNI growth will not deviate substantially from projected GDP growth (both expressed in constant 2005 ­ ecessary, US$).a In any country case study, it is good practice to compare different projections and, if n refine what is available. a. As indicated by the names of the terms, GDP is primarily a measure of production while GNI is an income measure, more specifically GNI = GDP plus net receipts from abroad of primary income (compensation of employees and property income). For most countries, the two measures are highly correlated; among low- and middle-income countries, they tend to diverge most strongly in countries where (net) FDI over time has represented a substantial share of total private investment, often in natural resource sectors, generating substan- tial profit remittances to the foreign investors. If additional information is available on how future GNI and GDP growth may differ for a country, then such information should be reflected in the GNI projections. Figure 1.3  Uganda—Historical Data and influenced by policy in the short to medium terms. Projections for Real GDP per Capita The purpose is to assess the feasibility of policy (2011=100) changes that accelerate SDG  progress and make more ambitious targets possible. Policies may 250 influence SDGs in two ways, by: (a) raising the 200 level of GNI per capita, which in turn, through various channels, affects SDGs, and (b) improving 150 country SDG outcomes relative to what is expected 100 given its GNI per capita. To illustrate, if a country underperforms in 50 both an SDG and its more important determi- 0 nants, then policy actions may be both feasible and rewarding. Examples include government 1990 1992 1994 1996 20 8 2000 2002 2004 2006 2008 2010 2012 2014 2016 20 8 2020 2022 2024 2026 2028 30 9 1 19 IASA OECD CEPII Historical spending in various areas and the related provi- Sources: WDI, IIASA, OECD, and CEPII. sion of inputs crucial to SDG progress. Such poli- cies may have an influence directly (by having a direct bearing on specific services—for example, 4. Step Three: Benchmarking health services targeted to reduce maternal mor- Determinants and Identifying tality) and/or indirectly (by contributing to Spending Priorities capacity-creating economic growth). The ­ ­ discussion of major policy changes has direct Current Performance of implications for costs and financing needs. Determinants The determinants—in our cross-country In Step 3, we regress SDG determinants against database represented by over 200 indicators— GNI per capita (in Step 1, we did this for SDG indi- may be classified according to which of the cators; cf. box 1.1). The identification of determi- ­ following four areas they impact: economic nants is guided by previous country and growth, education, health, and climate change. cross-country research, limited to indicators that In the fifth area that our approach covers— are available in cross-country databases. We SDGs related to access to infrastructure—the emphasize those determinants that may be basic approach is simpler: deviations are viewed 10 Framework for Country Development Diagnostics Post-2015 Table 1.1  Uganda—Historical and Projected Growth from Various Sources Average annual Source Time period Indicator (real values) Comment growth (%) WDI 3.3 1990–2012 GDP per capita Data used in figure 1.3 for period up to 2012 WDI 3.2 1990–2011 GNI per capita GDP per capita growth for 1990–2011 was 3.5 percent CEPII 4.0 2013–30 GDP per capita OECD 3.8 2013–30 GDP per capita IIASA 2.5 2013–30 GDP per capita IMF (2013b) 3.7 2013–30 GDP per capita Including oil revenues, adjusted for population growth Hausmann et al. 3.3 2009–20 GDP per capita Based on the Economic Complexity Index (2014) Republic of Uganda 5.6 2014–40 GDP per capita Calculation based on data for GDP growth and 2014, pp. 27, 30, 53 population in Uganda’s Vision 2040 BOX 1.3  SDG Business-as-Usual Projections for 2030 If the fit between GNI per capita and an SDG indicator is reasonably tight (which tends to be the case), the results of a cross-country regression permits us to compute projected business-as-usual 2030 values.a A loose relationship suggests that progress in the indicator is primarily a reflection of country-specific factors and that it should not be expected to respond strongly or systematically to changes in GNI per capita. When the relationship to GNI per capita is loose the coefficients are typically small (in absolute ­ terms); given this, the “expected” values for a recent year are close to the average for all low- and ­middle-income countries.b a. A tight enough relationship is defined as an R  2  > 0.3 (tight) or 0.3 > R  2  > 0.1 (moderately tight), while R  2 <0.1 are defined as loose. b. In addition, the confidence interval is wide in the case of a loose relationship, suggesting that any conclusion on over- or ­ under­performance is made with wide margins. Statistically, even though their confidence intervals are wide, as long as the estimated coefficient linking GNI per capita to the SDG indicator is nonzero, these values are closer than the cross-country average to what is ­ expected for the specific country. The same observation applies to expected values for fiscal space indicators. Table 1.2  Uganda—SDG Projections for 2030 SDG Recent value BAU projection for 2030 Poverty rate at $1.25 a day (PPP) (% of population) 38.0 11.5 Malnutrition (weight for age: % of children under 5) 14.1 8.8 Income share, bottom 40% (% of total income) 15.5 — Gini Index 44.3 — Access to improved sanitation (% of population) 33.9 44.8 Access to improved water (% of population) 74.8 80.7 Access to electricity (% of population) 14.6 31.0 Road density (km road per 100 sq. km of land area) 32.2 35.8 Internet use (% of population) 14.7 — Mobile cellular subscriptions (% of population) 45.0 — Net enrollment, pre-primary (%) 13.6 20.4 Net enrollment, primary (%) 90.9 — Primary completion rate (%) 53.1 66.1 Gross enrollment, secondary (%) 27.6 41.6 Secondary completion rate (%) 9.4 — table continues next page Framework for Country Development Diagnostics Post-2015 11 Table 1.2  continued SDG Recent value BAU projection for 2030 Maternal mortality (modeled estimate, per 100,000 live births) 310.0 146.3 Under 5 mortality (per 1,000 live births) 68.9 42.7 Prevalence of HIV total (% of population ages 15–49) 7.2 — Malaria reported 7.3 1.3 Prevalence of tuberculosis 175.0 109.0 CO2 emissions per capita 0.11 0.39 Note: Green = currently significantly overperforming; red = currently significantly underperforming; black = performing as expected; — = no projection because indicator has too loose a relationship with GNI per capita. Whether a specific deviation (positive or negative) reflects a stronger or weaker performance varies across indicators. For example, a positive deviation reflects weaker performance for poverty but stronger performance for water access. The terms overperformance and underperformance are used normatively; for example, with regards to the maternal mortality rate, a lower-than-expected rate is reflected as overperformance. mainly as indicating insufficient levels of effi- (­ figure  1.5). As for secondary schools, the cient investments. Shared prosperity is not ­ xpenditures per student are as expected but the e addressed in a separate section but rather high- pupil-teacher ratio is lower than expected. The lighted throughout. Wherever data allow, the fact that the completion rate is as expected results of the sample of the bottom 40 percent while  the enrollment rate is below expectations are presented, and indicators such as those (both rates are computed relative to the total pop- related to education and health, access to ulation in relevant age groups) suggests that the finance, and secondary road infrastructure are system performs relatively well for its ­ ­ spending given special attention. It is important to note level in bringing enrolled students to completion. that some determinants influence several SDGs, A more detailed investigation is needed to assess and that SDGs may be determinants of other the room available for efficiency improvements. SDGs.8 Of course, the fact that cross-country Table 1.3 presents findings for a longer list of analysis has shown that a certain determinant determinants, chosen from those that are directly matters for an outcome does not necessarily policy relevant, not only for education but also mean that it is important in a specific country for other SDGs, giving a flavor of the type of setting; conversely, a lack of evidence on the determinants that may be analyzed in a more cross-country level does not necessarily mean a detailed study. In addition to the determinants determinant is unimportant for a specific coun- in the table, household incomes per capita try. In order to arrive at more definitive conclu- (highly correlated with GNI per capita) and sions for a given country, it is necessary to assess some of the other SDGs, including those related and enrich the findings of our analysis, drawing to infrastructure—for example, access to safe on additional country information. water affecting health indicators—may also mat- To demonstrate this step, we look at expendi- ter. For those in red text, performance is signifi- tures per student at the primary and secondary cantly weaker than expected relative to Uganda’s school levels, highlighting data for Uganda GNI per capita, suggesting that improvements in figure 1.4): at the primary school level, spending (­ policies and outcomes in these areas may be is significantly lower than expected while, at the most feasible. secondary school level, it is within the expected range. These findings may help to explain the enrollment-completion puzzle presented in Identifying Spending Priorities Step  1:  Uganda’s lower than expected p ­rimary A cross-country perspective can shed useful ­ completion  rate may be due to lower-than- light  on spending decisions, which are espe- expected expenditure per student and, as a related cially difficult when made in a situation such as matter, a higher-than-expected pupil-teacher ratio Uganda’s, where large unmet needs coexist with a 12 Framework for Country Development Diagnostics Post-2015 Figure 1.4  Uganda—Expenditure per Primary Student and GNI per Capita (Left); Expenditure per Secondary Student and GNI per Capita (Right) 4.0 5 Expenditure per student, secondary Expenditure per student, primary 3.5 (% of GDP per capita) (% of GDP per capita) 4 2.5 UGA 3 3.0 UGA 2 2.0 1.5 1 150 400 1,100 2,980 8,100 150 400 1,100 2,980 8,100 GNI per capita (constant 2005 US$) GNI per capita (constant 2005 US$) ln(DET) = 1.581*** + .138*** ln(INC); R 2 = .09 ln(DET) = 3.476*** −.082 ln(INC); R 2 = .027 Sources: EdStats, World Bank. Figure 1.5  Uganda—Primary Pupil-Teacher Ratio and GNI per Capita (Left); Secondary Pupil-Teacher Ratio, Secondary and GNI per Capita (Right) 4.5 4.5 Pupil−teacher ratio, secondary Pupil−teacher ratio, primary 4.0 UGA 4.0 3.5 3.5 3.0 3.0 UGA 2.5 2.5 2.0 2.0 150 400 1,100 2,980 8,100 150 400 1,100 2,980 8,100 GNI per capita (constant 2005 US$) GNI per capita (constant 2005 US$) ln(DET) = 5.561*** −.315*** ln(INC); R = .508 2 ln(DET) = 4.751*** −.247*** ln(INC); R 2 = .408 Sources: EdStats, World Bank. Framework for Country Development Diagnostics Post-2015 13 Table 1.3  Uganda—Policy-Relevant SDG However, while spending per student as percent Determinants of GDP is less than expected, its spending on pri- SDG Recent mary education as percent of GDP is as expected. value The reason for this seeming contradiction is that Government consumption (% of GDP) 11.3 enrollment is relatively high, largely due to high Public investment (% of GDP) 6.7 rates of repetition and enrollment of students who Logistic performance index 2.8 are older than the expected age for their grade. If Ease of doing business rank 132.0 repetition rates can be reduced and completion Public expenditure per student, primary 7.6 (% of GDP per capita) rates increased—something that may require Public expenditure per student, secondary 20.7 more spending per student—the GDP share for (% of GDP per capita) primary spending required to offer services simi- Public expenditure per student, tertiary (% of GDP per capita) 45.6 lar to those of other countries will eventually Public expenditure, primary (% of GDP) 1.8 decline as students graduate from the primary Public expenditure, secondary (% of GDP) 0.8 level. All things considered, an initial jump in the Public expenditure, tertiary (% of GDP) 0.4 GDP spending share to 2.5 percent of GDP (com- Pupil-teacher ratio, primary 47.8 pared to the current 1.8 percent of GDP) would Pupil-teacher ratio, secondary 18.5 raise spending to the expected level. However, Public health expenditures (% of GDP) 2.5 even though such increased spending would raise Contraceptive use (% of population) 30.0 per-student resources to what is typical for coun- Physicians (per 1,000 people) 0.12 tries at Uganda’s GNI per capita, it still remains far Skilled staff at birth (% of births) 57.4 below what may be needed to offer a quality pri- Adolescent fertility rate (per 1,000 girls 15–19) 131.0 mary education.9 For secondary education, the Fertility rate (births per woman, 15+ years of age) 6.1 enrollment rate and spending as percent of GDP Note: Green = currently significantly overperforming; are both lower than expected while completion red = currently significantly underperforming; black = performing as expected. The terms overperformance and underperformance rates (measured relative to the population in the are used normatively; for example, with regards to the maternal mortality rate, a lower-than-expected rate is referred to as relevant age cohorts) and spending per student as overperformance. percent of GDP are as expected. As Uganda in the future meets the challenge of increasing the num- ber of entrants that proceed from primary, the constrained capacity to scale up spending with demands for ­ public spending on secondary edu- retained efficiency. Naturally though, spending cation will increase. As a result of expansion at priorities need to discuss the cross-country result lower levels, the demand for tertiary education keeping country-specific conditions and recent will also increase, albeit with a lag. In 2011, public developments in mind. spending on tertiary education was 0.4 percent of At the aggregate level, Uganda’s spending- GDP, less than expected. Like primary education, to-GDP ratio is low relative to its GNI per capita keeping spending per student as percent of GDP for aggregate public consumption (at 11.3 percent at expected levels may not be sufficient to offer a of  GDP in 2011, falling short by 2 percentage quality education.10 points) and, to a lesser extent, for aggregate pub- In addition to education, health and infra- lic investment, suggesting that some expansion structure are two major SDG-related spending would not put excessive pressures on financing priorities for a low-income country like Uganda. or institutional capacity. In health, key indicators such as under-five The above analysis focused mainly on primary and  maternal mortality rates, are at expected and secondary education. At the primary level, levels while total health spending is higher ­ Uganda’s government spent around 7.6 percent of than  expected (9.5 percent compared to an GDP per capita per student in 2011 (table 1.3), expected 5.9 percent of GDP). At a more disag- which is less than the expected 11.0 percent. gregated level, public spending is roughly as 14 Framework for Country Development Diagnostics Post-2015 expected (2.5  percent of GDP) and private 5. Step Four: Identifying ­ spending higher (7.0 percent of GDP compared Fiscal Space to an expected level of 3.0 percent) (Gable, Lofgren, and Osorio-Rodarte 2014). In the short The level and efficiency of public spending are to medium runs, the ability of the public health typically among the determinants of the develop- sector to absorb additional spending while ment of SDGs and their determinants. However, ­ maintaining efficiency is severely constrained by it is important to keep in mind that any given a lack of qualified manpower, while waste is level of spending may take place within a wide ­ substantial, estimated at 13 percent of spending range of policy frameworks, among other things, for 2005–06 (Okwero at al. 2010, pp. 47, 65–68). with varying roles for public and private service Meanwhile, the level of spending on current delivery. Also, the means by which resources are health Millennium Development Goals (MDGs) mobilized makes a difference to outcomes—for is well below the recommended minimum— example, the effects of additional aid are different US$54 per capita at 2005 prices (Task Force on from the effects of additional taxes.11 Innovative International Financing for  Health Here we primarily address fiscal space from a Systems 2009, p. 11; WHO 2010, pp.  36–37); if budgetary perspective since, by definition, bud- projected growth rates are achieved, Uganda’s get resources are most directly controlled by pol- total health spending would not reach this level icy makers. However, as will be noted, financing until about 2020. In other words, further financ- from NGOs and private investors may play an ing for increased health services will be a high important complementary role. Our framework priority, especially if the government managed to is comprehensive, analyzing the scope for creat- overcome the m ­ anpower and other constraints to ing additional fiscal space from taxes, fossil fuel increased absorptive capacity in the health sector. subsidy cuts, overseas development assistance Regarding infrastructural development, (ODA—that is, grants and concessional loans), investments, and spending on operations and and other borrowing (domestic or foreign). It is maintenance (in such sectors as water, sanita- also important to bring government efficiency tion,  roads, electricity, and information and into the analysis: if it is low initially, then communications technology, or ICT), are crucial improvements may release substantial resources for Uganda’s SDG agenda. But, despite having for additional high-priority spending without spent heavily on infrastructure during 2001– additional financing. If efficiency initially is high, 09—at slightly above 10 percent of GDP, or US$1 then this source of fiscal space is less important. billion per year—Uganda still lags behind com- However, if so, the government is likely in a bet- parator countries in electricity supply, is severely ter position to use additional financing to scale challenged in achieving universal access to sani- up services and investments in  priority areas tation and considerably lacking in provision of while maintaining acceptable efficiency.12 running water and other services. According to Drawing on the summary in table 1.4, among Ranganathan and Foster (2012, p. 42), a program the potential sources of fiscal space for priority for accelerated (but still not unreasonable) prog- spending, we find the following: ress may require annual spending of an addi- • Nonoil taxes. Tax revenues are the main tional US$400 million per year (in 2011 US$) source  of government financing in Uganda. through 2015, corresponding to around 2.4 Figure 1.6 shows how they have evolved since percent of GDP. Given the importance of infra- 1990, and benchmarks their current GDP structure access within the SDG agenda, and its share against those of other countries.13 As key role in raising growth and contributing to a shown, Uganda’s tax revenue, at 13 percent of wide range of development goals, it would be GDP in 2011, is as expected. The relationship crucial to continue to improve services in this with GNI per capita is not tight enough to area up to 2030. Framework for Country Development Diagnostics Post-2015 15 project future changes on the basis of pro- production ends and reserves are depleted; jected income growth. If nonoil tax policy for  the period 2016–30, oil revenues may were to change, then it would be important to amount to an average of roughly 4.9 percent of consider the detailed design and likely effects GDP per year (IMF 2013b, p. 57). on the SDG agenda of such changes, compar- • Fossil fuel subsidies. Currently Uganda’s ing the benefits from additional spending to ­subsidy level is at around 1.3 percent of GDP. the costs related to a reduction of the resources Subsidy reduction is thus a potential source controlled by households and enterprises.14 of fiscal space and would contribute posi- • Oil taxes. While considerable uncertainty is tively to the climate change agenda. It is dif- related to the oil sector—currently, 2018 is the ficult to assess the likelihood of reforms in expected starting year for production—it is this area. likely that the sector will generate a substantial • Official Development Assistance (ODA). increase in tax revenues. According to one set Uganda’s net ODA is at around 10.1 percent of of projections, the tax revenues from oil will GNI (9.4 percent of GDP), also roughly at reach 8 percent of GDP by 2023, after which the expected level (11.1 percent of GNI). The they will decline gradually until 2045, when cross-country relationship between GNI per Table 1.4  Government Fiscal Space—Recent Indicators and Future Directions of Change Income and efficiency indicators Recent value Impact on future fiscal space Comment Taxes (% of GDP) 13.0 + Likely increase (mainly due to revenues from oil sector) Fuel subsidies (% of GDP) 1.3 + Potential (and desirable) decrease ODA (% of GNI) 10.1 – Likely decrease External debt stocks (% of GNI) 22.5 + Potential room to increase borrowing Government efficiency + Potential (and desirable) increase Figure 1.6  Uganda—Tax Revenues 1990–2011 (% of GDP) (Left); Tax Revenues (% of GDP) versus GNI per Capita (Right) 14 4 12 UGA Tax revenue (% of GDP) 10 2 8 0 6 4 −2 2 0 −4 150 400 1,100 2,980 8,100 11 10 01 07 02 04 05 03 06 08 09 00 20 20 20 20 20 20 20 20 20 20 20 20 GNI per capita (constant 2005 US$) Tax revenue (% of GDP) ln(DET) = 1.794*** + .12 ln(INC); R 2 = .027 Sources: WDI, World Bank. 16 Framework for Country Development Diagnostics Post-2015 capita and ODA (as percent of GNI, or GDP) Debt Sustainability Analysis (DSA) considers suggests that Uganda’s ODA will decline as sustainable an increase in Uganda’s exter- relative to both GNI and GDP (figure 1.7, ­ nal  public or publicly guaranteed debt from left  panel) while remaining constant in per 16  percent of GDP in 2012 to 22 percent capita terms. The likely advent of large oil in  2033; this permits additional annual ­ revenues may lead to further cuts as donors borrowing of roughly 0.3 percent of GDP. In ­ turn to countries with more severe fiscal con- the DSA, it was assumed that other debt straints. The projected 2030 level of ODA for stocks—public domestic and external private Uganda—taking only the increased GNI per non-guaranteed—would not change from capita into ­account—is as low as 4.2 percent of their current GDP shares of  13  percent and GDP or, in an average year during 2016–30, 10 percent, respectively (IMF 2013b). around 6.1 percent of GDP, that is, a loss of • Government efficiency. A number of 3.4 percentage points. To limit this loss, it may government efficiency measures are available ­ be possible to tap into global initiatives, such (box 1.4). According to both the health and the as the Global Fund to Fight AIDS, Tuberculosis education indexes, Uganda’s performance is and Malaria. below the expected levels; among these two • Borrowing. Uganda’s external debt stocks have indexes, GNI per capita is strongly correlated decreased substantially, not least following the with the education index but largely uncor- HPIC initiative, and the current 22.5 percent related with the health index. Uganda is of GNI is lower than expected. Again, the rela- performing as expected in terms of the more ­ tionship to GNI per capita is not tight enough general Public Investment Management Index to make projections based on cross-country and better than expected according to the results. However, a recent IMF-World Bank World Bank Governance Indicators. Given that Figure 1.7  Uganda—ODA (% of GNI) versus GNI per Capita (Left); ODA (per Capita) versus GNI per Capita (Right) 50 3,000 Net ODA received per capita (current US$) UGA, 2012 20 1,000 400 Net ODA received (% of GNI) 7 3 150 1 50 log scale log scale 20 UGA, 2012 7 3 1 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) Function: In(SDG) = 8.35*** + –1.05*** x In(GNI pc); Function: In(SDG) = 4.61*** + –.09 x In(GNI pc); R 2 = .303 R 2 = .004 Sources: WDI, World Bank. Framework for Country Development Diagnostics Post-2015 17 the different indexes measure d ­ ifferent aspects GDP while maintaining acceptable efficiency. If of government performance, such mixed find- it were achieved, then gains in the SDG area ings may not be inconsistent. Among other could be considerable. For the sake of efficiency, country-specific sources, scattered survey evi- if spending is to be increased, it may be wise to dence also points to inefficiencies. For exam- do so gradually and seek guidance from fre- ple, on any given day, roughly 15–20 percent quent impact assessments. of the teachers (including head teachers with It is important to note that trade-offs are supervisory responsibilities) are absent, with involved, to varying degrees, when fiscal space is illness accounting for  an almost-negligible freed up and spending is increased according to share of absences (UNESCO 2014a, pp. 31 and priorities: policy makers need to think through 267–268). Similarly, an analysis of local scenarios for Uganda with and without major governments suggests, if all districts could be ­ policy changes, and the implications for the brought up to  the health and education SDG agenda. The trade-offs may be least severe outcome-­ to-spending ratios of the best per- for ­success in raising government efficiency and forming districts, then about one-third of ODA. For alternatives with different tax and sub- their budgets could be saved (World Bank sidy policies, the net short- and long-run impacts 2013c, p.  xiii). In  sum, even though they are on different population groups should be consid- unpredictable, efficiency gains have the poten- ered. Additional borrowing increases the risk of tial to add considerable fiscal space. unsustainable future debt levels. On balance, this information suggests the ­ scal space for SDG priority spending could fi increase by roughly 4–5 percent of GDP.15 6. Conclusions However, the extent of the increase is highly In this chapter, we present the Country uncertain, not least due to uncertainty regard- Development Diagnostics Post-2015 framework ing the future of the oil sector. In addition to for analyzing the implications for the SDG the sources included in the table, it may be pos- agenda at the level of individual low- and mid- sible to attract additional external private dle-income countries. The framework that we financing, especially for infrastructure invest- present is divided into a sequence of distinct ments, leveraged by additional government steps; each step is illustrated here with selected spending in this area. To provide context, findings from a more detailed country diagnostic according to recent figures, total government of Uganda (Gable, Lofgren, and Osorio-Rodarte spending amounts to around 20 percent of 2014). The fact that, in spite of accelerating prog- GDP (IMF 2013b, p. 28); it would be a severe ress, most countries will not achieve most of the challenge to raise spending by 4–5 percent of MDG targets by the 2015 deadline indicates that BOX 1.4  Measures of Government Effectiveness On the basis of relationships between inputs and outputs, Grigoli and Kapsoli (2013) and Grigoli (2014) constructed indexes for government efficiency in health and education spending; Dabla-Norris et al. (2011) developed a Public Investment Management Index (PIMI) that reflects actual practices in four areas (appraisal, selection, implementation, and evaluation). In addition, the World Bank Governance Indicators provide cross-country data on rule of law, government effectiveness, control of corruption, political stability and absence of violence, quality of regulations, and voice and accountability. 18 Framework for Country Development Diagnostics Post-2015 this is an important undertaking: while ambi- economy like Uganda’s that is expected to grow at tions should be global, in order to be effectively a relatively rapid pace and have access to addi- embraced, strategies and targets in individual tional foreign exchange resources (from oil). In countries should be locally owned and anchored other words, business as usual clearly is insuffi- in individual country realities and priorities.16 cient to achieve the global SDG ambitions. To The findings for Uganda—illustrating the accelerate progress, policy makers and country nature of country-specific insights that the leaders will have to prioritize government effec- framework may lead to—reveal a mixed picture tiveness and efficiency and ensure that develop- of how the country is performing compared to ment spending is raised and allocated to areas what is expected at its GNI per capita. The fact critical to the SDG agenda. that the country underperformed in various The Country Development Diagnostics Post- indicators may set off alarms and prompt more 2015 framework is intended to give analysts in detailed analysis, with the initial hypothesis that developing countries and the broader interna- improvements are clearly attainable in those tional community a useful starting point for areas. The analysis suggests that in some areas assessing policy priorities, targets, and financing certain linkages are at work (for example, options for virtually any low- or middle-income between relatively weak primary education out- country. The framework does not say what policy comes and the allocation of relatively few makers should do but it should help them pose resources per primary student). With regard to important questions and find answers, also draw- the SDG agenda, the results suggest that substan- ing on more detailed, country-specific studies.17 tial yet only moderate progress should realisti- Together, this information should provide helpful cally be expected by 2030. This is true even for an guidance for stronger SDG accomplishments. Notes 1. “The SDGs and targets are integrated and context. However, as noted by many (for exam- indivisible, global in nature and universally ­ ple, ADB  2006), cross-country regressions are applicable, taking into account different national often unable, for various interrelated reasons, to realities, capacities and levels of development successfully address the role of different deter- and respecting national policies and priorities. minants, severely limiting the usefulness of these Targets are defined as aspirational and global, with results to policy makers. More specifically, the each government setting its own national targets regressions tend to suffer from a lack of robust- guided by the global level of ambition but taking ness to different specifications; difficulty in assess- into account national circumstances. Each govern- ing the direction of causality between different ment will also decide how these aspirational and indicators (causality may often go in both direc- global targets should be incorporated in national tions); high correlations and complex interactions planning processes, policies and strategies. It is between determinants; variable relationships important to recognize the link between sustain- (across time and space); and imperfect indicators able development and other relevant ongoing pro- (for example, spending on human development is cesses in the economic, social and environmental an imperfect indicator of real services in human fields” (UN 2015; paragraph 55). development). 2. An initial version of the framework was applied in 4. In addition, the analysis may also review the evo- Gable, Lofgren, and Osario-Rodarte 2014. “Country lution of the SDG in recent decades as part of the Development Diagnostics Post-2015: Uganda.” assessment of initial country SDG performance. World Bank, Washington, DC. Compared to the In addition to benchmarking country perfor- Uganda chapter in this book, this chapter covers a mance against what is expected, it may also be larger part of the framework and more indicators. relevant to benchmark against top performance 3. Among the potential advantages is the abil- within countries that in other important respects ity to control for various alternative determi- remain similar to the case-study country. nants, and—when robust results are found—to 5. Uganda’s secondary completion rate is highly generalize results beyond the country-specific uncertain. Drawing on population, enrollment, Framework for Country Development Diagnostics Post-2015 19 and repetition data in EdStats, a rate of 9.4 percent ­ enefiting poor people in particular, is addressed b was calculated for 2011. in the seminal World Development Report of 6. We chose the projections of CEPII due to a 2004,  “Making Services Work for Poor People” c ombination of factors, including a transpar- ­ (World Bank 2003). According to the report, the ent model structure, clear documentation, and key  to improved service delivery is institutional comprehensive country coverage. See (http://​ changes that strengthen relationships of account- www.cepii.fr​/­C EPII/en/bdd_modele/bdd.asp). ability between policy makers, providers, and Note that the projected growth rate is from the citizens. A large body of research stimulated by most current projections at the time of the this report suggests that such institutional changes in-depth Uganda country study, which differs are possible but not easily implemented, largely from the projection in the Uganda brief pre- because politicians in many settings may be able sented later in this book. to resist accountability to citizens (Devarajan 7. Given that (a) SDGs have extreme values (such 2014; see also Overseas Development Institute as  100 percent for improved water access) and 2014). (b)  the current SDG level never is exactly as 13. Figure 1.6 suggests, interestingly, that ODA expected relative to GNI per capita, it is necessary per capita is unrelated to GNI per capita—that is, to incorporate convergence toward the expected there is no significant tendency to give higher aid value into the projections. It is here assumed that per capita to the countries where needs are highest. such convergence is gradual. For example, for a 14. IMF (2013b) suggests that, by 2018, an increase country that overperforms in water access, as GNI of 1.5 percentage points of GDP for nonoil would per capita increases the extent of overperformance be feasible; Uganda would still remain within its gradually declines, so that when the expected value expected range. is 100, overperformance has reached zero. 15. Using figures from the preceding discussion, a 8. For example, access to electricity is an SDG in its high estimate of the fiscal space increase may be own right and is likely also to influence both edu- as follows (all percent of GDP for an average year cation and health SDGs. 2016–30): 4.9 (oil taxes) + 1.5 (nonoil taxes) + 1.3 9. In 2011, at PPP in constant 2010 US$, average pub- (fuel subsidy cuts)—3.4 (ODA) + 0.3 (foreign bor- lic spending per primary student in ­ low-income, rowing) = 4.6. In addition, the government may middle-income, and high-income countries ­ be able to raise efficiency. However, as noted, the was US$94, US$554, and US$6,353, respectively changes for individual items are uncertain, diffi- (UNESCO 2014a, p. 383; UNESCO 2014b, table 11). cult to bring about, and/or subject to drawbacks 10. For Uganda and many other low-income coun- (especially if increased spending is not efficient). tries, the education quality gap and challenge 16. On the basis of data for 2010, Uganda seemed on is particularly strong at the primary level. This track to achieve the MDGs for extreme poverty, is because enrollment is higher at this level and education gender parity, under-five (and infant) spending per student tends to grow faster than mortality, and water access. On the other hand, GDP per capita (raising the value for spending Uganda was off track for undernourishment, per student as percent of GDP per capita), reflect- ­ primary completion, maternal mortality, and ing initial overenrollment relative to resources. ­ sanitation access (World Bank 2014a). At higher levels of education it is easier to manage 17. Such studies may be sector-focused or economy-­ the challenge: enrollment is smaller while growth wide. An economy wide approach is needed to in spending per student tends to be slower than consider the many interactions between policies, growth in GDP per capita. financing, growth, and SDG outcomes. MAMS 11. See World Bank (2013a) for a broader discussion (Maquette for MDG Simulations), initially devel- on financing instruments for the Post-2015 agenda. oped at the World Bank for analysis of MDG strat- 12. The challenges of raising government efficiency egies, is an example of such an approach. For more in  service delivery in general, and for services on MAMS, visit www.worldbank.org​ mams. /­ 20 Framework for Country Development Diagnostics Post-2015 Country Briefs Chapter 2 Ethiopia 1. Introduction Sections 2 and 3 address SDGs and fiscal space, respectively, while findings are summarized in Ethiopia, a land-locked, low-income country in Section 4. Africa, has experienced a remarkably strong eco- The analysis is done from a cross-country nomic growth driven in part by public infra- perspective: for the different indicators, Ethiopia’s structure investment. In coming years, it will be performance and prospects are benchmarked important  to monitor its sustainability, paying relative to other countries, considering its past, particular attention to infrastructure financing recent, and projected levels of GNI per capita.2 (see Moller and Wacker, forthcoming). During The latter variable tends to be highly correlated 2001–12, Ethiopia’s average growth rate for GDP with most of the SDGs and most of the factors per capita (at constant 2005 US$) was 6.0 percent. that determine their evolution; given this, it is Between 2000 and 2012, Ethiopia’s ranking used as a summary indicator of country capacity according to the UNDP Human Development to provide and efficiently utilize inputs that con- Index (among countries included in both years) tribute to SDGs (for example, health and educa- improved from the 1st to the 6th percentile. tion services) and to achieve SDG o ­ utcomes (like Hence, GDP growth and social improvements strong health and education results).3 have been impressive under the current eco- nomic model. However, judging from the devel- opment record of other successful developing 2. SDG Indicators: History and countries, Ethiopia still faces the long-run struc- tural challenge of finding a more prominent role Projections for the private sector.1 For 17 SDG indicators, table 2.1 summarizes This country-at-a-glance note is designed to data for Ethiopia: historical evolution, actual provide an initial picture of the challenges that and expected values for a recent year, and the Post-2015 agenda poses for Ethiopia; its find- projected 2030 values.4 In figure 2.1, data for ­ ings cannot guide policy on their own but should Ethiopia are shown in the context of the esti- be seen as an input into policy discussions. mated cross-­ country relationship between each The note may also serve as a starting point for a SDG ­ indicator and GNI per capita. For Ethiopia, more complete country development diagnostic the projected a ­verage annual rate of GNI per as well as a  more in-depth country-focused capita growth is 4.0 percent.5 The projected SDG analysis. The note is built around tables and ­ values reflect what can be expected given a coun- figures that provide data for a selection of SDG ­ try’s starting point, projected growth in GNI per target indicators and indicators related to fiscal capita, ­typical  rates of progress according to space—fiscal space matters since, while policy cross-country patterns, and a gradual conver- frameworks and the engagement of the private gence to close gaps between observed and sector may vary widely, rapid progress on the expected values.6 Projections for SDG indicators SDG agenda will require efficient and carefully are presented only when the cross-country rela- prioritized public spending. The note briefly tionship between the indicator and GNI per (a)  summarizes Ethiopia’s SDG progress since capita is classified as tight.7 A loose relationship 2000 and projects expected values for 2030; and suggests that progress in the indicator is primar- (b) assesses options for increasing fiscal space. ily a reflection of country-specific factors and Ethiopia 23 Table 2.1  Ethiopia—SDG Indicators: Evolution since 2000 and Projections to 2030 Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Poverty and shared prosperity Poverty, at $1.25 a day 54.6 30.7 68.4 10.3 By 2030, eradicate extreme poverty for all people everywhere, (PPP) (% of population) currently measured as people living on less than $1.25 a day. Shared prosperity: Income 22.3 20.4 17.6 — By 2030, progressively achieve and sustain income growth of share for lowest 40% the bottom 40 percent of the population at a rate higher than the national average. Education Pre-primary enrollment 1.6 26.1 11.9 31.4 By 2030, ensure that all girls and boys have access to quality (% gross) early childhood development, care, and pre-primary education so that they are ready for primary education. Primary completion 22.5 57.8 65.0 67.3 By 2030, ensure that all girls and boys complete free, equitable, (% gross) and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment 13.6 24.4 37.5 36.0 By 2030, ensure that all girls and boys complete free, equitable, (% gross) and quality primary and secondary education leading to relevant and effective learning outcomes. Primary completion, ratio of 54.8 97.9 90.8 — By 2030, ensure that all girls and boys complete free, equitable, females to males (%) and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment, ratio 66.7 89.8 79.1 93.6 By 2030, ensure that all girls and boys complete free, equitable, of females to males (%) and quality primary and secondary education leading to relevant and effective learning outcomes. Health Under-5 mortality (per 1,000 145.5 64.4 95.8 42.9 By 2030, end preventable deaths of newborns and children live births) under 5 years of age. Maternal mortality 990 420 546 231 By 2030, reduce the global maternal mortality ratio to less than (per 100,000 live births) 70 per 100,000 live births. Malaria cases 0.6 1.8 4.9 0.3 By 2030, end the epidemics of AIDS, tuberculosis, malaria, and (% of population) neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. HIV prevalence (% of 4.1 1.2 1.2 — By 2030, end the epidemics of AIDS, tuberculosis, malaria, and population ages 15–49) neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. Infrastructure Access to improved 8.2 23.6 22.3 32.4 By 2030, achieve access to adequate and equitable sanitation sanitation facilities and hygiene for all and end open defecation, paying special (% of population) attention to the needs of women and girls and those in vulnerable situations. Access to improved water 29.0 51.5 64.5 60.3 By 2030, achieve universal and equitable access to safe and source (% of population) affordable drinking water for all. Road density (km per 2.7 9.1 7.9 12.0 Develop quality, reliable, sustainable, and resilient 100 sq. km of land) infrastructure, including regional and transborder infrastructure, to support economic development and human well-being, with a focus on affordable and equitable access for all. Access to electricity 12.7 23.0 21.9 34.9 By 2030, ensure universal access to affordable, reliable, and (% of population) modern energy services. Internet users (per 1,000 0.0 1.9 4.4 — Significantly increase access to information and people) communications technology and strive to provide universal and affordable access to the Internet in least developed countries by 2020. Environment CO2 emissions (metric tons 0.09 0.07 0.12 0.29 Integrate climate change measures into national policies, per capita) strategies, and planning. table continues next page 24 Ethiopia Table 2.1  continued Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Memorandum item GNI per capita (constant 137 273 573 2005 US$) Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = currently significantly overperforming; red = currently significantly underperforming; black = performing as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). For Internet use, a projection is not made, despite a tight relationship, since the expected line shifts significantly even in the medium run due to external forces, such as technological development. Global ambition is from the Open Working Group (UN 2014). Figure 2.1  Ethiopia—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) a. Poverty (left), shared prosperity: income share of bottom 40% (right) 410 30 Poverty, at $1.25 a day (PPP) 2000 2011 Shared prosperity: Income 90 share for lowest 40% 20 (% of population) 2000 20 2011 4 2030 10 .90 9 .04 5 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 11.8*** + –1.3*** ln(GNI pc); R 2 = .487 ln(SDG) = 3.17*** + –.05** ln(GNI pc); R 2 = .046 b. Gross pre-primary enrollment (left), primary completion (right) 550 130 Pre-primary enrollment (% gross) Primary completion (% gross) 150 90 2030 40 2012 60 2011 2030 10 40 3 2000 20 2000 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = –0.73 + .570*** ln(GNI pc); R 2 = .348 ln(SDG) = 3.38*** + .14*** ln(GNI pc); R 2 = .372 figure continues next page Ethiopia 25 Figure 2.1  continued c. Gross secondary enrollment (left) 190 Secondary enrollment (% gross) 110 60 30 20 2012 2030 2000 6 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = 1.9*** + .31*** ln(GNI pc); R 2 = .557 d. Ratio of female to male primary completion (left), ratio of female to male secondary enrollment (right) 140 140 Secondary enrollment, ratio of 2012 2030 Primary completion, ratio of 2012 110 100 females to males (%) females to males (%) 90 70 70 2000 50 2000 40 30 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 4.31*** + .04*** ln(GNI pc); R 2 = .093 ln(SDG) = 3.82*** + .1*** ln(GNI pc); R 2 = .296 e. Under-5 mortality (left), maternal mortality (right) 500 2,000 2000 2000 (per 100,000 live births) 230 470 (per 1,000 live births) Maternal mortality Under−5 mortality 2013 110 100 2030 50 2013 20 20 2030 5 1 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 7.99*** + –.61*** ln(GNI pc); R = .559 2 ln(SDG) = 11.3*** + –.9*** ln(GNI pc); R 2 = .519 figure continues next page 26 Ethiopia Figure 2.1  continued f. Malaria cases (left), HIV prevalence (right) 90 30 Malaria cases (% of population) (% of population ages 15–49) 30 1.3 2012 HIV Prevalence 2000 9 .05 2000 2030 3 .90 2013 .10 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 14.5*** + –2.3*** ln(GNI pc); R 2 = .413 ln(SDG) = 1.87* + –.31** ln(GNI pc); R 2 = .044 g. Access to improved sanitation (left), access to improved water source (right) Access to improved sanitation facilities Access to improved water source 100 100 (% of population) (% of population) 60 70 2030 2012 30 2012 50 20 40 2030 2000 7 2000 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .61** + .45*** ln(GNI pc); R 2 = .534 ln(SDG) = 3.45*** + .13*** ln(GNI pc); R 2 = .435 h. Access to electricity (left), road density (right) Access to electricity (% of population) Road density (km per 100 sq. km of 400 330 100 90 30 2010 20 land) 2000 2030 2013 6 7 2030 2000 1.6 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .29 + .5*** ln(GNI pc); R = .459 2 ln(SDG) = −.64 + .48*** ln(GNI pc); R 2 = .189 figure continues next page Ethiopia 27 Figure 2.1  continued i. Internet users (left), CO2 emissions (right) 20 80 CO2 emissions (metric tons per capita) Internet users (per 1,000 people) 3 6 .80 0 2013 .20 2030 .40 2000 2010 .03 2000 0 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = –2.9*** + .78*** ln(GNI pc); R 2 = .567 ln(SDG) = –8.6*** + 1.16*** ln(GNI pc); R 2 = .676 Note: Highlighted observations are for Ethiopia at different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. that it should not be expected to respond strongly ­ olicy change are relatively high, a possibility p or systematically to changes in GNI per capita. that calls for further analysis. When the relationship to GNI per capita is loose As shown in figure 2.2, Ethiopia’s GNI the coefficients are typically  small (in  absolute per  capita percentile ranking among low- and terms); given this, the “expected” values for a middle-­income countries improved by 5 ­percentile recent year are close to the average for all low- points between 2000 and 2012 (from 1st to 6th and middle-income countries.8 percentile).11 Ethiopia’s percentile ranking In sum, among the 17 indicators, Ethiopia’s improved to roughly the same extent for 6 SDG current outcomes are better than expected (com- indicators (poverty, primary completion, access to pared to a typical country at the same GNI per improved water sources, access to electricity, capita level) for 7 (poverty, shared prosperity,9 Internet use, and CO2 emissions). For another 8, gross pre-primary enrollment, ratio of female the ranking improved even more than for GNI to  male primary completion, ratio of female to per capita (pre-primary enrollment, the ratio of male secondary enrollment, under-5 mortality, female to male primary completion rate, ratio of and CO2 emissions), while it falls short for 4 (pri- female to male secondary enrollment, under-5 mary completion, gross secondary enrollment, mortality, maternal mortality, HIV prevalence, access to improved water source, and Internet access to improved sanitation facilities, and use).10 For the remaining 6 indicators (maternal road density). For the remaining 3 indicators mortality, malaria, HIV prevalence, access to (shared prosperity, secondary enrollment, and electricity, access to improved sanitation, and malaria cases), the ranking deteriorated. Among road density), Ethiopia’s current outcomes are as these, the result is not entirely unexpected for expected. While underperformance for an indi- shared ­ prosperity given a weak inverse cator may be due to country-specific conditions cross-country correlation with GNI per capita that are difficult to change, it may alternatively (cf. figure 2.1a, right panel); however, this direc- point to areas in which payoffs from feasible tion of change is nevertheless problematic from an 28 Ethiopia Figure 2.2  Ethiopia—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012 GNI per capita (constant 2005 US$) 1 6 Shared prosperity: Income share for lowest 40% 94 78 Poverty at $1.25 a day (PPP) (% of population) 19 23 Pre-primary enrollment (% gross) 7 30 Primary completion (% gross) 4 8 Secondary enrollment (% gross) 7 3 Primary completion, ratio of females to males (%) 5 30 Secondary enrollment, ratio of females to males (%) 16 26 Under–5 mortality (per 1,000 live births) 17 27 Maternal mortality (per 100,000 live births) 8 17 Malaria cases (% of population) 41 27 HIV prevalence (% of population ages 15–49) 18 31 Access to improved sanitation facilities (% of population) 1 15 Access to improved water source (% of population) 2 6 Road density (km per 100 sq. km of land) 6 28 Access to electricity (% of population) 13 17 Internet users (per 1,000 people) 3 7 CO2 emissions (metric tons per capita) 89 94 0 10 20 30 40 50 60 70 80 90 100 2000 Recent Note: A high ranking signals strong performance; for the underlying indicator, this may correspond to a relatively high value (for example, for the secondary enrollment rate) or a relatively low value (for example, for the poverty rate). The rankings are based on all low- and middle-income countries (according to the 2013 classification) with data. The country samples vary across indicators but are always the ­ same for 2000 and recent for any given indicator. The ranking for an indicator is not reported if the available sample is less than 20 countries. Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. SDG perspective. For the other 3, a higher GNI performing better than expected), and road per capita is linked to improved performance; ­ density (now performing as expected). given this, these ranking declines are unex- By 2030, considerable improvements are pro- pected, suggesting that policies in countries that jected for most indicators (see table 2.1 and respec- otherwise are ­similar to Ethiopia are more apt to tive graphs in figure 2.1). However, compared to address these objectives. the global SDG agenda, also shown in  table  2.1, When comparing the results from regressions the improvements projected for Ethiopia are mod- on GNI per capita to changes in percentile erate. This is to a large extent due to a low initial rankings, a few patterns emerge. The largest drop ­ SDG level. For most indicators, this means that in ranking since 2000 was for shared prosper- the  realization of the global ambitions would ity  (although performance is still better than require a break in the current pattern, such as expected) and malaria (now performing as continued exceptional growth (beyond the pro- expected). The largest improvements in ranking jected real growth in GNI per capita of 4.0 are for pre-primary enrollment and ratio of percent) or a significant increase in external female to male primary completion (both now support. A break is also needed for indicators ­ Ethiopia 29 such as shared prosperity, for which a weak rela- table 2.2 and figure 2.3 summarize the historical tionship with GNI per capita precludes projec- evolution, actual and expected recent values, and, tions. Such a break would be facilitated by more when relevant, projected ­ values.13 When the rela- rapid and more inclusive growth for the popula- tionship is loose, projections are not made and the tion with lowest income, combined with SDG expected value  is in practice close to the average policies that benefit the disadvantaged. for the sample of all low- and middle-­income coun- discussion of expected values for SDG tries (cf.  ­ ­ indicators). The variables cover selected indicators related to three aspects of government activities: 3. Fiscal Space spending, receipts and debt, and governance  and In most countries, accelerated progress on the efficiency. While the findings of this country-at-a- SDG agenda will require efficient and grow- glance note cannot guide policy on their own, they ing  public spending in prioritized areas, most should be seen as an input into discussions about importantly human development and infrastruc- policy making. ture. Private spending is also of crucial importance, In general, room for additional priority both household spending on SDG-related services spending may be created by reducing low-priority and business investments in a wide range of areas spending, increasing current receipts, and/or (including but not limited to infrastructure).12 increasing borrowing. In terms of govern- With regard to Ethiopia’s fiscal space indicators, ment  spending, in areas that may support the Table 2.2  Ethiopia—Fiscal Space: Revenue, Spending, and Government Efficiency Actual Expected Projection Indicator 2000 Recent Recent 2030 Government spending Consumption (% of GDP) 22.9 8.3 12.2 — Investments (% of GDP) 12.2 7.0 — Primary education (% of GDP) 3.0 1.9 — Secondary education (% of GDP) 0.5 1.0 — Primary education, per student (% of GDP per capita) 19.2 10.9 — Secondary education, per student (% of GDP per capita) 10.4 20.8 — Health (% of GDP) 2.3 1.9 2.3 — Fuel subsidy (% of GDP) 1.4 3.6 — Government receipts and debts Tax revenue (% of GDP) 10.9 9.2 11.7 — Net ODA (% of GNI) 8.4 7.5 13.7 3.5 External debt (% of GNI) 68.6 26.8 27.7 — Governance and government efficiency Government effectiveness: percentile rank 15.6 35.9 12.2 38.9 Public investment management index 1.6 1.3 — Memorandum item GNI per capita (constant 2005 US$) 137 273 573 Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = current value significantly above the expected level; red = current value significantly below the expected level; black = current value as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). 30 Ethiopia Figure 2.3  Ethiopia—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting a. Government spending: consumption (left), investment (right) 130 50 60 20 Consumption (% of GDP) Investments (% of GDP) 2013 30 9 2000 10 4 2013 2 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.01*** + .09** ln(GNI pc); R 2 = .042 ln(SDG) = 2.34*** + −.07 ln(GNI pc); R 2 = .012 b. Government spending: primary education, per student (left), secondary education, per student (right) 50 90 Secondary education, per student Primary education, per student 30 50 (% of GDP per capita) (% of GDP per capita) 20 2010 20 8 10 2010 2 4 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.72*** + .12** ln(GNI pc); R 2 = .067 ln(SDG) = 3.51*** + −.09 ln(GNI pc); R 2 = .030 figure continues next page Ethiopia 31 Figure 2.3  continued c. Government spending: primary education (left), secondary education (right) 6 4 Secondary education (% of GDP) 4 Primary education (% of GDP) 2 2010 2 1.2 1.1 .60 2010 .40 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.07*** + −.08 ln(GNI pc); R 2 = .029 ln(SDG) = −.79* + .14*** ln(GNI pc); R 2 = .086 d. Government spending: health (left), fuel subsidy (right) 20 20 1.6 Fuel subsidy (% of GDP) 2011 Health (% of GDP) 6 .10 2000 1.6 2012 0 .50 150 400 1,000 3,000 8,000 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.05 + .16*** ln(GNI pc); R 2 = .093 ln(SDG) = 4.96*** + −.65*** ln(GNI pc); R 2 = .098 figure continues next page 32 Ethiopia Figure 2.3  continued e. Tax revenue (left), official development aid (right) 60 40 7 2000 Tax revenue (% of GDP) 30 Net ODA (% of GNI) 20 2012 2030 1.2 10 .20 1999 2011 3 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.75*** + .13 ln(GNI pc); R 2 = .027 ln(SDG) = 8.26*** + −1.0*** ln(GNI pc); R 2 = .308 f. External debt (left), government effectiveness (right) 2,000 Government effectiveness: percentile rank 80 2013 2030 550 30 External debt (% of GNI) 2000 160 10 50 2000 4 2013 10 1.3 .50 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.52*** + .14** ln(GNI pc); R 2 = .035 ln(SDG) = −.01 + .44*** ln(GNI pc); R 2 = .300 figure continues next page Ethiopia 33 Figure 2.3  continued g. Public investment management index 4 Public investment management index 2 2010 1.3 .80 .30 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = −.43 + .12** ln(GNI pc); R 2 = .078 Note: Highlighted observations are for Ethiopia at different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. SDG agenda, government spending is below the  government’s growth and transformation expected levels (compared to a typical country plan (Government of Ethiopia 2010, p. 1). One at the same GNI per capita level) for secondary apparent consequence of the emphasis on pub- education (as  a share of GDP and per student) ­ lic investment is the potential crowding out of and health. Total government consumption is private investment, for which the GDP share in lower  than expected (measured as a share of Ethiopia averaged only 6.9  percent of GDP in GDP) because the Government of Ethiopia is 2011 (World Bank 2013d). Ultimately, whether deliberately ­constraining consumption to finance adjustments are needed or not depends on public ­investment—a policy that has paid off in the  relative marginal returns in the form of terms of high economic growth. For total invest- social and economic progress from public ment and primary education (as a share of GDP investment. and per student), the government spending is Of the government receipts included in higher than expected. table  2.2, net Official Development Assistance Fuel subsidies are the most obvious case (ODA) and tax revenues are as expected.15 Note, of  low-priority spending from the post-2015 however, that the confidence interval for “expected agenda  perspective.14 However, Ethiopia’s tax revenues” is very broad and Ethiopia is on spending on fuel subsidies was relatively low in the  lower side at 9.2 percent of GDP. As further 2011 (1.4  percent of GDP) and since then oil shown, cross-country patterns suggest that, as prices have decreased while the government GNI per capita grows, net ODA tends to decline as has not fully passed on the decrease to consum- percent of GDP (without changing significantly in ers, suggesting that by now there is no effective per capita terms); in the case of Ethiopia this subsidy or even a net tax on fuel. Public invest- translates into a decrease from the current 7.5 ment is currently double the expected level percent of GNI to 3.5 percent in 2030. However, (12.2 percent of GDP in 2013 compared to the the fact that cross-country patterns point to a expected 7.1) reflecting its role as a pillar in likely decline in ODA does not mean that an 34 Ethiopia increase is excluded: it depends on the ­priorities of numerous factors that are well beyond the scope donors and their relationships with Ethiopia’s of this note, including government capacity in government. different areas and the scope to encourage com- The relationship between tax revenues and plementary private sector activities. From the GNI per capita, as well as the debt stock and GNI perspective of the SDG agenda and given strong per capita, is not tight enough for projections. linkages between private and government activi- However, World Bank and IMF documents stress ties and incomes, it is crucial that policies the need for increased tax revenues as a share of and  spending decisions promote a broad-based GDP.16 Higher taxes would reduce the resources change that encompasses services related to controlled by domestic households and firms, human development, infrastructure investments, pointing to the need to consider the combined and other measures in support of strong long- impact on SDGs and other indicators from run growth that is biased in favor of the less higher taxes and the spending increases that advantaged. are  financed by these taxes. According to the cross-country data, Ethiopia’s external debt stock is at expected levels. To finance planned public investments, the government is already projected 4. Conclusions in the short to medium term to increase both As summarized in table 2.3, Ethiopia’s current the  domestic and the external debt level, how- outcomes are better than expected (compared ever, this with the risk of entering moderate debt to  a typical country at the same GNI per distress.17 capita level) for 7 of the selected SDG indicators; Government efficiency is important to pro- poverty, shared prosperity, gross pre-primary tect and, if possible, increase in order to add to enrollment, ratio for female to male second- the room for priority spending and enhance its ary enrollment, ratio for female to male s ­ econdary impact on the SDG agenda. Ethiopia’s perfor- enrollment, under-5 mortality, and CO2 emis- mance is stronger than expected according to sions. However, for 4 indicators, including ­central both the World Bank Government Effectiveness education indicators such as primary completion indicator and the Public Investment Management and secondary enrollment, Ethiopia is under- Index (table 2.2). However, these are measures performing. For the remaining 6 indicators, at an aggregated level, and the picture may look Ethiopia’s current outcomes are as expected. different when assessing specific sectors. For ­ As also shown in table 2.3, for most of the example, primary education spending is higher indicators, the cross-country relationship with than expected while SDGs related to primary GNI per capita is relatively tight. Given this, by education are underperforming, suggesting that 2030, considerable improvements are projected spending on primary education could be more for the SDG indicators; however, compared to efficient. global ambitions, the improvements fall short In sum, while ODA as a share of GDP may for most indicators due to a combination of decrease as GNI per capita increases, Ethiopia low  initial SDG levels and projected future should be in a position to expand fiscal space growth rates that are lower than recent rates. via  some combination of higher tax revenues To  get closer to the realization of these ambi- and efficiency improvements in targeted sectors. tions, a break with such projections is needed. While there are discussions about increased for- Accelerated growth would raise the capacity to eign borrowing, further external borrowing on accelerate SDG progress. However, for some of nonconcessional terms may risk debt sustain- the SDG indicators, growth since 2000 has not ability. Decisions about the level and allocation been accompanied by equally strong progress, of government spending should be made in light as  indicated by the development of Ethiopia’s of government priorities and would depend on country rankings. This may in part be due to Ethiopia 35 Table 2.3  Ethiopia—Summary Results for SDG Indicators Cross-country relationship with Overperforming As expected Underperforming GNI/cap Tight •• Poverty (×) •• Maternal mortality (+) •• Primary completion (×) •• Gross pre-primary enrollment (+) •• Malaria (−) •• Gross secondary enrollment (−) •• Ratio of female to male secondary •• Access to electricity (×) •• Access to improved water source (×) enrollment (+) •• Access to improved sanitation (+) •• Internet users (×) •• Under-5 mortality (+) •• Road density (+) •• CO2 emissions (×) Loose •• Shared prosperity (−) •• HIV prevalence (+) •• Ratio of female to male primary completion (+) Note: (+) = larger country rank improvement 2000–12 than for GNI per capita; (−) = smaller country rank improvement (or deterioration) 2000–12 than for GNI per capita; (×) = the same country rank change 2000–12 as for GNI per capita (+/− 2 percentile points). lags in translating higher incomes to stronger as per capita incomes increase, but also for sec- SDG performance. However, it may also ­suggest ondary completion, for which the ranking has that, with improved policies, available resources deteriorated since 2000. could be employed more efficiently to improve While ODA as a share of GDP may decrease SDG performance. as GNI per capita increases, Ethiopia should be For three of the indicators in table 2.3, the in a position to expand fiscal space via some link to GNI per capita is weak, suggesting that combination of higher tax revenues and effi- growth matters less and that the role for policies ciency improvements. Moreover, the country that directly or indirectly influence these indica- should continue its search for the balance tors is  particularly important. This is most between the government and the private sector, emphatically the case for shared prosperity, which in virtually all countries plays the role as which, on balance, has a weak tendency to suffer the leading sector in the production sphere. 36 Ethiopia Annex 2A: Data Sources Indicator Source Comment GNI per capita (constant 2005 WDI. API ref: GNI per capita (constant 2005 US$) For Ethiopia, this indicator was defined on the US$) [NY.GNP.PCAP.KD] basis of WDI data for GNI per capita in 2005, and for other years calculated, applying GDP per capita real growth. SDG indicators Poverty, at $1.25 a day (PPP) WDI. API ref: poverty headcount ratio at $1.25 a day (% of population) (PPP) (% of population) [si.pov.dday] Shared prosperity: income share WDI. API ref: income share held by lowest 20% + for lowest 40% income share held by second 20% [SI.DST.FRST.20 + SI.DST.02ND.20] Pre-primary enrollment WDI. API ref: school enrollment, pre-primary 2012 data for Ethiopia from MoE Ethiopia (% gross) (% gross) [SE.PRE.ENRR] (2013). Primary completion (% gross) WDI. API ref: primary completion rate, total 2011 data for Ethiopia from MDG Data (% of relevant age group) [se.prm.cmpt.zs] Dashboards, World Bank. Secondary enrollment (% gross) WDI. API ref: school enrollment, secondary (% gross) 2012 data for Ethiopia from MoE Ethiopia [se.sec.enrr] (2013). Secondary completion (% gross) EdStats. API ref: DHS: secondary completion rate [hh.dhs.scr] Primary completion, ratio of WDI. API ref: primary completion rate, female (% of 2012 data for Ethiopia from Mo (ESAA). females to males (%) relevant age group)/primary completion rate, male (% of relevant age group) *100 [SE.PRM.CMPT.FE. ZS/SE.PRM.CMPT.MA.ZS*100] Secondary enrollment, ratio of WDI. API ref: ratio of female to male secondary 2012 data for Ethiopia from MoE Ethiopia females to males (%) enrollment (%) [se.enr.seco.fm.zs] (2013). Under-5 mortality (per 1,000 live WDI. API ref: mortality rate, under-5 (per 1,000 live births) births) [SH.DYN.MORT] Maternal mortality (per 100,000 WDI. API ref: maternal mortality ratio (modeled live births) estimate, per 100,000 live births) [sh.sta.mmrt] Malaria cases (% of population) HNP. API ref: malaria cases reported/population, total *100 [sh.sta.malr/SP.POP.TOTL*100] HIV prevalence (% of population WDI. API ref: prevalence of HIV, total (% of population ages 15–49) ages 15–49) [sh.dyn.aids.zs] Access to improved sanitation WDI. API ref: improved sanitation facilities facilities (% of population) (% of population with access) [sh.sta.acsn] Access to improved water WDI. API ref: improved water source (% of population source (% of population) with access) [sh.h2o.safe.zs] Road density (km per 100 sq. WDI. API ref: road density (km of road per 100 sq. km km of land) of land area) [Is.rod.dnst.k2] Access to electricity (% of WDI. API ref: access to electricity (% of population) population) [eg.elc.accs.zs] Internet users (per 1,000 WDI. API ref: Internet users (per 100 people) people) [IT.NET.USER.P2] CO2 emissions (metric tons per WDI. API ref: CO2 emissions (metric tons per capita) capita) [en.atm.co2e.pc] Fiscal space indicators Investment (% of GDP) WDI. API ref: gross fixed capital formation 2013 data for Ethiopia from World Bank (% of GDP)-­gross fixed capital formation, private (2014b). sector (% of GDP) [ne.gdi.ftot.zs]- [ne.gdi.fprv.zs] Consumption (% of GDP) WDI. API ref: general government final consumption expenditure (% of GDP) [NE.CON.GOVT.ZS] Primary education (% of GDP) EdStats. API ref: total expenditure on educational institutions and administration as a % of GDP. All sources. Primary [UIS.XGDP.1.FDINSTADM.FFD] annex continues next page Ethiopia 37 Indicator Source Comment Secondary education (% of EdStats. API ref: total expenditure on educational GDP) institutions and administration as a % of GDP. All sources. Secondary and post-secondary nontertiary [UIS.XGDP.234.FDINSTADM.FFD] Primary education, per student WDI. API ref: expenditure per student, primary (% of GDP per capita) (% of GDP per capita) [se.xpd.prim.pc.zs] Secondary education, per WDI. API ref: expenditure per student, secondary student (% of GDP per capita) (% of GDP per capita) [se.xpd.seco.pc.zs] Health (% of GDP) WDI. API ref: health expenditure, public (% of GDP) [sh.xpd.publ.zs] Fossil fuel subsidy (% of GDP) IMF (2013a). Total pretax subsidy (% of GDP) Subsidies are measured using a price-gap approach capturing both consumer (including implicit) and producer (except those that arise when suppliers are inefficient and make losses at benchmark prices) subsidies. Negative external effects are not included in the pre-tax subsidy. Tax revenue (% of GDP) WDI. API ref: tax revenue (% of GDP) [gc.tax.totl.gd.zs] Net ODA (% of GNI) WDI. API ref: net ODA received (% of GNI) [dt.oda.odat.gn.zs] External debt (% of GNI) WDI. API ref: external debt stocks (% of GNI) [DT.DOD.DECT.GN.ZS] Government effectiveness: WGI. API ref: government effectiveness: percentile Captures perceptions of the quality of public percentile rank rank [ge.per.rnk] services, the quality of the civil service, and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies. Grigoli education efficiency Grigoli (2014) Measure secondary education inefficiency in score terms of how much additional output could be achieved at current levels of spending. Grigoli health efficiency score Grigoli and Kapsoli (2013) Quantifies the inefficiency of public health expenditure using a stochastic frontier model that controls for the socioeconomic determinants of health. Public investment management Dabla-Norris et al. (2011) Four phases associated with public investment index management are covered: project appraisal, selection, implementation and evaluation. Note: WDI = World Development Indicators, World Bank; API ref = reference and code when using the World Bank Open Data; EdStats = Education statistics, World Bank; HNP = Health Nutrition and Population statistics, World Bank; WGI = Worldwide Governance Indicators, World Bank. Notes 1. For a discussion of issues related to the role of the contrary, a major challenge for policy makers is to public and private sectors in Ethiopia, see World identify policies that improve SDG performance Bank (2013d, pp. 11–19). In 2013, public and private relative to what is expected given the level of GNI investment were at 12.2 and 20.8  percent of GDP, per capita. A second challenge is to raise growth in respectively; however, the average for 2004–13 was GNI per capita as it indirectly influences country 14.7 percent for public investments and 13.4 percent SDG capacity. for private investments (World Bank 2014b, p. 53). 4. Sources for the indicators are presented in the 2. While a cross-country perspective provides an table in the annex 2A. Note that data for GNI per important complement to analysis that is centered capita in 2005 US$ is not available for Ethiopia in on an individual country, it is by definition lim- WDI for years other than 2005; the other years ited to analysis of variables that are available in are calculated on the basis of its 2005 value and cross-country databases. growth in GDP per capita (2005 US$). 3. This does not mean that GNI per capita is viewed 5. Projections for Ethiopia are based on real GDP as a direct determinant of SDG outcomes; on the per capita growth projections by the World Bank 38 Ethiopia country team. In the projections, it is assumed that 13. The treatment is the same as in table 2.1 and future GNI growth will coincide with future GDP related figures. That is, in table 2.2, projections growth (both expressed in constant 2005 US$). are only shown when the cross-country relation- 6. Given that (a) SDGs have extreme values (such ship between the indicator and GNI per capita is as 100 percent for improved water access) and (b) the considered tight enough. Due to data limitations, current SDG level never is exactly as expected given we focus on government spending indicators; GNI per capita, the projected ­ values gradually con- country-specific analysis is needed to consider ­ verge toward the expected values. For example, for policy in the context of the different roles of the a country that overperforms in water access, as GNI government and private services and spending. per capita increases the extent of overperformance 14. Fuel subsidies are detrimental to the climate and gradually declines, so that when the expected value encourage technologies that create less employ- is 100, overperformance has reached zero. ment for the growing labor force. 7. A tight enough relationship is defined as an 15. Net ODA is measured as a share of GNI, but net R2 > 0.3 (tight) or 0.3 > R2 > 0.1 (moderately tight), ODA per capita was also as expected. Total gov- while R2 > 0.1 are defined as loose. ernment revenues (excluding grants) was some- 8. In addition, the confidence interval is wide in the what lower than expected, suggesting that nontax case of a loose relationship, suggesting that any con- revenues are low in Ethiopia. clusion on over- or underperformance is made with 16. IMF (2014d, p. 30) projects tax revenues to wide margins. Statistically, even though their con- increase from 12.5 percent of GDP in 2012/13 fidence intervals are wide, as long as the estimated to 13.6  percent in 2018/19. World Bank (2013d, coefficient linking GNI per capita to the SDG indi- p. 18) is also stressing the importance of improved cator is nonzero, these values are closer than the tax revenues to cover the anticipated public cross-country average to what is expected for the investments. specific country. The same observation applies to 17. The current government strategy—the “Growth expected values for fiscal space indicators. and Transformation Plan” (GTP)—envisages a 9. In this note, “shared prosperity” is measured as significant part of investment to be undertaken the income share of the poorest 40 percent of the by public enterprises with average annual borrow- population. ing over the five-year period of some 15 percent 10. With regard to CO2, Ethiopia’s current and project of GDP, of which some two-thirds is to be bor- 2030 per capita emissions are 0.7 and 3.7 percent rowed externally (IMF 2014d, p. 12; Government of the current OECD average. Note that Ethiopia of Ethiopia 2010). IMF (2014d, p. 30) projects is overperforming for CO2 emissions (emitting the external debt to increase from 20.5 percent below expected levels) both when emissions are of GDP in 2012/13 to 27.7 percent in 2018/19, measured as a share of GDP and per capita. which together with the increased domestic debt 11. The ranking is based on data from 2000 and results in an increase of total public debt from 2012/2013, or the closest year with data (but only 37.4 to 56.0  percent of GDP. This is in line with if data no later than 1998 for “2000” or 2009 for the DSA (IMF 2014d, annex), which indicates that “recent” exist). Ethiopia’s overall public sector debt dynamics is 12. There are also cases where the solution to the low sustainable under the baseline scenario but vul- level of SDG is neither private nor public spending nerable to several alternative scenarios. See also but more efficient policies. World Bank (2014b, p. 5). Ethiopia 39 Chapter 3 Jamaica 1. Introduction performance and prospects are benchmarked relative to other countries, considering its past, Jamaica, an upper middle-­ income island country recent, and projected levels of GNI per capita. in the Caribbean, has been stuck in a negative The latter variable tends to be highly correlated spiral of low growth, high unemployment, high ­ with most of the SDGs and most of the fac- government debt, and uncertain fiscal finances tors  that determine their evolution; given this, (IMF 2014a, p. 1). During 2001–12, Jamaica’s it  is used as a summary indicator of country annual average growth in GDP per capita (at ­ capacity to provide and efficiently utilize inputs constant 2005 US$) was zero (0.014 percent). ­ that contribute to SDGs (for example, health During the same period, despite a slight improve- and education services) and to achieve SDG ment in the index score, Jamaica’s ranking accord- outcomes (like strong health and education ing to the UNDP Human Development Index results).2 (among countries included both in 2000 and 2012) deteriorated, from the 54th to the 50th per- centile. In an attempt to break with the past, in 2. SDG Indicators: History and 2013 the authorities embarked on an ambitious reform program. Projections This country-at-a-glance note, which is based For selected SDG indicators, table 3.1 summarizes on cross-country data and hence provides a data for Jamaica: historical evolution, actual and cross-country perspective, is designed to provide expected values for a recent year, and projected an initial picture of the challenges that the Post- 2030 values.3 In figure 3.1, data for Jamaica are shown 2015 agenda poses for Jamaica, serving as the in the context of the estimated c ­ross-country starting point for a more complete country devel- ­ relationship between each SDG indicator and GNI opment diagnostic as well as a more comprehen- per capita. For Jamaica, the projected average sive country-focused analysis.1 The note is built annual rate of GNI per capita growth is 1.1 around tables and figures that provide data for a percent.4 The projected SDG values reflect what selection of SDG target indicators and indicators can be expected given a country’s starting point, related to fiscal space—fiscal space matters since, projected growth in GNI per capita, typical rates of while policy frameworks and the engagement of progress according to cross-country patterns, and the private sector may vary widely, rapid progress a gradual convergence to close gaps between on the SDG agenda will require efficient and care- observed and expected values.5 Projections for fully prioritized public spending. Drawing on the SDG indicators are presented only when the information in these tables and figures, this note cross-country relationship between the indicator briefly (a) summarizes Jamaica’s SDG progress and GNI per capita is classified as tight.6 A loose since 2000 and projects expected values for 2030; relationship suggests that progress in the indicator and (b) assesses options for increasing fiscal is primarily a reflection of country-specific factors space. Sections 2 and 3 address SDGs and fiscal and that it should not be expected to respond space, respectively, while findings are summa- strongly or systematically to changes in GNI per rized in Section 4. capita. When the relationship to GNI per capita is The cross-country perspective is manifested loose the coefficients are typically small (in abso- in that, for the different indicators, Jamaica’s lute terms); given this the “expected” values for a Jamaica 41 recent year are close to the average for all low- and secondary enrollment,11 access to improved water middle-income countries.7 source, access to improved sanitation, access to In sum, Jamaica’s current outcomes are better electricity, Internet users, and CO2 emissions.12 than expected (compared to a typical country at its However, Jamaica is doing less well than expected level of GNI per capita) for 4 of the 16 indicators for 5 indicators: primary c ­ ompletion, secondary with data8: pre-primary enrollment, under-5 mor- enrollment, ratio of female to male primary com- tality, malaria cases,9 road density,10 and as expected pletion, maternal mortality, and HIV prevalence. for 7: secondary completion, ratio of female to male While underperformance for an indicator may be Table 3.1  Jamaica—SDG Indicators: Evolution since 2000 and Projections to 2030 Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Education Pre-primary enrollment 83.4 91.6 53.6 92.7 By 2030, ensure that all girls and boys have access to (% gross) quality early childhood development, care, and pre-primary education so that they are ready for primary education. Primary completion (% gross) 87.9 73.4 94.6 86.1 By 2030, ensure that all girls and boys complete free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment 86.7 77.8 84.2 84.9 By 2030, ensure that all girls and boys complete free, (% gross) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary completion 54.9 48.9 — By 2030, ensure that all girls and boys complete free, (% gross) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Primary completion, ratio of 105.7 96.3 99.7 — By 2030, ensure that all girls and boys complete free, females to males (%) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment, ratio of 102.1 103.8 102.6 105.3 By 2030, ensure that all girls and boys complete free, females to males (%) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Health Under-5 mortality 23.7 16.6 19.4 14.8 By 2030, end preventable deaths of newborns and (per 1,000 live births) children under 5 years of age. Maternal mortality 88.0 80.0 51.9 67.4 By 2030, reduce the global maternal mortality ratio to (per 100,000 live births) less than 70 per 100,000 live births. Malaria cases 0.000 0.001 0.007 0.000 By 2030, end the epidemics of AIDS, tuberculosis, (% of population) malaria, and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. HIV prevalence (% of 2.5 1.8 0.5 — By 2030, end the epidemics of AIDS, tuberculosis, population ages 15–49) malaria, and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. Infrastructure Access to improved sanitation 79.8 80.2 76.9 86.7 By 2030, achieve access to adequate and equitable facilities (% of population) sanitation and hygiene for all and end open defecation, paying special attention to the needs of women and girls and those in vulnerable situations. table continues next page 42 Jamaica Table 3.1  continued Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Access to improved water 93.4 93.1 91.7 95.2 By 2030, achieve universal and equitable access to source (% of population) safe and affordable drinking water for all. Road density (km per 190.7 201.3 138.5 221.3 Develop quality, reliable, sustainable, and resilient 100 sq. km of land) infrastructure, including regional and transborder infrastructure, to support economic development and human well-being, with a focus on affordable and equitable access for all. Access to electricity 86.7 92.0 82.3 96.0 By 2030, ensure universal access to affordable, (% of population) reliable, and modern energy services. Internet users (per 1,000 3.1 37.8 33.6 — Significantly increase access to information and people) communications technology and strive to provide universal and affordable access to the Internet in least developed countries by 2020. Environment CO2 emissions (metric tons 4.0 2.7 2.6 3.4 Integrate climate change measures into national per capita) policies, strategies, and planning. Memorandum item GNI per capita (constant 3,782 3,788 — 4,655 2005 US$) Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used, however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = currently significantly overperforming; red = currently significantly underperforming; black = performing as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). For Internet use, a projection is not made, despite a tight relationship, since the expected line shifts significantly even in the medium run due to external forces, such as technological development. Global ambition is from the Open Working Group (UN 2014). Figure 3.1  Jamaica—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) a. Gross pre-primary enrollment (left); primary completion (right) 170 120 2030 2013 Pre-primary enrollment (% gross) Primary completion (% gross) 80 2000 2030 40 2000 2010 50 10 40 3 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.64 + .560*** ln(GNI pc); R 2 = .325 ln(SDG) = 3.48*** + .13*** ln(GNI pc); R 2 = .358 figure continues next page Jamaica 43 Figure 3.1  continued b. Gross secondary enrollment (left); secondary completion (right) 110 90 2010 Secondary completion (% gross) Secondary enrollment (% gross) 60 2030 20 2000 2013 30 6 20 1.4 6 .09 150 400 1,000 3,000 8,000 150 400 1,000 3,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.94*** + .3*** ln(GNI pc); R 2 = .540 ln(SDG) = −.24 + .5* ln(GNI pc); R 2 = .109 c. Ratio of female to male primary completion (left); ratio of female to male secondary enrollment (right) 140 140 2000 2000 Secondary enrollment, ratio of 2013 Primary completion, ratio of 110 100 females to males (%) females to males (%) 2030 90 70 2013 70 50 40 30 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 4.37*** + .03*** ln(GNI pc); R 2 = .056 ln(SDG) = 3.88*** + .09*** ln(GNI pc); R 2 = .268 figure continues next page 44 Jamaica Figure 3.1  continued d. Under-5 mortality (left); maternal mortality (right) Maternal mortality (per 100,000 live births) 500 2,000 Under−5 mortality (per 1,000 live births) 230 470 110 2000 100 50 2030 2000 20 2013 20 2013 2030 5 1 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 7.98*** + −.61*** ln(GNI pc); R = .559 2 ln(SDG) = 11.3*** + −.9*** ln(GNI pc); R 2 = .518 e. Malaria cases (left); HIV prevalence (right) 40 90 HIV prevalence (% of population Malaria cases (% of population) 30 1.5 9 ages 15–49) .06 3 2000 2013 2009 .90 2030 2000 .10 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 13.9*** + −2.2*** ln(GNI pc); R = .341 2 ln(SDG) = 1.88* + −.31** ln(GNI pc); R 2 = .044 figure continues next page Jamaica 45 Figure 3.1  continued Access to improved sanitation facilities f. Access to improved sanitation (left); access to improved water source (right) 2030 100 2000 Access to improved water source 100 2012 2030 70 2012 60 (% of population) (% of population) 2000 50 30 40 20 7 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .61** + .45*** ln(GNI pc); R 2 = .534 ln(SDG) = 3.45*** + .13*** ln(GNI pc); R 2 = .435 g. Access to electricity (left); road density (right) Access to electricity (% of population) 400 330 Road density (km per 100 sq. km 2013 2030 2010 2030 2000 100 90 2000 of land) 30 20 7 6 1.6 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .29 + .5*** ln(GNI pc); R = .459 2 ln(SDG) = −.64 + .48*** ln(GNI pc); R 2 = .189 figure continues next page 46 Jamaica Figure 3.1  continued h. Internet users (left); CO2 emissions (right) 20 CO2 emissions (metric tons per capita) 80 2013 2000 Internet users (per 1,000 people) 3 2030 .80 2010 6 0.20 2000 0 0 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −2.9*** + .78*** ln(GNI pc); R 2 = .567 ln(SDG) = −8.6*** + 1.16*** ln(GNI pc); R 2 = .677 Note: Highlighted observations are for Jamaica at different years, whereas the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. due to country-specific conditions that are diffi- cases, access to improved water source, and cult to change, it may often point to areas in Internet use) the deterioration in ranking was which payoffs from feasible policy change are rel- even worse than for GNI per capita. atively high; a possibility that calls for further When comparing the results from regressions analysis. on GNI per capita to changes in percentile rank- Figure 3.2 shows that, between 2000 and 2012, ings, a few insights emerge. For example, primary Jamaica saw its ranking among low- and completion, secondary enrollment, and the ratio ­ middle-income countries deteriorate by as much of female to male primary completion are lower as 10 percentile points for GNI per capita.13 Only than expected and falling in ranking among for 2 of the SDG indicators, the ratio of female to other countries at a rate faster than the GNI per male secondary enrollment and CO2 emissions, capita ranking. On the other hand, maternal did its ranking improve, which in the case of CO2 mortality and HIV prevalence, both of which are emissions was expected because of the inverse underperforming at the current GNI per capita relationship to GNI per capita. In addition, level, are doing better than GNI per capita in 3 indicators more or less retained an unchanged terms of country ranking changes since 2000. ranking (HIV prevalence, road density, and By 2030, limited improvements are projected access to electricity). For 2 indicators (­pre-primary for most indicators, not least due to slow income enrollment and maternal mortality) the deterio- growth, and the ability to reach the post-2015 ration in ranking was less than for GNI per capita, global goals is uneven (shown in the last c ­ olumn and for another 2 indicators (under-5 mortality of table 3.1, and respective graphs in figure 3.1). and access to improved sanitation facilities), the Nevertheless, in part thanks to Jamaica’s strong deterioration in ranking was similar to GNI per initial conditions compared to other developing capita. However, for the other 6 indicators (pri- countries, the indicators for the ratio of female to mary completion, secondary enrollment, ratio of male secondary enrollment, maternal mortality, female to male primary completion, malaria malaria cases, and access to electricity are all Jamaica 47 Figure 3.2  Jamaica—Percentile Cross-Country Ranking for SDG Indicators since 2000 GNI per capita (constant 2005 US$) 84 74 Pre-primary enrollment (% gross) 94 88 Primary completion (% gross) 48 22 Secondary enrollment (% gross) 90 50 Secondary completion (% gross) 76 Primary completion, ratio of females to males (%) 89 23 Secondary enrollment, ratio of females to males (%) 60 68 Under-5 mortality (per 1,000 live births) 80 72 Maternal mortality (per 100,000 live births) 65 62 Malaria cases (% of population) 92 79 HIV prevalence (% of population ages 15–49) 25 25 Access to improved sanitation facilities (% of population) 70 59 Access to improved water source (% of population) 77 60 Road density (km per 100 sq. km of land) 100 100 Access to electricity (% of population) 65 63 Internet users (per 1,000 people) 80 66 CO2 emissions (metric tons per capita) 14 27 0 10 20 30 40 50 60 70 80 90 100 2000 Recent Note: A high ranking signals strong performance; for the underlying indicator, this may correspond to a relatively high value (for example, for the secondary enrollment rate) or a relatively low value (for example, for the poverty rate). The rankings are based on all low- and middle- income countries (according to the 2012 classification) with data. The country samples vary across indicators but are always the same for 2000 and recent for any given indicator. The ranking for an indicator is not reported if the available sample is less than 20 countries. Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used. However, the data are never older than 1998 for “2000” or 2009 for “recent.” Country-specific data years can be found in the respective graphs. projected to either realize or get close to realizing both household spending on SDG-related services the post-2015 global ambition. However, for other and business investments in a wide range of areas SDGs, to get closer to the realization of these (including but not limited to infrastructure). With ambitions, a break with the past seems needed. regard to Jamaica’s fiscal space indicators, table 3.2 Such a break would be facilitated by a combina- and figure 3.3 summarize the historical evolution, tion of more rapid growth and improvements in actual and expected recent values, and, when rele- policies that directly influence different SDGs. For vant, projected values.14 When the relationship is indicators, which are quite unrelated to GNI per loose, projections are not made and the expected capita, progress will depend on ­ country-specific value is in practice close to the average for the sam- policies, including those affecting links between ple of all low- and  middle-income countries (cf. income-generation in different sectors and the discussion of expected values for SDG indicators). distribution of income across households. The variables cover three aspects of government activities: spending, receipts and debt, and ­ governance and efficiency. While the findings of this country-at-a-glance note cannot guide policy 3. Fiscal Space on their own, they should be an input into discus- In most countries, accelerated progress on the sions about policy making. SDG agenda will require more efficient and grow- Room for additional priority spending may ing public spending in prioritized areas, most be created by reducing low-priority spending, importantly human development and infrastruc- increasing current receipts, and/or increasing ture. Private spending is also of crucial importance, borrowing. In terms of government spending in 48 Jamaica Table 3.2  Jamaica—Fiscal Space: Revenue, Spending, and Government Efficiency Actual Expected Projection Indicator 2000 Recent Recent 2030 Government spending Consumption (% of GDP) 14.3 16.3 15.3 — Investments (% of GDP) 5.0 2.8 5.8 — Primary education (% of GDP) 2.3 1.6 — Secondary education (% of GDP) 2.5 1.5 — Primary education, per student (% of GDP per capita) 22.5 15.5 — Secondary education, per student (% of GDP per capita) 29.4 17.0 — Health (% of GDP) 2.9 3.4 3.5 — Government receipts and debts Tax revenue (% of GDP) 23.9 16.1 — Net ODA (% of GNI) 0.10 0.51 0.85 0.42 External debt (% of GNI) 54.4 100.6 40.2 — Governance and government efficiency Government effectiveness: percentile rank 56.6 54.5 39.0 57.1 Grigoli education efficiency score 0.98 0.78 0.98 Grigoli health efficiency score 0.96 0.95 — Public investment management index 1.72 1.80 — Memorandum item GNI per capita (constant 2005 US$) 3,782 3,788 — 4,655 Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross- country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = currently significantly overperforming; red = currently significantly underperforming; black = performing as expected; — = no projection because the cross-country relationship was not tight enough (see criteria earlier in the note). areas that may support the SDG agenda, Jamaica’s while tax revenues are higher than expected. The values are as expected (compared to a typical cross-country pattern suggests that net ODA country at the same GNI per capita level) for will decline as a percent of GDP (without chang- total consumption and spending on health, ing significantly in per capita terms); in the case while public investment is lower than expected. of Jamaica, this translates into a minor decrease Government spending on education is higher (from 0.51 to 0.42 percent of GNI). The relation- than expected for both the primary and the sec- ship between tax revenues and GNI per capita ondary levels. Note that, for both education lev- and GNI per capita is not tight enough to proj- els, spending is measured in two ways: total as ect expected changes; the fact that Jamaica’s tax share of GDP and per student as share of GDP revenues are higher than expected suggests that per capita; for all of the four indicators that are the scope for future tax increases is limited.17 generated, spending is higher than expected. Finally, borrowing is an unlikely source of addi- From the perspective of the post-2015 agenda, tional fiscal space as the government intends to cuts in fuel subsidies are a top priority source of drastically reduce its total (external and domes- fiscal space15; however, for Jamaica, this is not rel- tic) debt as a share of GDP (IMF 2014a, pp. 16 evant given negligible—if any—subsidies and a and 28). The fact that Jamaica’s external public recent introduction of a gasoline tax.16 debt stock is much higher than expected sug- Among current receipts, net Official Devel­ gests that severe limits on borrowing may be opment Aid (ODA) is lower than expected, warranted.18 Jamaica 49 Figure 3.3  Jamaica—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting a. Government spending: consumption (left); investment (right) 130 60 80 20 Consumption (% of GDP) Investments (% of GDP) 30 2012 9 10 1997 4 2000 2012 2 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.01*** + .09** ln(GNI pc); R 2 = .042 ln(SDG) = 2.37*** + −.07 ln(GNI pc); R 2 = .013 b. Government spending: primary education, per student (left); secondary education, per student (right) 90 50 Secondary education, per student Primary education, per student 30 50 2012 (% of GDP per capita) (% of GDP per capita) 20 2011 20 8 10 2 4 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.72*** + .12*** ln(GNI pc); R 2 = .067 ln(SDG) = 3.51*** + –.09*** ln(GNI pc); R 2 = .030 figure continues next page 50 Jamaica Figure 3.3  continued c. Government spending: primary education (left); secondary education (right) 6 4 2013 Secondary education (% of GDP) 4 Primary education (% of GDP) 2013 2 2 1.2 1.1 .60 .40 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .99*** + −.06 ln(GNI pc); R 2 = .019 ln(SDG) = −.98** + .17*** ln(GNI pc); R 2 = .118 d. Government spending: health 20 Health (% of GDP) 6 2013 2030 2000 1.6 .50 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = .25 + .18*** ln(GNI pc); R 2 = .111 figure continues next page Jamaica 51 Figure 3.3  continued e. Tax revenue (left); official development aid (right) 60 50 8 30 Tax revenue (% of GDP) 2013 Net ODA (% of GNI) 20 1.2 10 .20 2030 2013 2000 3 0 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.69*** + .13* ln(GNI pc); R 2 = .030 ln(SDG) = 8.10*** + −1*** ln(GNI pc); R 2 = .298 f. External debt (left); government effectiveness (right) 2,000 2000 Government effectiveness: percentile rank 80 550 30 2030 External debt (% of GNI) 2013 160 2013 2000 10 50 4 10 1.3 .50 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.52*** + .14** ln(GNI pc); R 2 = .035 ln(SDG) = 0 + .44*** ln(GNI pc); R 2 = .300 figure continues next page 52 Jamaica Figure 3.3  continued g. Health expenditure efficiency (right); health expenditure efficiency (left) 1 2030 2010 Grigoli education efficiency score 2010 Grigoli health efficiency score 1 .60 .40 .90 .30 .10 0 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −4.0*** + .46*** ln(GNI pc); R = .630 2 ln(SDG) = −.09*** + .01 ln(GNI pc); R 2 = .031 h. Public investment management index 4 Public investment management index 2 1.3 2010 .80 .30 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = −.43 + .12** ln(GNI pc); R 2 = .079 Note: Highlighted observations are for Jamaica at different years, while the nonhighlighted country observations are the most recent observation for other low and middle-income countries. Jamaica 53 Government efficiency is important to protect expenditures would be difficult to find in the and, if possible, increase in order to add to the absence of higher growth. Any suggestions about room for priority spending and enhance its fiscal policy adjustments would require addi- impact on the SDG agenda. Table 3.2 displays tional country-specific i ­nformation that, from data for some measures of government efficiency. the perspective of the SDG agenda, would permit According to the measures at a more aggregated assessments of the benefits and the costs of efficiency level; Jamaica is performing better than feasible changes in the level and allocation of ­ expected in terms of the World Bank Government spending and taxation and/or point to areas for Effectiveness indicator and as expected for the efficiency improvements. Such adjustments Public Investment Management Index. Education should be part of a broader strategy for sustain- spending efficiency is higher than expected able growth, poverty reduction, and shared pros- according to the sector specific index, and health perity, among other factors considering Jamaica’s spending efficiency is as expected. That is, accord- sectoral production and trade structure, includ- ing to these indicators, government efficiency ing the importance of tourism and net private appears to be better than or as expected given transfers (primarily remittances from Jamaica’s Jamaica’s GNI per capita, an observation that population abroad)—in 2011/12, tourism does not negate that considerable efficiency gains amounted to 13.9 percent of GDP (IMF 2014a, still may be feasible in different areas. p. 32) and remittances to 15.0 percent of GDP in In sum, our cross-country results indicate 2013 (World Bank data). Given the important that government spending in terms of overall and very direct contribution of the remittances consumption and investment is at or below to the living standards of many Jamaicans, policy expected levels. Among the receipts, a compari- steps, fiscal and other, that encourage even higher son of Jamaican data to cross-country patterns remittances should be a high priority, including does not single out any of the categories for the identification of means of channeling them to which we have data (taxes, ODA, or borrowing) income-raising investments. as being easily tapped for additional fiscal space. Jamaica’s high debt burden (total public debt to GDP ratio was estimated at 137 percent at end- March 2015) and the need to maintain a high 4. Conclusions primary surplus (about 7 percent annually) to As summarized in table 3.3, Jamaica’s current out- achieve medium-term debt sustainability suggest comes are better than expected (compared to a that increased fiscal means for SDG related typical country at the same GNI per capita level) Table 3.3  Jamaica—Summary Results for SDG Indicators Cross-country relationship with Overperforming As expected Underperforming GNI/cap Tight •• Pre-primary school enrollment (+) •• Secondary completion •• Primary completion (–) •• Under-5 mortality (×) •• Ratio of female to male secondary •• Gross secondary enrollment (–) •• Malaria (–) enrollment (+) •• Maternal mortality (+) •• Road density (+) •• Access to improved water source (–) •• Access to improved sanitation (×) •• Access to electricity (+) •• CO2 emissions (+) •• Internet users (–) Loose •• Ratio of female to male primary completion (–) •• HIV prevalence (+) Note: (+) = country rank increase, 2000–12, or smaller deterioration than for GNI per capita; (–) = larger deterioration in country rank 2000–12 compared to GNI per capita; (×) = the same country rank change 2000–12 as for GNI per capita (+/– 2 percentile points). 54 Jamaica for 4 of the selected SDG indicators, as expected of resources to outcomes that may be addressed for another 7, and worse than expected for with policy adjustments. Moreover, among the 5,  including primary completion, secondary SDGs, some are only loosely related to GNI per ­ enrollment, ratio of female to male primary com- capita; this group includes ratio of female to male pletion, maternal mortality, and HIV prevalence. primary completion and HIV prevalence. This Projections for 2030 show mixed results, where loose relationship suggests that these indicators some indicators are expected to achieve or get close should not be expected to improve strongly or sys- to realizing the post-2015 global ambitions, while tematically to more rapid growth in GNI per for others a break with the past seems needed. capita but rather would require targeted policy Table 3.3 further shows that, for most indica- interventions. tors, the relationship to GNI per capita is tight, With regard to fiscal space, our cross-country suggesting that accelerated improvements in these perspective does not suggest that changes in any SDGs will likely follow from accelerated GNI per specific areas are obvious priorities since spend- capita growth and the related increases in resources ing is at or below the expected levels, efficiency is and capabilities. However, among this group of roughly as expected and the room to increase rev- indicators, primary completion and secondary enues is limited. Given this, increased support to enrollment are currently underperforming rela- SDG related expenditures will be difficult in the tive to expectations and have deteriorated more absence of increased economic growth, and pol- than GNI per capita since 2000. For these indica- icy directions for a future SDG agenda would tors, it seems particularly important for the gov- have to be guided by more detailed ­country-­specific ernment to identify inefficiencies in the translation information. Annex 3A: Data Sources Indicator Source Comment GNI per capita (constant 2005 WDI. API ref: GNI per capita (constant 2005 US$) WDI data for GNI per capita in 2005, and for US$) [NY.GNP.PCAP.KD] other years calculated, applying GDP per capita real growth. SDG indicators Poverty, at $1.25 a day (PPP) WDI. API ref: poverty headcount ratio at $1.25 a day No data beyond 2004 for Jamaica. (% of population) (PPP) (% of population) [SI.POV.DDAY] Shared prosperity: income share WDI. API ref: income share held by lowest 20% + No data beyond 2004 for Jamaica. for lowest 40% income share held by second 20% [SI.DST.FRST.20+SI.DST.02ND.20] Pre-primary enrollment (% WDI. API ref: school enrollment, pre-primary gross) (% gross) [SE.PRE.ENRR] Primary completion (% gross) WDI. API ref: primary completion rate, total (% of relevant age group) [SE.PRM.CMPT.ZS] Secondary enrollment (% gross) WDI. API ref: school enrollment, secondary (% gross) [SE.SEC.ENRR] Secondary completion (% EdStat. API ref: DHS: secondary completion rate For Jamaica, data from the World Bank country gross) [HH.DHS.SCR] team. Primary completion, ratio of WDI. API ref: primary completion rate, female females to males (%) (% of relevant age group)/primary completion rate, male (% of relevant age group) *100 [SE.PRM.CMPT.FE.ZS/SE.PRM.CMPT.MA.ZS*100] Secondary enrollment, ratio of WDI. API ref: ratio of female to male secondary females to males (%) enrollment (%) [SE.ENR.SECO.FM.ZS] annex continues next page Jamaica 55 Indicator Source Comment Under-5 mortality (per 1,000 live WDI. API ref: mortality rate, under-5 (per 1,000 live births) births) [SH.DYN.MORT] Maternal mortality (per 100,000 WDI. API ref: maternal mortality ratio (modeled live births) estimate, per 100,000 live births) [SH.STA.MMRT] Malaria cases (% of population) HNP. API ref: malaria cases reported/population, total *100 [SH.STA.MALR/SP.POP.TOTL *100] HIV prevalence (% of population WDI. API ref: prevalence of HIV, total (% of population ages 15–49) ages 15–49) [SH.DYN.AIDS.ZS] Access to improved sanitation WDI. API ref: improved sanitation facilities facilities (% of population) (% of population with access) [SH.STA.ACSN] Access to improved water WDI. API ref: improved water source (% of population source (% of population) with access) [SH.H2O.SAFE.ZS] Road density (km per 100 sq. WDI. API ref: road density (km of road per 100 sq. km km of land) of land area) [IS.ROD.DNST.K2] Access to electricity (% of WDI. API ref: access to electricity (% of population) population) [EG.ELC.ACCS.ZS] Internet Users (per 1,000 WDI. API ref: Internet users (per 100 people) people) [IT.NET.USER.P2] CO2 emissions (metric tons per WDI. API ref: CO2 emissions (metric tons per capita) capita) [EN.ATM.CO2E.PC] Fiscal space indicators Investment (% of GDP) WDI. API ref: gross fixed capital formation IMF (2014a). “Capital expenditures” (% of GDP)-gross fixed capital formation, private sector (% of GDP) [NE.GDI.FTOT.ZS]-[NE.GDI.FPRV. ZS] Consumption (% of GDP) WDI. API ref: general government final consumption expenditure (% of GDP) [NE.CON.GOVT.ZS] Primary education (% of GDP) EdStat. API ref: total expenditure on educational institutions and administration as a % of GDP. All sources. Primary [UIS.XGDP.1.FDINSTADM.FFD] Secondary education (% of EdStat. API ref: total expenditure on educational GDP) institutions and administration as a % of GDP. All sources. Secondary and post-secondary non-tertiary [UIS.XGDP.234.FDINSTADM.FFD] Primary education, per student WDI. API ref: expenditure per student, primary (% of GDP per capita) (% of GDP per capita) [SE.XPD.PRIM.PC.ZS] Secondary education, per WDI. API ref: expenditure per student, secondary student (% of GDP per capita) (% of GDP per capita) [SE.XPD.SECO.PC.ZS] Health (% of GDP) WDI. API ref: health expenditure, public (% of GDP) [SH.XPD.PUBL.ZS] Fossil fuel subsidy (% of GDP) IMF (2013a). Total pretax subsidy (% of GDP) Subsidies are measured using a price-gap approach capturing both consumer (including implicit) and producer (except those that arise when suppliers are inefficient and make losses at benchmark prices) subsidies. Negative external effects are not included in the pretax subsidy. Tax revenue (% of GDP) WDI. API ref: tax revenue (% of GDP) [GC.TAX.TOTL. For Jamaica, data from IMF (2015a). GD.ZS] Net ODA (% of GNI) WDI. API ref: net ODA received (% of GNI) [DT.ODA.ODAT.GN.ZS] External debt (% of GNI) WDI. API ref: external debt stocks (% of GNI) [DT.DOD.DECT.GN.ZS] annex continues next page 56 Jamaica Indicator Source Comment Government Effectiveness: WGI. API ref: government effectiveness: percentile Captures perceptions of the quality of public Percentile Rank rank [GE.PER.RNK] services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies. Grigoli education efficiency Grigoli (2014) Measure secondary education inefficiency in score terms of how much additional output could be achieved at current levels of spending. Grigoli health efficiency score Grigoli and Kapsoli (2013) Quantifies the inefficiency of public health expenditure using a stochastic frontier model that controls for the socioeconomic determinants of health. Public investment management Dabla-Norris et al. (2011) Four phases associated with public investment index management are covered: project appraisal, selection, implementation, and evaluation. Note: WDI = World Development Indicators, World Bank. API ref = reference and code when using the World Bank Open Data. EdStat = Education statistics, World Bank. HNP = Health Nutrition and Population statistics, World Bank. WGI = Worldwide Governance Indicators, World Bank. Notes 1. While a cross-country perspective is an important is approached. For example, for a country that complement to analysis that is centered on an indi- overperforms in water access, as GNI per capita vidual country, it is by definition limited to analy- increases the extent of overperformance gradually sis of variables that are available in cross-country declines, so that when the expected value is 100, databases. overperformance has reached zero. 2. This does not mean that GNI per capita is viewed as 6. A tight enough relationship is defined as an a direct determinant of SDG outcomes; on the con- R 2 > 0.3 (tight) or 0.3 > R 2 > 0.1 (moderately tight), trary, a major challenge for policy makers is to iden- while R 2 < 0.1 are defined as loose. tify policies that improve SDG performance relative 7. In addition, the confidence interval is wide in the to what is expected given the level of GNI per capita. case of a loose relationship, suggesting that any A second challenge is to raise growth in GNI per conclusion on over- or underperformance is made capita as it indirectly influences country SDG capacity. with wide margins. Statistically, even though their 3. Sources for the indicators are presented in the confidence intervals are wide, as long as the esti- table in the annex 3A. Note that data for GNI per mated coefficient linking GNI per capita to the SDG capita in 2005 US$ are not available for Jamaica indicator is nonzero, these values are closer than in WDI for years other than 2005; the other years the cross-country average to what is expected for are calculated on the basis of its 2005 value and the specific country. The same observation applies growth in GDP per capita (2005 US$). to expected values for fiscal space indicators. 4. Projections from CEPII are used for this and 8. Unfortunately there is no Jamaica-specific data for other Country Development Diagnostics appli- cross-country comparison for poverty or shared cations given their wide country coverage and prosperity beyond 2004. well-documented methodology; OECD data have ­ 9. Close to nonexistent in Jamaica. been used when projections have been missing. 10. However, note that the population density in In the projections, it is assumed that future GNI Jamaica is much higher than average. growth will coincide with future GDP growth (both 11. In the case of Jamaica, the ratio between female expressed in constant 2005 US$) given that that this and male gross secondary enrollment rates points is the variable that CEPII and other sources project. to a problem of gender imbalance that is the oppo- 5. Given that (a) SDGs have extreme values (such site of what typically is encountered in developing as 100 percent for improved water access) and countries (especially at the lower end of the income (b) the current SDG level never is exactly as spectrum): the boys are falling behind the girls. expected given GNI per capita, the projected val- 12. With regard to CO2, Jamaica’s current and pro- ues gradually converge toward the expected v ­ alues jected 2030 per capita emissions are 26.7 and 35.6 as the maximum or minimum feasible value percent of the current OECD average. Jamaica 57 13. If data are not available for 2000 or 2013, the clos- 16. Fuel subsidies, including the cost from external est earlier year with data is used; however, the effects when use of fossil fuels, amounted to merely data are never older than 1998 for “2000” or 2009 0.7 percent of GDP in 2011 (IMF 2013a). for “2013.” The year for Jamaica-specific data is 17. IMF projects a marginal increase of tax revenues reported in the graphs. from 24.0 in 2012/13 to 24.2 percentage of GDP 14. The treatment is the same as in table 3.1 and related in 2019/20 driven by an increase in the tax base; figures. That is, in table 3.2, projections are shown tax rates are projected to decrease (IMF 2014a, only when the cross-country relationship between p. 28). the indicator and GNI per capita is considered 18. According to the IMF (2014a, pp. 16 and 28), tight enough. Due to data limitations, we focus on the government has recently identified and government spending indicators; c ­ ountry-specific implemented policies that are consistent with its analysis is needed to consider policy in the context ­ objective of a reduction of its total public debt of the different roles of the government and pri- (direct and guaranteed) from 147 percent of GDP vate services and spending. in 2013 to 99 percent in 2020. As part of these 15. Fuel subsidies are detrimental to the climate and policies, the public external debt is projected to encourage technologies that create less employ- decrease from 63.5 percent of GDP in 2013 to 54.0 ment for the growing labor force. in 2020 (IMF 2014a, p. 50). 58 Jamaica Chapter 4 The Kyrgyz Republic 1. Introduction SDG progress since 2000 and projects expected values for 2030; and (b) assesses options for The Kyrgyz Republic, a land-locked and moun- increasing fiscal space. Sections 2 and 3 address tainous lower middle-income country in Central SDGs and fiscal space, respectively, while find- Asia, gained independence from the Russian ings are summarized in Section 4. Federation in 1991 and has since struggled with The analysis is done from a cross-country political and economic stability, not least result- perspective: for the different indicators, the ing from broken trade relations with Russia. Kyrgyz Republic’s performance and prospects After  becoming a parliamentary democracy in are benchmarked relative to other countries, 2010 a number of economic reforms have been considering its past, recent, and projected levels implemented, but the political situation remains of GNI per capita.1 The latter variable tends to be fragile. The Kyrgyz economy is the most open in highly correlated with most of the SDGs and the region, with strong reliance on gold exports most of the factors that determine their evolu- and remittances for its foreign exchange earnings. tion; given this, it is used as a summary indicator During 2001–12, the Kyrgyz Republic’s annual of country capacity to provide and efficiently uti- average growth rate for GNI per capita (at ­constant lize inputs that contribute to SDGs (for example, 2005 US$) was 3.1 percent, which may be com- health and education services) and to achieve pared to a developing (­­low- and middle-­ income) SDG ­ outcomes (like strong health and education country average of 3.0 percent. During the same results).2 period, despite a slight improvement in the index score, the Kyrgyz Republic’s ranking according to the UNDP Human Development Index (among 2. SDG Indicators: History and countries included both in 2000 and in 2012) deteriorated, from the 37th to the 32nd Projections percentile. For selected SDG indicators, table 4.1 summa- This country-at-a-glance note is designed to rizes data for the Kyrgyz Republic: historical provide an initial picture of the challenges that ­ evolution, actual and expected values for a recent the Post-2015 agenda poses for the Kyrgyz year, and projected 2030 values.3 In figure 4.1, Republic, serving as the starting point for a more data for the Philippines is shown in the context of complete country development diagnostic as the estimated cross-country relationship between well as a more comprehensive country-focused each SDG indicator and GNI per capita. For the analysis. The note is built around tables and Kyrgyz Republic, the projected average annual ­ figures that provide data for a selection of SDG rate of GNI per capita growth is 5.1 percent.4 target indicators and indicators related to fiscal The  projected SDG values reflect what can be space—fiscal space matters since, while policy expected given a country’s starting point, pro- frameworks and the engagement of the private jected growth in GNI per capita, typical rates of sector may vary widely, rapid progress on the progress according to cross-country patterns, and SDG agenda will require efficient and carefully a ­gradual convergence to close gaps between prioritized public spending. Drawing on the observed and expected values.5 Projections of the information in these tables and figures, this note SDG indicators are presented only when the briefly (a) summarizes the Kyrgyz Republic’s cross-country relationship between the indicator The Kyrgyz Republic 59 Table 4.1  The Kyrgyz Republic—SDG Indicators: Evolution since 2000 and Projections to 2030 Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Poverty and shared prosperity Poverty, at $1.25 a day 36.8 5.1 25.7 1.5 By 2030, eradicate extreme poverty for all people everywhere, (PPP) (% of population) currently measured as people living on less than $1.25 a day. Shared prosperity: 21.4 19.9 17.1 — By 2030, progressively achieve and sustain income growth of the Income share for bottom 40 percent of the population at a rate higher than the lowest 40% national average. Education Pre-primary enrollment 9.8 24.7 18.0 35.7 By 2030, ensure that all girls and boys have access to quality (% gross) early childhood development, care, and pre-primary education so that they are ready for primary education. Primary completion 93.4 97.7 72.7 98.5 By 2030, ensure that all girls and boys complete free, equitable, (% gross) and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment 84.3 88.2 46.0 91.3 By 2030, ensure that all girls and boys complete free, equitable, (% gross) and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary completion 72.0 18.8 — By 2030, ensure that all girls and boys complete free, equitable, (% gross) and quality primary and secondary education leading to relevant and effective learning outcomes. Primary completion, ratio 98.7 99.4 93.2 — By 2030, ensure that all girls and boys complete free, equitable, of females to males (%) and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment, 102.9 99.6 84.4 102.6 By 2030, ensure that all girls and boys complete free, equitable, ratio of females to and quality primary and secondary education leading to relevant males (%) and effective learning outcomes. Health Under-5 mortality 49.2 24.2 59.8 14.5 By 2030, end preventable deaths of newborns and children (per 1,000 live births) under 5 years of age. Maternal mortality 100 75 273 35 By 2030, reduce the global maternal mortality ratio to less than (per 100,000 live births) 70 per 100,000 live births. HIV prevalence (% of 0.1 0.2 0.9 — By 2030, end the epidemics of AIDS, tuberculosis, malaria, and population ages 15–49) neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. Infrastructure Access to improved 91.5 91.8 32.4 93.8 By 2030, achieve access to adequate and equitable sanitation and sanitation facilities hygiene for all and end open defecation, paying special attention (% of population) to the needs of women and girls and those in vulnerable situations. Access to improved water 78.7 87.6 71.7 91.4 By 2030, achieve universal and equitable access to safe and source (% of population) affordable drinking water for all. Road density (km per 9.3 17.2 11.2 — Develop quality, reliable, sustainable, and resilient infrastructure, 100 sq. km of land) including regional and transborder infrastructure, to support economic development and human well-being, with a focus on affordable and equitable access for all. Access to electricity 100.0 99.0 32.8 99.3 By 2030, ensure universal access to affordable, reliable, and (% of population) modern energy services. Internet users (per 1,000 1.0 23.4 8.0 — Significantly increase access to information and communications people) technology and strive to provide universal and affordable access to the Internet in least developed countries by 2020. table continues next page 60 The Kyrgyz Republic Table 4.1  continued Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Environment CO2 emissions (metric 0.9 1.2 0.3 1.7 Integrate climate change measures into national policies, tons per capita) strategies, and planning. Memorandum item GNI per capita (constant 392 563 1,402 2005 US$ Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used. However, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = currently significantly overperforming; red = currently significantly underperforming; black = performing as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). For Internet use, a projection is not made, despite a tight relationship, since the expected line shifts significantly even in the medium run due to external forces, such as technological development. Global ambition is from the Open Working Group (UN 2014). Figure 4.1  The Kyrgyz Republic—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) a. Poverty (left), shared prosperity: income share of bottom 40% (right) Shared prosperity: Income share for lowest 40% Poverty, at $1.25 a day (PPP) (% of population) 410 30 2011 90 2000 20 20 2000 4 2011 10 .90 2030 9 .04 5 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 11.9*** + −1.3*** ln(GNI pc); R = .493 2 ln(SDG) = 3.2*** + −.06** ln(GNI pc); R 2 = .054 figure continues next page The Kyrgyz Republic 61 Figure 4.1  continued b. Gross pre-primary enrollment (left), primary completion (right) 170 120 2030 2012 Pre-primary enrollment (% gross) 2000 Primary completion (% gross) 2012 80 40 2030 50 10 2000 40 3 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.64 + .560*** ln(GNI pc); R = .325 2 ln(SDG) = 3.48*** + .13*** ln(GNI pc); R 2 = .358 c. Gross secondary enrollment (left), secondary completion (right) 90 110 2030 Secondary completion (% gross) 2012 Secondary enrollment (% gross) 2000 2011 60 20 6 30 20 1.4 6 .09 150 400 1,000 3,000 8,000 150 400 1,000 3,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.94*** + .3*** ln(GNI pc); R 2 = .540 ln(SDG) = −.24 + .5* ln(GNI pc); R 2 = .109 figure continues next page 62 The Kyrgyz Republic Figure 4.1  continued d. Ratio of female to male primary completion (left), ratio of female to male secondary enrollment (right) 140 140 2030 Secondary enrollment, ratio of 2011 Primary completion, ratio of 110 2000 100 2000 females to males (%) females to males (%) 90 2012 70 70 50 40 30 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 4.37*** + .03*** ln(GNI pc); R = .056 2 ln(SDG) = 3.88*** + .09*** ln(GNI pc); R 2 = .268 e. Under-5 mortality (left), maternal mortality (right) Maternal mortality (per 100,000 live births) Under−5 mortality (per 1,000 live births) 500 2,000 230 470 110 100 2000 50 2013 2000 20 2030 20 2013 2030 5 1 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 7.98*** + −.61*** ln(GNI pc); R 2 = .559 ln(SDG) = 11.3*** + −.9*** ln(GNI pc); R 2 = .518 figure continues next page The Kyrgyz Republic 63 Figure 4.1  continued f. HIV prevalence HIV prevalence (% of population ages 15–49) 90 30 9 3 .90 2013 2000 .10 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = 1.88* + −.31** ln(GNI pc); R 2 = .044 g. Access to improved sanitation (left), access to improved water source (right) Access to improved water source (% of population) 2030 Access to improved sanitation facilities 100 2012 100 2030 2000 2012 2000 70 60 (% of population) 50 30 40 20 7 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .61** + .45*** ln(GNI pc); R 2 = .534 ln(SDG) = 3.45*** + .13*** ln(GNI pc); R 2 = .435 figure continues next page 64 The Kyrgyz Republic Figure 4.1  continued h. Access to electricity (left), road density (right) 400 330 Road density (km per 100 sq. km of land) Access to electricity (% of population) 2013 2030 100 90 2000 30 2030 2013 20 7 6 2000 1.6 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .29 + .5*** ln(GNI pc); R 2 = .459 ln(SDG) = −.64 + .48*** ln(GNI pc); R 2 = .189 i. Internet users (left), CO2 emissions (right) 20 CO2 emissions (metric tons per capita) 80 Internet users (per 1,000 people) 3 2030 2013 2010 .80 2000 6 .20 0 2000 0 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −2.9*** + .78*** ln(GNI pc); R 2 = .567 ln(SDG) = −8.6*** + 1.16*** ln(GNI pc); R 2 = .677 Note: Highlighted observations are for the Kyrgyz Republic at different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. and GNI per capita is classified as tight.6 A loose capita is loose the coefficients are typically small relationship suggests that progress in the indica- (in absolute terms); given this, the “expected” tor is primarily a reflection of country-specific ­ values for a recent year are close to the average factors and that it should not be expected to for all low- and middle-income countries.7 respond strongly or systematically to changes in In sum, the Kyrgyz Republic’s current out- GNI per capita. When the relationship to GNI per comes are better than expected (compared to a The Kyrgyz Republic 65 typical country at the same GNI per capita level) low- and middle-income countries stay more or for almost all of the selected SDG indicators less the same for GNI per capita (a slight improve- (poverty,8 shared prosperity, gross pre-primary ment by 2 percentile points). Compared to GNI enrollment, primary completion, gross second- per capita, the progress in ranking was stronger ary enrollment, secondary completion, ratio of for 4  indicators (poverty, under-5 mortality, female to male primary completion, ratio of access to improved water source, and road den- female to male secondary enrollment, under-5 sity) and similar for 2 (maternal mortality and mortality, maternal mortality, HIV prevalence, malaria). For the remaining 11 SDGs, while the access to improved water source, access to Kyrgyz Republic’s performance still remained improved sanitation, road density, access to strong given its GNI per capita, its ranking electricity, and Internet users).9 Only for one ­ deteriorated (HIV prevalence, shared prosperity, ­ indicator, CO2 emissions, is the Kyrgyz Republic pre-primary enrollment, primary completion, doing worse than expected.10 secondary enrollment, ratio of female to male in Figure 4.2 shows that, between 2000 and primary completion, ratio of female to male in 2012, the Kyrgyz Republic saw its ranking among secondary school enrollment, access to improved Figure 4.2  The Kyrgyz Republic—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012 GNI per capita (constant 2005 US$) 19 21 Poverty, at $1.25 a day (PPP) (% of population) 32 52 Shared prosperity: Income share for lowest 40% 86 75 Primary completion (% gross) 65 60 Primary completion, ratio of females to males (%) 56 44 Secondary enrollment, ratio of females to males (%) 63 55 Pre-primary enrollment (% gross) 30 27 Secondary enrollment (% gross) 82 68 Secondary completion (% gross) 86 Under-5 mortality (per 1,000 live births) 51 57 Maternal mortality (per 100,000 live births) 61 63 HIV prevalence (% of population ages 15–49) 99 82 Access to improved sanitation facilities (% of population) 86 80 Access to improved water source (% of population) 40 45 Road density (km per 100 sq. km of land) 39 47 Access to electricity (% of population) 94 75 Internet users (per 1,000 people) 57 52 CO2 emissions (metric tons per capita) 52 51 0 10 20 30 40 50 60 70 80 90 100 2000 Recent Note: A high ranking signals strong performance; for the underlying indicator, this may correspond to a relatively high value (for example, for the secondary enrollment rate) or a relatively low value (for example, for the poverty rate). The rankings are based on all low- and middle- income countries (according to the 2012 classification) with data. The country samples vary across indicators but are always the same for 2000 and recent for any given indicator. The ranking for an indicator is not reported if the available sample is less than 20 countries. Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. 66 The Kyrgyz Republic sanitation, access to electricity, Internet use, and 5.3 percent already is relatively high) and CO2 emissions). improvements in policies that directly influence When comparing the results from regressions different SDGs. on GNI per capita to changes in percentile rank- ings, a few patterns emerge. While the Kyrgyz Republic is performing better than expected for almost all SDGs, despite a more or less constant 3. Fiscal Space GNI per capita ranking, its ranking among other In most countries, accelerated progress on the low- and middle-income countries has deterio- SDG agenda will require efficient and growing rated for the majority of them; shared prosperity, public spending in prioritized areas, most impor- secondary enrollment, and HIV prevalence expe- tantly human development and infrastructure. rienced the largest drops in ranking. The reasons Private spending is also of crucial importance, behind this relative decline are not apparent from both household spending on SDG-related ser- this analysis but may be related to institutional vices and business investments in a wide range of factors (for example, limited capacity to reach out areas (including but not limited to infrastruc- to the relatively small disadvantaged groups that ture). With regard to the Kyrgyz Republic’s fiscal do not have access to various types of services space indicators, table 4.2 and figure 4.3 summa- and infrastructure). On the other hand, poverty, rize the historical evolution, actual and expected under-5 mortality, and road density are not only recent values, and, when relevant, projected overperforming but have improved in country ­values.12 When the relationship is loose, projec- ranking beyond the improvement in ranking for tions are not made and the expected value is in GNI per capita. For CO2 emissions, the only SDG practice close to the average for the sample of all for which the country is underperforming, the low- and middle-income countries (cf. discus- Kyrgyz Republic’s ranking has worsened slightly, sion of expected values for SDG indicators). The an outcome that is not surprising given the variables cover three aspects of government improved GNI per capita ranking; nevertheless, activities: spending, receipts and debt, and gov- the Kyrgyz Republic’s emissions remain high ernance and efficiency. While the findings of this considering its GNI per capita, suggesting that country-at-a-glance note cannot guide policy on the potential for improvements is high.11 their own, they should be seen as an input into By 2030, considerable improvements are pro- thinking about policy making. jected for most indicators (see table 4.1 and In terms of government spending in areas respective graphs in figure 4.1). Poverty, primary that may support the SDG agenda, the Kyrgyz completion rate, ratio of female to male primary Republic’s values are as expected (compared to a completion, ratio of female to male secondary typical country at the same GNI per capita level) enrollment, maternal mortality, malaria, HIV for total public investment, and above the prevalence, and access to electricity are all pro- expected level for total consumption, primary jected to either realize or get close to realizing education (per obvious case of low-priority the post-2015 global ambition (shown in the last spending from the post-2015 agenda perspec- column of table 4.1). However, for other SDGs, ­ tive).13 Cuts in this area seem particularly urgent to get closer to the realization of these ambi- for the Kyrgyz Republic given that both its spend- tions, a break with the past seems needed. This is ing (as high as 8.9 percent of GDP) and its CO2 also true for indicators such as shared prosper- emissions are above the expected levels. ity, for which a weak relationship with GNI per Among current receipts, Official Development capita precludes projections. Such a break would Assistance (ODA) is within the expected range, be facilitated by a combination of more rapid while tax revenues are higher than expected. growth (which for the Kyrgyz Republic may be As  further shown in table 4.2, cross-country difficult given that the projected growth rate of ­ patterns suggest that net ODA will decline as a The Kyrgyz Republic 67 Table 4.2  The Kyrgyz Republic—Fiscal Space: Revenue, Spending, and Government Efficiency Actual Expected Projection Indicator 2000 Recent Recent 2030 Government spending Consumption (% of GDP) 20.0 18.1 13.1 — Investments (% of GDP) 7.8 9.2 6.6 — Primary education, per student (% of GDP per capita) 17.1 11.9 — Secondary education, per student (% of GDP per capita) 25.5 19.4 — Health (% of GDP) 2.1 4.3 2.6 — Fuel subsidy (% of GDP) 8.9 2.4 — Government receipts and debts Tax revenue (% of GDP) 11.7 18.1 12.7 — Net ODA (% of GNI) 16.7 7.3 5.9 2.9 External debt (% of GNI) 150.5 98.4 30.9 — Governance and government efficiency Government effectiveness: percentile rank 33.2 28.7 17.2 35.4 Grigoli health efficiency score 0.9 0.9 — Public investment management index 1.4 1.4 — Memorandum item GNI per capita (constant 2005 US$) 392 563 1,402 Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = currently significantly overperforming; red = currently significantly underperforming; black = performing as expected; — = no projection because the cross-country relationship was not tight enough (see criteria earlier in the note). Figure 4.3  The Kyrgyz Republic—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting a. Government spending: consumption (left), investment (right) 130 50 60 20 Consumption (% of GDP) Investments (% of GDP) 30 2000 2013 9 2013 2000 10 4 2 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.01*** + .09** ln(GNI pc); R 2 = .042 ln(SDG) = 2.37*** + −.07 ln(GNI pc); R 2 = .013 figure continues next page 68 The Kyrgyz Republic Figure 4.3  continued b. Government spending: primary education, per student (left), secondary education, per student (right) 90 50 Secondary education, per student Primary education, per student 30 50 2013 (% of GDP per capita) (% of GDP per capita) 20 20 8 10 2 4 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.72*** + .12** ln(GNI pc); R = .067 2 ln(SDG) = 3.51*** + –.09 ln(GNI pc); R 2 = .030 c. Government spending: health (left), fuel subsidy (right) 20 20 2011 1.6 Fuel subsidy (% of GDP) Health (% of GDP) 6 2013 2030 .10 2000 1.6 .00 .50 150 400 1,000 3,000 8,000 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.25 + .18*** ln(GNI pc); R 2 = .111 ln(SDG) = 4.96*** + −.65*** ln(GNI pc); R 2 = .098 figure continues next page The Kyrgyz Republic 69 Figure 4.3  continued d. Tax revenue (left), official development aid (right) 60 60 2000 2013 8 30 2030 Tax revenue (% of GDP) Net ODA (% of GNI) 20 2012 1.2 2000 10 .20 3 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.69*** + .13* ln(GNI pc); R = .030 2 ln(SDG) = 8.10*** + −1*** ln(GNI pc); R 2 = .298 e. External debt (left), government effectiveness (right) Government effectiveness: percentile rank 2,000 80 2030 650 2000 30 External debt (% of GNI) 2013 2000 160 2013 10 50 4 10 1.3 .50 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.52*** + .14** ln(GNI pc); R 2 = .035 ln(SDG) = 0 + .44*** ln(GNI pc); R 2 = .300 figure continues next page 70 The Kyrgyz Republic Figure 4.3  continued f. Health expenditure efficiency (left), public investment management index (right) 1 4 Public investment management index Grigoli health efficiency score 2 2010 1.3 2010 .90 .80 .80 .30 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.09*** + .01 ln(GNI pc); R = .031 2 ln(SDG) = −.43 + .12** ln(GNI pc); R 2 = .079 Note: Highlighted observations are for the Kyrgyz Republic at different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. percent of GDP (without changing significantly in impact on the SDG agenda. Table 4.2 displays per capita terms), which in the case of the Kyrgyz data for some measures of government efficiency. Republic translates into a reduction from 7.3 to According to both the health index used in this 2.9  percent of GNI.14 However, the fact that study, the Kyrgyz Republic’s performance is cross-country patterns point to a likely decline in below the expected levels. However, the Kyrgyz ODA does not mean that an increase is excluded: it Republic is performing as expected in terms of depends on the priorities of donors and their rela- the more general World Bank Government tionships with the Krygyz Republic’s government. Effectiveness indicator and as expected for Public The relationship between tax revenues and Investment Management Index; that is according GNI per capita, as well as the debt stock and GNI to these indicators, government efficiency per capita, is not tight enough to project expected appears to be better than expected given the GNI changes. The higher than expected tax revenues per capita of the Kyrgyz Republic, an observation in the Kyrgyz Republic suggest that it is unlikely that does not negate that considerable efficiency taxes will be a major contributor to increased fis- gains still may be feasible in different areas. For cal space.15 This is also true for the currently example, the higher than expected health expen- higher than expected external debt stock, which ditures, together with below expected health limits the likelihood of increased external bor- expenditure efficiency, suggests that health effi- rowing. The fiscal impact could be substantial if ciency improvements should be considered. borrowing goes beyond what is consistent with In sum, our cross-country results indicate debt sustainability.16 that government spending already is at or above Government efficiency is important to pro- expected levels while it also seems difficult to sig- tect and, if possible, increase in order to add to nificantly raise revenues from any major sources the room for priority spending and enhance its (taxes, ODA, or borrowing). However, on the The Kyrgyz Republic 71 spending side, there may be room for cuts in and channeled into spending in support of the low-priority areas, including, most obviously, SDG agenda, including growth promotion, then spending on fuel subsidies; resource-saving effi- payoffs could be substantial. More broadly, from ciency gains may also be possible in various the perspective of this agenda and given strong areas. Such adjustment would permit budgetary linkages between private and government activi- changes that, directly or indirectly, promote the ties and incomes, it is crucial that policies and SDG agenda; the details should depend on gov- spending decisions promote a broad-based change ernment priorities and additional information. that encompasses services related to human devel- For example, while road density is better than opment, infrastructure investments, and other expected, the Kyrgyz Republic infrastructure measures in support of strong long-run growth assets are in very poor condition, suffering from that is biased in favor of the less advantaged. Given critical investment shortage and insufficient and the Kyrgyz Republic’s impressive performance in inadequate maintenance (including electricity, terms of human development and infrastructure urban heating, roads, and irrigation). Likewise, (considering its low GNI per capita), the potential while enrollment rates are met or exceeded, the for strong future growth and additional progress education system does not produce the quality on the SDG agenda should be very strong if the learning necessary to spur growth. Most impor- country manages to put in place appropriate poli- tantly, if the government, on the margin, would cies and ­create a supportive business climate.17 be able to expand human development and infra- structure services with sufficient efficiency, then spending increases in these areas may be advis- able; if not, it may be better for the SDG agenda 4. Conclusions to selectively reduce taxation. As summarized in table 4.3, the Kyrgyz Republic’s Beyond the government, it is important to note current outcomes are better than expected (com- that, in 2013, remittances accounted for as much pared to a typical country at the same GNI per as 31.5 percent of GDP (World Bank); if a substan- capita level) for all of the selected SDGs except tial part of these flows could be further encour- one, CO2 emissions, for which performance is aged (for example, via reduced transactions costs) worse than expected. By 2030, many of the SDGs Table 4.3  The Kyrgyz Republic—Summary Results for SDG Indicators Cross-country relationship with Overperforming As expected Underperforming GNI/cap Tight •• Poverty (+) •• Road density (+) •• CO2 emissions (–) •• Pre-primary school enrollment (–) •• Primary completion (–) •• Secondary school enrollment (–) •• Ratio of female to male secondary enrollment (–) •• Under-5 mortality (+) •• Maternal mortality (x) •• Malaria (x) •• Access to improved water source (x) •• Access to improved sanitation (–) •• Internet users (–) Loose •• Shared prosperity (–) •• Secondary completion rate •• Ratio of female to male primary completion (–) •• HIV prevalence (–) Note: (+) = larger country rank increase, 2000–12, or smaller deterioration than for GNI per capita; (–) = smaller increase or a decreased country rank 2000–12 compared to GNI per capita; (×) = the same country rank change 2000–12 as for GNI per capita (+/– 2 percentile points). 72 The Kyrgyz Republic are projected to either realize or come close to rankings for shared prosperity, ratio of female to realizing the global goals. However, for some male in primary completion, and HIV preva- SDGs (such as pre-primary enrollment, second- lence have all worsened since 2000. ary school enrollment, under-5 mortality, access The only area in which the Kyrgyz Republic to improved water source, and access to improved currently is underperforming is CO2 emission. sanitation), to get closer to the realization of these The presence of a tight negative relationship to ambitions, a break with the past is needed. GNI per capita suggests that future improvements Table 4.3 further shows that, for most indica- strongly depend on policies that keep emissions tors, the relationship to GNI per capita is tight in check. and the Kyrgyz Republic is doing better than In the fiscal area, cross-country data suggest expected. Improvements in these SDGs will most that spending is relatively high while the room to likely continue along with GNI per capita growth raise revenues is limited. Given this, improved and increases in resources and capabilities. spending efficiency should be a high priority, However, among the currently overperforming including cuts in fuel subsidies. Potential uses of SDGs, some are only loosely related to GNI per any new fiscal space include selective spending capita; this group includes shared prosperity, increases and tax cuts; the preferred path should secondary completion, ratio of female to male ­ depend on government priorities and the ability in  primary completion, and HIV prevalence. of the government to provide SDG-related ser- The  fact that these relationships are loose vices with sufficient efficiency. suggests  that these indicators should not be ­ Finally, the Kyrgyz Republic’s strong position expected to improve strongly or systematically to in terms of human development and infrastruc- more rapid growth in GNI per capita but rather ture puts the country in a favorable position to would depend on ­ country-specific ­ conditions bring about strong growth and SDG performance and policies. Among these indicators, the during the period up to 2030. Annex 4A: Data Sources Indicator Source Comment GNI per capita (constant WDI. API ref: GNI per capita (constant 2005 US$) 2005 US$) [NY.GNP.PCAP.KD] SDG indicators Poverty, at $1.25 a day (PPP) WDI. API ref: poverty headcount ratio at $1.25 a day (PPP) (% of population) (% of population) [si.pov.dday] Shared prosperity: income share WDI. API ref: income share held by lowest 20% + Income for lowest 40% share held by second 20% [SI.DST.FRST.20+SI.DST.02ND.20] Pre-primary enrollment (% gross) WDI. API ref: school enrollment, pre-primary (% gross) [SE.PRE.ENRR] Primary completion (% gross) WDI. API ref: primary completion rate, total (% of relevant age group) [se.prm.cmpt.zs] Secondary enrollment (% gross) WDI. API ref: school enrollment, secondary (% gross) [se.sec.enrr] Secondary completion (% gross) EdStats. API ref: DHS: secondary completion rate [hh.dhs.scr] Primary completion, ratio of WDI. API ref: primary completion rate, female (% of relevant females to males (%) age group)/primary completion rate, male (% of relevant age group) *100 [SE.PRM.CMPT.FE.ZS/SE.PRM.CMPT. MA.ZS*100] Secondary enrollment, ratio of WDI. API ref: ratio of female to male secondary enrollment females to males (%) (%) [se.enr.seco.fm.zs] annex continues next page The Kyrgyz Republic 73 Indicator Source Comment Under-5 mortality (per 1,000 live WDI. API ref: mortality rate, under-5 (per 1,000 live births) births) [SH.DYN.MORT] Maternal mortality (per 100,000 WDI. API ref: maternal mortality ratio (modeled estimate, live births) per 100,000 live births) [sh.sta.mmrt] Malaria cases (% of population) HNP. API ref: malaria cases reported/population, total *100 [sh.sta.malr/SP.POP.TOTL *100] HIV prevalence (% of population WDI. API ref: prevalence of HIV, total (% of population ages ages 15–49) 15–49) [sh.dyn.aids.zs] Access to improved sanitation WDI. API ref: improved sanitation facilities (% of population facilities (% of population) with access) [sh.sta.acsn] Access to improved water WDI. API ref: improved water source (% of population with source (% of population) access) [sh.h2o.safe.zs] Road density (km per 100 sq. WDI. API ref: road density (km of road per 100 sq. km of km of land) land area) [is.rod.dnst.k2] Access to electricity WDI. API ref: access to electricity (% of population) (% of population) [eg.elc.accs.zs] Internet users (per 1,000 people) WDI. API ref: Internet users (per 100 people) [IT.NET.USER.P2] CO2 emissions (metric tons per WDI. API ref: CO2 emissions (metric tons per capita) capita) [en.atm.co2e.pc] Fiscal space indicators Investment (% of GDP) WDI. API ref: gross fixed capital formation (% of GDP)-gross fixed capital formation, private sector (% of GDP) [ne.gdi.ftot.zs]-[ne.gdi.fprv.zs] Consumption (% of GDP) WDI. API ref: general government final consumption expenditure (% of GDP) [NE.CON.GOVT.ZS] Primary education (% of GDP) EdStats. API ref: total expenditure on educational institutions and administration as a % of GDP. All sources. Primary [UIS.XGDP.1.FDINSTADM.FFD] Secondary education EdStats. API ref: total expenditure on educational (% of GDP) institutions and administration as a % of GDP. All sources. Secondary and post-secondary non-tertiary [UIS.XGDP.234. FDINSTADM.FFD] Primary education, per student WDI. API ref: expenditure per student, primary (% of GDP per capita) (% of GDP per capita) [se.xpd.prim.pc.zs] Secondary education, per WDI. API ref: expenditure per student, secondary student (% of GDP per capita) (% of GDP per capita) [se.xpd.seco.pc.zs] Health (% of GDP) WDI. API ref: health expenditure, public (% of GDP) [sh.xpd.publ.zs] Fossil fuel subsidy (% of GDP) IMF (2013a). Total pretax subsidy (% of GDP) Subsidies are measured using a price-gap approach capturing both consumer (including implicit) and producer (except those that arise when suppliers are inefficient and make losses at benchmark prices) subsidies. Negative external effects are not included in the pretax subsidy. Tax revenue (% of GDP) WDI. API ref: tax revenue (% of GDP) [gc.tax.totl.gd.zs] Net ODA (% of GNI) WDI. API ref: net ODA received (% of GNI) [dt.oda.odat.gn.zs] External debt (% of GNI) WDI. API ref: external debt stocks (% of GNI) [DT.DOD.DECT.GN.ZS] Government effectiveness: WGI. API ref: government effectiveness: percentile rank Captures perceptions of the quality of percentile rank [ge.per.rnk] public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies. annex continues next page 74 The Kyrgyz Republic Indicator Source Comment Grigoli education efficiency Grigoli (2014) Measure secondary education score inefficiency in terms of how much additional output could be achieved at current levels of spending. Grigoli health efficiency score Grigoli and Kapsoli (2013) Quantifies the inefficiency of public health expenditure using a stochastic frontier model that controls for the socioeconomic determinants of health. Public investment management Dabla-Norris et al. (2011) Four phases associated with public index investment management are covered: project appraisal, selection, implementation, and evaluation. Note: WDI = World Development Indicators, World Bank; API ref = reference and code when using the World Bank Open Data; EdStats = Education statistics, World Bank; HNP = Health Nutrition and Population statistics, World Bank; WGI = Worldwide Governance Indicators, World Bank. Notes 1. While a cross-country perspective provides an 7. In addition, the confidence interval is wide in the important complement to analysis that is centered case of a loose relationship, suggesting that any on an individual country, it is by definition lim- conclusion on over- or underperformance is made ited to analysis of variables that are available in with wide margins. Statistically, even though their cross-country databases. confidence intervals are wide, as long as the esti- 2. This does not mean that GNI per capita is viewed mated coefficient linking GNI per capita to the SDG as a direct determinant of SDG outcomes; on the indicator is nonzero, these values are closer than contrary, a major challenge for policy makers is to the cross-country average to what is expected for identify policies that improve SDG performance the specific country. The same observation applies relative to what is expected given the level of GNI to expected values for fiscal space indicators. per capita. A second challenge is to raise growth in 8. Poverty measured as share of population living on GNI per capita as it indirectly influences country less than $1.25 a day dropped sharply in early 2000s SDG capacity. and was 5.1 percent in 2011. However, poverty 3. Sources for the indicators are presented in annex 4A. according to the national poverty line remains very 4. Projections from CEPII are used for this and pervasive in the Kyrgyz Republic, affecting over 38 other Country Development Diagnostics appli- percent of the population in 2012, and on the rise cations given their wide country coverage and since 2009 as a result of the ongoing economic slow- well-documented methodology; OECD data down and the protracted impacts of the Russian cri- have been used when projections have been sis and economic contraction (WDI, World Bank). missing. In the projections, it is assumed that 9. With the relatively low population density in the future GNI growth will coincide with future GDP Kyrgyz Republic, the higher than expected road growth (both expressed in constant 2005 US$) density is even more noteworthy. given that that this is the variable that CEPII and 10. With regard to CO2, the Kyrgyz Republic’s current other sources project. and project 2030 per capita emissions are 11.6 and 5. Given that (a) SDGs have extreme values (such as 19.1 percent of the current OECD average. 100 percent for improved water access) and (b) the 11. The main reason behind high CO2 emissions is current SDG level never is exactly as expected given high electricity consumption per capita, not the GNI per capita, the projected values gradually con- sources of energy—the share of electricity from verge toward the expected values. For example, for fossil fuels is somewhat lower than expected while a country that overperforms in water access, as GNI the share from hydroelectric plants is higher. per capita increases the extent of overperformance 12. The treatment is the same as in table 4.1 and related gradually declines, so that when the expected value figures. That is, in table 4.2, projections are shown is 100, overperformance has reached zero. only when the cross-country relationship between 6. A tight enough relationship is defined as an the indicator and GNI per capita is considered R2 > 0.3 (tight) or 0.3 > R2 > 0.1 (moderately tight), tight enough. Due to data limitations, we focus on while R2 < 0.1 are defined as loose. government spending indicators; c ­ ountry-specific The Kyrgyz Republic 75 analysis is needed to consider policy in the context 16. External public debt in the Kyrgyz Republic of the different roles of the government and pri- remains at a moderate risk of debt distress and is vate services and spending. projected to decrease from 46 percent of GDP in 13. Fuel subsidies are detrimental to the climate and 2012 to 36 in 2018 (IMF 2013, pp. 8, 23). encourage technologies that are less labor inten- 17. The Kyrgyz Republic scores 61 in ease of doing sive, tending to generate employment for fewer business (100 is best practice country), while the workers at lower wages. regional average for Europe and Central Asia is 14. Reducing the reliance of external aid is part of the 67 (Doing Business 2014, p. 8). Moreover, the governments’ agenda for macroeconomic stability Kyrgyz Republic is underperforming in terms (IMF 2013d). of business climate according to results from 15. IMF (2013, p. 23) suggests a marginal increase of regressing ease of doing business scores on tax revenues from 26.0 in 2012 to 26.6 percentage GNI  per capita for the full sample of low- and of GDP in 2018. middle-income countries. 76 The Kyrgyz Republic Chapter 5 Liberia 1. Introduction carefully prioritized public spending. Drawing on the information in these tables and figures, this Liberia, which is classified as a low-­income coun- note briefly (a) summarizes Liberia’s SDG progress try, is highly dependent on mining resources for since 2000 and projects expected values for 2030; foreign exchange earnings. A violent civil war and (b) assesses options for increasing fiscal space. ended with a peace agreement in 2003. More Sections 2 and 3 address SDGs and fiscal space, recently, it has been at the center of the Ebola epi- respectively, while findings are summarized in demic, the impact of which is not reflected in the Section 4. data that were available for this note. For Liberia, The analysis is done from a cross-country social and economic data are uncertain and perspective: for the different indicators, Liberia’s ­ lacking. Nevertheless, available data indicate performance and prospects are benchmarked that, during the period 2001–12, the country’s relative to other countries, considering its past, average growth rate for GNI per capita (at con- recent and projected levels of GNI per capita.2 stant 2005 US$) was 4.0 percent; higher than the The latter variable tends to be highly correlated developing (low- and middle-income) country with most of the SDGs and most of the factors average of 3.0 percent; however, the initial income that determine their evolution; given this, it is level was very low and the country is still one of used as a summary indicator of country capacity the poorest in the world. After a period of high to provide and efficiently utilize inputs that con- volatility during the conflict years, growth has tribute to SDGs (for example health and educa- stabilized since 2004; for the period ­ 2005–13, the tion services) and to achieve SDG outcomes average was about 10 percent, mainly based on (like strong health and education results).3 nonmining activities and presumably reflecting a post-conflict recovery.1 Between 2000 and 2012, despite a slight improvement in its score accord- 2. SDG Indicators: History and ing to the UNDP Human Development Index, Liberia’s ranking according to the index (among Projections countries included both in 2000 and 2012) stayed For selected SDG indicators, table 5.1 summarizes more or less the same (deteriorated from the 6th data for Liberia: historical evolution, actual and to the 4th percentile). expected values for a recent year, and projected This country-at-a-glance note is designed to 2030 values.4 In figure 5.1, data for Liberia are provide an initial picture of the challenges that the shown in the context of the estimated cross-­ Post-2015 agenda poses for Liberia, serving as the country relationship between each SDG indicator starting point for a more complete country and GNI per capita. For Liberia, the projected development diagnostic as well as a more in-depth ­ average annual rate of GNI per capita growth is country-focused analysis. The note is built 4.1  percent.5 The projected SDG values reflect around  tables and figures that provide data for a what  can be expected given a country’s starting selection of SDG target indicators and indicators point, projected growth in GNI per capita, typical related to fiscal space—fiscal space matters since, rates of  progress according to cross-­ country while policy frameworks and the engagement of ­ patterns, and a gradual convergence to close the private ­sector may vary widely, rapid progress gaps  between observed and expected values.6 on the SDG  agenda will require efficient and Projections for SDG indicators are presented only Liberia 77 when the cross-country relationship between the country-specific conditions that are difficult to indicator and GNI per capita is classified as tight.7 change, it may often point to areas in which pay- A loose relationship suggests that progress in the offs from feasible policy change are relatively indicator is primarily a reflection of country-­ high, a possibility that calls for further analysis. specific factors and that it should not be expected For the 13 indicators with sufficient data, to respond strongly or systematically to changes in ­ figure  5.2 shows Liberia’s changes in percentile GNI per capita. When the relationship to GNI per rankings among low- and middle-income coun- capita is loose the coefficients are typically small tries between 2000 and 2012.10 The country’s GNI (in absolute terms); given this, the “expected” per capita ranking stayed the same (2nd percentile ­ values for a recent year are close to the average for in both 2000 and 2012), and so did the rankings all low- and middle-income countries.8 for 4 of the SDGs: maternal mortality, malaria In sum, among the selected 15 indicators in cases, access to electricity, and CO2 emissions. The table 5.1, Liberia’s current outcomes are better rankings improved for 4 indicators—under-5 than expected (compared to a typical country at mortality, HIV prevalence, access to improved the same, very low, GNI per capita level) for 5 water sources, and Internet use—but deteriorated indicators: poverty, pre-primary enrollment, sec- for 5: pre-primary enrollment, secondary enroll- ondary enrollment, under-5 mortality, and access ment, ratio of female to male primary completion, to improved water source. The country falls short ratio of female to male secondary enrollment, and for 6 indicators: primary completion, ratio of access to improved sanitation. female to male primary completion, malaria When comparing the results from regressions cases, access to improved sanitation, access to on GNI per capita to changes in percentile rank- electricity, and CO2 emissions.9 Outcomes are as ings, a few patterns emerge; unfortunately, some expected for 4 of the indicators: ratio of female to of these are discouraging: Liberia’s rankings are male secondary enrollment, maternal mortality, deteriorating among the SDGs for which it has HIV prevalence, and Internet use. While under- better than expected performance (secondary performance for an indicator may be due to enrollment and the ratio of female to male  iberia—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Figure 5.1  L Setting (Log Scale) a. Poverty Poverty, at $1.25 a day (PPP) (% of population) 410 90 2010 20 2030 4 .90 .04 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = 11.8*** + −1.3*** ln(GNI pc); R 2 = .487 figure continues next page 78 Liberia Figure 5.1  continued b. Gross pre-primary enrollment (left), Primary completion (right) 170 120 Pre-primary enrollment (% gross) Primary completion (% gross) 2013 2030 80 2000 40 2030 50 2013 10 40 3 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.41 + .53*** ln(GNI pc); R 2 = .301 ln(SDG) = 3.47*** + .13*** ln(GNI pc); R 2 = .363 c. Gross secondary enrollment 190 110 Secondary enrollment (% gross) 2030 60 2011 2000 30 20 6 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = 1.9*** + .31*** ln(GNI pc); R 2 = .558 figure continues next page Liberia 79 Figure 5.1  continued d. Ratio of female to male primary completion (left), Ratio of female to male secondary enrollment (right) Secondary enrollment, ratio of females to males (%) Primary completion, ratio of females to males (%) 140 140 110 110 2011 90 90 2030 2000 2000 2013 70 70 40 40 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 4.38*** + .03*** ln(GNI pc); R 2 = .055 ln(SDG) = 3.82*** + .1*** ln(GNI pc); R 2 = .297 e. Under-5 mortality (left), maternal mortality (right) 600 2,000 Maternal mortality (per 100,000 live births) 2000 Under−5 mortality (per 1,000 live births) 230 2000 470 2013 2030 110 100 2013 50 2030 20 20 5 1 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 7.99*** + −.61*** ln(GNI pc); R 2 = .660 ln(SDG) = 11.3*** + −.9*** ln(GNI pc); R 2 = .519 figure continues next page 80 Liberia Figure 5.1  continued f. Malaria cases (left), HIV prevalence (right) HIV prevalence (% of population ages 15–49) 30 2012 90 1998 2030 30 Malaria cases (% of population) 1.3 9 .05 2000 3 .90 2013 .10 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 14.5*** + −2.3*** ln(GNI pc); R 2 = .412 ln(SDG) = 1.88* + −.31** ln(GNI pc); R 2 = .044 g. Access to improved sanitation (left), Access to improved water source (right) Access to improved sanitation facilities 100 Access to improved water source 100 2012 2030 70 (% of population) (% of population) 60 2000 50 30 2030 40 20 2012 2000 7 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .61** + .45*** ln(GNI pc); R 2 = .536 ln(SDG) = 3.45*** + .13*** ln(GNI pc); R 2 = .436 figure continues next page Liberia 81 Figure 5.1  continued h. Access to electricity 400 Access to electricity (% of population) 100 30 2030 6 2010 1.6 2000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = .29 + .5*** ln(GNI pc); R 2 = .458 i. Internet users (left), CO2 emissions (right) 20 80 CO2 emissions (metric tons per capita) Internet users (per 1,000 people) 3 2013 6 .80 2030 2010 .20 .40 2000 .03 2000 0 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −2.9*** + .78*** ln(GNI pc); R 2 = .568 ln(SDG) = −8.6*** + 1.16*** ln(GNI pc); R 2 = .677 Note: Highlighted observations are for Liberia at different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. 82 Liberia Table 5.1  Liberia—SDG Indicators: Evolution since 2000 and Projections to 2030 Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Poverty and shared prosperity Poverty, at $1.25 a day (PPP) 45.1 — 13.2 By 2030, eradicate extreme poverty for all people (% of population) everywhere, currently measured as people living on less than $1.25 a day. Education Pre-primary enrollment (% gross) 61.1 72.8 12.3 74.5 By 2030, ensure that all girls and boys have access to quality early childhood development, care, and pre-primary education so that they are ready for primary education. Primary completion (% gross) 59.0 66.0 66.7 By 2030, ensure that all girls and boys complete free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment (% gross) 35.2 45.2 35.2 53.7 By 2030, ensure that all girls and boys complete free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Primary completion, ratio of 84.1 85.7 92.5 — By 2030, ensure that all girls and boys complete free, females to males (%) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary completion, ratio of 72.7 82.0 78.9 87.6 By 2030, ensure that all girls and boys complete free, females to males (%) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Health Under-5 mortality (per 1,000 175 71 101 46 By 2030, end preventable deaths of newborns and live births) children under 5 years of age. Maternal mortality (per 100,000 1,100 640 592 341 By 2030, reduce the global maternal mortality ratio to live births) less than 70 per 100,000 live births. Malaria cases (% of population) 30.4 29.0 3.6 5.8 By 2030, end the epidemics of AIDS, tuberculosis, malaria, and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. HIV prevalence (% of population 3.1 1.1 1.2 — By 2030, end the epidemics of AIDS, tuberculosis, malaria, ages 15–49) and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. Infrastructure Access to improved sanitation 13.6 16.8 21.6 26.2 By 2030, achieve access to adequate and equitable facilities (% of population) sanitation and hygiene for all and end open defecation, paying special attention to the needs of women and girls and those in vulnerable situations. Access to improved water source 61.2 74.6 63.9 79.2 By 2030, achieve universal and equitable access to safe (% of population) and affordable drinking water for all. Access to electricity 0.6 4.1 18.8 16.5 By 2030, ensure universal access to affordable, reliable, (% of population) and modern energy services. Internet users (per 1,000 people) 0.0 4.6 4.1 — Significantly increase access to information and communications technology and strive to provide universal and affordable access to the Internet in least developed countries by 2020. Environment CO2 emissions (metric tons 0.2 0.2 0.1 0.4 Integrate climate change measures into national policies, per capita) strategies, and planning. Memorandum item GNI per capita (constant 145 231 491 2005 US$) Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = currently significantly overperforming; red = currently significantly underperforming; black = performing as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). For Internet use, a projection is not made, despite a tight relationship, since the expected line shifts significantly even in the medium run due to external forces, such as technological development. Global ambition is from the Open Working Group (UN 2014). Liberia 83 Figure 5.2  Liberia—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012 GNI per capita (constant 2005 US$) 2 2 Poverty, at $1.25 a day (PPP) (% of population) 24 Pre-primary enrollment (% gross) 30 19 Primary completion (% gross) 32 Secondary enrollment (% gross) 30 19 Primary completion, ratio of females to males (%) 25 13 Secondary enrollment, ratio of females to males (%) 22 14 Under-5 mortality (per 1,000 live births) 8 23 Maternal mortality (per 100,000 live births) 6 7 Malaria cases (% of population) 1 1 HIV prevalence (% of population ages 15–49) 24 36 Access to improved sanitation facilities (% of population) 12 9 Access to improved water source (% of population) 23 26 Access to electricity (% of population) 1 2 Internet users (per 1,000 people) 4 15 CO2 emissions (metric tons per capita) 84 83 0 10 20 30 40 50 60 70 80 90 100 2000 Recent Note: A high ranking signals strong performance; for the underlying indicator, this may correspond to a relatively high value (for example, for the secondary enrollment rate) or a relatively low value (for example, for the poverty rate). The rankings are based on all low- and middle-income countries (according to the 2012 classification) with data. The country samples vary across indicators but are always the same for 2000 and recent for any given indicator. The ranking for an indicator is not reported if the available sample is less than 20 countries. Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. secondary enrollment); for two other indicator By 2030, considerable improvements are (ratio of female to male primary completion and ­ projected for most indicators (see table 5.1 and access to improved sanitation facilities) Liberia is respective graphs in figure 5.1).11 However, com- both underperforming and experiencing a dete- pared to global ambitions, also shown in table 5.1, riorating ranking. the improvements are moderate, not least due to 84 Liberia very low initial levels, which means that the reali- relevant, projected values.13 When the relationship zation of such ambitions would require a dra- is loose, projections are not made and the expected matic break with the past. Such a break would be value is in practice close to the average for the facilitated by more rapid and more inclusive sample of all low- and middle-income countries ­ growth combined with SDG policies that benefit (cf. discussion of expected values for SDG indica- the disadvantaged. tors). In addition, they should be interpreted with a lot of caution given large off-budget donor spend- ing in different areas (not reflected in government spending data) and uncertain macro (GDP and 3. Fiscal Space GNI) data—as a result, the values for spending In most countries, accelerated progress on the SDG indicators expressed as a share of GDP could be agenda will require efficient and growing public misleading. The variables cover selected indicators spending in prioritized areas, most importantly related to three aspects of government activities: human development and infrastructure. Private spending, receipts and debt, and governance and spending is also of crucial importance, both house- efficiency. While the findings of this country-at-a- hold spending on SDG-related services and busi- glance note cannot guide policy on their own, they ness investments in a wide range of areas (including should be an input into the discussion on policy but not limited to infrastructure).12 With regard to making. Liberia’s fiscal space indicators, table 5.2 and In general terms, room for additional priority figure  5.3 summarize the historical evolution, ­ spending may be created by reducing low-priority actual and expected recent values, and, when spending, increasing current receipts, and/or Table 5.2  Liberia—Fiscal Space: Revenue, Spending, and Government Efficiency Actual Expected Projection Indicator 2000 Recent Recent 2030 Government spending Consumption (% of GDP) 7.5 15.5 12.0 — Investments (% of GDP) 6.3 7.6 7.2 — Primary education (% of GDP) 1.7 1.9 — Secondary education (% of GDP) 1.3 0.9 — Primary education, per student (% of GDP per capita) 17.0 10.5 — Secondary education, per student (% of GDP per capita) 61.0 21.7 — Health (% of GDP) 1.4 3.6 2.1 3.9 Government receipts and debts Tax revenue (% of GDP) 20.9 11.1 — Net ODA (% of GNI) 13.1 16.7 14.2 7.9 External debt (% of GNI) 551.7 30.9 27.3 — Governance and government efficiency Government effectiveness: Percentile rank 8.6 11.7 13.0 Grigoli health efficiency score 0.97 0.94 — Memorandum item GNI per capita (constant 2005 US$) 145 231 491 Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = current value significantly above the expected level; red = current value significantly below the expected level; black = current value as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). Liberia 85 increasing borrowing. Available data indicate this will hold depends on the specifics of Liberia’s that for government spending (expressed as a evolving relationship with donors and a possible percent of GDP), Liberia performs as expected shift from peace-keeping support to more devel- (compared to a typical country at the same opment-oriented support.16 If support to UN GNI per capita level) for total public investment, peace-keeping forces is included, the current level while government total consumption and of net ODA is as high as 36.1 percent. The rela- health  spending is higher than expected. For tionship between tax revenues and GNI per capita education-related spending, the government is ­ is not tight enough to project future values. spending as expected for primary when mea- However, even though the tax-GDP ratio is high, sured in percentage of GDP, but more than IMF projects an increase in tax revenues.17 Higher expected when measured per student; secondary taxes would reduce the resources controlled by spending is more than expected no matter if it is domestic households and firms, pointing to the measured as a share of GDP or per student.14 need to consider the combined impact on SDGs Of the government receipts included in and other indicators from higher taxes and the table  5.2, net Official Development Assistance spending increases that are financed by these (ODA) is as expected and tax revenues are higher taxes. Liberia’s external debt stock is at the than expected.15 As further shown, cross-country expected level, but the Liberia Debt Sustainability patterns suggest that, as GNI per capita grows, net Analysis (DSA) suggest there may be room for ODA declines as a percent of GDP (without increased external borrowing consistent with debt changing significantly in per capita terms). In the sustainability.18 case of Liberia this suggests that net ODA, Government efficiency is important to pro- excluding support to UN peace-keeping forces, is ­ tect and, if possible, increase in order to add to expected to decrease from the current 16.7 percent the room for priority spending and enhance its of GNI to 7.7 percent in 2030; however, whether impact on the SDG agenda. Table 5.2 displays Figure 5.3  Liberia—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting a. Government spending: consumption (left), investment (right) 130 50 80 Consumption (% of GDP) 20 Investments (% of GDP) 30 9 2012 2012 2000 10 4 2000 2 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.01*** + .09** ln(GNI pc); R 2 = .042 ln(SDG) = 2.34*** + –.07 ln(GNI pc); R 2 = .012 figure continues next page 86 Liberia Figure 5.3  continued b. Government spending: primary education, per student (left), secondary education, per student (right) 90 50 Secondary education, per student Primary education, per student 30 2010 50 (% of GDP per capita) (% of GDP per capita) 20 2010 20 8 10 2 4 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.67*** + .13*** ln(GNI pc); R = .076 2 ln(SDG) = 3.52*** + –.08 ln(GNI pc); R 2 = .029 c. Government spending: Primary education (left), Secondary education (right) 6 4 Secondary education (% of GDP) 4 Primary education (% of GDP) 2 2 2010 1.2 2010 1.1 .60 .40 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .940*** + –.06 ln(GNI pc); R 2 = .016 ln(SDG) =–.93** + .16*** ln(GNI pc); R 2 = .113 figure continues next page Liberia 87 Figure 5.3  continued d. Government spending: Health 20 6 Health (% of GDP) 2030 2013 1.6 2000 .50 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = −.25 + .18*** ln(GNI pc); R 2 = .111 e. Tax revenue (left), official development aid (right) 60 50 2012 8 2000 30 Tax revenue (% of GDP) Net ODA (% of GNI) 2030 20 2012 1.2 10 .20 3 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.75*** + .13 ln(GNI pc) ; R 2 = .027 ln(SDG) = 8.10*** + –1*** ln(GNI pc); R 2 = .298 figure continues next page 88 Liberia Figure 5.3  continued f. External debt (left), Government effectiveness (right) 2,000 Government effectiveness: Percentile rank 80 650 2000 30 External debt (% of GNI) 160 10 2030 50 2013 2013 4 10 1.3 .50 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.51*** + .14** ln(GNI pc); R 2 = .036 ln(SDG) = 0 + .44*** ln(GNI pc); R 2 = .300 g. Grigoli health efficiency score 1 Grigoli health efficiency score 2010 .90 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = −.09*** + .01 ln(GNI pc); R 2 = .031 Note: Highlighted observations are for Ethiopia at different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. Liberia 89 Table 5.3  Liberia—Summary Results for SDG Indicators Cross-country relationship Overperforming As expected Underperforming with GNI/cap Tight • Poverty • Ratio of female to male • Primary completion • Gross pre-primary enrollment (−) secondary enrollment (−) • Malaria cases (×) • Gross secondary enrollment (−) • Maternal mortality (×) • Access to improved sanitation (−) • Under-5 mortality (+) • Internet Users (+) • Access to electricity (×) • Access to improved water source • CO2 emissions (×) (+) Loose • HIV prevalence (+) • Ratio of female to male primary completion (−) Note: (+) = larger country rank improvement 2000–12 than for GNI per capita; (−) = smaller country rank improvement (or deterioration) 2000–12 than for GNI per capita; (×) = the same country rank change 2000–12 as for GNI per capita (+/− 2 percentile points). data for a limited number of government effi- 4. Conclusions ciency measures. Liberia’s performance is worse than expected according to the World Bank As summarized in table 5.3, Liberia’s current Government Effectiveness indicator and better ­ outcomes are better than expected (compared to than expected according to the health spending a typical country at the same GNI per capita efficiency index. Efficiency improvements should level) for 5 of the selected SDGs, while it falls be a high priority not only because overall assess- short for 6.19 For the other four SDG indicators ment indicators suggest that Liberia is falling Liberia’s current outcomes are as expected. behind, but also given the fact that, at very low As further shown in table 5.3, for most of the levels of GNI per capita, they are potentially large indicators, the cross-country relationship with also when performance is as expected. GNI per capita is relatively tight. Given this, by In sum, given the combination of currently 2030, considerable improvements are projected; higher than expected ODA, when including UN however, compared to global ambitions, the peace-keeping forces, and an anticipated ODA improvements fall short for most indicators. This decline as GNI per capita increases, the issue of means that, to get closer to the realization of how additional fiscal space may be created these ambitions, a break with the past is needed. becomes central. While adding to tax revenues Accelerated growth would raise the capacity to may be important, foreign borrowing may also be accelerate progress in these SDGs; however, in forthcoming. Opportunities to improve govern- the case of Liberia, the starting point is very low, ment efficiency need to be pursued. Decisions suggesting that additional support from the about the level and allocation of government global community may be necessary. spending should be made in light of government In addition, for indicators such as the ratio of priorities and would depend on numerous factors female to male primary completion and access to that are well beyond the scope of this note, includ- improved sanitation facilities, Liberia is under- ing government capacity in different areas and the performing and/or falling in country ranking, scope to encourage complementary private sector suggesting that the payoffs from targeted policies activities. From the perspective of the SDG agenda that manage to reverse these trends may be sub- and given strong linkages between private and stantial. Targeted policies are in general impor- government activities and incomes, it is crucial tant for indicators with a weak relationship to that policies and spending decisions promote a GNI per capita, since these variables should not broad-based change that encompasses services be expected to improve strongly or systemati- related to human development, infrastructure cally due to indirect effects of more rapid growth investments, and other measures in support of in GNI per capita. strong long-run growth that is biased in favor of The decline in total net ODA is mostly due the less advantaged. to the expected decline in UN peace-keeping 90 Liberia forces but also to the typical decline in net tax increases, fiscal space may be increased ODA  as a country’s income per capita grows. via  improvements in government efficiency Given this and the large needs to reach the (a  top priority) and increased borrowing (as global SDG ambitions, Liberia need to consider long as it does not violate debt sustainability making additional fiscal efforts. In addition to constraints). Annex 5A: Data Sources Indicator Source Comment GNI per capita (constant WDI. API ref: GNI per capita (constant 2005 US$) 2005 US$) [NY.GNP.PCAP.KD] SDG indicators Poverty, at $1.25 a day (PPP) WDI. API ref: poverty headcount ratio at $1.25 a day 2010 value from the World Bank country (% of population) (PPP) (% of population) [SI.POV.DDAY] team Shared prosperity: income share WDI. API ref: income share held by lowest 20% + for lowest 40% income share held by second 20% [SI.DST.FRST.20+SI. DST.02ND.20] Pre-primary enrollment (% gross) WDI. API ref: school enrollment, pre-primary (% gross) [SE.PRE.ENRR] Primary completion (% gross) WDI. API ref: primary completion rate, total (% of relevant age group) [SE.PRM.CMPT.ZS] Secondary enrollment (% gross) WDI. API ref: school enrollment, secondary (% gross) [SE.SEC.ENRR] Secondary completion (% gross) EdStat. API ref: DHS: secondary completion rate [HH.DHS.SCR] Primary completion, ratio of WDI. API ref: primary completion rate, female females to males (%) (% of relevant age group)/primary completion rate, male (% of relevant age group) *100 [SE.PRM.CMPT.FE.ZS/ SE.PRM.CMPT.MA.ZS*100] Secondary enrollment, ratio of WDI. API ref: ratio of female to male secondary females to males (%) enrollment (%) [SE.ENR.SECO.FM.ZS] Under-5 mortality (per 1,000 live WDI. API ref: mortality rate, under-5 (per 1,000 live births) births) [SH.DYN.MORT] Maternal mortality (per 100,000 WDI. API ref: maternal mortality ratio (modeled estimate, live births) per 100,000 live births) [SH.STA.MMRT] Malaria cases (% of population) HNP. API ref: malaria cases reported/population, total *100 [SH.STA.MALR/SP.POP.TOTL *100] HIV prevalence (% of population WDI. API ref: prevalence of HIV, total (% of population ages 15–49) ages 15–49) [SH.DYN.AIDS.ZS] Access to improved sanitation WDI. API ref: improved sanitation facilities facilities (% of population) (% of population with access) [SH.STA.ACSN] Access to improved water source WDI. API ref: improved water source (% of population (% of population) with access) [SH.H2O.SAFE.ZS] Road density (km per 100 sq. km WDI. API ref: road density (km of road per 100 sq. km of of land) land area) [IS.ROD.DNST.K2] Access to electricity WDI. API ref: access to electricity (% of population) (% of population) [EG.ELC.ACCS.ZS] Internet users (per 1,000 people) WDI. API ref: Internet users (per 100 people) [IT.NET.USER.P2] CO2 emissions (metric tons per WDI. API ref: CO2 emissions (metric tons per capita) capita) [EN.ATM.CO2E.PC] annex continues next page Liberia 91 Indicator Source Comment Fiscal space indicators Investment (% of GDP) WDI. API ref: gross fixed capital formation (% of GDP)-gross fixed capital formation, private sector (% of GDP) [NE.GDI.FTOT.ZS]-[NE.GDI.FPRV.ZS] Consumption (% of GDP) WDI. API ref: general government final consumption expenditure (% of GDP) [NE.CON.GOVT.ZS] Primary education (% of GDP) EdStat. API ref: total expenditure on educational institutions and administration as a % of GDP. All sources. Primary [UIS.XGDP.1.FDINSTADM.FFD] Secondary education (% of GDP) EdStat. API ref: total expenditure on educational institutions and administration as a % of GDP. All sources. Secondary and post-secondary non-tertiary [UIS.XGDP.234.FDINSTADM.FFD] Primary education, per student WDI. API ref: expenditure per student, primary (% of GDP per capita) (% of GDP per capita) [SE.XPD.PRIM.PC.ZS] Secondary education, per student WDI. API ref: expenditure per student, secondary (% of GDP per capita) (% of GDP per capita) [SE.XPD.SECO.PC.ZS] Health (% of GDP) WDI. API ref: health expenditure, public (% of GDP) [SH.XPD.PUBL.ZS] Fossil fuel subsidy (% of GDP) IMF (2013a). Total pretax subsidy (% of GDP) Subsidies are measured using a price-gap approach capturing both consumer (including implicit) and producer (except those that arise when suppliers are inefficient and make losses at benchmark prices) subsidies. Negative external effects are not included in the pretax subsidy. Tax revenue (% of GDP) WDI. API ref: tax revenue (% of GDP) [GC.TAX.TOTL.GD.ZS] Net ODA (% of GNI) WDI. API ref: net ODA received (% of GNI) Support to UN peace-keeping forces [DT.ODA.ODAT.GN.ZS] excluded using data from IMF (2012, p. 43). External debt (% of GNI) WDI. API ref: external debt stocks (% of GNI) [DT.DOD.DECT.GN.ZS] Government effectiveness: WGI. API ref: government effectiveness: percentile rank Captures perceptions of the quality of percentile rank [GE.PER.RNK] public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies. Grigoli education efficiency score Grigoli (2014) Measure secondary education inefficiency in terms of how much additional output could be achieved at current levels of spending. Grigoli health efficiency score Grigoli and Kapsoli (2013) Quantifies the inefficiency of public health expenditure using a stochastic frontier model that controls for the socioeconomic determinants of health. Public investment management Dabla-Norris et al. (2011) Four phases associated with public index investment management are covered: project appraisal, selection, implementation, and evaluation. Note: WDI = World Development Indicators, World Bank; API ref = reference and code when using the World Bank Open Data; EdStat = Education statistics, World Bank; HNP = Health Nutrition and Population statistics, World Bank; WGI = Worldwide Governance Indicators, World Bank. 92 Liberia Notes 1. This report does not cover the economic impact of observation applies to expected values for fiscal the 2014 Ebola crisis (see World Bank 2014d). space indicators. 2. While a cross-country perspective provides an 9. With regard to CO2, Liberia’s current and project important complement to analysis that is centered 2030 per capita emissions are 2.0 and 3.8 percent on an individual country, it is by definition lim- of the current OECD average. However, note that ited to analysis of variables that are available in Liberia is underperforming for CO2 emissions cross-country databases. (emitting above expected levels) both when emis- 3. This does not mean that GNI per capita is viewed sions are measured as a share of GDP and per as a direct determinant of SDG outcomes; on the capita. contrary, a major challenge for policy makers is 10. The ranking is based on data from 2000 and to identify policies that improve SDG perfor- 2012/2013, or the closest year with data. The year mance by influencing direct determinants (such for Liberia data is reported in the graphs. of access to relevant services or household per 11. Note that this analysis does not take into account capita consumption for different groups) relative the 2014 Ebola crisis. to what is expected given the level of GNI per 12. There are also cases where the solution to the low capita. A second challenge is to raise growth in level of SDG is neither private nor public spend- GNI per capita as it directly influences country ing but more efficient policies or complementary SDG capacity. policies. 4. Sources for the indicators are presented in 13. The treatment is the same as in table 5.1 and related annex 5A. figures. That is, in table 5.2, projections are shown 5. Projections from CEPII are used for this and only when the cross-country relationship between other Country Development Diagnostics appli- the indicator and GNI per capita is considered cations given their wide country coverage and tight enough. Due to data limitations, we focus on well-­documented methodology; OECD data have government spending indicators; c ­ ountry-specific been used when projections have been missing. analysis is needed to consider policy in the c ­ ontext In the projections, it is assumed that future GNI of the different roles of the government and pri- growth will coincide with future GDP growth (both vate services and spending. expressed in constant 2005 US$) given that that this 14. In many countries, significant savings from is the variable that CEPII and other sources project. reduced fuel subsidies is an obvious potential 6. Given that (a) SDGs have extreme values (such as source of fiscal space given their impact on the 100 percent for improved water access) and (b) environment, income distribution, and technol- the current SDG level never is exactly as expected ogy choice (penalizing labor-intensive technologies). given GNI per capita, the projected values grad- However, Liberia’s government does not subsidize ually converge toward the expected values. For fuels to any notable extent. There are only data on example, for a country that overperforms in water posttax fuel subsidies (including estimates of neg- access, as GNI per capita increases the extent of ative external effects) and those are significantly overperformance gradually declines, so that when lower than expected in the case of Liberia (IMF the expected value is 100, overperformance has 2013a). reached zero. 15. Net ODA is higher than expected both when 7. A tight enough relationship is defined as an ­ measured as a share of GNI and per capita. R2 > 0.3 (tight) or 0.3 > R2 > 0.1 (moderately tight), 16. Data on support to UN military forces are from while R2 < 0.1 are defined as loose. IMF (2012, p. 43). Note that this analysis does not 8. In addition, the confidence interval is wide in take into account any additional aid due to the 2014 the case of a loose relationship, suggesting that Ebola crisis, or potential aid increases designed to any conclusion on over- or underperformance help Liberia realize global SDG ambitions. is made with wide margins. Statistically, even 17. IMF (2012, p. 43) projects tax income to increase though their confidence intervals are wide, as from 23.6 percent of GDP in 2011 to 25.6 in 2015. long as the estimated coefficient linking GNI 18. The (pre-Ebola) IMF-World Bank DSA for per capita to the SDG indicator is nonzero, these Liberia indicates that Liberia continues to have values are closer than the cross-country aver- a low risk of debt distress, following assump- age to what is expected for Liberia. The same tions underpinned by developments in the iron Liberia 93 ore sector and, in the near term, foreign financed while remaining sustainable (IMF 2012, DSA investment. DSA projections suggest that the annex, p. 13). public debt (fully external) of 11.7 percent of 19. With regard to CO2, Liberia’s current and project GDP in 2012 may increase to 38.8 percent (of 2030 per capita emissions are 2.0 and 3.8 percent which 36.1 percent of GDP external) in 2018 of the current OECD average. 94 Liberia Chapter 6 Nigeria 1. Introduction relative to other countries, considering its past, recent, and projected levels of GNI per capita.1 Economic growth in Nigeria, a low-income The latter variable tends to be highly correlated country in Western Africa, has been ­remarkable— with most of the SDGs and most of the factors even in nonoil sectors—mainly driven by agri- that determine their evolution; given this, it is culture, trade, and services. During 2001–12, used as a summary indicator of country capacity Nigeria’s average growth rate for GNI per capita to provide and efficiently utilize inputs that con- (at constant 2005 US$) was 6.4  percent, which tribute to SDGs (for example, health and educa- may be compared to a developing (­ low- and tion services) and to achieve SDG outcomes (like ­ middle-income) country average of 3.0 percent. strong health and education results).2 While aggregate growth has been impressive, Nigeria still faces structural challenges such as income inequality, high unemployment, and 2. SDG Indicators: History and growth disparities among regions. A lack of security in the Northern part of the country Projections makes it more difficult to tackle these Aggregate growth projections covering most challenges. countries are produced by various international This country-at-a-glance note is designed to organizations.3,4 In the case of Nigeria, the pro- provide an initial picture of the challenges that jections to 2030 result in an average annual the Post-2015 agenda poses for Nigeria, serving GNI per capita growth of 2.6 percent. The levels as the starting point for a more complete country of selected SDGs are then projected to 2030, development diagnostic as well as a more in-depth based on the GNI per capita projections. These country-focused analysis. The note is built around business-as-usual projections of the SDGs ­ tables and figures that provide data for a selection reflect what can be expected given a country’s of SDG target indicators and indicators related to initial conditions, projected growth in GNI per fiscal space—fiscal space matters since, while pol- capita, typical rates of progress according to icy frameworks and the engagement of the pri- cross-country patterns, and a gradual conver- ­ vate sector may vary widely, rapid progress on the gence to close gaps between observed and SDG agenda will require efficient and carefully expected values.5 prioritized public spending. Drawing on the For selected SDG indicators, table 6.1 summa- information in these tables and figures, this note rizes data for Nigeria: historical evolution, actual briefly (a) summarizes Nigeria’s SDG progress and expected values for a recent year, and pro- since 2000 and projects expected values for 2030; jected 2030 values.6 In figure 6.1, data for Nigeria and (b) assesses options for increasing fiscal are shown in the context of the estimated space. Sections 2 and 3 address SDGs and fiscal cross-country relationship between each SDG space, respectively, while findings are summa- indicator and GNI per capita. Projections are pre- rized in Section 4. sented only when the cross-country r ­ elationship The analysis is done from a cross-country between the indicator and GNI per capita is clas- perspective: for the different indicators, Nigeria’s sified as tight.7 A loose relationship suggests that performance and prospects are benchmarked progress in the indicator is primarily a reflection Nigeria 95 Table 6.1  Nigeria—SDG Indicators: Evolution since 2000 and Projections to 2030 Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Poverty and shared prosperity Poverty, at $1.25 a day 68.7 62.0 11.6 29.5 By 2030, eradicate extreme poverty for all people (PPP) (% of population) everywhere, currently measured as people living on less than $1.25 a day. Shared prosperity: Income 14.1 15.0 16.6 — By 2030, progressively achieve and sustain income growth share for lowest 40% of the bottom 40 percent of the population at a rate higher than the national average. Education Pre-primary enrollment 13.4 24.2 23.4 By 2030, ensure that all girls and boys have access to (% gross) quality early childhood development, care, and pre-primary education so that they are ready for primary education. Primary completion 76.0 78.4 83.0 By 2030, ensure that all girls and boys complete free, (% gross) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment 24.5 43.8 55.0 56.3 By 2030, ensure that all girls and boys complete free, (% gross) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary completion 65.3 23.0 — By 2030, ensure that all girls and boys complete free, (% gross) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Primary completion, ratio of 89.1 95.0 — By 2030, ensure that all girls and boys complete free, females to males (%) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment, ratio 85.0 88.8 89.3 93.7 By 2030, ensure that all girls and boys complete free, of females to males (%) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Health Under-5 mortality (per 1,000 187.7 117.4 43.6 90.0 By 2030, end preventable deaths of newborns and children live births) under 5 years of age. Maternal mortality (per 950 560 171 379 By 2030, reduce the global maternal ratio to less than 70 100,000 live births) per 100,000 live births. Malaria cases (% of 2.0 0.3 0.2 0.1 By 2030, end the epidemics of AIDS, tuberculosis, malaria, population) and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. HIV prevalence (% of 3.5 3.2 0.8 — By 2030, end the epidemics of AIDS, tuberculosis, malaria, population ages 15–49) and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. Infrastructure Access to improved 32.5 27.8 41.8 39.9 By 2030, achieve access to adequate and equitable sanitation facilities sanitation and hygiene for all and end open defection, (% of population) paying special attention to the needs of women and girls and those in vulnerable situations. Access to improved water 54.8 64.0 77.1 71.5 By 2030, achieve universal and equitable access to safe source (% of population) and affordable drinking water for all. Road density (km per 100 10.9 13.0 16.1 Develop quality, reliable, sustainable, and resilient sq. km of land) infrastructure, including regional and transborder infrastructure, to support economic development and human well-being, with a focus on affordable and equitable access for all. table continues next page 96 Nigeria Table 6.1  continued Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Access to electricity 44.9 48.0 41.1 59.3 By 2030, ensure universal access to affordable, reliable, (% of population) and modern energy services. Internet users (per 1,000 0.1 38.0 11.9 — Significantly increase access to information and people) communications technology and strive to provide universal and affordable access to the Internet in least developed countries by 2020. Environment CO2 emissions (metric tons 0.6 0.5 0.5 1.0 Integrate climate change measures into national policies, per capita) strategies, and planning. Memorandum item GNI per capita (constant 467 986 1,570 2005 US$) Note: Recent refers to the latest year with data, typically 2012 or 2013. If data are not available for 2000 or 2012/2013, the closest year with data is used. If the closest year is more than two years away from the target year (for example, for 1997 or 2003; or 2009), the actual years are reported in the table in the Reference section. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. Green = currently significantly overperforming; red = currently significantly underperforming; black = performing as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). For Internet use, a projection is not made, despite a tight relationship, since the expected line shifts significantly even in the medium run due to external forces, such as technological development. Global ambition is from the Open Working Group (UN 2014). Figure 6.1  Nigeria—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) a. Poverty (left), shared prosperity: income share of bottom 40% (right) Shared prosperity: income share for lowest 40% Poverty, at $1.25 a day (PPP) (% of population) 410 30 90 1996 2010 20 20 2030 2010 1996 4 10 .90 9 .04 5 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 11.8*** + −1.3*** ln(GNI pc); R 2 = .482 ln(SDG) = 3.16*** + −.05** ln(GNI pc); R 2 = .043 figure continues next page Nigeria 97 Figure 6.1  continued b. Gross pre-primary enrollment (left), primary completion (right) 550 130 Pre-primary enrollment (% gross) Primary completion (% gross) 150 90 2030 2010 40 60 2030 10 2010 40 3 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −1.0* + .61*** ln(GNI pc); R 2 = .380 ln(SDG) = 3.35*** + .15*** ln(GNI pc); R 2 = .388 c. Gross secondary enrollment (left), secondary completion (right) 190 90 2008 110 Secondary completion (% gross) Secondary enrollment (% gross) 20 60 2030 6 2010 30 2000 1.4 20 6 .09 150 400 1,000 3,000 8,000 150 400 1,000 3,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.93*** + .3*** ln(GNI pc); R 2 = .556 ln(SDG) = −.1 + .48* ln(GNI pc); R 2 = .096 figure continues next page 98 Nigeria Figure 6.1  continued d. Ratio of female to male primary completion (left), ratio of female to male secondary enrollment (right) 140 140 Secondary enrollment, ratio of females Primary completion, ratio of females 110 100 90 2030 2010 2030 to males (%) to males (%) 2000 2010 70 70 50 40 30 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 4.28*** + .04*** ln(GNI pc); R 2 = .11 ln(SDG) = 3.78*** + .1*** ln(GNI pc); R 2 = .319 e. Under-5 mortality (left), maternal mortality (right) 500 2,000 Maternal mortality (per 100,000 live births) Under−5 mortality (per 1,000 live births) 2000 230 2000 2013 470 2030 2013 110 2030 100 50 20 20 5 1 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 8.01*** + −.61*** ln(GNI pc); R 2 = .570 ln(SDG) = 11.3*** + −.9*** ln(GNI pc); R 2 = .531 figure continues next page Nigeria 99 Figure 6.1  continued f. Malaria cases (left), HIV prevalence (right) HIV Prevalence (% of population ages 15–49) 90 30 Malaria cases (% of population) 30 1.3 2000 2010 9 2030 .05 2000 3 2013 .90 .10 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 14.5*** + −2.3*** ln(GNI pc); R 2 = .415 ln(SDG) = 1.96* + −.32** ln(GNI pc); R 2 = .048 g. Access to improved sanitation (left), access to improved water source (right) Access to improved sanitation facilities 100 Access to improved water source 100 70 2030 60 (% of population) (% of population) 2012 2000 2030 50 30 2000 2012 40 20 7 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .61** + .45*** ln(GNI pc); R 2 = .533 ln(SDG) = 3.45*** + .13*** ln(GNI pc); R 2 = .432 figure continues next page 100 Nigeria Figure 6.1  continued h. Access to electricity (left), road density (right) 330 Road density (km per 100 sq. km of land) Access to improved sanitation facilities 100 90 60 (% of population) 2030 20 30 2030 2000 2012 2006 20 7 7 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .61** + .45*** ln(GNI pc); R = .535 2 ln(SDG) = –.73 + .49*** ln(GNI pc); R 2 = .198 i. Internet users (left), CO2 emissions (right) 20 80 CO2 emissions (metric tons per capita) 2013 Internet users (per 1,000 people) 3 2030 8 .80 2000 2010 .20 .40 2000 0 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −3.0*** + .8*** ln(GNI pc); R 2 = .697 ln(SDG) = −8.6*** + 1.17*** ln(GNI pc); R 2 = .680 Note: Highlighted observations are for Nigeria at different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. Nigeria 101 of country-specific factors and that it should not CO2 ­ emissions), Nigeria’s current outcomes are be expected to respond strongly or systematically as expected. While underperformance for an to changes in GNI per capita. Note also that the indicator may be due to country-specific condi- “expected” values in the cases of a loose relation- tions that are difficult to change, it may often ship to GNI per capita should be taken with point to areas in which payoffs from feasible pol- ­ caution and interpreted only as best guesses. icy change are relatively high; a possibility that In sum, Nigeria’s current outcomes are better calls for further analysis. than expected (compared to a typical country at As shown in figure 6.2, Nigeria’s GNI per the same GNI per capita level) for 3 of the indica- capita ranking among low- and middle-income tors (secondary completion, access to electricity, countries improved by as much as 11 percentile and Internet use), while it falls short for 10 points between 2000 and 2012. Compared to ­ (poverty, shared prosperity, gross pre-primary GNI per capita, the progress in ranking among enrollment, gross secondary enrollment, ratio of SDG indicators was stronger only for Internet female to male primary completion, under-5 use. For 2 of the indicators the improvement mortality, maternal mortality, HIV prevalence, was similar to GNI per capita (shared prosper- access to improved water source, and access to ity and malaria). The remaining SDGs either improved sanitation).8 For the other 5 indicators improved their ranking by less than GNI per (primary completion, ratio of female to male sec- capita (as for 2 indicators; under-5 mortality ondary enrollment, malaria, road density, and and CO2 emissions), stayed at the same level of Figure 6.2  Nigeria—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012 GNI per capita (constant 2005 US$) 24 35 Poverty, at $1.25 a day (PPP) (% of population) 12 7 Shared prosperity: Income share for lowest 40% 33 42 Pre-primary enrollment (% gross) 35 Primary completion (% gross) 56 Secondary enrollment (% gross) 21 16 Secondary completion (% gross) 82 Primary completion, ratio of females to males (%) 29 Secondary enrollment, ratio of females to males (%) 35 24 Under-5 mortality (per 1,000 live births) 4 6 Maternal mortality (per 100,000 live births) 10 10 Malaria cases (% of population) 30 39 HIV prevalence (% of population ages 15–49) 20 16 Access to improved sanitation facilities (% of population) 28 19 Access to improved water source (% of population) 18 15 Road density (km per 100 sq. km of land) 26 Access to electricity (% of population) 35 33 Internet users (per 1,000 people) 12 67 CO2 emissions (metric tons per capita) 61 69 0 10 20 30 40 50 60 70 80 90 100 2000 Recent Note: A high ranking signals strong performance; for the underlying indicator, this may correspond to a relatively high value (for example, for the secondary enrollment rate) or a relatively low value (for example, for the poverty rate). The rankings are based on all low- and middle- income countries (according to the 2013 classification) with data. The country samples vary across indicators but are always the same for 2000 and recent for any given indicator. The ranking for an indicator is not reported if the available sample is less than 20 countries. 102 Nigeria ranking (maternal mortality), or even deterio- 3. Fiscal Space rated in ranking for 6 indicators (poverty, ratio of female to male secondary enrollment, gross In most countries, accelerated progress on the secondary enrollment, HIV prevalence, access SDG agenda will require efficient and growing to improved water source, access to improved public spending in prioritized areas, most impor- sanitation, and access to electricity). tantly human development and infrastructure. When comparing the results from regres- Private spending is also of crucial importance, sions on GNI per capita to changes in percentile both household spending on SDG-related ser- rankings, a few insights emerge. For example, vices and business investments in a wide range of while virtually all countries made major gains areas (including but not limited to infrastruc- in terms of Internet use since 2000, Nigeria out- ture).9 With regard to Nigeria’s fiscal space indica- performed other countries and is now perform- tors, table 6.2 and figure 6.3 summarize the ing better than expected for its GNI per capita. historical evolution, actual and expected recent However, for all other SDGs, Nigeria’s ranking values, and, when relevant, projected values.10 The is not keeping pace with GNI per capita variables cover three aspects of government activ- (improving by less than GNI per capita or even ities: spending, receipts and debt, and governance falling); among these, the current outcomes are and efficiency. While the content in this country- below expectations for the majority. The perfor- at-a-glance note is too limited for policy conclu- mance seems most alarming for poverty, ratio sions, the indicators have been selected to inform of female to male secondary enrollment, and thinking about the need and the ability of the gov- access to sanitation facilities. The reasons ernment to adjust spending in priority areas, tak- behind these results may be driven by a time lag ing efficiency into consideration. Again, the exact in income growth translating into SDG effects, level of the expected value should be interpreted or limited government success in e ­ nsuring that with caution when the relationship between the the broad masses share equitably in the benefits fiscal space indicator and GNI per capita is loose. from oil (which in 2012 accounted for 73 percent Room for additional priority spending may be of ­ government revenue and 36 percent of GDP) created by reducing low-priority spending, increas- (IMF 2014b, p. 24). ing current receipts, and/or increasing borrowing. By 2030, considerable improvements are In terms of the selected government spending in projected for most indicators (see table 6.1 and areas that may support the SDG agenda, Nigeria respective graphs in figure 6.1). However, com- performs as expected (compared to a typical coun- pared to global ambitions, also shown in try at the same GNI per capita level) for total public table  6.1, the improvements are mostly investments and fuel subsidies, but it is below the ­ moderate. Possible exceptions are malaria expected level for total consumption and public cases, the female to male rates for both primary health. Spending on fuel subsidies (2.7 percent of completion and secondary enrollment: for GDP in Nigeria) is the most obvious case of these Nigeria may get close to achieving global ­ low-priority spending from the post-2015 agenda ambitions. Nevertheless, for most indicators, perspective.11 While there is no specific data for this means that the realization of such ambi- educational spending on primary and secondary tions would require a break with the past. This level in Nigeria, other country data suggest there is is also true for indicators, such as shared pros- a general underspending at least in primary educa- perity, for which a weak relationship with GNI tion.12 In the absence of any overspending in the per capita precludes projections. Such a break totals of public consumption and investment, any would be facilitated by more rapid and more additional gains in fiscal space would likely have to inclusive growth combined with SDG policies come from higher revenues and/or improved that benefit the disadvantaged. spending efficiency (including ­ reallocation of spending to ­ high-priority areas). Nigeria 103 Table 6.2  Nigeria—Fiscal Space: Revenue, Spending, and Government Efficiency Actual Expected Projection Indicator 2000 Recent Recent 2030 Government spending Consumption (% of GDP) 8.3 8.1 13.7 — Investments (% of GDP) 6.2 6.4 — Health (% of GDP) 1.5 1.9 2.9 — Fuel subsidy (% of GDP) 2.7 1.7 — Government receipts and debts Tax revenue (% of GDP) 20.0 13.7 — Net ODA (% of GNI) 0.4 0.4 3.3 0.3 External debt (% of GNI) 78.5 2.8 33.3 — Governance and government efficiency Government effectivness: Percentile rank 14.6 16.3 21.6 21.2 Public investment management index 1.1 1.5 — Memorandum item GNI per capita (constant 2005 US$) 467 986 1,570 Note: Recent refers to the latest year with data, typically 2012 or 2013. If data is not available for 2000 or 2012/2013, the closest year with data is used. If the closest year is more than two years away from the target year (for example, for 1997 or 2003; or 2009), the actual years are reported in the table in the Reference section. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. Green = current value significantly above the expected level; red = current value significantly below the expected level; black = current value as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). Figure 6.3  Nigeria—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting a. Government spending: consumption (left), investment (right) 130 50 60 Consumption (% of GDP) Investments (% of GDP) 20 30 9 2012 10 2000 4 2013 2 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.22*** + .06 ln(GNI pc); R 2 = .021 ln(SDG) = 2.34*** + −.07 ln(GNI pc); R 2 = .012 figure continues next page 104 Nigeria Figure 6.3  continued b. Government spending: health (left), fuel subsidy (right) 20 20 2011 1.5 Fuel subsidy (% of GDP) 8 Health (% of GDP) .10 2012 1.6 2000 0 .50 150 400 1,000 3,000 8,000 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = -.04 + .16*** ln(GNI pc); R = .090 2 ln(SDG) = 4.73*** + -.62** ln(GNI pc); R 2 = .090 c. Tax revenue (left), official development aid (right) 60 40 30 7 Tax revenue (% of GDP) Net ODA (% of GNI) 20 2013 1.2 10 2000 2012 .20 2030 3 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.82*** + .12 ln(GNI pc); R 2 = .024 ln(SDG) = 8.26*** + −1.0*** ln(GNI pc); R 2 = .308 figure continues next page Nigeria 105 Figure 6.3  continued d. External debt (left), government effectiveness (right) 2,000 Government effectiveness: percentile rank 80 550 30 External debt (% of GNI) 2030 160 2000 2012 2000 10 50 4 10 2012 1.3 .50 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.82*** + .13** ln(GNI pc); R 2 = .032 ln(SDG) = .15+ .43*** ln(GNI pc); R 2 = .288 e. Public investment management index 4 Public investment management index 2 1.3 2010 .80 .30 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = −.41 + .12** ln(GNI pc); R 2 = .074 Note: Highlighted observations are for Nigeria at different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. 106 Nigeria Of the government receipts included in long-term decline in oil revenue will necessitate table  6.2, net Official Development Assistance a ­ dditional fiscal effort. (ODA) is lower than expected while tax reve- Government efficiency is important to pro- nues are higher than expected. Note, however, tect and, if possible, increase in order to add to that total tax revenues were 20 percent of GDP the room for priority spending and enhance its in 2013, but only 4 percent of these were nonoil impact on the SDG agenda. Table 6.2 displays tax revenue.13 As further shown, c ­ ross-country data for a limited number of government effi- patterns suggest that, as GNI per capita grows, ciency measures. Nigeria’s performance is net ODA declines as percent of GDP (without weaker than expected according to both the changing significantly in per capita terms). If World Bank Government Effectiveness indica- Nigeria were to gradually converge to expected tor and the Public Investment Management levels given its GNI per capita, then ODA Index. Hence, even though they are unpredict- would be expected to decrease from the current able, efficiency gains could potentially add 0.4 percent of GNI to 0.3 percent in 2030. It is ­ considerable fiscal space. difficult to project future ODA for Nigeria In sum, the fact that the decline in oil reve- given government access to high but uncertain nues is projected to be larger than the increase oil revenues, and the fact that the country at in nonoil revenues, the issue of how additional this point receives much less ODA than fiscal space may be created becomes central. expected given its GNI per capita. Moreover, Decreased spending on fuel subsidies is one the fact that cross-country patterns point to a potential source and—depending on future likely decline in ODA does not mean that an decisions of the government, donors, and the increase is excluded: It depends on the priori- actors of the international financial system— ties of donors and their relationships with increases in ODA and/or foreign borrowing Nigeria’s government. may also be forthcoming. Opportunities to The relationship between tax revenues and improve government efficiency should also be GNI per capita, as well as the debt stock and pursued. Decisions about the level and alloca- GNI per capita, is not tight enough for project- tion of government spending should be made in ing expected changes. However, IMF projects a light of government priorities and would depend decline in total revenues, with the decline in oil on numerous factors that are well beyond the revenues being larger than the increase in ­nonoil scope of this note, including government capac- revenues.14,15 Higher taxes would reduce the ity in different areas and the scope to encourage resources controlled by domestic households complementary private sector activities. For and firms, pointing to the need to consider the example, remittances from the country’s work- combined impact on SDGs and other indicators ers abroad was a ­ non-negligible 4.0 percent of from higher taxes and the spending increases GDP in 2013 (above what is expected for that are financed by these taxes. The cross-​ Nigeria’s GNI per capita level), and depending ­country data suggest that Nigeria’s external debt on how these flows are channeled, the contribu- stock is low and the latest Debt Sustainability tion to the SDG agenda will differ. From the Analysis concludes Nigeria remains as low risks perspective of the SDG agenda and given strong given current macroeconomic assumptions.16 linkages between private and ­ government activ- In sum, some fiscal space may be added through ities and incomes, it is crucial that policies and a decrease in fuel subsidies, a marginal increase spending decisions promote a broad-based in aid, and, more tentatively, increased debt. change that encompasses services related to However, although oil revenues are expected to human development, infrastructure invest- increase in the short to medium run (not least ments, and other measures in support of strong by addressing oil theft and production losses as long-run growth that is biased in favor of the well as o­ il-revenue management), the projected less advantaged. Nigeria 107 Table 6.3  Nigeria—Summary Results for SDG Indicators Cross-country relationship with Overperforming As expected Underperforming GNI/cap Tight • Internet users (+) • Primary completion • Poverty (–) • Access to electricity (-) • Ratio of female to male • Gross pre-primary enrollment secondary enrollment (–) • Gross secondary enrollment (–) • Malaria (x) • Under-5 mortality (–) • Road density • Maternal mortality (–) • CO2 emissions (–) • Access to improved water source (–) • Access to improved sanitation (–) Loose • Secondary completion • Shared prosperity (x) • Ratio of female to male primary completion • HIV prevalence (–) Note: (+) = larger country rank improvement 2000–12/13 than for GNI per capita; (−) = smaller country rank improvement (or deterioration) 2000–12/13 than for GNI per capita; (×) = the same country rank change 2000–12/13 as for GNI per capita (+/− 2 percentile points). 4. Conclusions potentially strong payoffs. If such policies are put in place, Nigeria may during the next As summarized in table 6.3, Nigeria’s current out- decades improve its percentile rankings com- comes are better than expected (compared to a pared to other countries with similar growth typical country at the same GNI per capita level) rates for GNI per capita. Policies would also be for 2 indicators (secondary completion and important for two indicators with a weak rela- Internet use) and as expected for another 6. tionship to GNI per capita: shared prosperity However, outcomes fall short of expectations for and HIV prevalence as they should not be 10 of the covered indicators. In terms of percentile expected to improve strongly or systematically rankings among all low- and ­ middle-income in response to more rapid growth in GNI per countries with data, the outcomes have deterio- ­ capita. However, more rapid growth would rated for most indicators and improved only for promise to contribute to more rapid growth in one (Internet use). the level of per capita incomes for the bottom As shown in table 6.3, for most of the indica- 40 percent. tors, the cross-country relationship with GNI The projected long-term decline in oil reve- per capita is relatively tight. Given this, by 2030, nue will necessitate additional fiscal effort. In considerable improvements are projected; how- addition, through increased nonoil taxes, ­ fiscal ever, compared to global ambitions, the improve- space may be increased via a combination of ments fall short for most indicators. This means decreased fuel subsidies, increased aid, govern- that, to get closer to the realization of these ment efficiency, and increased borrowing (as ambitions, a break with the past is needed. long as it does not violate debt sustainability con- Accelerated growth would raise the capacity to straints). The net impact on SDG progress from accelerate SDG progress. higher taxes would depend on marginal govern- However, for Nigeria, the fact that the coun- ment efficiencies both in spending and in try is underperforming for most indicators taxation. Finally, even though they are unpre- ­ suggests a crucial role for policies that, directly dictable, efficiency gains could potentially add or indirectly, promote the SDG agenda with considerable fiscal space. 108 Nigeria Annex 6A: Data Sources Indicator Source Comment GNI per capita (constant 2005 US$) WDI. API ref: GNI per capita (constant 2005 US$) [NY.GNP.PCAP.KD] SDG indicators Poverty, at $1.25 a day (PPP) WDI. API ref: poverty headcount ratio at $1.25 a (% of population) day (PPP) (% of population) [si.pov.dday] Shared prosperity: income share for WDI. API ref: income share held by lowest 20% lowest 40% + income share held by second 20% [SI.DST.FRST.20+SI.DST.02ND.20] Pre-primary enrollment (% gross) WDI. API ref: school enrollment, pre-primary (% gross) [SE.PRE.ENRR] Primary completion (% gross) WDI. API ref: primary completion rate, total (% of relevant age group) [se.prm.cmpt.zs] Secondary enrollment (% gross) WDI. API ref: school enrollment, secondary (% gross) [se.sec.enrr] Secondary completion (% gross) EdStat. API ref: DHS: secondary completion rate [hh.dhs.scr] Primary completion, ratio of females WDI. API ref: primary completion rate, female to males (%) (% of relevant age group)/primary completion rate, male (% of relevant age group) *100 [SE. PRM.CMPT.FE.ZS /SE.PRM.CMPT.MA.ZS*100] Secondary enrollment, ratio of WDI. API ref: ratio of female to male secondary females to males (%) enrollment (%) [se.enr.seco.fm.zs] Under-5 mortality (per 1,000 live WDI. API ref: mortality rate, under-5 (per 1,000 births) live births) [SH.DYN.MORT] Maternal mortality (per 100,000 live WDI. API ref: maternal mortality ratio (modeled births) estimate, per 100,000 live births) [sh.sta.mmrt] Malaria cases (% of population) HNP. API ref: malaria cases reported/ Population, total *100 [sh.sta.malr/ SP.POP. TOTL *100] HIV prevalence (% of population WDI. API ref: prevalence of HIV, total (% of ages 15–49) population ages 15–49) [sh.dyn.aids.zs] Access to improved sanitation WDI. API ref: improved sanitation facilities facilities (% of population) (% of population with access) [sh.sta.acsn] Access to improved water source WDI. API ref: improved water source (% of (% of population) population with access) [sh.h2o.safe.zs] Road density (km per 100 sq. km of WDI. API ref: road density (km of road per 100 land) sq. km of land area) [is.rod.dnst.k2] Access to electricity (% of WDI. API ref: access to electricity population) (% of population) [eg.elc.accs.zs] Internet users (per 1,000 people) WDI. API ref: Internet users (per 100 people) [IT.NET.USER.P2] CO2 emissions (metric tons per WDI. API ref: CO2 emissions (metric tons per capita) capita) [en.atm.co2e.pc] Fiscal space indicators Investment (% of GDP) WDI. API ref: gross fixed capital formation (% of GDP)-gross fixed capital formation, private sector (% of GDP) [ne.gdi.ftot. zs]-[ne.gdi.fprv.zs] Consumption (% of GDP) WDI. API ref: general government final consumption expenditure (% of GDP) [NE.CON.GOVT.ZS] annex continues next page Nigeria 109 Indicator Source Comment Primary education (% of GDP) EdStat. API ref: total expenditure on educational institutions and administration as a % of GDP. All sources. Primary [UIS.XGDP.1.FDINSTADM. FFD] Secondary education (% of GDP) EdStat. API ref: total expenditure on educational institutions and administration as a % of GDP. All sources. Secondary and post-secondary non-tertiary [UIS.XGDP.234.FDINSTADM.FFD] Primary education, per student (% WDI. API ref: expenditure per student, primary of GDP per capita) (% of GDP per capita) [se.xpd.prim.pc.zs] Secondary education, per student WDI. API ref: expenditure per student, (% of GDP per capita) secondary (% of GDP per capita) [se.xpd. seco.pc.zs] Health (% of GDP) WDI. API ref: health expenditure, public (% of GDP) [sh.xpd.publ.zs] Fossil fuel subsidy (% of GDP) IMF (2013a). Total pretax subsidy (% of GDP) Subsidies are measured using a price-gap approach capturing both consumer (including implicit) and producer (except those that arise when suppliers are inefficient and make losses at benchmark prices) subsidies. Negative external effects are not included in the pretax subsidy. Tax revenue (% of GDP) WDI. API ref: tax revenue (% of GDP) [gc.tax.totl.gd.zs] Net ODA (% of GNI) WDI. API ref: net ODA received (% of GNI) [dt.oda.odat.gn.zs] External debt (% of GNI) WDI. API ref: external debt stocks (% of GNI) [DT.DOD.DECT.GN.ZS] Government effectiveness: WGI. API ref: government effectiveness: Captures perceptions of the quality of public percentile rank percentile rank [ge.per.rnk] services, the quality of the civil service, and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies. Grigoli education efficiency score Grigoli (2014) Measure secondary education inefficiency in terms of how much additional output could be achieved at current levels of spending. Grigoli health efficiency score Grigoli and Kapsoli (2013) Quantifies the inefficiency of public health expenditure using a stochastic frontier model that controls for the socioeconomic determinants of health. Public investment management Dabla-Norris et al. (2011) Four phases associated with public investment index management are covered: project appraisal, selection, implementation, and evaluation. Note: WDI = World Development Indicators, World Bank; API ref = reference and code when using the World Bank Open Data; EdStats = Education statistics, World Bank; HNP = Health Nutrition and Population statistics, World Bank; WGI = Worldwide Governance Indicators, World Bank. Notes 1. While a cross-country perspective provides an identify policies that improve SDG performance important complement to analysis that is centered relative to what is expected given the level of GNI on an individual country, it is by definition lim- per capita. A second challenge is to raise growth in ited to analysis of variables that are available in GNI per capita as it indirectly influences country cross-country databases. SDG capacity. 2. This does not mean that GNI per capita is viewed 3. Projections from CEPII (or OECD when data as a direct determinant of SDG outcomes; on the is missing) have been used for the Country contrary, a major challenge for policy makers is to Development Diagnostics applications, but to 110 Nigeria what extent they seem realistic should be set 11. Fuel subsidies are detrimental to the climate and against other projections, historic growth, and encourage technologies that are less labor inten- recent developments. sive, tending to generate employment for fewer 4. Given the fact that available sources only project workers at lower wages. GDP whereas this book uses GNI data, we have to 12. Preliminary data from the World Bank ­ country assume, for most countries quite reasonably, that team suggest that spending on basic education projected GNI growth will not deviate substan- (from preschool to lower secondary school) tially from projected GDP growth (both expressed amounts to 0.9 percent of GDP or, when measured in constant 2005 US$). per student, to 5.0 percent of GDP per capita. This 5. Given that (a) SDGs have extreme values (such as can be compared to an expected level of spending 100 percent for improved water access) and (b) only for primary education of about 1.7 percent of the current SDG level never is exactly as expected GDP or 13 percent of GDP per capita when mea- relative to GNI per capita, it is necessary to incor- sured per student. porate convergence toward the expected value 13. Data from http://www.premiumtimesng.com​/­bus into the projections. It is here assumed that such iness/158283-nigerias-tax-revenue-ratio-gdp-dro convergence is gradual. For example, for a coun- ps-okonjo-iweala-says.html. try that overperforms in water access, as GNI per 14. In 2012, nonoil revenues and grants in 2012 capita increases the extent of overperformance amounted to 6.8 percent of GDP and o ­ il ­revenues gradually declines, so that when the expected to 18.5 percent. IMF (2014b, p. 24) projects that, value is 100, overperformance has reached zero. by 2018, nonoil revenues and grants will be 8.3 6. Sources for the indicators are presented in the and oil revenues 10.0 percent. An increase in table in the annex 6A. nonoil tax revenues is expected mainly through 7. A tight enough relationship is defined as an improved tax base and administration rather R2 > 0.3 (tight) or 0.3 > R2 > 0.1 (moderately tight), than the tax rate (IMF 2014b, p. 14). These pro- while R2 > 0.1 are defined as loose. jections were made before the dramatic inter- 8. With regard to CO2, Nigeria’s current and project national oil price decline in the fall of 2014 and 2030 per capita emissions are 4.9 and 12.0 percent early 2015. of the current OECD average. 15. Note, however, that these projections of oil 9. There are also cases where the solution to the low revenues are based on the oil prices, which are ­ level of SDG is neither private nor public spend- much higher than today’s prices and probably ing but more efficient policies or complementary much higher than the long–term equilibrium policies. price. 10. The treatment is the same as in table 6.1 and related 16. According to IMF-World Bank Debt Sustainability figures. That is, in table 6.2, projections are shown Analysis (DSA) for Nigeria (IMF 2014b, annex), only when the cross-country relationship between Nigeria remains at a low risk of debt distress under the indicator and GNI per capita is considered current macroeconomic assumptions. However, tight enough. Due to data limitations, we focus on without significant or timely compensating pol- government spending indicators; c ­ ountry-specific icy measures, a prolonged negative oil price shock analysis is needed to consider policy in the context or a fiscal revenue shortfall could undermine the of the different roles of the government and pri- recent progress in achieving macroeconomic and vate services and spending. debt sustainability. Nigeria 111 Chapter 7 Pakistan 1. Introduction spending. Drawing on the information in these tables and figures, this note briefly (a) summa- Pakistan is a  lower middle-income country in rizes Pakistan’s SDG progress since 2000 and South Asia. After gaining independence from projects expected values for 2030; and (b) assesses Britain in 1947, Pakistan has lived through peri- options for increasing fiscal space. Sections 2 and ods of military rule, political instability, and con- 3 address SDGs and fiscal space, respectively, flicts with neighboring India, as well as ongoing while findings are summarized in Section 4. conflict in its border areas with Afghanistan. The analysis is done from a cross-country Even though Pakistan has realized substantial perspective: for the different indicators, Pakistan’s social and economic progress in recent decades, performance and prospects are benchmarked compared to most other developing countries, relative to other countries, considering its past, economic growth has been relatively unstable, recent, and projected levels of GNI per capita.1 making progress on SDG indicators more The latter variable tends to be highly correlated difficult. During 2001–12, Pakistan’s average ­ with most of the SDGs and most of the factors growth rate for GNI per capita (at constant 2005 that determine their evolution; given this, it is US$) was 2.7 percent, which may be compared to used as a summary indicator of country capacity a developing (­ low- and middle-­ income) country to provide and efficiently utilize inputs that con- average of 3.0 percent. During the same period, tribute to SDGs (for example, health and educa- Pakistan’s ranking according to  the UNDP tion services) and to achieve SDG outcomes (like Human Development Index (among countries strong health and education results).2 included both in 2000 and 2012) remained in the 22nd percentile. However, recently, Pakistan has turned a situation of declining private invest- 2. SDG Indicators: History and ment, a weak external position, and a widening fiscal deficit into one of increased GDP growth Projections (4.1 percent in 2013/14), declining fiscal deficit, For selected SDG indicators, table 7.1 summa- decelerating inflation, and improved interna- rizes data for Pakistan: historical evolution, tional reserves (IMF 2015d, p. 18). actual and expected values for a recent year, and This country-at-a-glance note is designed to projected 2030 values.3 In figure 7.1, data for provide an initial picture of the challenges that Pakistan are shown in the context of the esti- the Post-2015 agenda poses for Pakistan, serving mated cross-country relationship between each as the starting point for a more complete country SDG indicator and GNI per capita. For Pakistan, development diagnostic as well as a more com- the projected average annual rate of GNI per prehensive country-focused analysis. The note is capita growth is 2.5 percent.4 The projected SDG built around tables and figures that provide data values reflect what can be expected given a for a selection of SDG target indicators and country’s starting point, projected growth in ­ ­ indicators related to fiscal space—fiscal space GNI per capita, typical rates of progress accord- matters since, while policy frameworks and the ing to cross-country patterns, and a gradual con- engagement of the private sector may vary vergence to close gaps between observed and widely,  rapid  progress on the SDG agenda will expected values.5 Projections for SDG indicators require efficient and carefully prioritized public are presented only when the cross-country Pakistan 113 Table 7.1  Pakistan—SDG Indicators: Evolution since 2000 and Projections to 2030 Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Poverty and shared prosperity Poverty, at $1.25 a day (PPP) 29.0 12.7 14.9 6.8 By 2030, eradicate extreme poverty for all people (% of population) everywhere, currently measured as people living on less than $1.25 a day. Shared prosperity: Income 21.1 22.7 16.7 — By 2030, progressively achieve and sustain income share for lowest 40% growth of the bottom 40 percent of the population at a rate higher than the national average. Education Pre-primary enrollment 62.8 82.1 22.1 83.6 By 2030, ensure that all girls and boys have access to (% gross) quality early childhood development, care, and pre-primary education so that they are ready for primary education. Primary completion 71.9 76.6 77.8 By 2030, ensure that all girls and boys complete free, (% gross) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment 36.6 52.3 46.6 By 2030, ensure that all girls and boys complete free, (% gross) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Primary completion, ratio of 85.7 94.5 — By 2030, ensure that all girls and boys complete free, females to males (%) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment, ratio 73.5 87.9 79.9 By 2030, ensure that all girls and boys complete free, of females to males (%) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Health Under-5 mortality 113 86 49 66 By 2030, end preventable deaths of newborns and (per 1,000 live births) children under 5 years of age. Maternal mortality 280 170 205 116 By 2030, reduce the global maternal mortality ratio to less (per 100,000 live births) than 70 per 100,000 live births. Malaria cases 0.06 0.16 0.30 0.06 By 2030, end the epidemics of AIDS, tuberculosis, malaria, (% of population) and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. HIV prevalence 0.10 0.10 0.84 — By 2030, end the epidemics of AIDS, tuberculosis, malaria, (% of population ages and neglected tropical diseases and combat hepatitis, 15–49) water-borne diseases, and other communicable diseases. Infrastructure Access to improved 37.4 47.6 38.1 54.7 By 2030, achieve access to adequate and equitable sanitation facilities sanitation and hygiene for all and end open defecation, (% of population) paying special attention to the needs of women and girls and those in vulnerable situations. Access to improved water 88.3. 91.4 75.1 92.9 By 2030, achieve universal and equitable access to safe source (% of population) and affordable drinking water for all. Road density (km per 31.2 33.0 12.7 35.9 Develop quality, reliable, sustainable, and resilient 100 sq. km of land) infrastructure, including regional and transborder infrastructure, to support economic development and human well-being, with a focus on affordable and equitable access for all. Access to electricity 79.5 91.4 37.2 92.7 By 2030, ensure universal access to affordable, reliable, (% of population) and modern energy service. Internet users 0.1 10.9 10.2 — Significantly increase access to information and (per 1,000 people) communications technology and strive to provide universal and affordable access to the Internet in least developed countries by 2020. table continues next page 114 Pakistan Table 7.1  continued Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Environment CO2 emissions 0.74 0.93 0.42 1.23 Integrate climate change measures into national polices, (metric tons per capita) strategies, and planning. Memorandum item GNI pre capita 589 807 1,267 (constant 2005 US$) Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = currently significantly overperforming; red = currently significantly underperforming; black = performing as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). For Internet use, a projection is not made, despite a tight relationship, since the expected line shifts significantly even in the medium run due to external forces, such as technological development. Global ambition is from the Open Working Group (UN 2014). Figure 7.1  Pakistan—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) a. Poverty (left), shared prosperity: income share of bottom 40% (right) 410 30 2011 Shared prosperity: income share 90 1999 Poverty, at $1.25 a day (PPP) 20 (% of population) 20 for lowest 40% 1999 4 2011 2030 10 .90 9 .04 5 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 11.8*** + −1.3*** ln(GNI pc); R 2 = .487 ln(SDG) = 3.17*** + −.05** ln(GNI pc); R 2 = .046 figure continues next page Pakistan 115 Figure 7.1  continued b. Gross pre-primary enrollment (left), primary completion (right) 550 130 Pre-primary enrollment (% gross) Primary completion (% gross) 150 90 2012 2030 2000 2012 2030 40 60 10 40 3 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.73 + .570*** ln(GNI pc); R 2 = .348 ln(SDG) = 3.38*** + .14*** ln(GNI pc); R 2 = .372 c. Gross secondary enrollment 190 Secondary enrollment (% gross) 110 60 2030 30 2012 20 6 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = 1.9*** + .31*** ln(GNI pc); R 2 = .558 figure continues next page 116 Pakistan Figure 7.1  continued d. Ratio of female to male primary completion (left), ratio of female to male secondary enrollment (right) 140 140 Secondary enrollment, ratio of Primary completion, ratio of 110 100 females to males (%) females to males (%) 90 2030 2012 70 2012 70 50 40 30 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 4.31*** + .04*** ln(GNI pc); R = .093 2 ln(SDG) = 3.82*** + .1*** ln(GNI pc); R 2 = .297 e. Under-5 mortality (left), maternal mortality (right) 500 2,000 Maternal mortality (per 100,000 live births) Under–5 mortality (per 1,000 live births) 230 470 2000 2000 110 2013 100 2013 2030 2030 50 20 20 5 1 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 7.99*** + −.61*** ln(GNI pc); R 2 = .560 ln(SDG) = 11.3*** + −.9*** ln(GNI pc); R 2 = .519 figure continues next page Pakistan 117 Figure 7.1  continued f. Malaria cases (left), HIV prevalence (right) HIV prevalence (% of population ages 15–49) 90 30 Malaria cases (% of population) 30 1.3 9 2012 .05 2000 2030 3 .90 2000 2013 .10 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 14.5*** + −2.3*** ln(GNI pc); R 2 = .412 ln(SDG) = 1.88* + −.31** ln(GNI pc); R 2 = .044 g. Access to improved sanitation (left), access to improved water source (right) Access to improved sanitation facilities 100 2012 2030 Access to improved water source 2000 100 2030 70 60 (% of population) (% of population) 2012 2000 50 30 40 20 7 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .61** + .45*** ln(GNI pc); R = .535 2 ln(SDG) = 3.45*** + .13*** ln(GNI pc); R 2 = .436 figure continues next page 118 Pakistan Figure 7.1  continued h. Access to electricity (left), road density (right) 400 330 Road density (km per 100 sq. km of land) Access to electricity (% of population) 2010 2030 90 100 2000 2011 2000 2030 30 20 7 6 1.6 .50 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .29 + .5*** ln(GNI pc); R 2 = .458 ln(SDG) = −.63 + .48*** ln(GNI pc); R 2 = .189 i. Internet users (left), CO2 emissions (right) 20 80 CO2 emissions (metric tons per capita) Internet Users (per 1,000 people) 3 2030 2010 6 2000 2013 .80 .20 .40 1999 0 0 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −2.9*** + .78*** ln(GNI pc); R 2 = .568 ln(SDG) = −8.6*** + 1.16*** ln(GNI pc); R 2 = .677 Note: Highlighted observations are for Pakistan at different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. Pakistan 119 relationship between the indicator and GNI per completion rate, secondary school enrollment, capita is classified as tight.6 A loose relationship ratio of female to male primary completion, ratio suggests that progress in the indicator is ­primarily of female to male secondary enrollment, under-5 a reflection of country-specific factors and that it mortality, Internet users, and CO2 emission).9 should not be expected to respond strongly or While underperformance for an indicator may systematically to changes in GNI per capita. be  due to country-specific conditions that are When the relationship to GNI per capita is loose difficult to change, it may alternatively point to ­ the coefficients are typically small (in absolute areas in which payoffs from feasible policy change terms); given this, the “expected” values for a are relatively high, a possibility that calls for fur- recent year are close to the average for all low- ther analysis. and middle-income countries.7 Figure 7.2 shows that, between 2000 and 2012, In sum, Pakistan’s current outcomes are better Pakistan’s ranking among low- and middle-­ than expected (compared to a typical country at the income countries stayed more or less the same in same GNI per capita level) for 8 of the indicators terms of GNI per capita (a deterioration by (shared prosperity, pre-primary school e ­ nrollment, percentile point). During the same period, 1  ­ maternal mortality, HIV prevalence, access to among the 13 SDG indicators with rankings for improved water source, access to improved sanita- both years, 4 improved their rankings (shared tion, access to electricity, and  road density).8 prosperity, access to sanitation facilities, access For  another 2 (poverty and malaria), current to  electricity, and Internet use (strongly).10 The outcomes are as expected. The country falls short ­ rankings were roughly unchanged for 4 indicators of expectations for the remaining 7 (primary (poverty, maternal mortality, HIV prevalence, Figure 7.2  Pakistan—Percentile Cross-Country Ranking for SDG Indicators since 2000 GNI per capita (constant 2005 US$) 32 31 Shared prosperity: Income share for lowest 40% 84 96 Poverty, at $1.25 a day (PPP) (% of population) 39 41 Pre-primary enrollment (% gross) 84 81 Primary completion (% gross) 39 Secondary enrollment (% gross) 31 Primary completion, ratio of females to males (%) 24 Secondary enrollment, ratio of females to males (%) 27 Under-5 mortality (per 1,000 live births) 25 15 43 Maternal mortality (per 100,000 live births) 43 Malaria cases (% of population) 68 46 99 HIV prevalence (% of population ages 15–49) 99 Access to improved sanitation facilities (% of population) 32 35 Access to improved water source (% of population) 61 54 Road density (km per 100 sq. km of land) 81 75 Access to electricity (% of population) 57 63 Internet users (per 1,000 people) 11 29 CO2 emissions (metric tons per capita) 57 55 0 10 20 30 40 50 60 70 80 90 100 2000 Recent Note: A high ranking signals strong performance; for the underlying indicator, this may correspond to a relatively high value (for example, for the secondary enrollment rate) or a relatively low value (for example, for the poverty rate). The ranking for an indicator is not reported if the available sample is less than 20. Recent refers to the latest year with data (typically 2011 or 2012). If data is not available for 2000 or 2013, the closest earlier year with data is used, however, the data is never older than 1998 for “2000” or 2009 for “recent.” The year for country- specific data can be found in the respective graphs. 120 Pakistan and  CO2 emissions), while they deteriorated However, compared to global ambitions, also for 5 ­ indicators (gross pre-primary enrollment, shown in table 7.1, the improvements are m ­ oderate. under-5 mortality, malaria cases, access to This means that, to get closer to the realization of improved water source, and road density). these ambitions, a break with the past is needed, When comparing the results from regressions something that would be facilitated by a combina- on GNI per capita to changes in percentile tion of more rapid growth and improvements in ­rankings, a few insights emerge. Under-5 ­mortality policies that directly influence different SDGs. stands out as a particularly problematic area since it combines underperformance with a strong dete- rioration in ranking compared to GNI per capita. This is also the case for malaria that, although cur- 3. Fiscal Space rent performance is as expected, Pakistan’s rank- In most countries, accelerated progress on the ing has significantly deteriorated. For the SDG SDG agenda will require efficient and growing indicator with the strongest ranking improve- public spending in prioritized areas, most ment, Internet use, Pakistan nevertheless remains importantly human development and infra- an underperformer. For shared prosperity, the structure. Private spending is also of crucial country’s ranking has improved considerably and importance, both household spending on SDG- the indicator value is now better than expected. related services and business investments in a By 2030, considerable improvements are wide range of areas (including but not limited projected for most of the selected indicators ­ to  infrastructure).11 With regard to Pakistan’s (see table 7.1 and respective graphs in figure 7.1). fiscal  space indicators, table 7.2 and figure 7.3 Table 7.2  Pakistan—Fiscal Space: Revenue, Spending, and Government Efficiency Actual Expected Projection Indicator 2000 Recent Recent 2030 Government spending Consumption (% of GDP) 8.6 11.0 13.4 — Investments (% of GDP) 5.6 3.5 6.5 — Primary education (% of GDP) 0.8 1.7 — Secondary education (% of GDP) 0.6 1.2 — Secondary education, per student (% of GDP per capita) 10.4 18.7 — Health (% of GDP) 0.4 1.0 2.6 1.2 Fuel subsidy (% of GDP) 4.0 1.9 — Government receipts and debts Tax revenue (% of GDP) 10.6 9.9 13.1 — Net ODA (% of GNI) 1.0 0.9 4.0 0.6 External debt (% of GNI) 45.1 22.8 32.3 — Governance and government efficiency Government effectiveness: percentile rank 31 23 20 27 Public investment management index 1.6 1.5 — Memorandum item GNI per capita (constant 2005 US$) 589 807 — 1,267 Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = current value significantly above the expected level; red = current value significantly below the expected level; black = current value as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). Pakistan 121 Figure 7.3  Pakistan—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting a. Government spending: consumption (left), investment (right) 130 50 60 Consumption (% of GDP) 20 Investments (% of GDP) 30 9 10 2013 2000 4 2000 2013 2 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.01*** + .09** ln(GNI pc); R 2 = .042 ln(SDG) = 2.37*** + −.07 ln(GNI pc); R 2 = .013 b. Government spending: primary school (left), secondary school (right) 6 4 Secondary education (% of GDP) 4 Primary education (% of GDP) 2 2 1.2 1.1 2013 .60 2013 .40 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .950** + −.06 ln(GNI pc); R 2 = .016 ln(SDG) = −.98** + .17*** ln(GNI pc); R 2 = .118 figure continues next page 122 Pakistan Figure 7.3  continued c. Government spending: health (left), fuel subsidy (right) 20 20 2011 1.6 Fuel subsidy (% of GDP) Health (% of GDP) 6 .10 1.6 0 2030 2013 .50 2000 150 400 1,000 3,000 8,000 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.25 + .18*** ln(GNI pc); R 2 = .111 ln(SDG) = 4.96*** + −.65*** ln(GNI pc); R 2 = .098 d. Tax revenue (left), official development aid (right) 60 40 7 30 Tax revenue (% of GDP) Net ODA (% of GNI) 20 2000 1.2 2012 2030 10 .20 2000 2013 3 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.69*** + .13* ln(GNI pc); R 2 = .030 ln(SDG) = 8.26*** + −1.0*** ln(GNI pc); R 2 = .308 figure continues next page Pakistan 123 Figure 7.3  continued e. External debt (left), government effectiveness (right) 2,000 Government effectiveness: percentile rank 80 2000 550 30 External debt (% of GNI) 2030 160 2013 2000 10 50 4 10 2013 1.3 .50 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.51*** + .14** ln(GNI pc); R = .035 2 ln(SDG) = 0 + .44*** ln(GNI pc); R 2 = .300 f. Public investment management index 4 Public investment management index 2 1.3 2010 .80 .30 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = −.43 + .12** ln(GNI pc); R 2 = .078 Note: Highlighted observations are for Pakistan in different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. summarize the historical evolution, actual and indicators). The variables cover selected indica- expected recent values, and, when relevant, pro- tors related to three aspects of government jected values.12 When the relationship is loose, activities: spending, receipts and debt, and gov- projections are not made and the expected ernance and efficiency. While the findings of value is in practice close to the average for the this country-at-a-glance note cannot guide sample of all low- and m ­ iddle-income coun- ­ policy on their own, they should be used as an tries (cf. discussion of expected values for SDG input into thinking about policy making. 124 Pakistan Room for additional priority spending may capita of Pakistan.19 However, other assessments be created by reducing low-priority spending, suggest considerable efficiency gains may be increasing current receipts, and/or increasing feasible.20 borrowing. Among the included government In general, fiscal policies in support of the SDG spending indicators, Pakistan is below expected agenda, including decisions about the level and levels (compared to a typical country at the same allocation of government spending, should be GNI per capita level) for total public investment, made in light of government priorities and addi- total public consumption, primary education, tional data that this note does not consider, includ- secondary education, and health, but it spends ing information about government capacity to more than expected on fuel subsidies.13 Fuel sub- efficiently expand activities in different areas and sidies, in 2011 as high as 4.0 percent of GDP, are the scope to encourage complementary private the most obvious case of low-priority spending sector activities. The above information suggests from the post-2015 agenda perspective.14 These that the most obvious opportunities for additional subsidies have since been on a declining path fiscal space would be decreased fuel subsidies and (IMF 2015d, p. 15). increased taxes. The fact that cross-country pat- Among current receipts, taxes and net terns point to a likely decline in ODA does not Official Development Assistance (ODA) are mean that an increase is excluded: it depends on lower than expected.15 As further shown in the priorities of donors and their relationships table  7.2, cross-country patterns suggest that, with Pakistan’s government. Opportunities to as GNI per capita grows, net ODA will decline improve government efficiency should also be as a percent of GDP (without changing signifi- pursued. The fact that, from a cross-country per- cantly in per capita terms), which in the case of spective, government spending is relatively low, Pakistan would translate into a decline from both in general and in priority areas, while effi- the recent 0.9 percent of GDP to 0.5 percent in ciency indicators are as expected, suggests that, in 2030.16 The relationships between tax revenues the long run, additional prioritized spending and GNI per capita is not tight enough to proj- could yield positive payoffs; however, such long- ect expected changes. However, cross-country run considerations need to be balanced against data suggest that the tax intake is lower than the more immediate priority of reducing the fiscal expected and recent World Bank and IMF anal- deficit (it has already declined from 8.3 percent of yses suggest that the tax to GDP ratio should be GDP [excluding grants] in 2012/13 to 5.5 percent increased.17 Pakistan’s public external debt in 2013/14). Beyond issues directly related to the stock (relative to GDP) is below the expected government budget, remittances from Pakistan’s level; nevertheless, the recent IMF/World Bank workers abroad are somewhat higher than debt-sustainability analysis (which considers expected (6.5 percent of GDP in 2013/14), signifi- additional information) indicates a future cantly raising the living standards of many decline in foreign borrowing but an increase in Pakistanis; measures that encourage remittances domestic borrowing.18 and channel them to income-raising investments Government efficiency is important to pro- may have high trade-offs for the SDG agenda. tect and, if possible, increase in order to add to More broadly, from the perspective of this agenda the room for priority spending and enhance its and given strong linkages between private and impact on the SDG agenda. Pakistan is perform- government activities and incomes, it is crucial ing somewhat better than expected according to that policies and spending decisions promote a the World Bank Government Effectiveness indi- broad-based change that encompasses services cator and as expected in terms of the Public related to human development, infrastructure Investment Management Index. That is, accord- investments, and other measures in support of ing to these indicators, government efficiency strong long-run growth that is biased in favor of appears to be as expected given the GNI per the less advantaged. Pakistan 125 4. Conclusions and malaria, for which Pakistan’s ranking among low- and middle-income countries has deterio- As summarized in table 7.3, among the SDG rated significantly more than for GNI per capita indicators that this note covers, Pakistan’s cur- ranking. rent outcomes are better than expected (com- For indicators with a weak relationship with pared to a typical country at the same GNI per GNI per capita, including our measure of shared capita level) for eight indicators, as expected for prosperity (the income share of the bottom two, and below expectations for seven. By 2030, 40 percent), strong and systematic improvements even though considerable improvements are should not be expected to accompany economic projected for most of the indicators, they are still growth; hence, policies affecting such indicators likely to fall short of the global ambitions, indi- require additional attention. cating that a break from business as usual and The cross-country perspective of this note current trends would be needed to realize such suggests that government spending is relatively ambitions. low, that government efficiency is not necessarily Table 7.3 further shows that for most indica- lagging, although additional fiscal space may tors, the relationship to GNI per capita is tight. mainly be created through increased tax ­revenues. Improvements in these SDGs will likely continue Given this, increased government spending in along with GNI per capita growth and increases in priority areas may be possible and yield positive resources and capabilities. Note, however, that payoffs in terms of SDGs. Detailed country-specific many of these SDGs are currently underperform- analysis and actions are needed to develop and ing, suggesting that additional gains are feasible. implement a strategy that brings Pakistan onto a The importance of reviewing current policies path that promises to realize the global ambitions seems especially acute for under-5 mortality of the SDG agenda. Table 7.3  Pakistan—Summary of Results for SDG Indicators Cross-country Overperforming As expected Underperforming relationship with GNI/cap Tight •• Pre-primary school enrollment (×) •• Poverty (×) •• Primary completion rate •• Maternal mortality (×) •• Malaria (−) •• Secondary school enrollment •• Access to improved water source (−) •• Ratio of female to male primary completion •• Access to improved sanitation (+) •• Ratio of female to male secondary enrollment •• Access to electricity (+) •• Under-5 mortality (−) •• Road density (−) •• Internet users (+) •• CO2 emissions (×) Loose •• Shared prosperity (+) •• HIV prevalence (×) Note: (+) = larger country rank improvement 2000–12 than for GNI per capita; (−) = drop in ranking 2000–12 or smaller improvement than for GNI per capita; (−) = the same country rank change 2000–12 as for GNI per capita. 126 Pakistan Annex 7A: Data Sources Indicator Source Comment GNI per capita WDI. API ref: GNI per capita (constant 2005 US$) (constant 2005 US$) [NY.GNP.PCAP.KD] SDG indicators Poverty, at $1.25 a day (PPP) WDI. API ref: poverty headcount ratio at $1.25 a day (PPP) For Pakistan; 2011 data from the (% of population) (% of population) [si_pov_dday] World Bank’s MDG data website. Shared prosperity: income share WDI. API ref: income share held by lowest 20% + income for lowest 40% share held by second 20% [SI.DST.FRST.20+SI. DST.02ND.20] Pre-primary enrollment (% gross) WDI. API ref: School enrollment, pre-primary (% gross) [SE.PRE.ENRR] Primary completion (% gross) WDI. API ref: primary completion rate, total (% of relevant age group) [se_prm_cmpt_zs] Secondary enrollment (% gross) WDI. API ref: school enrollment, secondary (% gross) [se_sec_enrr] Secondary completion (% gross) EdStat. API ref: DHS: secondary completion rate [hh_dhs_scr] Primary completion, ratio of WDI. API ref: primary completion rate, female (% of females to males (%) relevant age group)/primary completion rate, male (% of relevant age group) *100 [SE.PRM.CMPT.FE.ZS/SE.PRM. CMPT.MA.ZS*100] Secondary enrollment, ratio of WDI. API ref: ratio of female to male secondary enrollment females to males (%) (%) [se_enr_seco_fm_zs] Under-5 mortality (per 1,000 live WDI. API ref: mortality rate, under-5 (per 1,000 live births) births) [SH.DYN.MORT] Maternal mortality (per 100,000 WDI. API ref: maternal mortality ratio (modeled estimate, live births) per 100,000 live births) [sh_sta_mmrt] Malaria cases (% of population) HNP. API ref: malaria cases reported/population, total *100 [sh_sta_malr/SP.POP.TOTL *100] HIV prevalence (% of population WDI. API ref: prevalence of HIV, total (% of population ages 15–49) ages 15–49) [sh_dyn_aids_zs] Access to improved sanitation WDI. API ref: improved sanitation facilities facilities (% of population) (% of population with access) [sh_sta_acsn] Access to improved water source WDI. API ref: improved water source (% of population (% of population) with access) [sh_h2o_safe_zs] Road density (km per 100 sq. WDI. API ref: road density (km of road per 100 sq. km of land) km of land area) [is_rod_dnst_k2] Access to electricity (% of WDI. API ref: access to electricity (% of population) population) [eg_elc_accs_zs] Internet users (per 1,000 people) WDI. API ref: Internet users (per 100 people) [IT.NET.USER.P2] CO2 emissions (metric tons WDI. API ref: CO2 emissions (metric tons per capita) per capita) [en_atm_co2e_pc] Fiscal space indicators Investment (% of GDP) WDI. API ref: gross fixed capital formation (% of GDP)-gross fixed capital formation, private sector (% of GDP) [ne_gdi_ftot_zs]-[ne_gdi_fprv_zs] Consumption (% of GDP) WDI. API ref: general government final consumption expenditure (% of GDP) [NE.CON.GOVT.ZS] Primary education (% of GDP) EdStat. API ref: total expenditure on educational For Pakistan, data from the Ministry of institutions and administration as a % of GDP. All sources. Finance. Primary [UIS.XGDP.1.FDINSTADM.FFD] Secondary education (% of GDP) EdStat. API ref: total expenditure on educational For Pakistan, data from the Ministry of institutions and administration as a % of GDP. All sources. Finance. Secondary and post-secondary non-tertiary [UIS.XGDP.234.FDINSTADM.FFD] annex continues next page Pakistan 127 Indicator Source Comment Primary education, per student WDI. API ref: expenditure per student, primary (% of GDP per capita) (% of GDP per capita) [se_xpd_prim_pc_zs] Secondary education, per WDI. API ref: expenditure per student, secondary student (% of GDP per capita) (% of GDP per capita) [se_xpd_seco_pc_zs] Health (% of GDP) WDI. API ref: health expenditure, public (% of GDP) [sh_xpd_publ_zs] Fossil fuel subsidy (% of GDP) IMF (2013a). Total pretax subsidy (% of GDP) Subsidies are measured using a price-gap approach capturing both consumer (including implicit) and producer (except those that arise when suppliers are inefficient and make losses at benchmark prices) subsidies. Negative external effects are not included in the pretax subsidy. Tax revenue (% of GDP) WDI. API ref: tax revenue (% of GDP) [gc_tax_totl_gd_zs] Net ODA (% of GNI) WDI. API ref: net ODA received (% of GNI) [dt_oda_odat_gn_zs] External debt (% of GNI) WDI. API ref: external debt stocks (% of GNI) [DT.DOD.DECT.GN.ZS] Government effectiveness: WGI. API ref: government effectiveness: percentile rank Captures perceptions of the quality of public percentile rank [ge_per_rnk] services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies. Grigoli education efficiency score Grigoli (2014) Measure secondary education inefficiency in terms of how much additional output could be achieved at current levels of spending. Grigoli health efficiency score Grigoli and Kapsoli (2013) Quantifies the inefficiency of public health expenditure using a stochastic frontier model that controls for the socioeconomic determinants of health. Public investment management Dabla-Norris et al. (2011) Four phases associated with public index investment management are covered: project appraisal, selection, implementation, and evaluation. Note: WDI = World Development Indicators, World Bank; API ref = reference and code when using the World Bank Open Data; EdStat = Education statistics, World Bank; HNP = Health Nutrition and Population statistics, World Bank; WGI = Worldwide Governance Indicators, World Bank. Notes 1. While a cross-country perspective provides an 4. Projections from CEPII are used for this and other important complement to analysis that is centered Country Development Diagnostics applications on an individual country, it is by definition lim- given its wide country coverage and ­well-documented ited to analysis of variables that are available in methodology; OECD data have been used when cross-country databases. projections have been missing. In the projections, 2. This does not mean that GNI per capita is viewed it is assumed that future GNI growth will coincide as a direct determinant of SDG outcomes; on the with future GDP growth (both expressed in constant contrary, a major challenge for policy makers is to 2005 US$) given that that this is the variable that identify policies that improve SDG performance CEPII and other sources project. relative to what is expected given the level of GNI 5. Given that (a) SDGs have extreme values (such as per capita. A second challenge is to raise growth in 100 percent for improved water access) and (b) the GNI per capita as it indirectly influences country current SDG level never is exactly as expected given SDG capacity. GNI per capita, the projected values gradually con- 3. Sources for the indicators are presented in the verge toward the expected values. For example, for table in the annex 7A. a country that overperforms in water access, as GNI 128 Pakistan per capita increases the extent of overperformance 14. Fuel subsidies are detrimental to the climate and gradually declines, so that when the expected value encourage technologies that are less labor inten- is 100, overperformance has reached zero. sive, tending to generate employment for fewer 6. A tight enough relationship is defined as an workers at lower wages. The disaggregated ­ levels R 2 > 0.3 (tight) or 0.3 > R 2 > 0.1 (moderately tight), for Pakistan in 2011 were petroleum products while R 2 > 0.1 are defined as loose. (0.13% of GDP), electricity (1.31% of GDP), 7. In addition, the confidence interval is wide in the ­ natural gas (2.54% of GDP), and coal (0.00% of case of a loose relationship, suggesting that any GDP) (IMF 2013a). conclusion on over- or underperformance is made 15. Net ODA is lower than expected both per capita with wide margins. Statistically, even though their and as a percent of GDP. Total government reve- confidence intervals are wide, as long as the esti- nues (excluding grants) as a share of GDP are also mated coefficient linking GNI per capita to the SDG lower than expected. indicator is nonzero, these values are closer than 16. Apart from a temporary increase in the beginning the cross-country average to what is expected for of the 2000s, net ODA as a percentage of GNI has the specific country. The same observation applies been less than expected for Pakistan since the to expected values for fiscal space indicators. mid-1990s. 8. For Pakistan, the very high population density 17. World Bank (2013e, p. 13) suggests that Pakistan raises the need for and makes it easier to achieve should aim for tax revenues of 14–15 percent of high road density. GDP by 2018. IMF (2014f, p. 23) concludes that 9. With regard to CO2, Pakistan’s current and pro- a broadening of the revenue base and better tax jected 2030 per capita emissions are 9.2 and 14.0 administration would be essential to continued percent of the current OECD average. Pakistan is improvement in the fiscal situation. IMF (2015d, underperforming (emitting above expected levels) p. 27) suggests that Pakistan’s tax revenues may both when CO2 emission are measured per capita increase from its 2012/13 level of 9.9 percent of and as a share of GDP. GDP to 12.9 percent in 2019/20. 10. If data for 2013 were not available, data for 2012, 18. The Pakistan DSA (IMF 2014f; annex II) finds 2011, or 2010 were used. that reduced foreign debt and fiscal strength- 11. There are also cases where the solution to the low ening are needed for Pakistan to become resil- level of SDG is neither private nor public spending ient to standard size shocks. The external debt but more efficient policies. has already decreased from 34.7 in 2010 to 26.2 12. The treatment is the same as in table 7.1 and in 2013 and is projected to decrease further to related  figures. That is, in table 7.2, projections 23.3 percent in 2020 (IMF 2015d, p. 45). Given are shown only when the cross-country r ­ elationship strong reliance on short-term debt, the projected between the indicator and GNI per capita is consid- decline would be sensitive to a large interest rate ered tight enough. Due to data limitations, we focus shock. on government spending indicators; country-spe- 19. Pakistan is also performing as expected for the cific analysis is needed to consider policy in the World Bank’s Governance Indicator “Regulatory context of the different roles of the government and quality” and “Rule of Law.” private services and spending. 20. World Bank (2013e, pp. 14, 60) argues that 13. Although educational expenditures have increased “addressing the drag imposed on the private in absolute levels, they have not increased as a ­ sector  by an underperforming and overpaid ­ proportion of GDP. public sector is essential.” ­ Pakistan 129 Chapter 8 Peru 1. Introduction to a strong policy framework that engages the private sector. The note briefly (a) summarizes ­ Peru, an upper middle-income country in South Peru’s SDG progress since 2000 and projects America, has been one of the best macroeconomic expected values for 2030; and (b) assesses options performers in Latin America over the past decade for increasing fiscal space. Sections 2 and 3 address in terms of GDP growth, macroeconomic manage- SDGs and fiscal space, respectively, while findings ment, poverty reduction, and s ­hared-prosperity. are summarized in Section 4. However, much is still to be achieved to address The analysis and the SDG projections are done the basic needs of the population and converge to from a cross-country perspective: for the different the living standards of OECD countries. indicators, Peru’s performance and prospects are According to 2013 Peru data from the World benchmarked relative to other countries, consid- Development Indicator (WDI) Database, the ering its past, recent, and projected levels of GNI economy is mainly based on services, which per capita.1 Since a country’s GNI tends to be accounts for 58 percent of GDP, while industry highly correlated with both SDGs and the factors represents 36 percent and agriculture a mere that determine their evolution, it is used as a sum- 6  percent (World Bank 2013a). Minerals and mary indicator of country capacity to provide and hydrocarbons accounted for 13 percent of GDP, efficiently utilize inputs that contribute to SDGs 14 percent of government revenues and 68 (for example, health and education services) and percent of total exports (IMF 2014c, p. 6). to achieve SDG ­ outcomes (like strong health and During 2001–12, Peru’s average growth rate for education results).2 GNI per capita (at constant 2005 US$) was 4.1 percent, which may be compared to a devel- oping (low- and middle-income) country aver- 2. SDG Indicators: History and age of 3.0  percent. During the same period, Peru’s ranking according to the UNDP Human Projections Development Index (among countries included For the selected SDG indicators, table 8.1 summa- both in 2000 and 2012) remained at more or less rizes data for Peru: historical evolution, actual and the same level (deteriorated slightly from the expected values for a recent year, and projected 58th to the 54th percentile). 2030 values.3 In figure 8.1, data for Peru are shown This country-at-a-glance note is designed to in the context of the estimated cross-country provide an initial picture of the challenges that the relationship between each SDG indicator and GNI ­ Post-2015 SDG agenda might pose for Peru; which per capita. For Peru, the projected average annual should be seen as an input to the policy discus- rate of GNI per capita growth is 4.2 percent.4 The sions around this agenda, together with a study on projected SDG values reflect what can be expected Peru’s productivity (to be issued in 2015) and a given a country’s starting point, projected growth Systematic Country Diagnostic (forthcoming in in GNI per capita, typical rates of progress accord- 2016). The note is built around data for a selection ing to cross-country patterns, and a gradual con- of SDG target indicators and indicators related to vergence to close gaps between observed and fiscal space—fiscal space matters since rapid prog- expected values.5 Projections of the SDG indicators ress on the SDG agenda will require efficient and are presented only when the cross-country rela- carefully prioritized public spending—in addition tionship between the indicator and GNI per capita Peru 131 Table 8.1  Peru—SDG Indicators: Evolution since 2000 and Projections to 2030 Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Poverty and share prosperity Poverty, at $1.25 a day (PPP) 12.5 2.9 1.8 1.3 By 2030, eradicate extreme poverty for all people (% of population) everywhere, currently measured as people living on less than $1.25 a day. Shared prosperity: income 11.1 13.4 15.4 — By 2030, progressively achieve and sustain income share for lowest 40% growth of the bottom 40 percent of the population at a rate higher than the national average. Education Pre-primary enrollment 58.9 86.0 53.5 91.2 By 2030, ensure that all girls and boys have access to (% gross) quality early childhood development, care, and pre-primary education so that they are ready for primary education. Primary completion (% gross) 101.9 92.5 94.9 100.0 By 2030, ensure that all girls and boys complete free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment (% gross) 84.8 94.0 84.4 99.6 By 2030, ensure that all girls and boys complete free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary completion 87.9 48.1 — By 2030, ensure that all girls and boys complete free, (% gross) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Primary completion, ratio of 96.7 100.1 99.7 — By 2030, ensure that all girls and boys complete free, females to males (%) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment, ratio of 93.1 97.8 102.6 106.1 By 2030, ensure that all girls and boys complete free, females to males (%) equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Health Under-5 mortality (per 1,000 live 39.8 16.7 19.3 12.1 By 2030, end preventable deaths of newborns and births) children under 5 years of age. Maternal mortality (per 100,000 160.0 89.0 51.6 55.7 By 2030, reduce the global maternal mortality ratio to live births) less than 70 per 100,000 live births. Malaria cases (% of population) 0.3 0.1 0.0 0.0 By 2030, end the epidemics of AIDS, tuberculosis, malaria, and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. HIV prevalence (% of 0.5 0.4 0.5 — By 2030, end the epidemics of AIDS, tuberculosis, population ages 15–49) malaria, and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. Infrastructure Access to improved sanitation 63.2 73.1 75.7 97.1 By 2030, achieve access to adequate and equitable facilities (% of population) sanitation and hygiene for all and end open defecation, paying special attention to the needs of women and girls and those in vulnerable situations. Access to improved water 80.6 86.8 91.3 97.2 By 2030, achieve universal and equitable access to source (% of population) safe and affordable drinking water for all. Road density (km per 100 sq. 6.1 10.0 25.7 19.1 Develop quality, reliable, sustainable, and resilient km of land) infrastructure, including regional and transborder infrastructure, to support economic development and human well-being, with a focus on affordable and equitable access for all. Access to electricity (% of 71.9 85.1 76.6 100.0 By 2030, ensure universal access to affordable, population) reliable, and modern energy services. table continues next page 132 Peru Table 8.1  continued Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Internet users (per 1,000 3.1 39.2 33.8 — Significantly increase access to information and people) communications technology and strive to provide universal and affordable access to the Internet in least developed countries by 2020. Environment CO2 emissions (metric tons 1.2 2.0 2.2 4.5 Integrate climate change measures into national per capita) policies, strategies and planning. Memorandum item GNI per capita (constant 2005 2,267 3,656 6,538 US$ Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = currently significantly overperforming; red = currently significantly underperforming; black = performing as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). For Internet use, a projection is not made, despite a tight relationship, since the expected line shifts significantly even in the medium run due to external forces, such as technological development. Global ambition is from the Open Working Group (UN 2014). Figure 8.1  Peru—SDG Indicators (Log Scale) versus GNI per Capita in a Cross-Country Setting (Log Scale) a. Poverty (left), Shared prosperity: income share of bottom 40% (right) Shared prosperity: Income share for lowest 40% 410 30 90 Poverty, at $1.25 a day (PPP) 20 (% of population) 20 2000 2012 4 2012 2030 10 2000 .90 9 .04 5 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 11.9*** + −1.3*** ln(GNI pc); R 2 = .493 ln(SDG) = 3.2*** + −.06** ln(GNI pc); R 2 = .054 figure continues next page Peru 133 Figure 8.1  continued b. Gross pre-primary enrollment (left), primary completion (right) 170 120 2000 2013 Pre-primary enrollment (% gross) Primary completion (% gross) 2013 2030 2000 2030 80 40 50 10 40 3 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.41 + .53*** ln(GNI pc); R 2 = .301 ln(SDG) = 3.47*** + .13*** ln(GNI pc); R 2 = .363 c. Gross secondary enrollment (left), secondary completion (right) 90 2012 110 2000 Secondary completion (% gross) Secondary enrollment (% gross) 2030 20 60 2013 6 30 20 1.4 6 .09 150 400 1,000 3,000 8,000 150 400 1,000 3,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.94*** + .3*** ln(GNI pc); R 2 = .540 ln(SDG) = −.24 + .5* ln(GNI pc); R 2 = .109 figure continues next page 134 Peru Figure 8.1  continued d. Ratio of female to male primary completion (left), ratio of female to male secondary enrollment (right) Secondary enrollment, ratio of females to males (%) Primary completion, ratio of females to males (%) 140 140 110 100 2013 2030 90 2000 2000 2013 70 70 50 40 30 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 4.38*** + .03*** ln(GNI pc); R 2 = .055 ln(SDG) = 3.88*** + .09*** ln(GNI pc); R 2 = .268 e. Under-5 mortality (left), maternal mortality (right) 500 2,000 Maternal mortality (per 100,000 live births) Under−5 mortality (per 1,000 live births) 230 470 110 2000 2013 100 2030 50 2000 20 20 2013 2030 5 1 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 7.98*** + −.61*** ln(GNI pc); R 2 = .559 ln(SDG) = 11.3*** + −.9*** ln(GNI pc); R 2 = .518 figure continues next page Peru 135 Figure 8.1  continued f. Malaria cases (left), HIV prevalence (right) HIV prevalence (% of population ages 15−49) 40 90 30 Malaria cases (% of population) 1.5 2000 2013 9 .06 2030 3 .90 2000 2013 .10 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 13.9*** + −2.2*** ln(GNI pc); R 2 = .341 ln(SDG) = 1.88* + −.31** ln(GNI pc); R 2 = .044 g. Access to improved sanitation (left), access to improved water source (right) 2030 2030 100 100 Access to improved sanitation Access to improved water facilities (% of population) source (% of population) 2000 70 2000 2012 60 2012 50 30 40 20 7 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .61** + .45*** ln(GNI pc); R 2 = .534 ln(SDG) = 3.45*** + .13*** ln(GNI pc); R 2 = .435 figure continues next page 136 Peru Figure 8.1  continued h. Access to electricity (left), road density (right) 330 400 Road density (km per 100 sq. km of land) Access to electricity (% of population) 90 100 2010 2030 30 2000 20 2030 7 2011 6 2000 1.6 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.64 + .48*** ln(GNI pc); R 2 = .189 ln(SDG) = .29 + .5*** ln(GNI pc); R 2 = .459 i. Internet users (left), CO2 emissions (right) 20 CO2 emissions (metric tons per capita) 80 0 Internet users (per 1,000 people) 3 2030 2013 2010 .80 2000 6 .20 2000 .00 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −2.9*** + .78*** ln(GNI pc); R 2 = .567 ln(SDG) = −8.6*** + 1.16*** ln(GNI pc); R 2 = .677 Note: Highlighted observations are for Peru at different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. Peru 137 is classified as tight.6 A very loose relationship source) and for others more significantly (­primary ­ suggests that progress in the indicator is a reflec- completion, malaria cases, Internet users, and tion of other country-specific factors and that it CO2 emissions). Among these, the result is not should not be expected to respond strongly or sys- entirely unexpected for CO2 emissions, which has tematically to changes in GNI per capita. When an inverse correlation with GNI per capita the relationship to GNI per capita is loose the coef- (cf. figure 8.1, right panel). ficients are in practice also small (in absolute When comparing the results from regres- terms); given this the “expected” values for a recent sions on GNI per capita to changes in percentile year are close to the average for all low- and rankings, a few insights emerge. For example, ­middle-income countries.7 Peru is doing as expected for primary comple- Of the 18 SDG indicators, Peru’s current tion but its ranking has deteriorated more than ­ outcomes are better than expected (compared to any other indicator, in spite of Peru’s improved a typical country at the same GNI per capita level) GNI per capita ranking. On the other hand, for 3; pre-primary school enrollment, ­ secondary Peru’s maternal mortality and road density school enrollment, and under-5 m ­ortality. are  lower than expected but its ranking has However, it falls short for 7 indicators; poverty, improved by more than GNI per capita, sug- shared prosperity, ratio of female to male second- gesting that policies have been successful ary enrollment, maternal mortality, malaria inci- beyond what is expected from increases in GNI dence, access to improved water source, and road per capita. Finally, for poverty, malaria, and density.8 For the remaining 8 indicators (primary access to water, Peru’s underperformance is school completion, secondary school completion, combined with a performance that is weaker ratio of female to male primary school comple- than GNI per capita in terms of changes in per- tion, HIV prevalence, access to improved sanita- centile ranking.11 tion, access to electricity, Internet use, CO2 By 2030, considerable improvements are pro- emissions), Peru’s current outcomes are as jected for most indicators if the assumption on expected. While underperformance for an indi- GNI growth materializes as expected (see table 8.1 cator may be due to country-specific conditions and respective graphs in figure 8.1). In such a that are difficult to change, it may often point to case, poverty, primary completion rate, second- areas in which payoffs from feasible policy change ary school enrollment, ratio of female to male are relatively high; a possibility that calls for fur- primary completion rate, ratio of female to male ther analysis. secondary enrollment rate, maternal mortality, ­ ndicators Figure 8.2 shows that, for the 17 SDG i malaria cases, access to improved sanitation, with enough data, between 2000 and 2012, Peru access to improved water, and access to electric- improved its GNI per capita ranking among low- ity are projected to either realize or get close to and middle-income countries by 4 percentile realizing the global ambition (shown in the last points.9 Peru’s percentile ranking improved to column of table 8.1). However, for the other roughly the same extent for 4 SDG indicators SDGs to get closer to the realization of these (shared prosperity,10 pre-primary enrollment, ambitions, a break with the past is needed. This is ratio of female to male secondary enrollment, and also true for indicators, such as shared prosperity, access to electricity). For another 4, the ranking for which a weak relationship with GNI per improved more than for GNI per capita (under-5 capita precludes projections. Such a break would mortality, maternal mortality, HIV prevalence, be facilitated by a combination of more rapid and road density). For the remaining 9 indicators growth and improvements in overall and sectoral the ranking deteriorated; for some marginally policies that directly influence different SDGs; (poverty, secondary enrollment, ratio of female to keeping in mind that a growth rate of 4.2 percent male primary completion, access to improved until 2030 has already been assumed for the sanitation facilities, and access to improved water ­business-as-usual projections. 138 Peru Figure 8.2  Peru—Percentile Cross-Country Ranking for SDG Indicators since 2000 GNI per capita (constant 2005 US$) 69 73 Poverty, at $1.25 a day (PPP) (% of population) 63 61 Shared prosperity: Income share for lowest 40% 20 23 Pre-primary enrollment (% gross) 78 85 Primary completion (% gross) 90 47 Secondary enrollment (% gross) 85 80 Secondary completion (% gross) 93 51 Primary completion, ratio of females to males (%) 47 42 Secondary enrollment, ratio of females to males (%) 44 Under-5 mortality (per 1,000 live births) 59 72 Maternal mortality (per 100,000 live births) 48 57 Malaria cases (% of population) 56 45 HIV prevalence (% of population ages 15–49) 60 67 Access to improved sanitation facilities (% of population) 54 51 Access to improved water source (% of population) 48 43 Road density (km per 100 sq. km of land) 20 33 Access to electricity (% of population) 55 Internet users (per 1,000 people) 79 70 CO2 emissions (metric tons per capita) 43 35 0 10 20 30 40 50 60 70 80 90 100 2000 Recent Note: A high ranking signals strong performance; for the underlying indicator, this may correspond to a relatively high value (for example, for the secondary enrollment rate) or a relatively low value (for example, for the poverty rate). The rankings are based on all low- and middle-income countries (according to the 2013 classification) with data. The country samples vary across indicators but are always the same for 2000 and recent for any given indicator. The ranking for an indicator is not reported if the available sample is less than 20 countries. Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. 3. Fiscal Space private nor public spending but more efficient policies. With regard to Peru’s additional fiscal In most countries, accelerated progress on the effort to reach the SDGs, table 8.2 and figure 8.3 SDG agenda will require efficient and possibly summarize the historical evolution, actual and growing public spending in prioritized areas, expected (given GNI per capita) recent values, most importantly human development and and, when relevant, projected values.12 The vari- infrastructure. Private spending is also of crucial ables cover selected indicators related to three importance, both household spending on SDG- aspects of government activities: spending, related services and business investments to receipts governance, and efficiency. In addition, profitably provide services in a wide range of we address debt sustainability. While these find- areas related to the SDGs (including but not lim- ings on their own would be insufficient for policy ited to infrastructure). There are also cases where design, they provide an interesting input to the solution to the low level of SDG is neither Peru 139 Table 8.2  Peru—Fiscal Space: Revenue, Spending, and Government Efficiency Actual Expected Projection Indicator 2000 Recent Recent 2030 Government spending Consumption (% of GDP) 11.6 11.2 15.4 — Investments (% of GDP) 4.1 5.8 5.8 — Primary education (% of GDP) 1.4 1.3 1.6 — Secondary education (% of GDP) 0.9 0.9 1.5 — Primary education, per student (% of GDP per capita) 8.3 11.1 15.4 — Secondary education, per student (% of GDP per capita) 10.2 10.4 16.9 — Health (% of GDP) 2.7 3.1 3.5 3.5 Government receipts and debts Tax revenue (% of GDP) 12.8 16.5 16.0 — Net ODA (% of GNI) 0.80 0.19 0.85 0.11 External debt (% of GNI) 58.5 29.0 40.3 — Governance and government efficiency Government effectiveness: Percentile rank 52.7 48.8 39.2 57.4 Grigoli education efficiency score 0.90 0.73 1.00 Grigoli health efficiency score 0.98 0.95 — Public investment management index 2.16 1.77 — Memorandum item GNI per capita (constant 2005 US$) 2,267 3,656 6,538 Note: Recent refers to the latest year with data (typically 2011–13). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = current value significantly above the expected level; red = current value significantly below the expected level; black = current value as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). Figure 8.3  Peru—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting a. Government spending: consumption (left), investment (right) 130 50 60 20 Consumption (% of GDP) Investments (% of GDP) 30 9 10 2013 2013 4 2000 2000 2 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.01*** + .09** ln(GNI pc); R = .042 2 ln(SDG) = 2.37*** + −.07 ln(GNI pc); R 2 = .013 figure continues next page 140 Peru Figure 8.3  continued b. Government spending: primary education, per student (left), secondary education, per student (right) 90 50 Secondary education, per student Primary education, per student 30 50 (% of GDP per capita) (% of GDP per capita) 20 20 8 2012 1998 2012 10 1998 2 4 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.72*** + .12** ln(GNI pc); R 2 = .067 ln(SDG) = 3.51*** + −.09 ln(GNI pc); R 2 = .030 c. Government spending: primary education (left), secondary education (right) 6 4 Secondary education (% of GDP) 4 Primary education (% of GDP) 2 2 1.2 2000 2013 2013 1.1 2000 .60 .40 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .940*** + −.06 ln(GNI pc); R 2 = .016 ln(SDG) = −.93** + .16*** ln(GNI pc); R 2 = .113 figure continues next page Peru 141 Figure 8.3  continued d. Government spending: health 20 Health (% of GDP) 6 2030 2000 2013 1.6 .50 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = –.25 + .18*** ln(GNI pc); R 2 = .111 e. Tax revenue (left), official development aid (right) 60 50 8 30 Tax revenue (% of GDP) Net ODA (% of GNI) 20 2012 1.2 2000 10 2000 .20 2013 2030 3 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.69*** + .13* ln(GNI pc); R 2 = .030 ln(SDG) = 8.10*** + −1*** ln(GNI pc); R 2 = .298 figure continues next page 142 Peru Figure 8.3  continued f. External debt (left), government effectiveness (right) 2,000 2000 2013 Government effectiveness: percentile rank 80 550 30 2030 External debt (% of GNI) 160 10 50 2000 4 2013 10 1.3 .50 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.52*** + .14** ln(GNI pc); R = .035 2 ln(SDG) = 0 + .44*** ln(GNI pc); R 2 = .300 g. Education expenditure efficiency (left), health expenditure efficiency (right) 1 2010 Grigoli education efficiency score 2010 Grigoli health efficiency score 1 2030 .60 .40 .90 .30 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −4.0*** + .46*** ln(GNI pc); R 2 = .630 ln(SDG) = −.09*** + .01 ln(GNI pc); R 2 = .031 figure continues next page Peru 143 Figure 8.3  continued h. Public investment management index 4 Public investment management index 2010 2 1.3 .80 .30 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = −.43 + .12** ln(GNI pc); R 2 = .079 Note: Highlighted observations are for Peru in different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. policy discussions around the size of the govern- Among current receipts, tax revenues are ment that could be needed to fulfill the SDGs. within the expected range while net Overseas In terms of government spending in areas Development Assistance (ODA) is lower than that may support the SDG agenda, Peru spends expected, compared to other countries at the same as expected (compared to a typical country at income per capita level. As also shown in table 8.2, the same GNI per capita level) for total public cross-country patterns suggest that net ODA will investment, but below the expected level for decline as a percent of GDP (without changing total public consumption, primary education, significantly in per capita terms). The relationship secondary and secondary education, and between tax revenues and GNI per capita and health.13 Note that, for primary and secondary GNI per capita is not tight enough to project education, spending is measured in two ways: expected changes. Still, if recent changes to tax total as share of GDP and per student as share of rates are taken into account, it is likely that tax rev- GDP per capita; for all four indicators, spending enues will be below peers and, thus, become an is lower than expected—something that the important contributor of an increased fiscal authorities are well alert to, and efforts are effort.16 In addition, in comparison with OECD ongoing to create fiscal space for education. countries, Peru’s tax revenues are low. The public Fuel subsidies are the most obvious case of debt position is very robust due to commendable low-priority spending from the post-2015 macroeconomic as well as asset and liability man- agenda perspective.14 However, Peru’s spending agement policies.17 In 2014, Peru’s total public on fuel subsidies appears to be very low.15 Since debt stood at US$39 billion, or 20 percent of GDP, Peru might need to increase spending in key down from 42 percent in December 2005 and was SDG related areas, resources will need to be among the lowest relative to other middle-income mobilized either from spending reductions in countries in the LAC region. Debt reduction fol- non-SDG related areas, from higher revenues, lows a decade of high real GDP growth accompa- or from higher spending efficiency. nied by overall fiscal surpluses.18 144 Peru Enhancing government’s spending and over- 4. Conclusions all efficiency will increase the room for priority spending and enhance its impact on the SDG As summarized in table 8.3, Peru’s current SDG agenda. Table 8.2 displays data for some mea- outcomes are better than expected (compared to sures of government efficiency. According to a typical country at the same GNI per capita both the health and the education indexes used level) for three indicators, while it falls short in in this study, Peru’s performance is better than terms of poverty, shared prosperity, maternal expected; among these two indexes, GNI per mortality, malaria, access to improved water capita is strongly correlated with the education source, and road density. For the other eight index but largely uncorrelated with the health indicators, Peru’s current SDG outcomes are as index. Peru is also performing better than expected. By 2030, considerable improvements expected in terms of the more general Public are projected for most indicators if the country Investment Management Index and the World manages to sustain a pace of growth of 4.2 Bank Government Effectiveness indicator. Given percent on average for GNI per capita. For the that the different indexes measure different majority of SDG outcomes, this means the post- aspects of government performance, it seems 2015 global ambitions will be realized or close to reasonable to conclude that, at the aggregate being realized. level, it is unlikely that improvements in the Table 8.3 further shows that, for about half of efficiency of spending will generate significant ­ the SDG indicators, the relationship to GNI per fiscal space for resources to flow toward SDG- capita is tight and Peru is doing as expected or related spending. However, efficiency still needs better than expected. Improvements in these to be a central part of any public expenditure SDGs will most likely continue along with GNI strategy, and—as for any country—is expected to per capita growth and increases in resources and improve as GNI per capita improves. capabilities. Note, however, that, among these In sum, our cross-country results indicate that relatively successful SDGs, there are those government spending in terms of overall con- ­ indicators where Peru’s percentile ranking has sumption and investment is at or below peers. deteriorated despite its significantly improved Among the receipts, a comparison of Peru’s data to GNI per capita rank. For these SDGs, the effi- cross-country patterns combined with recent eco- ciency of policies through which resources are nomic studies do not single out any of the catego- translated into SDG outcomes may need special ries for which we have data (taxes, ODA, or attention. borrowing) as being easily tapped for additional The areas in which Peru currently is under- fiscal space. Any suggestions about fiscal policy performing relative to peer countries differ among adjustments would require additional ­ country​-​ each other in terms of how well they are explained ­ specific information that, from the perspective of by GNI per capita—that is, the “tightness” of their the SDG agenda, would permit assessments of the relationship to GNI per capita. The relationship is benefits and the costs of feasible changes in the relatively tight for poverty, ratio of female to male level and allocation of spending and taxation and/ secondary enrollment, maternal mortality, or point to areas for efficiency improvements. Such malaria incidence, access to improved water adjustments should be part of a broader strategy source, and road density. The presence of a tight for sustainable growth, poverty ­ reduction, and relationship to GNI per capita suggests that future shared prosperity.19 With that in mind, it is impor- improvements depend on a combination of sus- tant to remember the strong growth performance tained rapid growth, better policies, and a larger of Peru, and that many of the global SDG goals but well-focused fiscal effort. Peru is also under- seem within reach for Peru when based on busi- performing for an indicator with a weak relation- ness-as-usual GNI per capita projections—that is, ship to GNI per capita: the income share of the without any major adjustments in fiscal space. bottom 40 percent. The loose relationship to GNI Peru 145 Table 8.3  Peru—Summary Results for SDG Indicators Cross-country relationship with Overperforming As expected Underperforming GNI/cap Tight •• School enrollment, pre-primary (´) •• Primary completion (−) •• Poverty (–) •• School enrollment, secondary (−) •• Ratio of female to male primary •• Ratio of female to male •• Under-5 mortality (+) completion (−) secondary enrollment (×) •• Access to improved sanitation (−) •• Maternal mortality (+) •• Access to electricity (×) •• Malaria cases (–) •• Internet users (−) •• Access to improved water (–) •• CO2 emissions (−) •• Road density (+) Loose •• Secondary completion •• Income share for lowest 40% (×) •• HIV Prevalence (+) Note: (+) = larger country rank improvement 2000–12 than for GNI per capita; (−) = smaller country rank improvement (or deterioration) 2000–12 than for GNI per capita; (×) = the same country rank change 2000–12 as for GNI per capita (+/− 2 percentile points). A tight enough relationship is defined as an R  2 > 0.3 (tight) or 0.3 > R  2 > 0.1 (moderately tight), while R  2 > 0.1 are defined as loose. per capita suggests that this indicator should not the room to increase revenues is limited when be expected to improve strongly or systematically taking additional analyses into account. However, in response to more rapid growth in GNI per a detailed, forward-looking Public Expenditure capita; it would rather require targeted policy Review will be needed to provide better insights interventions. However, more rapid growth in this area to the country. Given this, policy would clearly promise to raise growth in the level directions for a future SDG agenda would have of per capita incomes for the bottom 40 percent. to be guided by more detailed country-specific With regard to the likely required fiscal effort information. However, even without any assump- to achieve the SDGs, our cross-country perspec- tions about major increases in fiscal effort, tive does not suggest any obvious priorities since Peru is projected to reach a number of the global spending is at or below the expected levels, SDG goals if it maintains the pace of strong spending efficiency is roughly as expected and ­economic growth. Annex 8A: Data Sources Indicator Source Comment GNI per capita (constant 2005 WDI. API ref: GNI per capita (constant 2005 US$) US$) [NY.GNP.PCAP.KD] SDGs Poverty, at $1.25 a day (PPP) WDI. API ref: poverty headcount ratio at $1.25 a day (PPP) (% of population) (% of population) [si.pov.dday] Shared prosperity: income WDI. API ref: income share held by lowest 20% + income share for lowest 40% share held by second 20% [SI.DST.FRST.20+SI.DST.02ND.20] Pre-primary enrollment WDI. API ref: school enrollment, pre-primary (% gross) (% gross) [SE.PRE.ENRR] Primary completion (% gross) WDI. API ref: primary completion rate, total (% of relevant age group) [se.prm.cmpt.Zs] Secondary enrollment WDI. API ref: school enrollment, secondary (% gross) (% gross) [se.sec.enrr] Secondary completion EdStat. API ref: DHS: secondary completion rate (% gross) [hh.dhs.scr] annex continues next page 146 Peru Indicator Source Comment Primary completion, ratio of WDI. API ref: primary completion rate, female (% of relevant females to males (%) age group)/primary completion rate, male (% of relevant age group) *100 [SE.PRM.CMPT.FE.ZS /SE.PRM.CMPT.MA.ZS*100] Secondary enrollment, ratio of WDI. API ref: ratio of female to male secondary enrollment (%) females to males (%) [se.enr.seco.fm.zs] Under-5 mortality (per 1,000 WDI. API ref: mortality rate, under-5 (per 1,000 live births) live births) [SH.DYN.MORT] Maternal mortality (per WDI. API ref: maternal mortality ratio (modeled estimate, 100,000 live births) per 100,000 live births) [sh.sta.mmrt] Malaria cases (% of HNP. API ref: malaria cases reported/population, total *100 population) [sh.sta.malr/SP.POP.TOTL *100] HIV prevalence (% of WDI. API ref: prevalence of HIV, total (% of population ages population ages 15–49) 15–49) [sh.dyn.aids.zs] Access to improved sanitation WDI. API ref: improved sanitation facilities (% of population facilities (% of population) with access) [sh.sta.acsn] Access to improved water WDI. API ref: improved water source (% of population with source (% of population) access) [sh.h2o.safe.zs] Road density (km per 100 sq. WDI. API ref: road density (km of road per 100 sq. km of land km of land) area) [is.rod.dnst.k2] Access to electricity WDI. API ref: access to electricity (% of population) (% of population) [eg.elc.accs.Zs] Internet users (per 1,000 WDI. API ref: Internet users (per 100 people) [IT.NET.USER.P2] people) CO2 emissions (metric tons WDI. API ref: CO2 emissions (metric tons per capita) per capita) [en.atm.co2e.pc] Fiscal space indicators Investment (% of GDP) WDI. API ref: gross fixed capital formation (% of GDP)-gross fixed capital formation, private sector (% of GDP) [ne.gdi.ftot.Zs]-[ne.gdi.fprv.zs] Consumption (% of GDP) WDI. API ref: general government final consumption expenditure (% of GDP) [NE.CON.GOVT.ZS] Primary education (% of GDP) EdStat. API ref: total expenditure on educational institutions and administration as a % of GDP. All sources. Primary [UIS.XGDP.1.FDINSTADM.FFD] Secondary education EdStat. API ref: total expenditure on educational institutions (% of GDP) and administration as a % of GDP. All sources. Secondary and post-secondary non-tertiary [UIS.XGDP.234.FDINSTADM.FFD] Primary education, per student WDI. API ref: expenditure per student, primary (% of GDP per capita) (% of GDP per capita) [se.xpd.prim.pc.zs] Secondary education, per WDI. API ref: expenditure per student, secondary student (% of GDP per capita) (% of GDP per capita) [se.xpd.seco.pc.zs] Health (% of GDP) WDI. API ref: health expenditure, public (% of GDP) [sh.xpd.publ.zs] Fossil fuel subsidy (% of GDP) IMF (2013a). Total pretax subsidy (% of GDP) Subsidies are measured using a price-gap approach capturing both consumer (including implicit) and producer (except those that arise when suppliers are inefficient and make losses at benchmark prices) subsidies. Negative external effects are not included in the pretax subsidy. Tax revenue (% of GDP) WDI. API ref: tax revenue (% of GDP) [gc.tax.totl.gd.Zs] Net ODA (% of GNI) WDI. API ref: net ODA received (% of GNI) [dt.oda.odat.gn.zs] External debt (% of GNI) WDI. API ref: external debt stocks (% of GNI) [DT.DOD.DECT.GN.ZS] annex continues next page Peru 147 Indicator Source Comment Government effectiveness: WGI. API ref: government effectiveness: percentile rank Captures perceptions of the quality of percentile rank [ge.per.rnk] public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies. Grigoli education efficiency Grigoli (2014) Measure secondary education score inefficiency in terms of how much additional output could be achieved at current levels of spending. Grigoli health efficiency score Grigoli and Kapsoli (2013) Quantifies the inefficiency of public health expenditure using a stochastic frontier model that controls for the socioeconomic determinants of health. Public investment Dabla-Norris et al. (2011) Four phases associated with public management index investment management are covered: project appraisal, selection, implementation, and evaluation. Note: PPP = purchasing power parity; WDI = World Development Indicators, World Bank; API ref = reference and code when using the World Bank Open Data; EdStats = Education statistics, World Bank; HNP = Health Nutrition and Population statistics, World Bank; WGI = Worldwide Governance Indicators, World Bank. Notes 1. While a cross-country perspective provides an 5. Given that (a) SDGs have extreme values (such as important complement to analysis that is centered 100 percent for improved water access) and (b) on an individual country, it is by definition lim- the current SDG level never is exactly as expected ited to analysis of variables that are available in given GNI per capita, the projected values grad- cross-country databases. ually converge toward the expected values. For 2. This does not mean that GNI per capita is viewed example, for a country that overperforms in water as a direct determinant of SDG outcomes; on the access, as GNI per capita increases the extent of contrary, a major challenge for policy makers is to overperformance gradually declines, so that when identify policies that improve SDG performance the expected value is 100, overperformance has relative to what is expected given the level of GNI reached zero. per capita. A second challenge is to raise growth in 6. A “tight” enough relationship is defined as one GNI per capita as it indirectly influences country with  an R2 greater than 0.3, while a “moderately SDG capacity. tight” is defined as one with an R2 between 0.3 3. Sources for the indicators are presented in the and 0.1, and a “loose” ­relationship as one with an R2 table in the annex 8A. smaller than 0.1. 4. Projections from CEPII (v 2.3) are used for this 7. In addition, the confidence interval is wide in the and other Country Development Diagnostics case of a loose relationship, suggesting that any applications given their wide country coverage conclusion on over- or underperformance is made and well-documented methodology; OECD data with wide margins. Statistically, even though their have been used when projections have been miss- confidence intervals are wide, as long as the esti- ing. In the projections, it is assumed that future mated coefficient linking GNI per capita to the GNI growth will coincide with future GDP growth SDG indicator is nonzero, these values are closer (both expressed in constant 2005 US$) given that than the cross-country average to what is expected that this is the variable that CEPII and other for Senegal. The same observation applies to sources project. As a comparison, IMF (2015b, expected values for fiscal space indicators. p. 7) projects a 4.5 percent medium-term growth 8. With regard to CO2, Peru’s current and projected potential for real GDP in Peru by 2020. 2030 per capita emissions are 19.5 and 51.0 percent 148 Peru of the current OECD average. Also, with a rela- 15. There are no data on “pretax fuel subsidies” for Peru tively low population density, a low road density (there are no taxes on either petroleum products, may still reflect fair access to roads. coal, natural gas, or electricity), but in terms of “post- 9. The ranking is based on data from 2000 and tax subsidies,” which also measures the costs through 2012/2013, or the closest year with data (but only associated negative external effects, Peru’s level is if data no later than 1998 for “2000” or 2009 for significantly lower than expected (IMF 2013a). ­ “recent” exist). 16. The authorities reduced corporate and personal 10. In this note, “shared prosperity” is measured as income tax rates in 2014 to align rates in Peru with the income share of the poorest 40 percent of the the rest of the region and to support private invest- population. ment and growth. IMF (2015b, p. 28) projections 11. Note, however, IMF (2015b, p. 1) reports that suggest tax revenues will be 15.6 percent of GDP there are visible, absolute improvements in pov- by 2020, that is, lower than in 2013. erty reduction in Peru. 17. Peru is considered to be debt sustainable even when 12. The treatment is the same as in table 8.1 and looking at total (public and private) debt. IMF proj- related figures. That is, in table 8.2, projections ects an increase of total external debt from 27.4 are shown only when the cross-country rela- percent of GDP in 2012 to 31.6 percent in 2016, tionship between the indicator and GNI per followed by a decline to 26.7 percent in 2020 (IMF capita is considered tight enough. Due to data 2015b, p. 26). Such an increase in the debt stock limitations, we focus on government spending would still leave Peru below what is expected. indicators; country-specific analysis is needed 18. There has also been a change in the debt structure: to consider policy in the context of the different domestic currency denominated debt represents roles of the government and private services and more than half of the total public debt (up from spending. close to zero a decade earlier). Markets have rec- 13. Note that in Peru most major roads are under PPPs ognized the country’s prudent fiscal and macro- and, thus, are financed by users to a large extend economic policies by continuously upgrading (outside the budget framework, that is, through Peru’s sovereign debt ratings. concession contracts). Peru is the third-largest 19. In 2014, the government of Peru has implemented attractor of private capital to infrastructure in the a series of fiscal and structural packages, includ- world according to the WB PPI data base: http:// ing tax cuts, increases in fiscal spending, and ppi​.worldbank.org/. structural measures to support investment, con- 14. Fuel subsidies are detrimental to the climate and sumption, and growth. See IMF (2015b, annex) encourage technologies that create less employ- on measures taken by the government to promote ment for the growing labor force. financial and social inclusion in Peru. Peru 149 Chapter 9 The Philippines 1. Introduction require efficient and carefully prioritized public spending. Drawing on the information in these The Philippines is a lower middle-income ­country tables and figures, this note briefly (a) summarizes in Southeast Asia made up of more than 7,000 the Philippines’ SDG progress since 2000 and islands. Over time, the importance of services has projects expected values for 2030; and (b) assesses increased at the expense of agriculture and options for increasing fiscal space. Sections 2 and manufacturing, while remittances amount to ­ 3 address SDGs and fiscal space, respectively, 9.8 percent of GDP (in 2013, World Bank data).1 while findings are summarized in Section 4. During 2001–12, the Philippines’ average annual The analysis is done from a cross-country growth rate for GNI per capita (at constant 2005 perspective: for the different indicators, the US$) was 3.1 percent, which may be compared to Philippines’ performance and prospects are a developing (low- and ­ middle-income) country benchmarked relative to other countries, consid- average of 3.0 percent. During the same period, ering its past, recent, and projected levels of GNI the Philippines’ ranking according to the UNDP per capita.2 The latter variable tends to be highly Human Development Index (among countries correlated with most of the SDGs and most of the included both in 2000 and in 2012) deteriorated factors that determine their evolution; given from the 42nd to the 37th percentile. Although this, it is used as a summary indicator of country growth accelerated in the past decade, unemploy- capacity to provide and efficiently utilize inputs ment and poverty have been slow to decline. In that contribute to SDGs (for example, health and addition, problems of poor infrastructure, lim- education services) and to achieve SDG out- ited competition, and governance issues have cre- comes (like strong health and education results).3 ated a climate that is not conducive to productive sector investments but instead encourages over- seas employment that leads to remittance inflows 2. SDG Indicators: History and real exchange rate appreciation. This country-at-a-glance note is for most of and Projections the SDG indicators based on data covering the For selected SDG indicators, table 9.1 summa- period until 2012. It is designed to provide an ini- rizes data for the Philippines: historical evolution, tial picture of the challenges that the Post-2015 actual and expected values for a recent year, and agenda poses for the Philippines; its findings can- projected 2030 values.4 In figure 9.1, data for the not guide policy on their own but should be seen Philippines are shown in the context of the esti- as an input into policy discussions. The note may mated cross-country relationship between each also serve as the starting point for a more com- SDG indicator and GNI per capita. For the plete country development diagnostic as well as a Philippines, the projected average annual rate of more comprehensive country-focused analysis. GNI per capita growth is 3.5 percent.5 The pro- The note is built around tables and figures that jected SDG values reflect what can be expected provide data for a selection of SDG ­ target indica- given a country’s starting point, projected growth tors and indicators related to fiscal space—fiscal in GNI per capita, typical rates of progress accord- space matters since, while policy frameworks and ing to cross-country patterns, and a gradual con- the engagement of the private sector may vary vergence to close gaps between observed and widely, rapid progress on the SDG agenda will expected values.6 Projections of the SDG The Philippines 151 Table 9.1  The Philippines—SDG Indicators: Evolution since 2000 and Projections to 2030 Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Poverty and shared prosperity Poverty, at $1.25 a day 24.6 15.4 4.8 6.7 By 2030, eradicate extreme poverty for all people everywhere, (PPP) (% of ­population) currently measured as people living on less than $1.25 a day. Shared prosperity: income 14.1 15.3 16.0 — By 2030, progressively achieve and sustain income growth of share for lowest 40% the bottom 40 percent of the population at a rate higher than the national average. Education Pre-primary enrollment 25.5 51.5 34.5 64.9 By 2030, ensure that all girls and boys have access to quality (% gross) early childhood development, care, and pre-primary education so that they are ready for primary education. Primary completion 86.4 91.3 85.6 97.0 By 2030, ensure that all girls and boys complete free, equitable, (% gross) and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment 74.4 81.4 66.9 89.2 By 2030, ensure that all girls and boys complete free, equitable, (% gross) and quality primary and secondary education leading to relevant and effective learning outcomes. Primary completion, ratio of 106.5 102.0 97.2 — By 2030, ensure that all girls and boys complete free, equitable, females to males (%) and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment, 109.8 102.0 95.0 105.1 By 2030, ensure that all girls and boys complete free, equitable, ratio of females to and quality primary and secondary education leading to males (%) relevant and effective learning outcomes. Health Under-5 mortality (per 39.9 29.9 29.8 21.0 By 2030, end preventable deaths of newborns and children 1,000 live births) under 5 years of age. Maternal mortality (per 120.0 120.0 97.8 71.3 By 2030, reduce the global maternal ratio to less than 70 per 100,000 live births) 100,000 live births. Malaria cases (% of 0.05 0.01 0.05 0.00 By 2030, end the epidemics of AIDS, tuberculosis, malaria, and population) neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. Infrastructure Access to improved 65.5 74.3 54.7 84.2 By 2030, achieve access to adequate and equitable sanitation sanitation facilities (% of and hygiene for all and end open defecation, paying special population) attention to the needs of women and girls and those in vulnerable situations. Access to improved water 87.6 91.8 83.3 95.2 By 2030, achieve universal and equitable access to safe and source (% of population) affordable drinking water for all. Access to electricity (% of 71.3 83.3 57.8 92.5 By 2030, ensure universal access to affordable, reliable, and population) modern energy services. Internet users (per 1,000 2.0 37.0 19.4 — Significantly increase access to information and communications people) technology and strive to provide universal and affordable access to the Internet in least developed countries by 2020. Environment CO2 emissions (metric tons 0.9 0.9 1.2 2.3 Integrate climate change measures into national policies, per capita) strategies and planning. Memorandum item GNI per capita (constant 1,247 1,789 3,389 2005 US$) Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = currently significantly overperforming; red = currently significantly underperforming; black = performing as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). For Internet use, a projection is not made, despite a tight relationship, since the expected line shifts significantly even in the medium run due to external forces, such as technological development. Global ambition is from the Open Working Group (UN 2014). 152 The Philippines Figure 9.1  The Philippines—SDG Indicators (Log Scale) versus GNI per Capita in a Cross- Country Setting (Log Scale) a. Poverty (left), shared prosperity: income share of bottom 40% (right) 410 30 Shared prosperity: income share Poverty, at $1.25 a day (PPP) 90 2000 20 2012 (% of population) 20 for lowest 40% 2012 2030 2000 4 10 .90 9 .04 5 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 11.8*** + −1.3*** ln(GNI pc); R 2 = .487 ln(SDG) = 3.17*** + −.05** ln(GNI pc); R 2 = .046 b. Gross pre-primary enrollment (left), primary completion (right) 550 130 2009 2030 Pre-primary enrollment (% gross) Primary completion (% gross) 150 2030 90 1999 2009 40 60 2000 10 40 3 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.73 + .570*** ln(GNI pc); R 2 = .348 ln(SDG) = 3.38*** + .14*** ln(GNI pc); R 2 = .372 figure continues next page The Philippines 153 Figure 9.1  continued c. Gross secondary enrollment 190 Secondary enrollment (% gross) 2012 2030 110 60 1999 30 20 6 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = 1.9*** + .31*** ln(GNI pc); R 2 = .558 d. Ratio of female to male primary completion (left), ratio of female to male secondary enrollment (right) 140 140 2030 2012 Secondary enrollment, ratio of Primary completion, ratio of 110 1999 1999 100 females to males (%) females to males (%) 90 2012 70 70 50 40 30 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 4.31*** + .04*** ln(GNI pc); R = .093 2 ln(SDG) = 3.82*** + .1*** ln(GNI pc); R 2 = .297 figure continues next page 154 The Philippines Figure 9.1  continued e. Under-5 mortality (left), maternal mortality (right) 500 2,000 Under−5 mortality (per 1,000 live births) Maternal mortality (per 100,000 230 470 110 2013 100 2030 live births) 2000 50 2000 20 2013 20 2030 5 1 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 7.99*** + −.61*** ln(GNI pc); R 2 = .560 ln(SDG) = 11.3*** + −.9*** ln(GNI pc); R 2 = .519 f. Malaria cases 30 Malaria cases (% of population) 1.3 .05 2000 2012 2030 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = 14.5*** + −2.3*** ln(GNI pc); R 2 = .412 figure continues next page The Philippines 155 Figure 9.1  continued Access to improved sanitation facilities g. Access to improved sanitation (left), access to improved water source (right) 2000 2012 2030 100 Access to improved water source 2030 100 2000 2012 70 (% of population) (% of population) 60 50 30 40 20 7 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .61** + .45*** ln(GNI pc); R 2 = .535 ln(SDG) = 3.45*** + .13*** ln(GNI pc); R 2 = .436 h. Access to electricity 400 Access to electricity (% of population) 2000 2010 2030 100 30 6 1.6 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = .29 + .5*** ln(GNI pc); R 2 = .458 figure continues next page 156 The Philippines Figure 9.1  continued i. Internet users (left), CO2 emissions (right) 20 CO2 emissions (metric tons per capita) 0 80 Internet users (per 1,000 people) 2013 3 2000 2030 .80 2010 6 .20 2000 .0 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −2.9*** + .78*** ln(GNI pc); R 2 = .568 ln(SDG) = −8.6*** + 1.16*** ln(GNI pc); R 2 = .677 Note: Highlighted observations are for Philippines at different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. indicators are presented only when the The  country falls short of expectations for the cross-country relationship between the indicator remaining 2 indicators; poverty and, although and GNI per capita is classified as tight.7 A loose only slightly, for maternal mortality. While relationship suggests that progress in the indica- underperformance for an indicator may be due tor is primarily a reflection of country-specific to country-specific conditions that are difficult factors and that it should not be expected to to change, it may often point to areas in which respond strongly or systematically to changes in payoffs from feasible policy change are relatively GNI per capita. When the relationship to GNI high, a possibility that calls for further analysis. per capita is loose the coefficients are typically Figure 9.2 shows that, between 2000 and 2012, small (in absolute terms); given this, the the Philippines’ ranking among low- and “expected” values for a recent year are close to the middle-income countries deteriorated by 5 per- ­ average for all low- and middle-income centile points in terms of GNI per capita. During countries.8 the same period, the Philippines’ ranking rose In sum, the Philippines’ current outcomes are for  4 of the selected SDGs: shared prosperity, better than expected (compared to a typical pre-primary school enrollment, access to elec- country at the same GNI per capita level) for tricity, and CO2 emissions; the ranking decline 10 indicators (pre-primary enrollment, primary for malaria was with less than for GNI per capita. completion, secondary enrollment, ratio of The ranking for another 5 indicators (primary female to male primary completion, ratio of completion, under-5 mortality, maternal mortal- female to male secondary enrollment, malaria, ity, access to improved sanitation f ­ acilities, access access to improved water source, access to to improved water source) deteriorated to roughly improved sanitation, access to electricity, the same extent as GNI per capita, while it Internet users, and CO2 emissions).9,10,11 For 2 of dropped even further for the remaining 5 indica- the indicators (under-5 mortality and shared tors (poverty, secondary enrollment, ratio of prosperity), current outcomes are as expected. female to make primary completion, ratio of The Philippines 157 Figure 9.2  The Philippines—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012 GNI per capita (constant 2005 US$) 55 50 Poverty, at $1.25 a day (PPP) (% of population) 46 36 Shared prosperity: Income share for lowest 40% 33 48 Pre-primary enrollment (% gross) 44 54 Primary completion (% gross) 47 43 Secondary enrollment (% gross) 66 50 Primary completion, ratio of females to males (%) 92 73 Secondary enrollment, ratio of females to males (%) 84 64 Under-5 mortality (per 1,000 live births) 59 52 Maternal mortality (per 100,000 live births) 59 52 Malaria cases (% of population) 72 70 Access to improved sanitation facilities (% of population) 57 52 Access to improved water source (% of population) 61 55 Access to electricity (% of population) 50 55 Internet users (per 1,000 people) 71 63 CO2 emissions (metric tons per capita) 51 55 0 10 20 30 40 50 60 70 80 90 100 2000 Recent Note: A high ranking signals strong performance; for the underlying indicator, this may correspond to a relatively high value (for example, for the secondary enrollment rate) or a relatively low value (for example, for the poverty rate). The ranking for an indicator is not reported if the available sample is less than 20. Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however the data are never older than 1998 for “2000” or 2009 for “recent.” Country-specific data years can be found in the respective graphs. female to male secondary enrollment, and By 2030, considerable improvements are pro- Internet use). jected for most indicators (see table 9.1 and respec- When comparing the results from regressions tive graphs in figure 9.1). Primary completion, on GNI per capita to changes in percentile rank- ratio of female to male primary completion, ratio ings, a few patterns emerge. First of all, for the of female to male secondary enrollment, maternal majority of the indicators, the Philippines is per- mortality, malaria, and access to improved water forming better than expected, but, since 2000, source are all projected to either realize or get close progress has been much weaker than for GNI per to realizing the post-2015 global ambition (shown capita in terms of percentile ranking for several in the last column of table 9.1). However, for other indicators: secondary enrollment, ratio of female SDGs to get closer to the realization of these ambi- to male primary completion, ratio of female to tions, a break with the past is needed. This is also male secondary enrollment, and Internet use. true for indicators, such as shared prosperity, for The SDG that has improved its ranking the most which a weak relationship with GNI per capita is shared prosperity, for which the Philippines precludes projections. Such a break would be facil- now is performing as expected. However, for itated by a combination of more rapid growth poverty, underperformance is combined with a (beyond the projected average growth of performance that was weaker than GNI per 3.5  percent) and improvements in policies that capita in terms of changes in percentile ranking. directly influence different SDGs. 158 The Philippines 3. Fiscal Space values for SDG indicators). The variables cover three aspects of government activities: spending, In most countries, accelerated progress on the receipts and debt, and governance and efficiency. SDG agenda will require efficient and growing In terms of government spending in areas that public spending in prioritized areas, most impor- may support the SDG agenda, the Philippines is tantly human development and infrastructure. below expected levels (compared to a typical coun- Private spending is also of crucial importance, try at the same GNI per capita level) for all reported both household spending on SDG-related ser- categories; total public investments, total public vices and business investments in a wide range of consumption, secondary education, and health, areas (including but not limited to infrastruc- except for primary education where spending is as ture).12 With regard to the Philippine fiscal space expected.14,15 However, when measuring education indicators, table 9.2 and figure 9.3 summarize the spending per student (as share of GDP per capita) historical evolution, actual and expected recent then spending is lower than expected for both pri- values, and, when relevant, projected values.13 mary and secondary education. Fuel subsidies are When the relationship is loose, projections are not the most obvious case of low-priority spending made and the expected value is in practice close to from the post-2015 agenda perspective; however, the average for the sample of all low- and middle-­ for the Philippines, the limited data that are avail- income countries (cf. discussion of expected able suggest that they are quite insignificant.16 Table 9.2  The Philippines—Fiscal Space: Revenue, Spending, and Government Efficiency Actual Expected Projection Indicator 2000 Recent Recent 2030 Government spending Consumption (% of GDP) 11.4 11.1 14.5 — Investments (% of GDP) 6.2 2.4 6.1 — Primary education (% of GDP) 2.0 1.6 1.6 — Secondary education (% of GDP) 0.9 1.3 — Primary education per student (% of GDP per capita) 12.0 11.8 13.8 — Secondary education per student (% of GDP per capita) 10.3 14.8 17.6 — Health (% of GDP) 1.5 1.7 3.1 — Government receipts and debts tax revenue (% of GDP) 12.8 12.9 14.7 — Net ODA (% of GNI) 0.60 0.06 1.70 0.03 External debt (% of GNI) 61.6 18.6 36.4 — Governance and government efficiency Government effectivness: percentile rank 50.2 56.9 28.5 62.0 Grigoli education efficiency score 0.72 0.56 0.85 Grigoli health efficiency score 0.96 0.95 — Public investment management index 1.85 1.65 — Memorandum item GNI per capita (constant 2005 US$) 1,247 1,789 3,389 Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = current value significantly above the expected level; red = current value significantly below the expected level; black = current value as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). The Philippines 159 Figure 9.3  The Philippines—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting a. Government spending: consumption (left), investment (right) 130 50 60 Consumption (% of GDP) 20 Investments (% of GDP) 30 9 2000 10 2000 4 2013 2013 2 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.01*** + .09** ln(GNI pc); R = .042 2 ln(SDG) = 2.34*** + −.07 ln(GNI pc); R 2 = .012 b. Government spending: primary education, per student (left), secondary education, per student (right) 90 50 Secondary education, per student Primary education, per student 30 50 (% of GDP per capita) (% of GDP per capita) 20 20 2000 8 2013 2013 10 2000 2 4 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.72*** + .12** ln(GNI pc); R = .067 2 ln(SDG) = 3.51*** + –.09 ln(GNI pc); R 2 = .030 figure continues next page 160 The Philippines Figure 9.3  continued c. Government spending: primary education (left), secondary education (right) 6 4 Secondary education (% of GDP) 4 Primary education (% of GDP) 2 2000 2 1.2 2013 1.1 2013 .60 .40 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.07*** + −.08 ln(GNI pc); R = .029 2 ln(SDG) = −.79** + .14*** ln(GNI pc); R 2 = .086 d. Government spending: health 20 Health (% of GDP) 6 1.6 2012 2000 .50 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = −.05 + .16*** ln(GNI pc); R 2 = .093 figure continues next page The Philippines 161 Figure 9.3  continued e. Tax revenue (left), official development aid (right) 60 40 7 Tax revenue (% of GDP) 30 Net ODA (% of GNI) 20 1.2 2000 10 2000 2012 .20 2013 2030 3 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.75*** + .13 ln(GNI pc); R 2 = .027 ln(SDG) = 8.26*** + −1.0*** ln(GNI pc); R 2 = .308 f. External debt (left), government effectiveness (right) 2,000 Government effectiveness: Percentile rank 80 2013 2030 2000 550 30 External debt (% of GNI) 160 2000 10 50 4 10 2013 1.3 .50 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.51*** + .14** ln(GNI pc); R 2 = .035 ln(SDG) = 0 + .44*** ln(GNI pc); R 2 = .300 figure continues next page 162 The Philippines Figure 9.3  continued g. Education expenditure efficiency (left), health expenditure efficiency (right) 1 Grigoli education efficiency score Grigoli health efficiency score 2030 2010 1 2010 .60 .40 .90 .30 .10 .80 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −4.0*** + .46*** ln(GNI pc); R 2 = .627 ln(SDG) = −.09*** + .01 ln(GNI pc); R 2 = .031 h. Public investment management index Public investment management index 4 2 2010 1.3 .80 .30 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = −.43 + .12** ln(GNI pc); R 2 = .079 Note: Highlighted observations are for the Philippines in different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. The Philippines 163 Among current receipts, taxes are within the In sum, the most obvious source of additional expected ranges, while net Official Development fiscal space is a higher tax intake. Depending on Assistance (ODA) is lower than expected. As fur- the future decisions of the government, donors, ther shown in table 9.2, cross-country patterns and the actors of the international financial sys- suggest that net ODA will decline as a percent of tem, increases in ODA and/or foreign borrowing GDP (without changing significantly in per capita may also be forthcoming. Opportunities to terms); however, for the Philippines, the level is improve government efficiency should be pur- already so low (0.06 percent of GNI in 2013) that sued, in particular given that higher taxes in this is of no significance.17 However, while the themselves typically would have a negative cross-country relationship is tight, this does not impact on most SDGs. Nevertheless, the fact mean that an increase in ODA is excluded for the that, from a cross-country perspective, govern- Philippines: it depends on the priorities of donors ment spending is relatively low, both in general and their relationships with the government. The and in priority areas, while efficiency indicators relationship between tax revenues and GNI per are relatively strong, suggests that additional pri- capita, as well as the debt stock and GNI per oritized spending would yield positive payoffs. capita, is not tight enough to project expected However, decisions about the level and alloca- changes. However, the current intention in the tion of government spending should be made in Philippines is to expand tax revenues.18 Higher light of government priorities and additional taxes or lower subsidies would both reduce the data that this note does not consider, including resources controlled by domestic households and information about government capacity to effi- firms, pointing to the need to consider the com- ciently expand activities in different areas and bined impact on SDGs and other indicators from the scope to encourage complementary pri- higher taxes and the spending increases that are vate  sector activities. Remittances from the financed by these taxes. The external debt stock is Philippines’ large stock of workers abroad, which below the expected level. However,  the current amounted to 9.8 percent of GDP in 2013 (World government goal is to lower borrowing further.19 Bank), is one example of activities outside the Government efficiency is important to protect government that may still have a strong impact and, if possible, increase in order to add room for on the SDG agenda. More broadly, from the per- priority spending and enhance its impact on the spective of this agenda and given strong linkages SDG agenda. Table 9.2 displays data for some between private and government activities and measures of government efficiency. According to incomes, it is crucial that policies and spending both the health and the education indexes used decisions promote a broad-based change that in this study, the Philippines’s performance is encompasses services related to human develop- above the expected levels; among these two ment, infrastructure investments, and other indexes, GNI per capita is strongly correlated measures in support of strong long-run growth with the education index but largely uncorrelated that is biased in favor of the less advantaged. with the health index. The Philippines is also per- forming better than expected according to the more general World Bank Government Effectiveness indicator, and as expected in terms 4. Conclusions of the Public Investment Management Index. As summarized in table 9.3, the current out- That is, according to these indicators, govern- comes of the Philippines are better than expected ment efficiency appears to be better than expected (compared to a typical country at the same GNI given the GNI per capita of the Philippines, per capita level) for most of the selected SDG an observation that does not negate that consid- indicators. For other indicators, they are as erable efficiency gains still may be feasible in expected (under-5 mortality and shared prosper- ­different areas. ity) or below expectations as is the case for 164 The Philippines poverty (using data until 2012) and, to a smaller future improvements depend on a combination extent, for maternal mortality. By 2030, consider- of rapid growth and better policies. able improvements are projected for most indi- Special attention to policies is needed when cators and, for several of these, post-2015 global the relationship between the SDG and GNI per ambitions are realized or close to being realized. capita is loose, which is the case for the ratio However, for others, such as poverty, a break female to male primary completion and shared with the past is needed. prosperity. Their loose relationships to GNI per Table 9.3 further shows that for most indica- capita suggest that these two indicators should tors, the relationship to GNI per capita is tight. not be expected to improve strongly or systemat- Improvements in these SDGs will likely continue ically to more rapid growth in GNI per capita but along with GNI per capita growth and increases rather primarily depend on country-specific in resources and capabilities. Note, however, that, conditions and policies. for some of these SDGs, the Philippines’ per- Given relatively low levels of government centile ranking has deteriorated more than for spending, both overall and in SDG areas with GNI per capita, suggesting that the efficiency of data, and levels of government efficiency that policies through which resources are translated seem to be relatively good, increased prioritized into SDG outcomes may be lagging. One indica- government spending may yield positive payoffs tor standing out is poverty; not only has the if sufficiently high efficiency can be maintained. country’s percentile ranking deteriorated since The main source of additional fiscal space may be 2000 but the outcome is now weaker than higher taxes, even though other sources, such expected given GNI per capita. The presence of a as increased ODA and efficiency improvements, tight relationship to GNI per capita suggests that also may contribute. Table 9.3  The Philippines—Summary of Results for SDG Indicators Cross-country relationship with Overperforming As expected Underperforming GNI/cap Tight •• Primary completion (×) •• Under-5 mortality (×) •• Poverty (−) •• Secondary school enrollment (−) •• Maternal mortality (×) •• Pre-primary school enrollment (+) •• Ratio of female to male secondary enrollment (−) •• Malaria (+) •• Access to improved water source (×) •• Access to improved sanitation (×) •• Access to electricity (+) •• Internet users (−) •• CO2 emissions (+) Loose •• Ratio of female to male primary •• Shared prosperity (+) completion (−) Note: (+) = larger country rank improvement (or smaller drop) 2000–12 than for GNI per capita; (−) = larger drop 2000–12 than for GNI per capita; (×) = the same country rank change 2000–12 as for GNI per capita. The Philippines 165 Annex 9A: Data Sources Indicator Source Comment GNI per capita (constant WDI. API ref: GNI per capita (constant 2005 US$) 2005 US$) [NY.GNP.PCAP.KD] Sustainable development goals indicators Poverty, at $1.25 a day (PPP) WDI. API ref: poverty headcount ratio at $1.25 a day 2012 data for the Philippines from the World (% of population) (PPP) (% of population) [SI.POV.DDAY] Bank country team. Shared prosperity: income WDI. API ref: income share held by lowest 20% + income share for lowest 40% share held by second 20% [SI.DST.FRST.20+SI. DST.02ND.20] Pre-primary enrollment WDI. API ref: school enrollment, pre-primary (% gross) (% gross) [SE.PRE.ENRR] Primary completion WDI. API ref: primary completion rate, total (% of relevant (% gross) age group) [SE.PRM.CMPT.ZS] Secondary enrollment WDI. API ref: school enrollment, secondary (% gross) 2012 data for the Philippines form DepEd (% gross) [SE.SEC.ENRR] BEIS. Secondary completion EdStats. API ref: DHS: secondary completion rate (% gross) [HH.DHS.SCR] Primary completion, ratio of WDI. API ref: primary completion rate, female (% of 2012 data for the Philippines form DepEd females to males (%) relevant age group)/primary completion rate, male (% of BEIS. relevant age group) *100 [SE.PRM.CMPT.FE.ZS /SE.PRM.CMPT.MA.ZS*100] Secondary enrollment, ratio WDI. API ref: ratio of female to male secondary enrollment 2012 data for the Philippines form DepEd of females to males (%) (%) [SE.ENR.SECO.FM.ZS] BEIS. Under-5 mortality (per 1,000 WDI. API ref: mortality rate, under-5 (per 1,000 live births) live births) [SH.DYN.MORT] Maternal mortality (per WDI. API ref: maternal mortality ratio (modeled estimate, 100,000 live births) per 100,000 live births) [SH.STA.MMRT] Malaria cases (% of HNP. API ref: malaria cases reported/population, population) total *100 [SH.STA.MALR / SP.POP.TOTL *100] HIV Prevalence (% of WDI. API ref: Prevalence of HIV, total (% of population population ages 15–49) ages 15–49) [SH.DYN.AIDS.ZS] Access to improved WDI. API ref: improved sanitation facilities (% of sanitation facilities (% of population with access) [SH.STA.ACSN] population) Access to improved water WDI. API ref: improved water source (% of population with source (% of population) access) [SH.H2O.SAFE.ZS] Road density (km per 100 WDI. API ref: road density (km of road per 100 sq. km of sq. km of land) land area) [IS.ROD.DNST.K2] Access to electricity (% of WDI. API ref: access to electricity (% of population) population) [EG.ELC.ACCS.ZS] Internet Users (per 1,000 WDI. API ref: Internet users (per 100 people) people) [IT.NET.USER.P2] CO2 emissions (metric tons WDI. API ref: CO2 emissions (metric tons per capita) per capita) [EN.ATM.CO2E.PC] Fiscal space indicators Investment (% of GDP) WDI. API ref: gross fixed capital formation (% of GDP)-gross fixed capital formation, private sector (% of GDP) [NE.GDI.FTOT.ZS]-[NE.GDI.FPRV.ZS] Consumption (% of GDP) WDI. API ref: general government final consumption expenditure (% of GDP) [NE.CON.GOVT.ZS] Primary education (% of EdStats. API ref: total expenditure on educational GDP) institutions and administration as a % of GDP. All sources. Primary [UIS.XGDP.1.FDINSTADM.FFD] annex continues next page 166 The Philippines Indicator Source Comment Secondary education EdStats. API ref: total expenditure on educational (% of GDP) institutions and administration as a % of GDP. All sources. Secondary and post-secondary non-tertiary [UIS.XGDP.234.FDINSTADM.FFD] Primary education, per WDI. API ref: expenditure per student, primary (% of GDP student (% of GDP per per capita) [SE.XPD.PRIM.PC.ZS] capita) Secondary education, per WDI. API ref: expenditure per student, secondary student (% of GDP per (% of GDP per capita) [SE.XPD.SECO.PC.ZS] capita) Health (% of GDP) WDI. API ref: Health expenditure, public (% of GDP) [SH.XPD.PUBL.ZS] Fossil fuel subsidy (% of IMF (2013a). Total pretax subsidy (% of GDP) Subsidies are measured using a price-gap GDP) approach capturing both consumer (including implicit) and producer (except those that arise when suppliers are inefficient and make losses at benchmark prices) subsidies. Negative external effects are not included in the pretax subsidy. Tax revenue (% of GDP) WDI. API ref: tax revenue (% of GDP) [GC.TAX.TOTL.GD.ZS] Net ODA (% of GNI) WDI. API ref: net ODA received (% of GNI) 2013 data for the Philippines from the World [DT.ODA.ODAT.GN.ZS] Bank country team. External debt (% of GNI) WDI. API ref: external debt stocks (% of GNI) [DT.DOD.DECT.GN.ZS] Government effectiveness: WGI. API ref: government effectiveness: percentile rank Captures perceptions of the quality of public percentile rank [GE.PER.RNK] services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies. Grigoli education efficiency Grigoli (2014) Measure secondary education inefficiency in score terms of how much additional output could be achieved at current levels of spending. Grigoli health efficiency Grigoli and Kapsoli (2013) Quantifies the inefficiency of public health score expenditure using a stochastic frontier model that controls for the socioeconomic determinants of health. Public investment Dabla-Norris et al. (2011) Four phases associated with public management index investment management are covered: project appraisal, selection, implementation, and evaluation. Note: WDI = World Development Indicators, World Bank; API ref = reference and code when using the World Bank Open Data; EdStats = Education statistics, World Bank; HNP = Health Nutrition and Population statistics, World Bank; WGI = Worldwide Governance Indicators, World Bank. Notes 1. In 2012, the GDP shares for agriculture, industry, ­ olicy makers is to identify policies that improve p and services were at 12, 31, and 57 percent, respec- SDG performance relative to what is expected tively (WDI, World Bank). given the level of GNI per capita. A second chal- 2. While a cross-country perspective provides an lenge is to raise growth in GNI per capita as it indi- important complement to analysis that is centered rectly influences country SDG capacity. on an individual country, it is by definition lim- 4. Sources for the indicators are presented in ited to analysis of variables that are available in annex 9A. cross-country databases. 5. Projections from CEPII are used for this and 3. This does not mean that GNI per capita is viewed other Country Development Diagnostics appli- as a direct or the only determinant of SDG out- cations given their wide country coverage and comes; on the contrary, a major challenge for well-documented methodology; OECD data The Philippines 167 have been used when projections have been rather improvements in how the money is spent missing. In the projections, it is assumed that (perhaps via a change in the policy mix). future GNI growth will coincide with future GDP 13. The treatment is the same as in table 9.1 and related growth (both expressed in constant 2005 US$) figures. That is, in table 9.2, projections are shown given that that this is the variable that CEPII and only when the cross-country relationship between other sources project. the indicator and GNI per capita is considered 6. Given that (a) SDGs have extreme values (such tight enough. Due to data limitations, we focus on as  100 percent for improved water access) and government spending indicators; country-­ specific (b)  the current SDG level never is exactly as analysis is needed to consider policy in the context expected given GNI per capita, the projected of the different roles of the government and pri- values gradually converge toward the expected ­ vate services and spending. values. For example, for a country that over- ­ 14. Interest costs have tended to absorb a sizable share performs in water access, as GNI per capita of revenue, compressing room to spend on health, increases the extent of over performance gradu- education, and infrastructure (IMF 2013, p. 12). ally declines, so  that when the expected value is 15. The Philippines government health expenditure is 100, over-­ performance has reached zero. now considerably above that which is recorded in 7. A tight enough relationship is defined as an latest WDI data, reflecting the fact that the bud- R2 > 0.3 (tight) or 0.3 > R2 > 0.1 (moderately tight), get of the Philippine Department of Health (at the while R2 < 0.1 are defined as loose. national level) almost doubled between 2012 and 8. In addition, the confidence interval is wide in 2014. When government national and local level the case of a loose relationship, suggesting that health spending is added, this translates into a any conclusion on over- or underperformance large increase in overall health spending. is made  with wide margins. Statistically, even 16. Fuel subsidies are detrimental to the climate and though their confidence intervals are wide, as long encourage technologies that are less labor inten- as the estimated coefficient linking GNI per capita sive, tending to generate employment for fewer to the SDG indicator is nonzero, these values are workers at lower wages. Pretax spending is not closer than the cross-country average to what is available for the Philippines but post tax fuel sub- expected for a specific country. The same obser- sidies (also including negative external effects) is vation applies to expected values for fiscal space currently at 1.1 percent of GDP, which is lower indicators. than expected. 9. These data are from 2009. However, the lat- 17. Net ODA decreased from about 3 percent in the est trend is in line with the better than expected beginning of the 1990s to levels below 0.5 percent result. From 2013—through the Early Years Act— starting from the mid-2000s. completion of kindergarten has become a require- 18. IMF (2013, p. 12) suggests that the Philippines ment for a student to enter Grade 1; hence, the tax revenues may be increased from its 2012 level national statistics recorded a substantial increase of 12.8 percent of GDP to 16.0 percent by 2016. in pre-primary enrollment in that year. However, This is in line with the government’s own develop- while these data are not computed as gross enroll- ment plan for 2011–16, where the overall strategy ment but rather net enrollment (for 5 year old in the fiscal sector is to increase the tax effort to children), as of 2013, the enrollment for kinder- reach 16.1 percent by 2016 (NEDA 2014). Such an garten for 5-year-old children has shot up to 77.4 increase would still leave the Philippines within its percent. expected range. 10. Note that in the Philippines, there is a problem 19. The 2013 IMF-World Bank Debt Sustainability with males falling significantly behind females in Analysis (DSA) (IMF 2013, annex 48) for the terms of education. Hence, “better than expected” Philippines finds the outlook for public debt results for the ratio of female to male primary dynamics favorable. Public debt is projected to completion and the ratio of female to male sec- continue a declining trend to 42 percent of GDP ondary enrollment may be misleading since it by 2017, which for the total external debt means a reflects a problematic gender disparity. decline from 34.1 percent of GDP in 2011 to 22.3 11. With regard to CO2, the Philippines’s current and percent in 2017 (IMF 2013, p. 37). This intention is projected 2030 per capita emissions are 8.6 and confirmed in NEDA (2014), which states that the 35.8 percent of the current OECD average. government aims to further decrease the debt to 12. There may also be cases where the solution to the 43.4 percent of GDP in 2016, from 45.8 percent as low level of SDG is not a spending increase but of June 2013. 168 The Philippines Chapter 10 Senegal 1. Introduction The analysis is done from a cross-country per- spective: for the different indicators, Senegal’s per- Senegal is a lower middle-income country in West formance and prospects are benchmarked relative Africa. Since 2006, growth has been slow partly to other countries, considering its past, recent, and because of a series of exogenous shocks, including projected levels of GNI per capita.1 The latter vari- spikes in food and fuel prices, the global financial able tends to be highly correlated with most SDGs crisis, regional droughts and floods, and, more and most of the factors that determine their evolu- recently, the spillovers from Ebola. During 2001–12, tion; given this, it is used as a summary indicator of Senegal’s average annual growth rate for GNI per country capacity to provide and efficiently utilize capita (at constant 2005 US$) was as low as 1.1 inputs that contribute to SDGs (for example, health percent, compared to a developing (low- and mid- and education services) and to achieve SDG out- dle-income) country average of 3.0 percent. During comes (like strong health and education results).2 the same period, Senegal’s ranking according to the UNDP Human Development Index (among coun- tries included both in 2000 and in 2012) was 2. SDG Indicators: History and unchanged at the 13th percentile). While there are numerous development constraints for Senegal, a Projections few stand out; the fiscal deficit, bottlenecks in For selected SDG indicators, table 10.1 summa- transport and energy, a poor business climate, stag- rizes data for Senegal: historical evolution, actual nation in ­ traditional exports, and inefficient gov- and expected values for a recent year, and ernment spending (IMF 2015c, p. 5). projected 2030 values.3 In figure 10.1, data for ­ This country-at-a-glance note is designed to Senegal are shown in the context of the estimated provide an initial picture of the challenges that cross-country relationship between each SDG the Post-2015 agenda poses for Senegal, serving indicator and GNI per capita. For Senegal, the as the starting point for a more complete country projected average annual rate of GNI per capita development diagnostic as well as a more com- growth is merely 1.2 percent.4 The projected SDG prehensive country-focused analysis. The note is values reflect what can be expected given a coun- built around tables and figures that provide data try’s starting point, projected growth in GNI per for a selection of SDG target indicators and indi- capita, typical rates of progress according to cators related to fiscal space—fiscal space matters cross-country patterns, and a gradual conver- since, while policy frameworks and the engage- gence to close gaps between observed and ment of the private sector may vary widely, expected values.5 Projections for SDG indicators rapid  progress on the SDG agenda will require are presented only when the cross-country rela- efficient and carefully prioritized public spend- tionship between the indicator and GNI per ing. Drawing on the information in these tables capita is classified as tight.6 A loose relationship and figures, this note briefly (a) summarizes suggests that progress in the indicator is primar- Senegal’s SDG progress since 2000 and projects ily  a reflection of country-specific factors and expected values for 2030; and (b) assesses options that  it should not be expected to respond for increasing fiscal space. Sections 2 and 3 strongly or systematically to changes in GNI per address SDGs and fiscal space, respectively, while capita. When the relationship to GNI per capita is findings are summarized in Section 4. loose the coefficients are in practice also small Senegal 169 Table 10.1  Senegal—SDG Indicators: Evolution since 2000 and Projections to 2030 Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Poverty and shared prosperity Poverty, at $1.25 a day (PPP) 47.3 34.1 15.0 25.5 By 2030, eradicate extreme poverty for all people (% of population) everywhere, currently measured as people living on less than $1.25 a day. Shared prosperity: Income 16.4 16.7 — By 2030, progressively achieve and sustain share for lowest 40% income growth of the bottom 40 per cent of the population at a rate higher than the national average. Education Pre-primary enrollment 2.8 14.3 21.8 17.2 By 2030, ensure that all girls and boys have (% gross) access to quality early childhood development, care, and pre-primary education so that they are ready for primary education. Primary completion (% 38.6 60.5 76.3 64.3 By 2030, ensure that all girls and boys complete gross) free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment 15.9 41.0 51.9 45.3 By 2030, ensure that all girls and boys complete (% gross) free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary completion 18.6 22.2 By 2030, ensure that all girls and boys complete (% gross) free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Primary completion, ratio of 73.3 108.8 94.4 — By 2030, ensure that all girls and boys complete females to males (%) free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment, ratio 65.1 91.2 87.7 92.7 By 2030, ensure that all girls and boys complete of females to males (%) free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Health Under-5 mortality (per 1,000 137.0 55.3 50.9 49.1 By 2030, end preventable deaths of newborns live births) and children under 5 years of age. Maternal mortality 480 320 215 269 By 2030, reduce the global maternal ratio to less (per 100,000 live births) than 70 per 100,000 live births. Malaria cases 0.5 2.1 0.3 1.3 By 2030, end the epidemics of AIDS, (% of population) tuberculosis, malaria, and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. HIV prevalence (% of 0.5 0.5 0.9 — By 2030, end the epidemics of AIDS, population ages 15–49) tuberculosis, malaria and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. Infrastructure Access to improved 42.7 51.9 37.7 54.7 By 2030, achieve access to adequate and sanitation facilities equitable sanitation and hygiene for all and end (% of population) open defection, paying special attention to the needs of women and girls and those in vulnerable situations. table continues next page 170 Senegal Table 10.1  continued Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Access to improved water 66.4 74.1 74.9 76.2 By 2030, achieve universal and equitable access source (% of population) to safe and affordable drinking water for all. Road density (km per 100 7.4 7.6 12.7 9.0 Develop quality, reliable, sustainable, and resilient sq. km of land) infrastructure, including regional and transborder infrastructure, to support economic development and human well-being, with a focus on affordable and equitable access for all. Access to electricity 36.8 56.5 37.6 59.4 By 2030, ensure universal access to affordable, (% of population) reliable, and modern energy services. Internet users (per 1,000 0.4 20.9 9.8 — Significantly increase access to information and people) communications technology and strive to provide universal and affordable access to the Internet in least developed countries by 2020. Environment CO2 emissions (metric tons 0.40 0.55 0.43 0.66 Integrate climate change measures into national per capita) policies, strategies and planning. Memorandum item GNI per capita (constant 689 787 956 2005 US$) Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent”. The year for country specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = currently significantly overperforming; red = currently significantly underperforming; black = performing as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). For Internet use, a projection is not made, despite a tight relationship, since the expected line shifts significantly even in the medium run due to external forces, such as technological development. Global ambition is from the Open Working Group (UN 2014). Figure 10.1  Senegal—SDG Indicators (Log Scale) versus GNI per Capita in a Cross- Country Setting (Log Scale) a. Poverty (left), shared prosperity: income share of bottom 40% (right) Shared prosperity: income share for lowest 40% Poverty, at $1.25 a day (PPP) (% of population) 410 30 90 1999 2011 20 20 2030 2011 4 10 .90 9 .04 5 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 11.8*** + −1.3*** ln(GNI pc); R 2 = .487 ln(SDG) = 3.17*** + −.05** ln(GNI pc); R 2 = .046 figure continues next page Senegal 171 Figure 10.1  continued b. Gross pre-primary enrollment (left), primary completion (right) 550 Pre-primary enrollment (% gross) Primary completion (% gross) 150 90 2030 40 60 2012 2030 10 2012 40 2000 3 2000 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.73 + .570*** ln(GNI pc); R 2 = .348 ln(SDG) = 3.38*** + .14*** ln(GNI pc); R 2 = .372 c. Gross secondary enrollment (left), secondary completion (right) 190 90 Secondary completion (% gross) Secondary enrollment (% gross) 110 20 60 2012 6 2030 2011 30 1.4 20 2000 6 .09 150 400 1,000 3,000 8,000 150 400 1,000 3,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.9*** + .31*** ln(GNI pc); R 2 = .558 ln(SDG) = −.2 + .49* ln(GNI pc); R 2 = .098 figure continues next page 172 Senegal Figure 10.1  continued Primary completion, ratio of females to males (%) d. Ratio of female to male primary completion (left), ratio of female to male secondary enrollment (right) 140 140 2012 Secondary enrollment, ratio of 110 100 2030 females to males (%) 2013 90 2000 70 70 2000 50 40 30 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 4.31*** + .04*** ln(GNI pc); R 2 = .093 ln(SDG) = 3.82*** + .1*** ln(GNI pc); R 2 = .297 e. Under-5 mortality (left), maternal mortality (right) 500 2000 Maternal mortality (per 100,000 live births) Under−5 mortality (per 1,000 live births) 2000 2013 230 470 2030 110 2000 2013 100 50 2030 20 20 5 1 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 7.99*** + −.61*** ln(GNI pc); R 2 = .560 ln(SDG) = 11.3*** + −.9*** ln(GNI pc); R 2 = .519 figure continues next page Senegal 173 Figure 10.1  continued f. Malaria cases (left), HIV prevalence (right) 90 HIV Prevalence (% of population ages 15–49) 30 2012 30 Malaria cases (% of population) 1.3 2030 2000 9 .05 3 .90 2000 2013 .10 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 14.5*** + −2.3*** ln(GNI pc); R 2 = .412 ln(SDG) = 1.88* + −.31** ln(GNI pc); R 2 = .044 g. Access to improved sanitation (left), access to improved water source (right) Access to improved sanitation facilities (% of population) Access to improved swater source (% of population) 100 100 2012 2030 70 2030 60 2000 2012 2000 50 30 40 20 7 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .61** + .45*** ln(GNI pc); R 2 = .535 ln(SDG) = 3.45*** + .13*** ln(GNI pc); R 2 = .436 figure continues next page 174 Senegal Figure 10.1  continued h. Access to electricity (left), road density (right) 400 330 Road density (km per 100 sq. km of land) Access to electricity (% of population) 2010 2030 100 90 30 2000 20 2030 7 2000 6 2011 1.6 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .29 + .5*** ln(GNI pc); R 2 = .458 ln(SDG) = −.63 + .48*** ln(GNI pc); R 2 = .189 i. Internet users (left), CO2 emissions (right) 20 80 CO2 emissions (metric tons per capita) 0 Internet users (per 1000 people) 2013 3 2030 .80 2010 2000 6 .20 2000 .40 .00 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −2.9*** + .78*** ln(GNI pc); R 2 = .568 ln(SDG) = −8.6*** + 1.16*** ln(GNI pc); R 2 = .677 Note: Highlighted observations are for Senegal at different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. Senegal 175 (in absolute terms); given this the “expected” val- school enrollment, maternal mortality, malaria, ues for a recent year are close to the average for road density, and CO2 emissions.8 While under- all low- and middle-income countries.7 performance for an indicator may be due to As indicated, Senegal’s current outcomes are country-specific conditions that are difficult to better than expected (compared to a typical change, it may often point to areas in which pay- country at the same GNI per capita level) for 6 offs from feasible policy change are relatively of the 18 indicators covered in table 10.1: ratio of high, a possibility that calls for further analysis. female to male primary completion, ratio of Figure 10.2 shows that, since 2000, Senegal’s female to male secondary enrollment, HIV prev- ranking among low- and middle-income countries alence, access to improved sanitation facilities, has deteriorated by 5 percentile points in terms of access to electricity, and Internet users. For 4 GNI per capita. Nevertheless, among the 15 indi- indicators—shared prosperity, secondary com- cators with sufficient data, Senegal’s ranking still pletion, under-5 mortality, and access to rose compared to 2000 for 7 of the indicators: improved water source—current outcomes are as pre-primary enrollment, gross secondary enroll- expected. The country falls short of expectations ment, ratio of female to male primary completion for 8 indicators: poverty, pre-primary school (for the latter quite dramatically, by as much as enrollment, primary completion rate, secondary 80  percentile points), ratio of female to male Figure 10.2  Senegal—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012 GNI per capita (constant 2005 US$) 35 30 Shared prosperity: Income share for lowest 40% 42 Poverty, at $1.25 a day (PPP) (% of population) 29 22 Pre-primary enrollment (% gross) 14 20 Primary completion (% gross) 17 10 Secondary enrollment (% gross) 10 15 Secondary completion (% gross) 32 Primary completion, ratio of females to males (%) 14 94 Secondary enrollment, ratio of females to males (%) 13 30 Under-5 mortality (per 1,000 live births) 17 30 Maternal mortality (per 100,000 live births) 33 29 Malaria cases (% of population) 50 25 HIV prevalence (% of population ages 15–49) 60 60 Access to improved sanitation facilities (% of population) 38 36 Access to improved water source (% of population) 28 23 Road density (km per 100 sq. km of land) 25 22 Access to electricity (% of population) 31 42 Internet users (per 1,000 people) 40 48 CO2 emissions (metric tons per capita) 68 67 0 10 20 30 40 50 60 70 80 90 100 2000 Recent Note: A high ranking signals strong performance; for the underlying indicator, this may correspond to a relatively high value (for example, for the secondary enrollment rate) or a relatively low value (for example, for the poverty rate). The ranking for an indicator is not reported if the available sample is less than 20. Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent”. The year for country- specific data can be found in the respective graphs. 176 Senegal secondary enrollment, under-5 mortality, access to rapid growth (beyond the projected 1.2 percent) electricity, and Internet users.9 In addition, 3 indi- and improvements in policies that directly influ- cators (HIV prevalence, access to improved sanita- ence different SDGs; such policies are particularly tion facilities, and CO2 emissions) stayed at the important for shared prosperity and other indica- same ranking or deteriorated less than GNI per tors that are largely unrelated to GNI per capita. capita. For 5 of the indicators the ranking deterio- rated roughly to the same extent as GNI per capita: poverty, primary completion, maternal mortality, ­ access to improved water source, and road density. 3. Fiscal Space Only for 1 indicator—malaria—was the decline In most countries, accelerated progress on the more severe than for GNI per capita. SDG agenda will require efficient and growing By 2030, considerable improvements are pro- public spending in prioritized areas, most impor- jected for most indicators (see table 10.1 and tantly human development and infrastructure. respective graphs in figure 10.1). However, com- Private spending is also of crucial importance, pared to global ambitions, the improvements for both household spending on SDG-related ser- most indicators are moderate. This means that, vices and business investments in a wide range of to get closer to the realization of these ambitions, areas (including but not limited to infrastruc- a break with the past is needed. Such a break ture).10 With regard to Senegal’s fiscal space indi- would be facilitated by a combination of more ­ gure 10.3 summarize the cators, table 10.2 and fi Table 10.2  Senegal—Fiscal Space: Revenue, Spending, and Government Efficiency Actual Expected Projection Indicator 2000 Recent Recent 2030 Government spending Consumption (% of GDP) 12.8 15.5 13.4 — Investments (% of GDP) 4.5 6.4 6.4 — Primary education (% of GDP) 1.3 2.2 1.7 — Secondary education (% of GDP) 0.8 1.6 1.2 — Primary education, per student (% of GDP per capita) 11.7 17.0 12.4 — Secondary education, per student (% of GDP per capita) 35.2 29.0 18.8 — Health (% of GDP) 1.9 2.8 2.8 — Fuel subsidy (% of GDP) 2.3 1.9 — Government receipts and debts Tax revenue (% of GDP) 19.2 13.2 — Net ODA (% of GNI) 9.4 7.8 4.2 6.3 External debt (% of GNI) 79.6 34.9 32.1 — Governance and government efficiency Government effectivness: percentile rank 51.7 37.8 19.3 39.1 Grigoli education efficiency score 0.27 0.38 0.31 Grigoli health efficiency score 0.94 0.94 — Public investment management index 0.9 1.5 — Memorandum item GNI per capita (constant 2005 US$) 689 787 956 Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent”. The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = current value significantly above the expected level; red = current value significantly below the expected level; black = current value as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). Senegal 177 Figure 10.3  Senegal—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting a. Government spending: consumption (left), investment (right) 130 50 60 20 Consumption (% of GDP) Investments (% of GDP) 30 2013 9 2013 10 2000 2000 4 2 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.01*** + .09** ln(GNI pc); R 2 = .042 ln(SDG) = 2.34*** + −.07 ln(GNI pc); R 2 = .012 b. Government spending: primary education, per student (left), secondary education, per student (right) 90 50 Secondary education, per student Primary education, per student 30 50 1998 (% of GDP per capita) (% of GDP per capita) 20 2010 2010 20 1998 8 10 2 4 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.72*** + .12** ln(GNI pc); R 2 = .067 ln(SDG) = 3.53*** + −.09 ln(GNI pc); R 2 = .032 figure continues next page 178 Senegal Figure 10.3  continued c. Government spending: primary education (left), secondary education (right) 6 4 4 Secondary education (% of GDP) Primary education (% of GDP) 2010 2 2010 2 1.2 1998 1.1 1998 .60 .40 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.07*** + −.08 ln(GNI pc); R 2 = .029 ln(SDG) = −.79* + .14*** ln(GNI pc); R 2 = .086 d. Government spending: health (left), fuel subsidy (right) 20 20 1.6 2011 Fuel subsidy (% of GDP) 6 Health (% of GDP) .10 2012 2000 1.6 0 .00 .50 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.05 + .16*** ln(GNI pc); R 2 = .093 ln(SDG) = 4.96*** + −.65*** ln(GNI pc); R 2 = .098 figure continues next page Senegal 179 Figure 10.3  continued e. Tax revenue (left), official development aid (right) 60 40 2000 2012 7 2030 30 Tax revenue (% of GDP) Net ODA (% of GNI) 2012 20 1.2 10 .20 3 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.75*** + .13 ln(GNI pc); R 2 = .027 ln(SDG) = 8.26*** + −1.0*** ln(GNI pc); R 2 = .308 f. External debt (left), government effectiveness (right) 2,000 80 Government effectiveness: percentile rank 2000 550 2013 30 2030 External debt (% of GNI) 160 2000 10 50 2013 4 10 1.3 .50 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.51*** + .14** ln(GNI pc); R 2 = .035 ln(SDG) = 0 + .44*** ln(GNI pc); R 2 = .300 figure continues next page 180 Senegal Figure 10.3  continued g. Education expenditure efficiency (left), health expenditure efficiency (right) 1 Grigoli education efficiency score 1 Grigoli health efficiency score .60 2010 .40 .90 2030 .30 2010 .10 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = –.04*** + .46*** ln(GNI pc); R 2 = .035 ln(SDG) = −.09*** + .01 ln(GNI pc); R 2 = .031 h. Public investment management index 4 Public investment management index 2 1.3 .90 2010 .80 .30 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = −.43 + .12** ln(GNI pc); R 2 = .078 Note: Highlighted observations are for Senegal in different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. Senegal 181 historical evolution, actual and expected recent the priorities of donors and their relationships values, and, when relevant, projected values.11 with Senegal’s government. When the relationship is loose, projections are The relationship between tax revenues and not made and the expected value is in practice GNI per capita is not tight enough to project close to the average for the sample of all low- expected changes. However, Senegal’s govern- and  middle-income countries (cf. discussion ment has signaled that its intention is to expand of  expected values for SDG indicators). The tax revenues.15 Higher taxes or lower subsidies variables cover selected indicators related to ­ would both reduce the resources controlled by three aspects of government activities: spending. domestic households and firms, pointing to the receipts and debt, and governance and efficiency. need to consider the combined impact on SDGs While the findings of this country-at-a-glance and other indicators from higher taxes and the note cannot guide policy on their own, they spending increases that are financed by these should be an input into the discussion about taxes. The level of public foreign debt (as a share ­policy making. of GDP) is as expected (with a weak and slightly In terms of government spending in areas positive relationship to GNI per capita). Senegal that may support the SDG agenda, Senegal is continues to face a low risk of debt distress, but above expected levels (compared to a typical risks have increased. Hence, with an external country at the same GNI per capita level) for debt stock at the expected level, increased total public consumption, primary education, external borrowing can only be expected to be ­ and secondary education. Spending on total marginal.16 public investment and health are as expected. Government efficiency is important to pro- Note that, for primary and secondary education, tect and, if possible, increase in order to add to spending is measured in two ways: total as share the room for priority spending and enhance its of GDP and per student as share of GDP per impact on the SDG agenda. Table 10.2 displays capita; for both levels, spending according to data for some measures of government efficiency, both indicators is higher than expected. Fuel with mixed results for Senegal. According to subsidies (at 2.3 percent of GDP in 2013)12 are the education spending efficiency index used in the most obvious case of low-priority spending this study, Senegal’s performance is below expec- from the post-2015 agenda perspective—­ tations, while it is doing as expected for the elimination of these subsidies should be a top health index; among these two indexes, GNI per priority.13 SDG-related areas in which the gov- capita is strongly correlated with the educa- ernment currently is overspending while the tion  index but largely uncorrelated with the associated SDG indicators are lagging need to be health index. Senegal is also performing below discussed in relation to improved efficiency and expectations in  terms of the more general impact. Public Investment Management Index, but better Among current receipts, tax revenues and net than expected according to the World Bank Official Development Assistance (ODA) as a Government Effectiveness index. Given that the share of GDP are both higher than expected.14 As different indexes measure different aspects of further shown in table 10.2, cross-country pat- government performance, such mixed findings terns suggest that net ODA will decline as a may not be inconsistent. The Senegal Public percent of GDP (without changing significantly Expenditure Review 2012 also suggests the inef- in per capita terms); for Senegal, this would mean ficiencies in Senegal’s public spending are more a decrease from the current level of 7.8 percent of severe than for many other countries and that GNI to 6.3 percent in 2030. However, the fact these inefficiencies have been increasing over that cross-country patterns point to a likely time (World Bank 2012, pp. 13–14). In addition, decline in ODA does not mean that such an out- more ­ qualitative evidence points to spending come should be taken for given: it depends on inefficiencies.17 182 Senegal In sum, while there are some areas standing spending decisions promote a broad-based change out, such as decreased spending on fossil fuel sub- that encompasses services related to human devel- sidies, creation of additional fiscal space would opment, infrastructure investments, and other probably require a combination of actions to measures in support of strong long-run growth decrease spending in areas of lesser priority, that are biased in favor of the less advantaged. improve government and spending efficiency, and, possibly, take measures to increase tax revenues— although they are already higher than expected.18 Among these, opportunities to improve govern- 4. Conclusions ment efficiency should in particular be pursued As summarized in table 10.3, Senegal’s current given that, other things being equal, higher taxa- outcomes are better than expected (compared to tion tends to have a negative impact on private a typical country at the same GNI per capita level) consumption, savings, and investment. In general, in some areas but below expectations in others, decisions about the level and allocation of govern- including poverty and a number of education ment spending should be made in light of govern- indicators (pre-primary enrollment, primary ment priorities and additional considerations completion, and secondary enrollment). By 2030, beyond the scope of this note, including assess- improvements are projected for most indicators ments about government capacity to efficiently and a few of the post-2015 global ambitions may expand activities in different areas and the scope to even be realized. However, for most, a break with encourage complementary private sector activities. the past would be needed. Beyond issues directly related to the government Table 10.3 further shows that, for most indi- budget, remittances from Senegal’s workers abroad cators, the relationship to GNI per capita is tight. were as high as 10.7 percent of GDP in 2013 (higher Given this, acceleration of growth, especially of than expected for Senegal’s GNI per capita level), the inclusive type, should be a top priority. significantly raising the living standards of many Improvements in these SDGs will likely continue Senegalese. Measures that encourage even higher along with GNI per capita growth and related levels of remittances and channel them to increases in resources and capabilities. From a income-raising investments may have high pay- multicountry perspective, performance has been offs for the SDG agenda. More broadly, from the particularly strong in terms or primary school perspective of this agenda and given strong link- gender equality (indicated by a strongly improved ages between private and government activities percentile ranking for the ratio of female to male and incomes, it is crucial that policies and primary completion). However, for malaria, Table 10.3  Senegal—Summary of Results for SDG Indicators Cross-country relationship with Overperforming As expected Underperforming GNI/cap Tight •• Ratio of female to male •• Under-5 mortality (+) •• Poverty (×) secondary enrollment (+) •• Access to improved water •• Pre-primary school enrollment (+) •• Access to improved source (×) •• Primary completion (×) sanitation (+) •• Gross secondary enrollment (+) •• Access to electricity (+) •• Maternal mortality (×) •• Internet users (+) •• Malaria (×) •• Road density (×) •• CO2 emissions (+) Loose •• Ratio of female to male primary •• Shared prosperity completion (+) •• Secondary completion •• HIV prevalence (+) Note: (+) = larger country rank improvement (or smaller drop) 2000–12 than for GNI per capita; (−) = larger drop 2000–12 than for GNI per capita; (×) = the same country rank change 2000–12 as for GNI per capita. Senegal 183 Senegal is not only underperforming but has over-all growth would clearly promise to raise deteriorated by more in ranking than GNI per growth in the level of per capita incomes for the capita, suggesting that special attention should bottom 40 percent even though the impact of the be given to the efficiency of related policies. growth acceleration on shared prosperity would For indicators with a weak relationship with depend on country and policy specifics. GNI per capita, strong and systematic improve- Given relatively high levels of government ments should not be expected to accompany eco- consumption and tax revenues, opportunities to nomic growth; hence, policies affecting such improve efficiency in government spending seem indicators require additional attention. This is, particularly important. The main source of addi- for example, the case for shared prosperity and tional fiscal space may be elimination of fossil secondary completion, both of which currently fuel subsidies and decreased spending in low-­ are at expected levels. However, more rapid priority areas. Annex 10A: Data Sources Indicator Source Comment GNI per capita (constant WDI. API ref: GNI per capita (constant 2005 US$) 2005 US$) [NY.GNP.PCAP.KD] SDG indicators Poverty, at $1.25 a day (PPP) WDI. API ref: poverty headcount ratio at $1.25 a day (PPP) (% of population) (% of population) [SI.POV.DDAY] Shared prosperity: income WDI. API ref: income share held by lowest 20% + income share for lowest 40% share held by second 20% [SI.DST.FRST.20+SI.DST.02ND.20] Pre-primary enrollment WDI. API ref: school enrollment, pre-primary (% gross) (% gross) [SE.PRE.ENRR] Primary completion (% gross) WDI. API ref: primary completion rate, total (% of relevant age group) [SE.PRM.CMPT.ZS] Secondary enrollment WDI. API ref: school enrollment, secondary (% gross) (% gross) [SE.SEC.ENRR] Secondary completion EdStat. API ref: DHS: secondary completion rate (% gross) [HH.DHS.SCR] Primary completion, ratio of WDI. API ref: primary completion rate, female (% of relevant females to males (%) age group)/primary completion rate, male (% of relevant age group) *100 [SE.PRM.CMPT.FE.ZS /SE.PRM.CMPT.MA.ZS*100] Secondary enrollment, ratio of WDI. API ref: ratio of female to male secondary enrollment (%) females to males (%) [SE.ENR.SECO.FM.ZS] Under-5 mortality (per 1,000 WDI. API ref: mortality rate, under-5 (per 1,000 live births) live births) [SH.DYN.MORT] Maternal mortality WDI. API ref: maternal mortality ratio (modeled estimate, per (per 100,000 live births) 100,000 live births) [SH.STA.MMRT] Malaria cases (% of HNP. API ref: malaria cases reported/population, total *100 population) [sh_sta_malr/ SP.POP.TOTL *100] HIV prevalence (% of WDI. API ref: prevalence of HIV, total (% of population population ages 15–49) ages 15–49) [SH.DYN.AIDS.ZS] Access to improved WDI. API ref: improved sanitation facilities (% of population sanitation facilities (% of with access) [SH.STA.ACSN] population) Access to improved water WDI. API ref: improved water source (% of population with source (% of population) access) [SH.H2O.SAFE.ZS] Road density (km per WDI. API ref: road density (km of road per 100 sq. km of land 100 sq. km of land) area) [IS.ROD.DNST.K2] Access to electricity WDI. API ref: access to electricity (% of population) (% of population) [EG.ELC.ACCS.ZS] annex continues next page 184 Senegal Indicator Source Comment Internet users (per WDI. API ref: Internet users (per 100 people) [IT.NET.USER.P2] 1,000 people) CO2 emissions (metric tons WDI. API ref: CO2 emissions (metric tons per capita) per capita) [EN.ATM.CO2E.PC] Fiscal space indicators Investment (% of GDP) WDI. API ref: gross fixed capital formation (% of GDP)-gross fixed capital formation, private sector (% of GDP) [NE.GDI.FTOT.ZS]-[NE.GDI.FPRV.ZS] Consumption (% of GDP) WDI. API ref: general government final consumption expenditure (% of GDP) [NE.CON.GOVT.ZS] Primary education (% of EdStat. API ref: total expenditure on educational institutions GDP) and administration as a % of GDP. All sources. Primary [UIS.XGDP.1.FDINSTADM.FFD] Secondary education EdStat. API ref: total expenditure on educational institutions (% of GDP) and administration as a % of GDP. All sources. Secondary and post-secondary non-tertiary [UIS.XGDP.234.FDINSTADM.FFD] Primary education, per WDI. API ref: expenditure per student, primary student (% of GDP per (% of GDP per capita) [SE.XPD.PRIM.PC.ZS] capita) Secondary education, per WDI. API ref: expenditure per student, secondary student (% of GDP per (% of GDP per capita) [SE.XPD.SECO.PC.ZS] capita) Health (% of GDP) WDI. API ref: health expenditure, public (% of GDP) [SH.XPD.PUBL.ZS] Fossil fuel subsidy (% of IMF (2013a). Total pretax subsidy (% of GDP) Subsidies are measured using a GDP) price-gap approach capturing both consumer (including implicit) and producer (except those that arise when suppliers are inefficient and make losses at benchmark prices) subsidies. Negative external effects are not included in the pretax subsidy Tax revenue (% of GDP) WDI. API ref: tax revenue (% of GDP) [GC.TAX.TOTL.GD.ZS] Net ODA (% of GNI) WDI. API ref: net ODA received (% of GNI) [DT.ODA.ODAT.GN.ZS] External debt (% of GNI) WDI. API ref: external debt stocks (% of GNI) [DT.DOD.DECT.GN.ZS] Government effectiveness: WGI. API ref: government effectiveness: percentile rank Captures perceptions of the quality of percentile rank [ge_per_rnk] public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies. Grigoli education efficiency Grigoli (2014) Measure secondary education score inefficiency in terms of how much additional output could be achieved at current levels of spending Grigoli health efficiency score Grigoli and Kapsoli (2013) Quantifies the inefficiency of public health expenditure using a stochastic frontier model that controls for the socioeconomic determinants of health. Public investment Dabla-Norris et al. (2011) Four phases associated with public management index investment management are covered: project appraisal, selection, implementation, and evaluation. Note: WDI = World Development Indicators, World Bank; API ref = reference and code when using the World Bank Open Data; EdStat = Education statistics, World Bank; HNP = Health Nutrition and Population statistics, World Bank; WGI = Worldwide Governance Indicators, World Bank. Senegal 185 Notes 1. While a cross-country perspective provides an of CO2 emissions per unit of GDP, Senegal’s are important complement to analysis that is centered also higher than expected. on an individual country, it is by definition lim- 9. If data are not available for 2000 or 2013, the ited to analysis of variables that are available in closest earlier year with data is used. The year for cross-country databases. country-specific data can be found in the respec- 2. This does not mean that GNI per capita is viewed tive graphs. as a direct determinant of SDG outcomes; on the 10. There are also cases where the solution to the low contrary, a major challenge for policy makers is to level of SDG is neither private nor public spend- identify policies that improve SDG performance ing but more efficient policies or complementary relative to what is expected given the level of GNI policies. per capita. A second challenge is to raise growth in 11. The treatment is the same as in table 10.1 and GNI per capita as it indirectly influences country related figures. That is, in table 10.2, projections SDG capacity. are shown only when the cross-country relation- 3. Sources for the indicators are presented in the ship between the indicator and GNI per capita is annex 10A. considered tight enough. Due to data limitations, 4. Projections from CEPII (v 2.3) are used for this we focus on government indicators; country-­ and other Country Development Diagnostics specific analysis is needed to consider policy in the applications given their wide country coverage context of the different roles of the government and well-documented methodology; OECD data and private services and spending. have been used when projections have been miss- 12. Estimate for 2014 suggests fuel subsidies will drop ing. In the projections, it is assumed that future under 1.4 percent of GDP, and it will be even less GNI growth will coincide with future GDP growth in 2015 due to the fall in oil prices (World Bank (both expressed in constant 2005 US$) given that country team). that this is the variable that CEPII and other 13. Fuel subsidies are detrimental to the climate and sources project. encourage technologies that are less labor inten- 5. Given that (a) SDGs have extreme values (such sive, tending to generate employment for fewer as  100 percent for improved water access) workers at lower wages. and (b)  the current SDG level never is exactly 14. Net ODA per capita is somewhat higher than as expected given GNI per capita, the pro- expected for Senegal. jected ­ values gradually converge toward the 15. IMF (2015c, p. 38) suggests that Senegal’s tax expected ­ values. For example, for a country that ­ revenues may be increased from its 2013 level overperforms in water access, as GNI per capita ­ of 18.4 percent of GDP to 20.1 percent in 2019. increases the extent of overperformance gradually The government has created a new tax code and declines, so that when the expected value is 100, continues to implement tax policy and revenue over-­performance has reached zero. administration measures (IMF 2015c, pp. 7, 44). 6. A tight enough relationship is defined as an 16. Starting from a level of 32.7 percent of GDP in R2 > 0.3 (tight) or 0.3 > R2 > 0.1 (moderately tight), 2013, the external public debt is projected to while R2 < 0.1 are defined as loose. be at 35.1 percent in 2019 and 30.1 in 2024; the 7. In addition, the confidence interval is wide in total (domestic and foreign) public debt is pro- the  case of a loose relationship, suggesting that jected to evolve from 47.1 percent of GDP in any  conclusion on over- or underperformance 2013 to 50.7 percent in 2019 and 40.9 in 2024 is made with wide margins. Statistically, even (IMF 2015c, p. 34, DSA annex). The Senegal debt though their confidence intervals are wide, as sustainability analysis (IMF 2015c, DSA annex) long as the estimated coefficient linking GNI shows that Senegal remains at a low risk of debt per capita to the SDG indicator is nonzero, these distress under the assumption of fiscal consolida- ­ values are closer than the cross-country average tion and a shift toward less concessional financ- to what is expected for Senegal. The same obser- ing. Recently the government decided to keep vation applies to expected values for fiscal space the total debt-to-GDP ratio below 52 percent indicators. (IMF 2015c, pp. 10, 44). 8. With regard to CO2, Senegal’s current and 17. The IMF and the government (IMF 2015c, projected 2030 per capita emissions are 5.4 and ­ pp.  11–12, 43) agree that successful develop- 8.2 percent of the current OECD average. In terms ment will require a significant improvement in 186 Senegal public spending efficiency. This includes substan- 18. World Bank (2014c, p. 8) suggests that, given tial improvements in the regulatory framework ­ continued struggles with regard to revenue and governance, as well as in the quality and effi- collection, the ability to create fiscal space will ­ ciency of public investment together with actions be driven by success in reducing low priority to limit public consumption. spending. Senegal 187 Chapter 11 Uganda 1. Introduction to provide and efficiently utilize inputs that ­ contribute to SDGs (for example, health and Uganda is a landlocked low-income country in education services) and to achieve SDG out- Sub-Saharan Africa. While suffering from a pro- comes (like strong health and education results).3 tracted civil war in the north, most of the coun- try has enjoyed relative political stability since 1986. During 2001–12, Uganda’s average growth rate for GNI per capita (at constant 2005 US$) 2. SDG Indicators: History and was 3.8 percent, which may be compared to a developing (low- and middle-­ income) country Projections average of 3.0 percent. During the same period, For selected SDG indicators, table 11.1 summa- Uganda’s ranking according to the UNDP rizes data for Uganda: historical evolution, actual Human Development Index (among countries and expected values for a recent year, and pro- included both in 2000 and 2012) remained jected 2030 values.4 In figure 11.1, data for Uganda unchanged (at the 11th percentile). are shown in the context of the estimated This country-at-a-glance note is designed to cross-country relationship between each SDG provide an initial picture of the challenges that indicator and GNI per capita. For Uganda, the the Post-2015 agenda poses for Uganda; its projected average annual rate of GNI per capita findings cannot guide policy on their own but growth is 2.7 percent.5 The projected SDG values should be seen as an input into policy discus- reflect what can be expected given a country’s sions. The note may also serve as a starting starting point, projected growth in GNI per capita, point for a more complete country develop- typical rates of progress according to c ­ ross-country ment diagnostic as well as more comprehensive patterns, and a gradual convergence to close gaps country-focused analysis. The note provides between observed and expected values.6 In data for a selection of SDG target indicators table 11.1, projections are presented only when the and indicators related to fiscal space,1 and cross-country relationship between the indicator briefly (a) summarizes Uganda’s SDG progress and GNI per capita is classified as tight.7 A very since 2000 and  projects expected values for loose relationship suggests that progress in the 2030; and (b)  assesses options for increasing indicator is primarily a reflection of ­country-specific fiscal space. Sections 2 and 3 address SDGs and factors and that it should not be expected to fiscal space, respectively, while findings are respond strongly or systematically to changes in summarized in Section 4. GNI per capita. When the relationship to GNI per The analysis is done from a cross-country capita is loose the coefficients are typically small perspective: for the different indicators, Uganda’s (in absolute terms); given this, the “expected” performance and prospects are benchmarked ­ values for a recent year are close to the average for relative to other countries, considering its past, all low- and middle-income countries.8 recent, and projected levels of GNI per capita.2 In sum, Uganda’s current outcomes are better The latter variable tends to be highly correlated than expected (compared to a typical country at with most of the SDGs and most of the factors the same GNI per capita level) in 6 cases; ratio that determine their evolution; given this, it is of  female to male primary completion, access to used as a summary indicator of country capacity improved water source, access to improved Uganda 189 Table 11.1  Uganda—SDG Indicators: Evolution since 2000 and Projections to 2030 Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Poverty and shared prosperity Poverty, at $1.25 a day (PPP) 59.4 37.8 34.2 20.1 By 2030, eradicate extreme poverty for all people (% of population) everywhere, currently measured as people living on less than $1.25 a day. Shared prosperity: Income 15.8 15.1 17.3 — By 2030, progressively achieve and sustain income share for lowest 40% growth of the bottom 40 percent of the population at a rate higher than the national average. Education Pre-primary enrollment 13.6 14.5 18.6 By 2030, ensure that all girls and boys have access (% gross) to quality early childhood development, care, and pre-primary education so that they are ready for primary education. Primary completion (% gross) 53.1 69.2 60.7 By 2030, ensure that all girls and boys complete free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Secondary enrollment 16.4 27.6 41.4 36.9 By 2030, ensure that all girls and boys complete (% gross) free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. (OWG) Secondary completion 10.5 9.4 15.8 — By 2030, ensure that all girls and boys complete (% gross) free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Primary completion, ratio of 97.8 92.1 — By 2030, ensure that all girls and boys complete females to males (%) free, equitable, and quality primary and secondary education leading to releavant and effective learning outcomes Secondary enrollment, ratio of 77.1 83.3 81.6 87.4 By 2030, ensure that all girls and boys complete females to males (%) free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes. Health Under-5 mortality (per 1,000 147.0 66.1 73.4 50.0 By 2030, end preventable deaths of newborns and live births) children under 5 years of age. Maternal mortality 650 360 369 239 By 2030, reduce the global maternal mortality ratio (per 100,000 live births) to less than 70 per 100,000 live births. Malaria cases (% of 14.6 7.3 1.6 2.3 By 2030, end the epidemics of AIDS, tuberculosis, population) malaria, and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. HIV prevalence (% of 7.3 7.4 1.0 — By 2030, end the epidemics of AIDS, tuberculosis, population ages 15–49) malaria, and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases. Infrastructure Access to improved sanitation 29.8 33.9 27.6 40.1 By 2030, achieve access to adequate and equitable facilities (% of population) sanitation and hygiene for all and end open defecation, paying special attention to the needs of women and girls and those in vulnerable situations. Access to improved water 56.5 74.8 68.5 78.4 By 2030, achieve universal and equitable access to source (% of population) safe and affordable drinking water for all. table continues next page 190 Uganda Table 11.1  continued Actual Expected Projection Indicator Global ambition 2030 2000 Recent Recent 2030 Road density (km per 100 sq. 32.2 8.8 34.8 Develop quality, reliable, sustainable and resilient km of land) infrastructure, including regional and transborder infrastructure, to support economic development and human well-being, with a focus on affordable and equitable access for all. Access to electricity 8.6 14.6 26.2 23.6 By 2030, ensure universal access to affordable, (% of population) reliable, and modern energy services. Internet users (per 1,000 0.2 16.2 6.1 — Significantly increase access to information and people) communications technology and strive to provide universal and affordable access to the Internet in least developed countries by 2020. Environment CO2 emissions (metric tons 0.1 0.1 0.2 0.3 Integrate climate change measures into national per capita) policies, strategies, and planning. Memorandum item GNI per capita (constant 264 410 643 2005 US$) Note: Recent refers to the latest year with data (typically 2011 or 2012). If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = currently significantly overperforming; red = currently significantly underperforming; black = performing as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). For Internet use, a projection is not made, despite a tight relationship, since the expected line shifts significantly even in the medium run due to external forces, such as technological development. Global ambition is from the Open Working Group (UN 2014). Figure 11.1  Uganda—SDG Indicators (Log Scale) versus GNI per Capita in a Cross- Country Setting (Log Scale) a. Poverty (left), shared prosperity: income share of bottom 40% (right) Shared prosperity: Income share for lowest 40% Poverty, at $1.25 a day (PPP) (% of population) 410 30 90 1999 2013 20 20 2030 1999 2013 4 10 .90 9 .04 5 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 11.8*** + −1.3*** ln(GNI pc); R 2 = .487 ln(SDG) = 3.17*** + −.05** ln(GNI pc); R 2 = .046 figure continues next page Uganda 191 Figure 11.1  continued b. Gross pre-primary enrollment (left), primary completion (right) 550 130 Pre-primary enrollment (% gross) Primary completion (% gross) 150 90 2030 40 60 2011 2010 2030 10 40 3 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.73 + .570*** ln(GNI pc); R 2 = .348 ln(SDG) = 3.38*** + .14*** ln(GNI pc); R 2 = .372 c. Gross secondary enrollment (left), secondary completion (right) 190 90 Secondary completion (% gross) 110 Secondary enrollment (% gross) 20 60 2000 2011 6 2030 30 2009 1.4 20 2000 6 .09 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.9*** + .31*** ln(GNI pc); R 2 = .558 ln(SDG) = −.2 + .49* ln(GNI pc); R 2 = .098 figure continues next page 192 Uganda Figure 11.1  continued d. Ratio of female to male primary completion (left), ratio of female to male secondary enrollment (right) Secondary enrollment, ratio of females to males (%) Primary completion, ratio of females to males (%) 140 140 110 2011 100 90 2030 2009 70 2000 70 50 40 30 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 4.31*** + .04*** ln(GNI pc); R 2 = .093 ln(SDG) = 3.82*** + .1*** ln(GNI pc); R 2 = .297 e. Under-5 mortality (left), maternal mortality (right) 500 2,000 Maternal mortality (per 100,000 live births) Under−5 mortality (per 1,000 live births) 230 2000 2000 470 2013 110 2030 100 50 2013 2030 20 20 5 1 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 7.99*** + −.61*** ln(GNI pc); R 2 = .560 ln(SDG) = 11.3*** + −.9*** ln(GNI pc); R 2 = .519 figure continues next page Uganda 193 Figure 11.1  continued f. Malaria cases (left), HIV prevalence (right) 2000 90 HIV Prevalence (% of population ages 15–49) 30 2012 2030 30 Malaria cases (% of population) 1.3 9 2000 2013 .05 3 .90 .10 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 14.5*** + −2.3*** ln(GNI pc); R 2 = .412 ln(SDG) = 1.88* + −.31** ln(GNI pc); R 2 = .044 g. Access to improved sanitation (left), access to improved water source (right) Access to improved sanitation facilities 100 2030 Access to improved water source 100 2012 70 60 (% of population) (% of population) 2030 2000 2012 50 30 2000 40 20 7 20 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = .61** + .45*** ln(GNI pc); R 2 = .535 ln(SDG) = 3.45*** + .13*** ln(GNI pc); R 2 = .436 figure continues next page 194 Uganda Figure 11.1  continued h. Road density (left), access to electricity (right) 330 400 Road density (km per 100 sq. km of land) Access to electricity (% of population) 90 100 2030 2008 30 20 2030 2010 7 2000 6 1.6 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.63 + .48*** ln(GNI pc); R 2 = .189 ln(SDG) = .29 + .5*** ln(GNI pc); R 2 = .458 i. Internet users (left), CO2 emissions (right) 20 80 CO2 emissions (metric tons per capita) Internet users (per 1,000 people) 3 2013 .80 6 .20 2030 2010 2000 .40 0 2000 0 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −2.9*** + .78*** ln(GNI pc); R 2 = .568 ln(SDG) = −8.6*** + 1.16*** ln(GNI pc); R 2 = .677 Note: Highlighted observations are for Uganda in different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. Uganda 195 sanitation, road density, Internet use, and CO2 Figure 11.2 shows that, between 2000 and emissions, while it falls short in 6 cases: shared 2012, Uganda’s GNI per capita ranking among prosperity, primary completion, gross secondary low- and middle-income countries stayed more enrollment, malaria, HIV prevalence, and access or less the same (improved by 2 percentile to electricity.9 For the other 6 indicators (poverty, points).10 During the same period, Uganda has gross pre-primary enrollment, secondary comple- seen its ranking improve more noticeably for 5 tion, ratio of female to male secondary enroll- indicators (poverty, maternal mortality, under-5 ment, under-5 mortality, and maternal mortality), mortality, access to improved water source, and Uganda’s current outcomes are as expected. While Internet use) while, for malaria the change is neg- underperformance for an indicator may be due ligible. Meanwhile, the rankings have deteriorated to  country-specific conditions that are difficult for 8 indicators (shared prosperity, gross second- to  change, it may often point to areas in which ary enrollment, secondary completion rate, ratio payoffs from feasible policy change are relatively of female to male in secondary enrollment, HIV high, a possibility that calls for further analysis. prevalence, access to electricity, access to improved Figure 11.2  Uganda—Percentile Cross-Country Ranking for SDG Indicators 2000 and 2012 GNI per capita (constant 2005 US$) 10 12 Shared prosperity: Income share for lowest 40% 54 45 Poverty, at $1.25 a day (PPP) (% of population) 14 19 Pre-primary enrollment (% gross) 41 Primary completion (% gross) 17 Secondary enrollment (% gross) 11 6 Secondary completion (% gross) 60 40 Primary completion, ratio of females to males (%) 52 Secondary enrollment, ratio of females to males (%) 26 16 14 Under-5 mortality (per 1,000 live births) 26 20 Maternal mortality (per 100,000 live births) 25 Malaria cases (% of population) 8 8 HIV prevalence (% of population ages 15–49) 12 9 Access to improved sanitation facilities (% of population) 28 24 18 Access to improved water source (% of population) 26 Road density (km per 100 sq. km of land) 53 Access to electricity (% of population) 10 9 Internet users (per 1,000 people) 26 41 CO2 emissions (metric tons per capita) 95 89 0 10 20 30 40 50 60 70 80 90 100 2000 Recent Note: A high ranking signals strong performance; for the underlying indicator, this may correspond to a relatively high value (for example, for the secondary enrollment rate) or a relatively low value (for example, for the poverty rate). The rankings are based on all low- and middle- income countries (according to the 2012 classification) with data. The country samples vary across indicators but are always the same for 2000 and recent for any given indicator. The ranking for an indicator is not reported if the available sample is less than 20 countries. Recent refers to the latest year with data. If data are not available for 2000 or 2013, the closest earlier year with data is used. Data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. 196 Uganda sanitation, and CO2 emissions). Among these, the Private spending is also of crucial importance, result is not entirely unexpected for CO2 emis- both household spending on SDG-related ser- sions and shared prosperity given an inverse vices and business investments in a wide range of cross-country correlation with GNI per capita areas (including, but not limited to, expected (only weak for shared prosperity, see figure 11.1, recent values and, when relevant, projected val- panels a and i); however, this direction of change ues for fiscal space indicators for Uganda).12 With is nevertheless problematic from the perspective regard to Uganda’s additional fiscal effort to reach of the twin goals. For the other five, a higher GNI the SDGs, table 11.2 and figure 11.3 summarize per capita is linked to improved performance; the historical evolution, actual, and expected given this, these ranking declines are unexpected, (given GNI per capita) recent values, and, when suggesting that policies in countries that other- relevant, projected values. When the relationship wise are similar to Uganda are more apt to address is loose, projections are not made and the these objectives. expected value is in practice close to the average When comparing the results from regressions for the sample of all low- and middle-­ income on GNI per capita to changes in percentile rank- countries (cf. discussion of expected v ­alues ings, a few patterns emerge. Although secondary for  SDG indicators). The variables cover three completion, ratio of female to male in secondary aspects of government activities: spending, enrollment, access to improved sanitation, and receipts and debt, and governance and efficiency. CO2 emissions are either overperforming or per- In terms of government spending as a percent- forming as expected, they have fallen further in age of GDP, in areas that may support the SDG ranking compared to GNI per capita.11 More agenda, Uganda performs as expected (compared alarmingly, for secondary enrollment, Uganda’s to a typical country at the same GNI per capita underperformance is combined with a perfor- level) for total public investment and primary edu- mance that is weaker than GNI per capita in cation, but is below the expected level for total terms of changes in percentile ranking. public consumption, secondary education, and By 2030, considerable improvements are pro- public health. None of the SDG spending indica- jected for most indicators (see table 11.1 and tors (included table 11.2) is above the expected respective graphs in figure 11.1). However, com- value. For secondary education spending, note pared to global ambitions, also shown in that, even though Uganda’s government falls below table 11.1, the improvements are moderate. This the expected value if spending is measured as share means that, to get closer to the realization of of GDP, spending is as expected if measured per these ambitions, a break with the past is needed. student (due to a relatively low ­ secondary school This is also true for indicators, such as shared enrollment rate). For primary, the opposite is true; prosperity, for which a weak relationship with as a share of GDP, spending is as expected, but per GNI per capita precludes projections. Such a student it is lower than expected (due to a relatively break would be facilitated by a combination of high primary school enrollment rate). Spending more rapid growth (beyond the projected a ­ verage on fossil fuel subsidies (currently 1.3  percent of growth rate of 2.7 percent) and improvements in GDP in Uganda)—although lower than expected— policies that directly influence different SDGs. is the most obvious case of low-priority spending from the post-2015 agenda perspective.13 Among current receipts, taxes and net Official Development Assistance (ODA) are both very 3. Fiscal Space important and within the expected ranges. As In most countries, accelerated progress on the further shown in table 11.2, cross-country pat- SDG agenda will require efficient and growing terns suggest that net ODA will decline as a public spending in prioritized areas, most impor- percent of GDP (without changing significantly tantly human development and infrastructure. in per capita terms), which in the case of Uganda Uganda 197 Table 11.2  Uganda—Fiscal Space: Revenue, Spending, and Government Efficiency Actual Expected Projection Indicator 2000 Recent Recent 2030 Government spending Consumption (% of GDP) 14.5 8.3 12.7 — Investments (% of GDP) 6.1 5.5 6.8 — Primary education (% of GDP) 1.8 1.8 — Secondary education (% of GDP) 0.8 1.1 — Primary education, per student (% of GDP per capita) 7.6 11.3 — Secondary education, per student (% of GDP per capita) 20.7 20.2 — Health (% of GDP) 1.8 1.9 2.5 — Fuel subsidy (% of GDP) 1.3 3.0 — Government receipts and debts tax revenue (% of GDP) 10.4 13.0 12.1 — Net ODA (% of GNI) 14.0 8.5 8.5 5.1 External debt (% of GNI) 58.1 21.0 29.4 — Governance and government efficiency Government effectivness: percentile rank 14.0 33.0 14.8 35.6 Grigoli education efficiency score 0.19 0.27 0.28 Grigoli health efficiency score 0.91 0.94 — Public investment management index 1.4 1.4 — Memorandum item GNI per capita (constant 2005 US$) 264 410 643 Note: Recent refers to the latest year with data. If data are not available for 2000 or 2013, the closest earlier year with data is used; however, the data are never older than 1998 for “2000” or 2009 for “recent.” The year for country-specific data can be found in the respective graphs. Expected refers to the expected level of the indicator at the country’s GNI per capita, given the cross-country pattern between the indicator and GNI per capita. If the relationship is loose, the confidence interval for the expectation in question is relatively wide. Green = current value significantly above the expected level; red = current value significantly below the expected level; black = current value as expected; — = no projection because the cross-country relationship is not considered sufficiently tight (see criteria earlier in the note). translates into a reduction from 8.5 to 5.1 percent Uganda’s external debt stock is relatively low, and of GNI.14 However, the fact that cross-country the latest Debt Sustainability Analysis (DSA) patterns point to a likely decline in ODA does not suggests there may be room for increased exter- mean that such an outcome should be taken for nal borrowing.16 The fiscal impact would be fairly given: it depends on the priorities of donors and limited, however, as long as borrowing is limited their relationships with Uganda’s government. to what is consistent with debt sustainability. The The relationship between tax revenues and only other major receipt change, which is subject GNI per capita, as well as the debt stock and GNI to considerable uncertainty, is related to the oil per capita, is not tight enough to project expected sector. According to one set of projections, with changes. However, tax revenues are currently production starting in 2018, tax revenues from within the expected range and the government is oil will reach 8 percent of GDP by 2023, after committed to increase tax revenues, a change which they will decline gradually until 2045, that the IMF also includes in its projections.15 when production ends and reserves are depleted; Higher taxes or lower subsidies would both for the period 2016–30, oil revenues may amount reduce the resources controlled by domestic to an average of roughly 4.9 percent of GDP per households and firms, pointing to the need to year (IMF 2013b, p. 57).17 The advent of large oil consider the combined impact on SDGs and revenues may lead to further aid cuts as donors other indicators from higher taxes and the spend- turn to countries with more severe fiscal con- ing increases that are financed by these taxes. straints. In sum, future fiscal space may be 198 Uganda Figure 11.3  Uganda—Fiscal Space Indicators and GNI per Capita in a Cross-Country Setting a. Government spending: consumption (left), investment (right) 130 50 60 Consumption (% of GDP) 20 Investments (% of GDP) 30 9 2000 2000 2013 10 4 2013 2 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.01*** + .09** ln(GNI pc); R 2 = .042 ln(SDG) = 2.34*** + −.07 ln(GNI pc); R 2 = .012 b. Government spending: primary education, per student (left), secondary education, per student (right) 6 4 Secondary education (% of GDP) 4 Primary education (% of GDP) 2 2 2011 1.2 1.1 2009 .60 .40 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.07*** + −.08 ln(GNI pc); R 2 = .029 ln(SDG) = −.8** + .14*** ln(GNI pc); R 2 = .086 figure continues next page Uganda 199 Figure 11.3  continued c. Government spending: primary education (left), secondary education (right) 90 50 Secondary education, per student Primary education, per student 30 50 (% of GDP per capita) (% of GDP per capita) 20 2009 20 8 2012 10 2 4 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.72*** + .12** ln(GNI pc); R 2 = .066 ln(SDG) = 3.53*** + −.09 ln(GNI pc); R 2 = .030 d. Government spending: health (left), fuel subsidy (right) 20 20 1.6 Fuel subsidy (% of GDP) 2011 Health (% of GDP) 6 .10 1.6 2000 2012 .00 .50 150 400 1,000 3,000 8,000 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −.05 + .16*** ln(GNI pc); R 2 = .093 ln(SDG) = 4.96*** + −.65*** ln(GNI pc); R 2 = .098 figure continues next page 200 Uganda Figure 11.3  continued e. Tax revenue (left), official development aid (right) 60 40 2000 7 2012 30 Tax revenue (% of GDP) 2030 Net ODA (% of GNI) 20 1.2 2012 10 2000 .20 3 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 1.75*** + .13 ln(GNI pc); R 2 = .027 ln(SDG) = 8.26*** + −1.0*** ln(GNI pc); R 2 = .308 f. External debt (left), government effectiveness (right) 2,000 Government effectiveness: percentile rank 80 550 2000 2013 30 2030 External debt (% of GNI) 160 2000 10 50 4 2013 10 1.3 .50 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = 2.51*** + .14** ln(GNI pc); R = .035 2 ln(SDG) = 0 + .44*** ln(GNI pc); R 2 = .300 figure continues next page Uganda 201 Figure 11.3  continued g. Education expenditure efficiency (left), health expenditure efficiency (right) 1 Grigoli education efficiency score 1 Grigoli health efficiency score .60 .90 2010 .40 .30 2030 2010 150 400 1,000 3,000 8,000 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) GNI per capita (2005 US$) ln(SDG) = −4.0*** + .46*** ln(GNI pc); R 2 = .627 ln(SDG) = −.09*** + .01 ln(GNI pc); R 2 = .031 h. Public investment management index 4 Public investment management index 2 1.3 2010 .80 .30 150 400 1,000 3,000 8,000 GNI per capita (2005 US$) ln(SDG) = −.43 + .12** ln(GNI pc); R 2 = .078 Note: Highlighted observations are for Uganda in different years, while the nonhighlighted country observations are the most recent observation for other low- and middle-income countries. boosted by some combination of higher taxes the room for priority spending and enhance its and external borrowing while likely ODA impact on the SDG agenda. Table 11.2 displays changes would dampen such gains, leaving oil data for some measures of government efficiency. revenues as the main source of potentially sig- According to both the health and the education nificant fiscal space changes. indexes used in this study, Uganda’s performance Government efficiency is important to pro- is below the expected levels; among these two tect and, if possible, increase in order to add to indexes, GNI per capita is strongly correlated 202 Uganda with the education index but largely uncorrelated 4. Conclusions with the health index. Uganda is performing as expected in terms of the more general Public As summarized in table 11.3, Uganda’s current Investment Management Index and better than outcomes are better than expected (compared to a expected according to the World Bank typical country at the same GNI per capita level) Governance Indicators. Given that the different for 6 indicators, as expected for another 6, while it indexes measure different aspects of government falls short for the remaining 6, the latter covering performance, such mixed findings may not be shared prosperity, primary school completion, inconsistent. In addition, scatted Uganda-specific secondary school enrollment, HIV prevalence, survey evidence also points to inefficiencies.18 and access to electricity. By 2030, considerable In  sum, even though they are unpredictable, improvements are projected for most indicators. efficiency gains could potentially add consider- ­ However, compared to the post-2015 global ambi- able fiscal space. tions, the improvements are moderate. This means Decisions about the level and allocation of that, to get closer to the realization of these ambi- government spending should be made in light of tions, a break with the past is needed. government priorities and would depend on Table 11.3 further shows that, for most of the numerous factors that are well beyond the scope indicators, the relationship to GNI per capita is of this note, including government capacity in tight. Improvements in these SDGs will likely different areas and the scope to encourage com- continue along with projected GNI per capita plementary private sector activities. For example, growth and increases in resources and capabili- remittances from the country’s workers abroad ties. However, among these SDGs, secondary was a non-negligible 4.3 percent of GDP in 2013 enrollment, ratio of female to male secondary (roughly as expected for Uganda’s GNI per enrollment, access to sanitation, access to elec- capita  level); measures that encourage even tricity, and CO2 emissions have since 2000 higher ­levels of remittances and channel them to declined in ranking compared to other coun- income-raising investments may have high pay- tries while the GNI per capita ranking has stayed offs for the SDG agenda. More generally, given more or less the same. For these SDGs, the effi- strong linkages between private and government ciency of policies through which resources are activities and incomes, it is crucial that policies translated into SDG outcomes may need special and spending decisions promote a broad-based attention. change that encompasses services related to Uganda is also underperforming for two indi- human development, infrastructure investments, cators with a weak relationship to GNI per capita: and other measures in support of strong long- shared prosperity and HIV prevalence. In both of run  growth that is biased in favor of the less these cases, the decline in ranking since 2000 has advantaged. been larger than for GNI per capita. In general, Table 11.3  Uganda—Summary Results for SDG Indicators Cross-country relationship Overperforming As expected Underperforming with GNI/cap Tight •• Access to improved water source •• Poverty (+) •• Primary completion (+) •• Gross pre-primary enrollment •• Gross secondary enrollment (–) •• Access to improved sanitation (–) •• Ratio of female to male secondary •• Malaria (×) •• Road density enrollment (–) •• Access to electricity (–) •• Internet Users (–) •• Under-5 mortality rate (+) •• CO2 emissions (–) •• Maternal mortality rate (+) •• Road density Loose •• Ratio of female to male primary •• Secondary completion (–) •• Shared prosperity (–) completion •• HIV prevalence (–) Note: (+) = larger country rank improvement 2000–12 than for GNI per capita; (–) = smaller country rank improvement 2000–12 than for GNI per capita; (×) = the same country rank improvement 2000–12 as for GNI per capita. Uganda 203 the loose relationship to GNI per capita suggests GDP) may be exacerbated by higher oil revenues that these indicators should not be expected to as donors direct more resources to countries improve strongly or systematically to more rapid under more severe fiscal constraints. There may growth in GNI per capita but rather would depend be some room to raise additional resources from on country-specific circumstances and policy external borrowing and, to a more significant interventions. extent, from higher taxes. However, the net In terms of fiscal space, the main potential but impact on SDG progress from higher taxes would uncertain gains are likely to stem from oil reve- be positive only if, on the margin, the government nues and improvements in government efficiency. increases spending and tax revenues with a suffi- The anticipated decline in ODA (as a share of ciently high efficiency. Annex 11A: Data Sources Indicator Source Comment GNI per capita (constant WDI. API ref: GNI per capita (constant 2005 US$) 2005 US$) [NY.GNP.PCAP.KD] SDG indicators Poverty, at $1.25 a day WDI. API ref: poverty headcount ratio at $1.25 a day (PPP) (PPP) (% of population) (% of population) [SI.POV.DDAY] Shared prosperity: income WDI. API ref: income share held by lowest 20% + income share share for lowest 40% held by second 20% [SI.DST.FRST.20+SI.DST.02ND.20] Pre-primary enrollment WDI. API ref: school enrollment, pre-primary (% gross) (% gross) [SE.PRE.ENRR] Primary completion WDI. API ref: primary completion rate, total (% of relevant age (% gross) group) [SE.PRM.CMPT.ZS] Secondary enrollment WDI. API ref: school enrollment, secondary (% gross) [SE.SEC. (% gross) ENRR] Secondary completion EdStats. API ref: DHS: secondary completion rate [HH.DHS. Drawing on population, enrollment, and (% gross) SCR] repetition data in EdStats, data for Uganda was calculated for 2011. Primary completion, ratio of WDI. API ref: primary completion rate, female (% of relevant females to males (%) age group)/primary completion rate, male (% of relevant age group) *100 [SE.PRM.CMPT.FE.ZS/SE.PRM.CMPT.MA.ZS*100] Secondary enrollment, ratio WDI. API ref: ratio of female to male secondary enrollment (%) of females to males (%) [SE.ENR.SECO.FM.ZS] Under-5 mortality (per 1,000 WDI. API ref: mortality rate, under-5 (per 1,000 live births) live births) [SH.DYN.MORT] Maternal mortality (per WDI. API ref: maternal mortality ratio (modeled estimate, per 100,000 live births) 100,000 live births) [SH.STA.MMRT] Malaria cases HNP. API ref: malaria cases reported/ Population, total *100 (% of population) [SH.STA.MALR/ SP.POP.TOTL *100] HIV prevalence WDI. API ref: prevalence of HIV, total (% of population ages (% of population ages 15–49) [SH.DYN.AIDS.ZS] 15–49) Access to improved WDI. API ref: improved sanitation facilities (% of population with sanitation facilities access) [SH.STA.ACSN] (% of population) Access to improved water WDI. API ref: improved water source (% of population with source (% of population) access) [SH.H2O.SAFE.ZS] Road density (km per 100 WDI. API ref: road density (km of road per 100 sq. km of land sq. km of land) area) [IS.ROD.DNST.K2] Access to electricity WDI. API ref: access to electricity (% of population) (% of population) [EG.ELC.ACCS.ZS] annex continues next page 204 Uganda Indicator Source Comment Internet users (per 1,000 WDI. API ref: Internet users (per 100 people) [IT.NET.USER.P2] people) CO2 emissions (metric tons WDI. API ref: CO2 emissions (metric tons per capita) per capita) [EN.ATM.CO2E.PC] Fiscal space indicators Investment (% of GDP) WDI. API ref: gross fixed capital formation (% of GDP)-gross fixed capital formation, private sector (% of GDP) [NE.GDI.FTOT.ZS]-[NE.GDI.FPRV.ZS] Consumption (% of GDP) WDI. API ref: general government final consumption expenditure (% of GDP) [NE.CON.GOVT.ZS] Primary education (% of EdStats. API ref: total expenditure on educational institutions GDP) and administration as a % of GDP. All sources. Primary [UIS.XGDP.1.FDINSTADM.FFD] Secondary education EdStats. API ref: total expenditure on educational institutions (% of GDP) and administration as a % of GDP. All sources. Secondary and post-secondary non-tertiary [UIS.XGDP.234.FDINSTADM.FFD] Primary education, per WDI. API ref: expenditure per student, primary (% of GDP per student (% of GDP per capita) [SE.XPD.PRIM.PC.ZS] capita) Secondary education, per WDI. API ref: expenditure per student, secondary (% of GDP student (% of GDP per per capita) [SE.XPD.SECO.PC.ZS] capita) Health (% of GDP) WDI. API ref: Health expenditure, public (% of GDP) [SH.XPD.PUBL.ZS] Fossil fuel subsidy (% of IMF (2013a). Total pretax subsidy (% of GDP) Subsidies are measured using a GDP) price-gap approach capturing both consumer (including implicit) and producer (except those that arise when suppliers are inefficient and make losses at benchmark prices) subsidies. Negative external effects are not included in the pretax subsidy. Tax revenue (% of GDP) WDI. API ref: tax revenue (% of GDP) [GC.TAX.TOTL.GD.ZS] Net ODA (% of GNI) WDI. API ref: net ODA received (% of GNI) [DT.ODA.ODAT.GN.ZS] External debt (% of GNI) WDI. API ref: external debt stocks (% of GNI) [DT.DOD.DECT.GN.ZS] Government effectiveness: WGI. API ref: government effectiveness: percentile rank Captures perceptions of the quality of percentile rank [GE.PER.RNK] public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies. Grigoli education efficiency Grigoli (2014) Measure secondary education score inefficiency in terms of how much additional output could be achieved at current levels of spending. Grigoli health efficiency Grigoli and Kapsoli (2013) Quantifies the inefficiency of public health score expenditure using a stochastic frontier model that controls for the socioeconomic determinants of health. Public investment Dabla-Norris et al. (2011) Four phases associated with public management index investment management are covered: project appraisal, selection, implementation, and evaluation. Note: WDI = World Development Indicators, World Bank; API ref = reference and code when using the World Bank Open Data; EdStats = Education statistics, World Bank; HNP = Health Nutrition and Population statistics, World Bank; WGI = Worldwide Governance Indicators, World Bank. Uganda 205 Notes 1. Fiscal space matters since, while policy frame- expected for the specific country. The same obser- works and the engagement of the private sec- vation applies to expected values for fiscal space tor may vary widely, rapid progress on the SDG indicators. agenda will require efficient and carefully priori- 9. With regard to CO2, Uganda’s current and project tized public spending. 2030 per capita emissions are 1.1 and 4.1 percent 2. While a cross-country perspective provides an of the current OECD average. For Uganda, the rel- important complement to analysis that is centered atively high population density raises the need for on an individual country, it is by definition lim- and makes it easier to achieve high road density. ited to analysis of variables that are available in Note that for CO2 emissions when measured per cross-country databases. unit of GDP, Uganda is still overperforming (that 3. This does not mean that GNI per capita is viewed is, emitting less than expected). as a direct determinant of SDG outcomes; on the 10. If data are not available for 2000 or 2013, the clos- contrary, a major challenge for policy makers is to est earlier year with data is used; however, the identify policies that improve SDG performance data are never older than 1998 for “2000” or 2009 relative to what is expected given the level of GNI for “recent.” The year for Uganda specific data is per capita. A second challenge is to raise growth in reported in the graphs. GNI per capita as it indirectly influences country 11. For CO2, lower emissions per capita than expected SDG capacity. signals overperformance; similarly, the lower the 4. Sources for the indicators are presented in annex per capita emissions, the higher the percentile 11A. ranking of a country. 5. Projections from CEPII (v. 2.3) are used for this 12. The treatment is the same as in table 11.1 and and other Country Development Diagnostics related figures. That is, in table 11.2, projections applications given their wide country coverage are made only when the cross-country relation- and well-documented methodology; OECD data ship between the indicator and GNI per capita is have been used when projections have been miss- considered tight enough. Due to data limitations, ing. In the projections, it is assumed that future we focus on government spending indicators; GNI growth will coincide with future GDP growth country-specific analysis is needed to consider (both expressed in constant 2005 US$) given that policy in the context of the different roles of the that this is the variable that CEPII and other government and private services and spending. sources project. 13. Fuel subsidies are detrimental to the climate and 6. Given that (a) SDGs have extreme values (such as encourage technologies that are less labor inten- 100 percent for improved water access) and (b) the sive, tending to generate employment for fewer current SDG level never is exactly as expected workers at lower wages. given GNI per capita, the projected values grad- 14. To limit the ODA loss, it may be possible to tap ually converge toward the expected values. For into global initiatives such as the Global Fund example, for a country that overperforms in water to Fight AIDS, Tuberculosis and Malaria. Note access, as GNI per capita increases the extent of that net ODA measured per capita is also as overperformance gradually declines, so that when expected. the expected value is 100, overperformance has 15. IMF (2014e, p. 20) suggests that, by 2018/19, tax reached zero. revenues of 15.0 percent of GDP, up from their 7. A tight enough relationship is defined as an 2012/13 level of 12.6 percent, would be feasi- R2 > 0.3 (tight) or 0.3 > R2 > 0.1 (moderately tight), ble. Such an increase would still leave Uganda while R2 < 0.1 is defined as loose. within its expected range. The FY2014/15 bud- 8. In addition, the confidence interval is wide in get, approved by parliament, removed many tax the case of a loose relationship, suggesting that exemptions, and the government reiterated its any conclusion on over- or underperformance commitment to increasing the tax-to-GDP ratio is made with wide margins. Statistically, even by 0.5 percent of GDP per year over the medium though their confidence intervals are wide, as long term (IMF 2014e, pp. 4, 7). as the estimated coefficient linking GNI per capita 16. According to the recently updated IMF-World to the SDG indicator is nonzero, these values are Bank Debt Sustainability Analysis (DSA), Uganda closer than the cross-country average to what is remains at low risk of debt distress. However, the 206 Uganda debt service-to-revenue ratio is high owing to the 18. On any given day, roughly 15–20 percent of the relatively low revenues and the short maturity of teachers (including head teachers with super- domestic debt, posing some sustainability risks visory responsibilities) are absent, with illness (IMF 2014e, annex I). Uganda’s external public or accounting for an almost negligible share of publicly guaranteed debt is projected to increase absences (UNESCO 2014a, pp. 31 and 267–268). from 17.6 percent of GDP in 2014 to 24.9 percent Similarly, an analysis of local governments sug- in 2020 after which it decreases, reaching to 20.9 gests, if all districts could be brought up to the percent in 2035. health and education outcome-to-spending ratios 17. These projections of oil revenues are based on oil of the best performing districts, then about one prices that are well above the levels of the spring third of their budgets could be saved (World Bank of 2015. 2013b, p. xiii). Uganda 207 References ADB (Asian Development Bank). 2006. Key Indicators Hausmann, R., J. Matovu, R. Osire, and K. Wyett. 2014. of Developing Asian and Pacific Countries. Manila, “How Should Uganda Grow?” ESID Working Paper Philippines. No. 30, Effective States and Inclusive Development Dabla-Norris, E., J. Brumby, A. Kyobe, Z. Mills, and Research Centre (ESID), Manchester, UK. C.  Papageorgiou. 2011. “Investing in Public IMF (International Monetary Fund). 2012. Liberia Investment: An Index of Public Investment 2012 Article IV Consultation. IMF Country Report Efficiency.” IMF Working Paper 11/37. International 12/340. Washington, DC: IMF. Monetary Fund, Washington, D.C. ———. 2013a. Energy Subsidy Reform: Lessons and Devarajan, S. 2014. “What the 2004 WDR Got Wrong.” Implications. Washington, DC: IMF. Blog. http://blogs.worldbank.org/futuredevelopment​ ———. 2013b. “Uganda: 2013 Article IV Consultation /what-2004-wdr-got-wrong. and Sixth Review under the Policy Support Doing Business. 2014. Doing Business 2015: Going Instrument, Request for a Three-Year Policy Beyond Efficiency. Economy Profile 2015. Kyrgyz Support Instrument and Cancellation of Current Republic. Washington, DC: World Bank. Policy Support Instrument—Staff Report.” Public Foster, V., and N. Pushak. 2011. “Nigeria’s Infrastructure Information Notice and Press Release on the a Continental Perspective.” Policy Research Working Executive Board Discussion; and Statement by the Paper 5686, World Bank, Washington, DC. Executive Director for Uganda, Country Report Fouré, J., A. Bénassy-Quéré, and L. Fontagné. 2010. 13/215, July, Washington, DC. “The Great Shift: Macroeconomic Projections for ———. 2013c. Philippines: Article IV Consultations. IMF the World Economy at the 2050 Horizon.” CEPII Country Report 13/102. Washington, DC: IMF. Working Papers 2012–03. ———. 2013d. Kyrgyz Republic: 2013 Article IV Gable, S., H. Lofgren, and I. Osario-Rodarte. 2014. Consultation and Fourth Review under the Three- Country Development Diagnostics Post-2015: Year Arrangement under the Extended Credit Uganda. Washington, DC: World Bank. Facility, Request for Waiver of Nonobservance of a Gable, S., H. Lofgren, and I. Osario-Rodarte. 2015. Performance Criterion, and Request for Modification Country Development Diagnostics Post-2015. of Performance Criteria—Staff Report; Public Washington, DC: World Bank. Information Notice and Press Release on the Government of Ethiopia. 2010. Growth and Executive Board Discussion; and Statement by the Transformation Plan (GTP) 2010/11-2014/15. Executive Director for the Kyrgyz Republic. IMF Ministry of Finance and Economic Development, Country Report 13/175. Washington, DC: IMF. Government of Ethiopia, Addis Ababa. ———. 2014a. Jamaica: 2014 Article IV Consultation Grigoli, F. 2014. “A Hybrid Approach to Estimating and Fourth Review under the Extended Arrangement the Efficiency of Public Spending on Education in under the Extended Fund Facility and Request for Emerging and Developing Economies.” IMF Modification of Performance Criteria—Staff Report; Working Paper 14/19, International Monetary Press Release; and Statement by the Executive Fund, Washington, DC. Director for Jamaica. IMF Country Report 14/169. Grigoli, F. and J. Kapsoli. 2013. “Waste Not, Want Not: Washington, DC: IMF. The Efficiency of Health Expenditure in Emerging ———. 2014b. Nigeria 2013 Article IV Consultation— and Developing Economies.” IMF Working Staff Report; Press Release and Statement by the Paper  13/187, International Monetary Fund, Executive Director for Nigeria. IMF Country Report Washington, DC. 14/103. Washington, DC: IMF. Hausmann, R., C. A. Hidalgo, S. Bustos, M. Coscia, ———. 2014c. Peru: Staff Report for the Article IV S. Chung, J. Jimenez, A. Simoes, and M. A. Yıldırım. Consultation. IMF Country Report 14/21. 2011. “The Atlas of Economic Complexity: Washington, DC: IMF. Mapping Paths to Prosperity.” Center for ———. 2014d. The Federal Democratic of Ethiopia; International Development; John F. Kennedy Article IV Consultation—Staff Report; Press Release School of Government, Harvard University; and Statement by the Executive Director for the and  Macro Connections, MIT Media Lab, Federal Democratic of Ethiopia. IMF Country Cambridge, MA. Report 14/303. Washington, DC: IMF. References 209 ———. 2014e. Uganda: Third Review under the Policy Republic of Uganda. 2014. Uganda Vision 2040. Support Instrument—Staff Report and Press Release. Kampala. Country Report 14/344. Washington, DC: IMF. Task Force on Innovative International Financing ———. 2014f. Pakistan: First Review under the for Health Systems. 2009. “More Money for Extended Arrangement under the Extended Fund Health, and More Health for the Money.” World Facility, Request for Waiver of Nonobservance of a Health Organization, United Nations, New York. Performance Criterion and Modification of UN (United Nations). 2013. A New Global Partnership: Performance Criteria. IMF Country Report 14/1. Eradicate Poverty and Transform Economies Washington, DC: IMF. through Sustainable Development. Report of the ———. 2015a. Jamaica: Seventh Review under the High-Level Panel of Eminent Persons on the Post- Extended Fund Facility and Request for Modification 2015 Development Agenda, UN, New York. of Performance Criteria—Staff Report; Press ———. 2014. “Open Working Group Proposal for Release; and Statement by the Executive Director for Sustainable Development Goals.” Open Working Jamaica. IMF Country Report 15/95, Washington, Group of the General Assembly on Sustainable DC: IMF. Development Goals, Document A/68/970, UN, ———. 2015b. Peru: Article IV Consultation—Press New York. Release; Staff Report; and Statement by the Executive ———. 2015. Transforming Our World: The 2030 Director for Peru. IMF Country Report 15/133. Agenda for Sustainable Development. Finalized text Washington, DC: IMF. for adoption. ———. 2015c. Senegal: Staff Report for the 2014 Article UNESCO (United Nations Educational, Scientific and IV Consultation and Eighth Review under the Policy Cultural Organization). 2014a. Education for All Support Instrument—Staff Report; Press Release; Global Monitoring Report 2013/2014. Paris. and Statement by the Executive Director for Senegal. ———. 2014b. Statistical Tables for UNESCO (2014a). IMF Country Report 15/2. Washington, DC: IMF. h t t p : / / w w w. u n e s c o . o r g / n e w / f i l e a d m i n​ ———. 2015d. Pakistan: Sixth Review under the /­MULTIMEDIA​/HQ/ED/GMR/pdf/gmr2013​/2013​ Extended Arrangement and Modification of _Report​_Statistical_tables-School_year_ending​ Performance Criteria; Staff Report, Press Release _2011.zip. and Statement by the Executive Director for WHO (World Health Organization). 2010. “Constraints Pakistan. IMF Country Report 15/96. Washington, to Scaling Up the Health Millennium Development DC: IMF. Goals: Costing and Financial Gap Analysis.” MoE Ethiopia. 2013. Education Statistics Annual Background Document for the Taskforce on Abstract: November 2005E.C (2012/13 G.C). Addis Innovative International Financing for Health Ababa: Ministry of Education, Government of Systems, International Health Partnership, WHO Ethiopia. and World Bank. Moller, L. and K. Wacker. Forthcoming. “Ethiopia’s World Bank. 2003. Making Services Work for Growth Acceleration and How to Sustain It— Poor  People. World Development Report 2004, Insights from a Cross-Country Regression Model.” Washington, DC. Policy Research Working Paper, World Bank. ———. 2012. Sénégal: Développer les outils des institu- NEDA (The National Economic Development tions de l’État pour une gestion plus efficiente de la Authority). 2014. “Philippine Development Plan dépense publique au Sénégal—Revue des Finances 2011–16, Midterm Update; With Revalidated Publiques. Rapport 59695-SN, Washington, DC. Results Matrices.” NEDA, Pasig City. ———. 2013a. Financing for Development Post-2015. Okwero, P., A. Tandon, S. Sparkes, J. McLaughlin, J. G. Washington, DC: World Bank. Hoogeveen. 2010. “Fiscal Space for Health in ———. 2013b. World Development Indicators (WDI) Uganda.” Working Paper 186, Africa Human Database http://data.worldbank.org/data-catalog​ Development Series, World Bank, Washington, DC. /­world-development-indicators. Overseas Development Institute (ODI). 2014. The ———. 2013c. Republic of Uganda: Service Delivery Impact of 2004 World Development Report. http:// with More Districts in Uganda; Fiscal Challenges www.odi.org/events/wdr-conference-making​ and Opportunities for Reforms. Public Expenditure -services-work-for-poor-people. Review, Report ACS4421, Washington, DC: Ranganathan, R., and V. Foster. 2012. “Uganda’s World Bank. Infrastructure: A Continental Perspective.” Policy ———. 2013d. Ethiopia Economic Update II: Laying Research Working Paper 5963, World Bank Africa the Foundation for Achieving Middle Income Status. Region and Sustainable Development Department. Report 78501. Washington, DC: World Bank. 210 References ———. 2013e. Pakistan: Finding the Path to Job- ———. 2014b. 3rd Ethiopia Economic Update: Enhancing Growth—A Country Economic Strengthening Export Performance through Memorandum. PREM South Asia, Report 75521. Improved Competitiveness. Washington, DC: World Washington, DC: World Bank. Bank. ———. 2013f. Republic of Uganda: Service Delivery ———. 2014c. Senegal Economic Update: Learning with More Districts in Uganda Fiscal Challenges from the Past for a Better Future. Washington, DC: and Opportunities for Reforms. Public Expenditure World Bank. Review, Report ACS4421, June, Washington, DC. ———. 2014d. The Economic Impact of the 2014 Ebola ———. 2014a. “MDG Dashboard.” (accessed July 3), Epidemic: Short and Medium Term Estimates for http://data.worldbank.org/mdgs/trends-and​ West Africa. Report 91219. Washington, DC: -­projections-of-each-mdg-indicator-for-each​-country. World Bank. References 211 In September 2015, the United Nations General Assembly adopted the 2030 Agenda for Sustainable Development. Individual countries face the challenge of implementing strategies that help realize the ambitions of this agenda, embodied in the Sustainable Development Goals (SDGs). This book presents a Country Development Diagnostics Post-2015 framework, designed to assess country-level implications of the 2030 Agenda for Sustainable Development, and applications of the framework to ten countries. The framework helps policy makers identify policies that may accelerate progress on the SDGs, and analyze sources of fiscal space to finance additional spending. Current levels of SDGs and their determinants are benchmarked on the basis of country per capita income, making it possible to compare the focus country to other countries. On the basis of current prospects for income growth, the framework also projects likely SDG outcomes for 2030 in the absence of accelerated progress. SKU K8430