Growing Smarter WO RLD BA N K E A S T A SIA AND PACIFIC REGIONAL REP O RT S Known for their economic success and dynamism, countries in the East Asia and Pacific region must tackle an increasingly complex set of challenges to continue on a path of sustainable devel- opment. Learning from others within the region and beyond can help identify what works, what doesn’t, and why, in the search for practical solutions to these challenges. This regional flagship series presents analyses of issues relevant to the region, drawing on the global knowledge and experience of the World Bank and its partners. The series aims to inform public discussion, policy formulation, and development practitioners’ actions to turn challenges into opportunities. T I T L ES IN TH E SE R IES Growing Smarter: Learning and Equitable Development in East Asia and Pacific Riding the Wave: An East Asian Miracle for the 21st Century Live Long and Prosper: Aging in East Asia and Pacific East Asia Pacific at Work: Employment, Enterprise, and Well-Being Toward Gender Equality in East Asia and the Pacific: A Companion to the World Development Report Putting Higher Education to Work: Skills and Research for Growth in East Asia All books in this series are available for free at https://openknowledge.worldbank.org / handle/10986/2147. World Bank East Asia and Pacific Regional Report Growing Smarter Learning and Equitable Development in East Asia and Pacific © 2018 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 21 20 19 18 This work is a product of the staff of The World Bank with external contributions. The fi ndings, interpre- tations, 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 copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution —Please cite the work as follows: World Bank. 2018. Growing Smarter: Learning and Equitable Development in East Asia and Pacific . World Bank East Asia and Pacific Regional Report. Washington, DC: World Bank. doi:10.1596/978-1-4648-1261-3. 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 offi cial World Bank translation. The World Bank shall not be liable for any content or error in this translation. 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 con- tained 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 World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. ISBN (paper): 978-1-4648-1261-3 ISBN (electronic): 978-1-4648-1269-9 DOI: 10.1596/978-1-4648-1261-3 Cover photo: Students in class at the Banteay Dek Primary School, Cambodia. © Global Partnership for Education / Livia Barton. Used with the permission of Global Partnership for Education / Livia Barton. Further permission required for reuse. Library of Congress Cataloging-in-Publication Data has been requested. Contents Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Policies that promote learning: analytical framework for this report . . . . . . . . . . . . . . . . . . 1 The state of education in East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Align institutions to ensure basic conditions for learning . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Concentrate effective, equity-minded public spending on basic education . . . . . . . . . . . . . 14 Select and support teachers throughout their careers to allow them to focus on the classroom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Ensure that children are ready to learn in school . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Assess students to diagnose issues and inform instruction . . . . . . . . . . . . . . . . . . . . . . . . . 23 Charting the course ahead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 1 High-Quality Schooling and Economic Growth in East Asia and Pacific . . . . . . . . . . . . . 33 The region generated unprecedented and transformative economic growth . . . . . . . . . . . . 33 How has sustained high growth changed the region? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 A sound macroeconomic environment allowed human capital to drive growth . . . . . . . . . 36 Successful economies systematically decreased the distance to the technological frontier . . . . 41 Successful economies prepared early for the next phase in becoming a knowledge economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Human capital protects people from falling back into poverty . . . . . . . . . . . . . . . . . . . . . 44 The region’s legacy of equitable growth is under threat . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 v vi CONTENTS More remains to be done if countries are to avoid the “middle-income trap” . . . . . . . . . . 47 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 2 The State of Education in East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 East Asia and Pacific is home to a quarter of the world’s school-age children, most of whom are enrolled in school . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Education systems fall into four groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Parts of the region are facing a learning crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 The region has more than its share of top performers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 The equity of learning outcomes is greater in East Asia than in the OECD . . . . . . . . . . . . 65 Educational systems in the region’s low- and middle-income countries serve learners from the bottom 40 percent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 An emerging literature on the roots of high performance highlights elements that promote learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 SPOTLIGHT 1 Education in the Pacific Island Countries: Achievements and Challenges . . . . 75 Role of the church . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Attainment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Achievement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Results from Early Grade Reading Assessments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Public expenditure on education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 3 Institutional Alignment for Policy Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 Institutional alignment is critical to ensure that students learn. . . . . . . . . . . . . . . . . . . . . . 81 Systemic reform of education for improved learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Aligning political support for investment in education with jobs and social mobility. . . . . 84 Sound administrative systems start by ensuring that basic conditions for learning are in place . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Providing clear guidance to teachers through national curricula and textbooks. . . . . . . . . 87 Sequenced reforms allowed more complex and ambitious learning goals to be achieved. . . . .89 Institutional alignment and sequenced reforms helped high-performing economies to reach critical milestones in expanding access and improving quality . . . . . . . . . . . . . 92 Note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 SPOTLIGHT 2 What Lessons Can Be Drawn from Top Performing Systems’ Experience with Technical and Vocational Education and Training? . . . . . . 97 TVET was central to national education policy and national economic development strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 TVET 2.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 4 Spending Public Resources on Education Effectively . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Trends in public spending on education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Prioritizing public spending on basic education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 High-performing systems manage teachers and school infrastructure efficiently . . . . . . . 111 CONTENTS vii Enhancing the equitable distribution of resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 Conclusions and recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 SPOTLIGHT 3 How Have Governments Used Public-Private Partnerships to Improve Education Outcomes? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Does private enrollment increase access? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Does private provision of education improve learning outcomes? . . . . . . . . . . . . . . . . . . . 134 Do public-private partnerships in education reduce inequity?. . . . . . . . . . . . . . . . . . . . . . 135 What can governments do to promote public-private partnerships that lead to positive outcomes? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 5 Developing Skilled Teachers and Supporting Effective Teaching . . . . . . . . . . . . . . . . . . . 139 Coherent systems of teacher recruitment, development, and support . . . . . . . . . . . . . . . . 140 Strong supporting elements are aligned in ways that make the teacher’s job easier . . . . . . 150 What happens in the classroom? Insights into effective teaching . . . . . . . . . . . . . . . . . . . 152 Challenges in teacher reform. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Conclusions and recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 6 Getting Children Ready to Learn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Top Performing Systems invested in readiness to learn for lasting returns . . . . . . . . . . . . 165 As low-income countries in the region strive to emulate the success of Top Performing Systems, they need to consider the status quo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Low-income countries in the region lack certain key packages of services . . . . . . . . . . . . 172 The costs of inaction are high, and countries can afford to act . . . . . . . . . . . . . . . . . . . . 178 Conclusions and recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 7 Using Student Assessments to Improve Instruction and Learning . . . . . . . . . . . . . . . . . . 189 Multilevel measuring and monitoring of learning outcomes are critical . . . . . . . . . . . . . . 190 Top Performing Systems in East Asia have linked assessments with teaching . . . . . . . . . . 195 Keeping pace with shifting labor market priorities and evolving beyond cognitive skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 Conclusions and recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202 8 Charting the Course Ahead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 What has worked? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 Aligning policies in five core areas is key to success . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 Increasing the role of the private sector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 Keeping up with rapid changes in technology and labor markets . . . . . . . . . . . . . . . . . . . 208 A way forward . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209 A map of countries’ current systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212 viii CONTENTS Appendix Education Systems in East Asia and Pacific at a Glance. . . . . . . . . . . . . . . . . . . 213 Cambodia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215 Hong Kong SAR, China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216 Indonesia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218 Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Republic of Korea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 Lao People’s Democratic Republic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 Macao SAR, China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 Malaysia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224 Mongolia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 Papua New Guinea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226 The Philippines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228 Singapore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 Taiwan, China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230 Thailand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231 Timor-Leste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 Vietnam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 Statistical websites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 Boxes O.1 Complementing the 2018 World Development Report with regional lessons of success . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 O.2 Elements of policies and practices that promote learning. . . . . . . . . . . . . . . . . . . . . . . . 4 O.3 Nomenclature in this report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 O.4 Education and the extraordinary record of growth in East Asia and Pacific . . . . . . . . 11 1.1 Growth has been slower in the Pacific Island countries . . . . . . . . . . . . . . . . . . . . . . . . 34 1.2 Learning-adjusted years of schooling can reveal how effective and efficient school systems are . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 1.3 Gains in poverty reduction are fragile, particularly for people with little human capital . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 2.1 Educational attainment, achievement, and learning in provinces in China that do not participate in international assessments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 2.2 Vietnam’s 2015 PISA performance and the Escuela Nueva initiative . . . . . . . . . . . . . . 66 2.3 Educational attainment gaps between majority and minority ethnic groups in China, the Lao People’s Democratic Republic, and Vietnam . . . . . . . . . . . . . . . . . . 68 2.4 Disability and inclusion in Vietnam’s education system . . . . . . . . . . . . . . . . . . . . . . . . 69 S1.1 Grouping Pacific Island countries by level of performance . . . . . . . . . . . . . . . . . . . . . . 79 3.1 The original elements of Vietnam’s Fundamental School Quality Level standards, as designed in 2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 3.2 Elements of policies and practices that promote learning. . . . . . . . . . . . . . . . . . . . . . . 92 S2.1 World Bank engagement in TVET in China: the Liaoning Urban Construction School’s eco-laboratory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 4.1 Government spending on education as a share of GDP is a better indicator of education spending than is spending as a share of total government spending . . . 104 4.2 Myanmar is gradually increasing public spending on education . . . . . . . . . . . . . . . . 105 CONTENTS ix 4.3 Inexpensive, small changes in classrooms can make big differences in student performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 4.4 Spending on private tutoring is on the rise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 S3.1 What are public-private partnerships in education? . . . . . . . . . . . . . . . . . . . . . . . . . . 132 5.1 Policy misalignment hurt Indonesia’s education system and teachers . . . . . . . . . . . . . 140 5.2 Models of teachers observing teachers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 5.3 Subject and pedagogical knowledge are critical for effective teaching . . . . . . . . . . . . 153 5.4 Teaching practices in Guangdong, China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 6.1 What is readiness to learn? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 6.2 The early years are the cornerstone of readiness to learn . . . . . . . . . . . . . . . . . . . . . . 166 6.3 Child care services can improve women’s labor force participation: evidence from Indonesia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 6.4 Improving literacy and readiness across the Pacific through the PEARL program . . . 182 7.1 Multiple types of assessments can be used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190 7.2 Balancing the advantages and disadvantages of high-stakes exams . . . . . . . . . . . . . . 192 7.3 What does the literature say about the use of formative assessments and student performance? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 7.4 Using SABER to benchmark assessment systems in Vietnam and Shanghai, China . . . 196 7.5 Using formative assessments to inform teacher development and policy making in Shanghai, China. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197 8.1 The future requires both strong fundamentals and a wider view of skills . . . . . . . . . 209 A.1 Radar charts in this appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213 Figures O.1 Five policy domains promote learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 O.2 Sixty percent of students in East Asia and Pacific are in education systems that are in crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 O.3 Students in China and Vietnam are among the top performers in developing East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 O.4 Vietnam and Beijing, Shanghai, Jiangsu, and Guangdong (China) have more top performers than the United States on PISA mathematics assessments . . . . . . . . . . . . . . 9 O.5 In Vietnam and Beijing, Shanghai, Jiangsu, and Guangdong (China), even poor students learn more than their OECD counterparts. . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 BO.4.1 Growth in East Asia and Pacific has exceeded global averages for decades, 1961–2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 O.6 Stunting remains prevalent in many countries in East Asia and Pacific, despite decades of improvement, 1986–2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 O.7 Families do not have consistent service coverage between pregnancy and preschool. . . . 22 O.8 Closing the achievement gap between socioeconomic groups is affordable . . . . . . . . . 23 1.1 Growth in East Asia and Pacific has exceeded global averages for decades, 1961–2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 1.2 Without sustained high growth since 1960, economic output in East Asia and Pacific would have been less than 10 percent of its current level . . . . . . . . . . . . . . . . . 35 1.3 The share of people in East Asia and Pacific working in agriculture has fallen, and the share working in services has risen, 1991–2011 . . . . . . . . . . . . . . . . . . . . . . . 35 1.4 Educational attainment in East Asia is many times higher than it was two generations ago . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 x CONTENTS 1.5 Fast-growing East Asia and Pacific economies generally score well above the OECD average on PISA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 B1.3.1 Both extreme and moderate poverty in East Asia and Pacific are decreasing . . . . . . . . 45 1.6 Inequality in East Asia has been rising since 1990 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 1.7 To achieve the level of GDP that Taiwan, China, reached in the early 2000s, Vietnam needs to grow at an average rate of 7 percent a year through 2035 . . . . . . . . 48 2.1 A quarter of the world’s 1.3 billion school-age children live in East Asia and Pacific . . . . 52 2.2 Upper-secondary enrollment rates are high in Top Performing and Above-Average Performing Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 2.3 The gross enrollment rate in preschools in East Asia and Pacific has surged since 1980 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 2.4 The increase in preschool enrollment has not been even everywhere for everyone . . . . 55 2.5 Average years of schooling among women in East Asia rose sharply between 1950 and 2010 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 2.6 Students in China and Vietnam are among the top performers in developing East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 B2.1.1 Estimated mean years of formal schooling in China among adults 25 and older are converging to world and regional averages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 2.7 Some economies in East Asia and Pacific outperform the OECD, but others face a learning crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 2.8 Sixty percent of students in East Asia and Pacific are in education systems that are in crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 2.9 East Asia and Pacific economies struggle to varying degrees to ensure that students reach basic proficiency levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 2.10 Vietnam and Beijing, Shanghai, Jiangsu, and Guangdong (China) have more top performers than the United States on PISA mathematics assessments . . . . . . . . . . 65 2.11 In Vietnam and Beijing, Shanghai, Jiangsu, and Guangdong (China), even poor students learn more than their OECD counterparts . . . . . . . . . . . . . . . . . . 66 2.12 On the 2015 PISA science assessment, gender gaps are smaller than other kinds of gaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 2.13 In several economies in East Asia and Pacific, more than a third of second graders cannot read a single word . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 S1.1 Primary enrollment rates vary in Pacific Island countries . . . . . . . . . . . . . . . . . . . . . . 76 S1.2 Some Pacific Island countries are succeeding in keeping children in school throughout primary school . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 3.1 In economies with Top Performing Systems, the share of the population with no schooling plummeted between 1950 and 2010. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 3.2 Basic inputs can help to enhance learning outcomes: an example from Vietnam . . . . . 87 3.3 Timeline of selected educational developments in East Asia and Pacific . . . . . . . . . . . 90 B4.1.1 Spending on education as a share of government expenditure is largely uncorrelated with spending as a share of GDP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 4.1 Per student spending as a share of GDP per capita does not correlate with 2015 PISA science scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 4.2 Japan prioritized basic education in its earlier years of economic growth, 1960–2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 4.3 Annual per student spending in Japan and the Republic of Korea rose at all levels of education between 2000 and 2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 4.4 Government spending on education as a share of GDP per capita varies widely . . . . 109 4.5 Returns to education differ by level of schooling and country income group. . . . . . . 110 CONTENTS xi 4.6 Indonesia spends more public funds on higher education than on upper-secondary education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 4.7 The composition of spending on primary education varies widely across countries . . . 112 4.8 In Singapore, the proportion of recurrent and development education spending fluctuated over time, 1981–2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 4.9 Salaries of primary school teachers in East Asia rise markedly with seniority . . . . . . 113 4.10 Higher teacher salaries relative to GDP per capita are correlated with student performance on the 2012 PISA among wealthier economies but not among less wealthy ones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 4.11 In most of the region, student-teacher ratios declined at the primary and secondary levels between 1998 and 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 4.12 In about half the economies in the region, the school-age population has decreased since 2000 and is projected to decrease over the next two decades . . . . . . 116 4.13 Class sizes are larger in East Asia and Pacific than in most PISA-participating economies, but the student-teacher ratios are not higher . . . . . . . . . . . . . . . . . . . . . . 117 4.14 Learning outcomes in the region’s Top Performing and Above-Average Performing Systems are consistently more equitable than in other economies . . . . . . 119 4.15 In almost all economies, wealthier students are more likely than poorer students to use private tutors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 S3.1 There is little correlation in East Asia and Pacific between the share of private enrollment and the gross enrollment rate for preprimary education . . . . . . . . . . . . . 133 S3.2 The private sector helped to increase preprimary enrollments, but it alone cannot guarantee access, particularly in lower-income countries . . . . . . . . . . . . . . . . . . . . . 134 S3.3 Private enrollment is not correlated with access at the primary level, but it is correlated at the upper-secondary and, to a less extent, lower-secondary levels . . . . 135 S3.4 Regional assessments in the Pacific show an unclear impact of the school authority (governmental, nongovernmental) on students’ numeracy and literacy proficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 5.1 A positive correlation exists between the value that society places on teachers, and student learning outcomes as measured by PISA . . . . . . . . . . . . . . . . . . . . . . . . . 142 5.2 Teacher modesty in self-assessment of skills may facilitate or complement continuous professional development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 5.3 Teachers who meet more frequently for professional collaboration tend to have higher levels of self-efficacy (belief in their own abilities) . . . . . . . . . . . . . . . . . 149 5.4 Considering the direct or indirect role played by teachers in implementation can make policies more effective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 5.5 Teachers in Top Performing East Asia and Pacific systems tend to spend a smaller proportion of total working hours teaching and more time preparing lessons compared with teachers in other countries . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 5.6 The Top Performing Systems in East Asia and Pacific tend to use more elaboration and less memorization in teaching math . . . . . . . . . . . . . . . . . . . . . . . . 155 5.7 Memorization is less useful, and elaboration strategies are more useful, as math problems become more difficult . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 B5.4.1 Teachers in Guangdong, China, perform on average at higher levels on CLASS studies than do teachers in other countries . . . . . . . . . . . . . . . . . . . . . . . . . . 157 6.1 There is a trend of increasing private ownership among early childhood institutions in Japan, 1999–2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 6.2 Stunting remains prevalent in many countries in East Asia and Pacific, despite decades of improvement, 1986–2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 xii CONTENTS 6.3 Large shares of second-grade students in East Asia and Pacific cannot read a single word . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 6.4 Poor children score lower than nonpoor children on tests of executive function, and, in the absence of any services, the difference can grow over time . . . . . . . . . . . . 172 6.5 Rates of functional illiteracy among 15-year-old students vary widely in East Asia . . . 172 6.6 Five packages of services for children under six improve readiness to learn . . . . . . . . 173 6.7 Families do not have consistent service coverage between pregnancy and preschool . . . 174 6.8 Children from the poorest households are far less likely than children from the richest households to be able to count from 1 to 10 . . . . . . . . . . . . . . . . . . . . . . . 175 6.9 Access to both preschool and high-quality home care shows wide wealth gaps . . . . . 175 6.10 Home environments explain a substantial portion of the wealth gap in ability to count from 1 to 10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 6.11 Both parenting practices and participation in early childhood education and development services affect the development of four-year-olds in Indonesia. . . . . 176 6.12 Both home engagement and time in kindergarten affect the development of five-year-olds in Mongolia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 6.13 The quality of preschools varies widely across settings, and average quality is low . . . . 177 6.14 About 80 percent of preschools in rural Indonesia meet local standards, and fewer meet the minimum score for quality as outlined in ECERS-R . . . . . . . . . . . . . . 178 6.15 Access to preschool is associated with a range of child development outcomes in the Lao People’s Democratic Republic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 6.16 Access to preschool is associated with a range of child development outcomes in Indonesia, particularly for poor children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 6.17 Closing the achievement gap between socioeconomic groups is affordable . . . . . . . . 181 7.1 Almost all PISA-participating East Asia and Pacific economies use teacher-developed tests and use standardized tests less regularly . . . . . . . . . . . . . . . . 196 7.2 The use of standardized assessments varies within East Asia and Pacific. . . . . . . . . . 198 7.3 Poorer children in East Asia and Pacific attend schools that are less likely to publicize achievement results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 7.4 Even when competing for students, many schools do not publicize their achievement data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 8.1 Five policy domains promote learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 A.1 Average years of educational attainment in Cambodia, 1950–2010 . . . . . . . . . . . . . 214 A.2 Selected development indicators in Cambodia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 A.3 Average years of educational attainment in China, 1950–2010 . . . . . . . . . . . . . . . . . 215 A.4 Selected development indicators in China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216 A.5 Average years of educational attainment in Hong Kong SAR, China, 1950–2010 . . . 217 A.6 Average years of educational attainment in Indonesia, 1950–2010 . . . . . . . . . . . . . . 218 A.7 Selected development indicators in Indonesia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218 A.8 Average years of educational attainment in Japan, 1950–2010 . . . . . . . . . . . . . . . . . 220 A.9 Average years of educational attainment in the Republic of Korea, 1950–2010. . . . . 221 A.10 Average years of educational attainment in the Lao People’s Democratic Republic, 1950–2010 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222 A.11 Selected development indicators in the Lao People’s Democratic Republic . . . . . . . . 222 A.12 Average years of educational attainment in Macao SAR, China, 1950–2010 . . . . . . 223 A.13 Average years of educational attainment in Malaysia, 1950–2010 . . . . . . . . . . . . . . 225 A.14 Selected development indicators in Malaysia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 A.15 Average years of educational attainment in Mongolia, 1950–2010 . . . . . . . . . . . . . . 226 A.16 Selected development indicators in Mongolia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226 A.17 Average years of educational attainment in Papua New Guinea, 1950–2010 . . . . . . 227 CONTENTS xiii A.18 Selected development indicators in Papua New Guinea . . . . . . . . . . . . . . . . . . . . . . . 227 A.19 Average years of educational attainment in the Philippines, 1950–2010 . . . . . . . . . . 228 A.20 Selected development indicators in the Philippines. . . . . . . . . . . . . . . . . . . . . . . . . . . 228 A.21 Average years of educational attainment in Singapore, 1950–2010 . . . . . . . . . . . . . . 229 A.22 Average years of educational attainment in Taiwan, China, 1950–2010 . . . . . . . . . . 231 A.23 Average years of educational attainment in Thailand, 1950–2010 . . . . . . . . . . . . . . 232 A.24 Selected development indicators in Thailand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 A.25 Selected development indicators in Timor-Leste . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233 A.26 Average years of educational attainment in Vietnam, 1950–2010 . . . . . . . . . . . . . . . 234 A.27 Selected development indicators in Vietnam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 Tables O.1 Education systems in East Asia and Pacific can be categorized into four performance groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 O.2 PISA scores on science in East Asia and Pacific are higher than predicted based on per capita income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 O.3 A coherent system covers all aspects of the teacher career cycle . . . . . . . . . . . . . . . . . . 19 O.4 The status of policies and practices that promote learning varies within the region . . . 27 O.5 Concerted policy action and continuity of implementation drive systems improvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 1.1 Success stories of sustained high growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 1.2 Rate of return to an additional year of schooling in selected economies in East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 1.3 Demand for human capital in East Asia has risen even faster than supply . . . . . . . . . . 41 1.4 China is quickly becoming the world’s largest producer of PhDs in sciences and engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 1.5 Education of the household head is related to selected agriculture indicators in the Lao People’s Democratic Republic, 2008 and 2013 . . . . . . . . . . . . . . . . . . . . . . . . 46 2.1 Education systems in East Asia and Pacific vary widely in size . . . . . . . . . . . . . . . . . . . 52 2.2 East Asia and Pacific has 26 percent of global enrollment but only 13 percent of the world’s out-of-school children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 2.3 Access and coverage are significantly lower in countries with Below-Average Performing and Emerging Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 2.4 East Asia and Pacific economies have participated in various international assessments since 1995 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 2.5 Economies in East Asia and Pacific routinely score above 500 on PISA . . . . . . . . . . . . 58 2.6 Education systems in East Asia and Pacific can be categorized into four performance groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 2.7 Enrollments in East Asia and Pacific, by level of education and type of system . . . . . . 59 2.8 PISA scores on science in East Asia and Pacific are higher than predicted based on per capita income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 2.9 Girls in East Asia generally outperform boys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 S1.1 Early reading assessments in Pacific Island countries . . . . . . . . . . . . . . . . . . . . . . . . . . 78 4.1 Government spending on education as a share of GDP varies widely, even among economies with similar levels of performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 4.2 Share of public expenditure on educational institutions in Japan, the Republic of Korea, and the OECD, by level of education, 2013 . . . . . . . . . . . . . . . . . . . . . . . . 110 4.3 Most East Asian countries spend only a small share of GDP on preprimary education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 xiv CONTENTS B4.4.1 Use of private tutoring is high across the world . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 B4.4.2 Effect of one additional hour of private tutoring on PISA math scores in select economies in East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 4.4 Steps toward achieving more effective spending of public resources on education . . . 125 SB3.1.1 Public-private partnerships in education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 S3.1 The share of private enrollment has changed over the last two decades, 1990–2014 . 132 5.1 A coherent system covers all aspects of the teacher career cycle . . . . . . . . . . . . . . . . . 141 5.2 Different systems use different methods of selecting and filtering teachers . . . . . . . . . 143 5.3 Top Performing East Asia and Pacific systems tend to focus learning on fewer content standards and topics in textbooks compared with other countries . . . . . . . . . 151 5.4 Content and pedagogy needed for lesson preparation and execution . . . . . . . . . . . . . 154 5.5 Steps toward attracting and developing skilled teachers and supporting effective teaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 6.1 Long-term gains in test scores underscore the importance of investments in readiness to learn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 6.2 Steps toward supporting children’s readiness to learn . . . . . . . . . . . . . . . . . . . . . . . . . 184 7.1 The use of national assessments varies within the region . . . . . . . . . . . . . . . . . . . . . . 194 7.2 Select countries have changed policy based on data from EGRAs . . . . . . . . . . . . . . . 195 7.3 Steps toward systematically strengthening assessments . . . . . . . . . . . . . . . . . . . . . . . . 201 8.1 Policy actions in moving from Emerging Systems to Top Performing Systems. . . . . . . 207 8.2 The status of policies and practices that promote learning varies within the region . . . . 211 A.1 Selected indicators for Cambodia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 A.2 Selected indicators for China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215 A.3 Selected indicators for Hong Kong SAR, China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216 A.4 Selected indicators for Indonesia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218 A.5 Selected indicators for Japan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 A.6 Selected indicators for the Republic of Korea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 A.7 Selected indicators for the Lao People’s Democratic Republic . . . . . . . . . . . . . . . . . . . 221 A.8 Selected indicators for Macao SAR, China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 A.9 Selected indicators for Malaysia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224 A.10 Selected indicators for Mongolia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 A.11 Selected indicators for Papua New Guinea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226 A.12 Selected indicators for the Philippines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228 A.13 Selected indicators for Singapore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 A.14 Selected indicators for Taiwan, China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230 A.15 Selected indicators for Thailand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231 A.16 Selected indicators for Timor-Leste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 A.17 Selected indicators for Vietnam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 Foreword S ince 1960, economies in the East Asia leading to learning. The World Development and Pacific region have had both faster Report 2018: Learning to Realize Education’s economic growth and greater human Promise focuses attention on the typical edu- capital accumulation than any other. They cation system in the developing world—where have made large investments in improving the inequalities in learning outcomes are wide amount and quality of schooling to promote and improvements in systemwide learning are rapid and continual economic progress. For often slow. These two reports complement a handful of the region’s economies, success each other, with the present report centering raised both the supply of and demand for on policies and practices that have led national skilled labor and transformed many into pros- education systems in East Asia and Pacific perous and inclusive middle-income societies. to produce graduates with consistently high Too many countries in the region, however, learning outcomes, and to do so equitably. have fallen short of their economic aspirations Education holds promise for macroeco- and have failed to take advantage of educa- nomic growth and for individuals’ opportuni- tion’s promise. ties, especially among the bottom 40 percent of Both groups are eager to learn how they income earners. Knowledge of successful poli- can do better. Understanding the elements of cies and practices is vitally important for the success is a critically important policy priority. World Bank’s Twin Goals of inclusive growth Countries wanting to learn so as to fuel eco- and poverty reduction. Growing Smarter: nomic growth ask themselves, What policies Learning and Equitable Development in East and practices help to promote superior learn- Asia and Pacific focuses on the lessons that ing outcomes? And, what can governments have allowed the region’s economies not only do to consistently and equitably raise aggre- to avoid learning crises but also to build and gate learning in their national school systems? maintain high-performing education systems. Growing Smarter: Learning and Equitable Development in East Asia and Pacific pro- vides answers to these questions. Victoria Kwakwa The developing world is in the midst of Vice President a global learning crisis: in an unacceptably East Asia and Pacific Region high number of countries, schooling is not The World Bank xv Acknowledgments T his report was prepared by a team led the guidance received at the Quality by Michael Crawford, Amer Hasan, Enhancement Review from Rabia Ali, Samer and Raja Bentaouet Kattan. The team Al-Samarrai, Amanda Devercelli, Andrew comprised Sachiko Kataoka, Andrew Ragatz, Mason, Keiko Miwa, Sudhir Shetty, and Andrew Coflan, Elaine Ding, Courtney Venkatesh Sundararaman. Marie-Helene Melissa Merchant, Elisabeth Sedmik, and Cloutier, Tsuyoshi Fukao, Javier Luque, Anny Wong. Xiaoyan Liang served as co- Lars Sondergaard, An Thi My Tran, Binh task team leader of the report at the concept Thanh Vu, and Noah Yarrow made helpful stage. Work was conducted under the overall suggestions. guidance of Sudhir Shetty and Harry Patrinos. This report draws on background papers The team is grateful to Neda Bostani, Paul by Jimmy Graham, Sean Kelly, and Anny Cahu, Huma Kidwai, Kevin MacDonald, Wong. Nozomi Nakajima, and Yilin Pan for their A team at Communications Development, inputs. It thanks peer reviewers Cristian Inc., led by Bruce Ross-Larson and including Aedo, Rodrigo Chaves, Deon Filmer, Elena Jonathan Aspin, Joe Caponio, Mike Crumplar, Glinskaya, and Venkatesh Sundararaman and John Wagley, edited the report. for their advice. It gratefully acknowledges xvii Abbreviations BKKBN National Board on Family Planning (Indonesia) B-S-J-G (China) Beijing, Shanghai, Jiangsu, and Guangdong (China) CHIP China Household Income Project CLASS Classroom Assessment Scoring System ECED early childhood education and development ECERS-R Early Childhood Environment Rating Scale-Revised EGMA Early Grade Mathematics Assessment EGRA Early Grade Reading Assessment ESC Educational Service Contracting FIMS First International Mathematics Study FISS First International Science Study FSQL Fundamental School Quality Level GDP gross domestic product HSEP High School Equalization Policy IDEO Intergenerational Deaf Education Outreach LUCS Liaoning Urban Construction School (China) MEXT Ministry of Education, Culture, Sports, Science, and Technology (Japan) MICS Multiple Indicator Cluster Survey OECD Organisation for Economic Co-operation and Development PASEC Programme d’Analyse des Systèmes Éducatifs de la CONFEMEN PEARL Pacific Early Age Readiness and Learning PEMANDU Performance Management and Delivery Unit (Malaysia) PETS-QSDS Public Education Expenditure Tracking and Quantitative Service Delivery Study (the Philippines) PILNA Pacific Islands Literacy and Numeracy Assessment PIRLS Progress in International Reading Literacy Study xix xx ABBREVIATIONS PISA Programme for International Student Assessment PPP purchasing power parity R&D research and development S&E science and engineering SABER Systems Approach for Better Education Results SABER-EPS Systems Approach for Better Education Results-Engaging the Private Sector SAR special administrative region SEA-PLM Southeast Asia Primary Learning Metrics SIMS Second International Mathematics Study SIRS Second International Reading Study SISS Second International Science Study STEP Skills Toward Employability and Productivity TALIS Teaching and Learning International Survey TIMSS Trends in International Mathematics and Science Study TVET technical and vocational education and training VNEN Vietnam Escuela Nueva Overview Introduction skills, but they do not fundamentally alter the basic need for foundational skills or the One-quarter of the world’s school-age children processes for acquiring them. Reading is still live in East Asia and Pacific. About 40 percent the foundation for acquiring all other types of the region’s students are in school systems of knowledge—the cognitive equivalent of that perform well and allow them to learn as the opposable thumb. Students still must much as or more than students anywhere in master the fundamentals of math, logic, and the world. But tens of millions of others are in data analysis. Being able to communicate school but not learning. Up to 60 percent of effectively requires mastery of grammar and students in the region are in poorly perform- vocabulary—and years of practice in oral and ing school systems where performance in key written expression. Behavioral skills and the subjects is either low or unknown. Many of ability to work in teams improve through these students have learning outcomes that are structured practice and feedback. Resilience below basic proficiency levels and are greatly and grit remain the glue that supports the disadvantaged as a result. ongoing acquisition of skills and their effec- The impressive achievements of some low- tive application in the workplace. and middle-income countries in the region show that schooling in resource-constrained contexts can lead to learning for all. The pol- Policies that promote learning: icy lessons from countries that have improved analytical framework for this education quality while expanding access are report relevant and valuable to low- and middle- What policies and practices promote learn- income countries—in East Asia and Pacific ing in schools? What should a country do if it and elsewhere—to ensure that their students wants to achieve high and equitable learning learn. These lessons are all the more relevant outcomes? No single explanation covers all given the learning crisis facing many countries cases, but when countries focus on five policy in the region and across the globe (box O.1). domains and align 15 elements within them Education remains a long-term process (figure O.1), learning improves most. These of acquiring knowledge, skills, habits, and policies and practices promote learning by behaviors. Current labor market condi- improving the teaching and learning experi- tions require new types of knowledge and ence in classrooms. 1 2 GROWING SMARTER BOX O.1 Complementing the 2018 World Development Report with regional lessons of success The World Development Report 2018: Learning to expanded schooling and learning, and it showcases Realize Education’s Promise (WDR 2018) documents systems that implemented reforms at scale. By exam- the dire state of learning in low- and middle-income ining these experiences, the report provides both countries, noting poor learning outcomes, high diagnosis and detailed prescriptions for improving inequality, and slow progress (World Bank 2018b). It education systems within the region and across the provides evidence on the myriad ways in which edu- globe. cation systems in low- and middle-income countries The WDR 2018 provides a three-tiered model of are failing to ensure that students learn and calls for the ways in which countries can address the learn- renewed efforts to assess learning, act on the evidence ing crisis: assessing learning, acting on evidence, that assessments provide, and align actors for change. and aligning actors to make the system work. This Several national and subnational education sys- report presents a complementary framework that tems in East Asia and Pacific fit the characterization covers five policy areas. It begins with an alignment of a learning crisis. But others offer clear examples of of institutions, which creates an enabling environ- performance at the highest international levels—far ment that cuts across the remaining policy areas: above their predicted performance given their levels effective and equity-minded public spending, selec- of gross domestic product (GDP) per capita. tion and support of teachers, preparation of students This report complements the WDR 2018 in two for learning, and systematic use of assessments to ways. It focuses on the experiences of countries that inform instruction. Institutional alignment achieve such alignment, but too often inputs do not make it into schools and classrooms. The level of institutional alignment—the Students who lack a desk or a textbook and coherence of objectives and responsibilities teachers who lack a coherent curriculum or a especially as they relate to public spending, chalkboard cannot be expected to engage in teachers, readiness to learn, and assessment— meaningful classroom interactions that pro- determines how completely and effectively duce learning. policies are designed, implemented, adjusted, When goals and incentives are not aligned, and evaluated. The experience of high- efforts to achieve learning are undermined. By performing education systems in the region contrast, when different aspects of the educa- underscores the critical role that institutional tion system are in alignment, well-designed alignment and sound administrative sys- reforms that focus on teaching and learning tems play in delivering good-quality educa- can raise learning outcomes. Progress may tion. Institutional alignment facilitates policy sometimes be slow, but East Asia and Pacific coherence and ensures that policies, goals, provides evidence that success can accrue if and incentives across key domains are in sync, reform efforts are sustained. so that education systems achieve their core task of producing graduates with relevant knowledge and skills rather than just creden- Public spending tials. The same reform may succeed in a coun- try where institutional alignment is strong but Effective spending means that resources are fail in another where it is weak. spent to produce expected outcomes. When Institutional alignment allows sound fewer resources can produce the same out- administrative systems to develop and comes, the spending is called cost-effective or deliver the basic inputs and infrastructure efficient. In education, effective spending is needed for schools to function well. It may about outcomes in terms of access, learning, seem intuitive that all school systems should and equity. Top Performing Systems in East OVERVIEW 3 FIGURE O.1 Five policy domains promote learning UT INSTIT IONS PUB ENT LIC SM S as da vel at ts. at ins duc oho -sca pati ing om ith bl ss en cla cti ion pr ass in ro w u e m i rn Co PE ive tru at r le on S ss on al s ogr ess fo form ery ose arge rtic lea Sp tra sic e urce icts hin nc on l re d d ling SE en ba so istr be en te du s t In ev n l l h p ark en m Ch se ta . ND d pub ca o s ss fro Di atio oug chm an eff li tio ch e t. a AS ne n fal ec c s n. oo rn th n t m tiv pe Be ING th ely nd a at r c . a ar n e g in e a g te in ls d. rm LEARNING o g e a er a te un b pr by r. wh an ss ea d te a ded in lea . th ss ea cla . ar ck ea on d th te g ed g rn Fo nit nd ildh pm act ices d es rly eve cro se o t g an in ed din fe in ch cla res e xts cle ba th ch ing ed nd ier nd rs che of y t in rs rs er vi ng rv co s a c cu ve im oo en ors . A te ad en s a eas s a che ea ss lit in ee he tt ro di se d g er le eri oal bs tice tea s a t vene an s o de pr d t se to bi tra s p ac clu y p ovi ob REA ric p s. m n c ve ove edu rv de nd i ex g rs’ j rac ew me ecti ur om er oo . ce a hi lop th ca ice liv e c ro rch ssr he p t n eco sel Co ld m e tio s. an ina nee um ti re e q ea oo po b the ul ac o d n’s nt ua n rd h lo ss rv r ise ph fro lity Ra DIN ys m of r ica bi e t ssr up l rth m o RS ak cla S c c ES . HE p TO ep C M S a Ke A er nt E ls. En LEA Ce ur T oo s e ch th at RN a ll s th n eb c ei a si c cond in pla iti o n s f o r le a r n i n g a r e Asia and Pacific all adopted three principles for respected, prepared, selected and advanced in effective spending of public resources: prioritiz- their careers on the basis of merit, have clear ing public spending on basic education, man- learning goals and performance expectations aging essential inputs efficiently, and enhancing for students, and are supported in their work. the equitable distribution of resources. Teachers are a core element of East Asia and Pacific’s Top Performing Systems, which have established competent, qualified, and moti- Teachers vated teaching forces that promote sustained A common theme across high-performing learning. East Asia and Pacific systems pro- education systems is their investment in and vide many lessons for teacher recruitment, focus on teachers. Over time, systems per- selection, support, retention, and professional form best when they have teachers who are development. 4 GROWING SMARTER Readiness to learn development, countries need to measure learning to ensure that the benefits of educa- Readiness to learn, a multifaceted construct, is tion reach all students. It is not enough that as much about children’s readiness for school children are in the classroom—it is imperative as it is about schools’ readiness for children. to make certain that they are learning. This This holistic concept is essential to a stu- domain entails not only using assessments, dent’s success not just in primary school but but also having the right policies and frame- throughout life. This domain encompasses works to support a system of assessments. both the supply and the quality of services The Top Performing Systems in East Asia for children’s physical and cognitive develop- and Pacific have systematically used a mix of ment. Strong support for families’ efforts to assessments and their data to develop their assist in their children’s academic and socio- education systems, placing significant value emotional development pays high dividends on obtaining and using information about at low cost. Top Performing Systems in East student learning and on employing multiple Asia and Pacific have increasingly focused methods to assess student learning. on children’s physical and cognitive develop- No single formula exists for how to achieve ment, assessed and improved the quality of success. But high-performing systems share the services they offer, and coordinated actors common elements and overlap in key areas to deliver needed services. in their approaches to and implementation of policies (box O.2). Increased student learning Assessment does not immediately follow from the mere presence of any or all of these elements— Because the quality, not just the quan- indeed, their quality and the degree to which tity, of schooling is crucial for growth and they are aligned are critical. BOX O.2 Elements of policies and practices that promote learning The success of some education systems in East Asia • Make teachers’ jobs easier by providing clear and Pacific shows that students learn most when learning goals and uncluttered texts. efforts focus on five policy domains and align 15 ele- • Keep experienced teachers in the classroom and ments. These domains and elements are as follows: leading as peers and researchers. • Center teacher training on classroom practice and Align institutions to ensure basic conditions for the ability to teach the curriculum. learning: • Ensure that the basic conditions for learning are Ensure that children are ready to learn in school: in place in all schools. • Focus on physical and cognitive development from birth. Concentrate effective, equity-minded public spending • Assess and improve the quality of early childhood on basic education: education and development services. • Spend effectively. • Coordinate across actors to deliver needed services. • Concentrate public spending on basic education. • Channel resources to schools and districts that Assess students to diagnose issues and inform are falling behind. instruction: • Benchmark learning through participation in Select and support teachers throughout their careers international large-scale assessments. to allow them to focus on the classroom: • Diagnose cohort progress at every educational • Raise the selectiveness of who becomes a teacher. sublevel. • Support new teachers by observing classroom • Inform instruction with data from formative practices and providing feedback. classroom assessment. OVERVIEW 5 This overview is structured as follows. Improvements in outcomes for women The next section analyzes the state of educa- have also been positive. In 1950, the average tion in East Asia and Pacific and describes woman in the region had completed less than performance on international assessments. a year of schooling—well below the world The sections that follow examine each of the average for women of 2.5 years. Six decades five framework domains, elaborating on the later, the population was more than double, experience of Top Performing Systems and and the average attainment of women had describing the challenges of other countries increased to 7.4 years of schooling, catching in the region. The last section discusses how up to the global average for women. Today, countries can translate these findings into girls in most countries in the region enroll and strategies and actions that improve learning. stay in school as long as or longer than most boys and learn as much or more on average. The state of education in Education systems fall into four groups East Asia and Pacific Discussions of education quality some- A quarter of the world’s school-age times rely on assessment scores as measures children live in East Asia and Pacific— of student learning. Of students in East and most of them are enrolled in school Asia and Pacific, 55 percent are enrolled in The 331 million school-age children in East countries or regions that have participated Asia and Pacific represent about a quarter in at least one international standardized of the world’s school-age population. Most assessment since 2000. The Programme for school-age children are enrolled in school. At International Student Assessment (PISA) of the primary level, the 6 million primary-age the Organisation for Economic Co-operation children not in school represent just 3 percent and Development (OECD) and the Trends in of all primary-age students. Out-of-school rates International Mathematics and Science Study are higher at the secondary level, and some (TIMSS) provide comparable information countries have troublingly high secondary- on learning outcomes in reading, math, and school dropout rates. But the region has made science. Early Grade Reading Assessments good progress in getting children into school. (EGRAs) provide information on children’s Just 13 percent of the world’s out-of-school ability to read, but their results are not com- children live in East Asia and Pacific. parable across countries. The largest education system in the region Countries can be divided into four perfor- mance groups (box O.3 and table O.1): is China’s, with 182 million students in basic education (National Bureau of Statistics of • Top Performing Systems consistently China 2016). National systems in five coun- score more than half a standard deviation tries (China, Indonesia, Japan, the Philippines, above the average score for OECD mem- and Vietnam) each enroll more than 10 mil- ber countries (equivalent to 1.6 years of lion students. Ten countries have systems with schooling). fewer than 100,000 students. Tuvalu has the • Above-Average Performing Systems con- fewest students, with just 3,000. sistently score up to half a standard devia- East Asia and Pacific has made considerable tion above the average score for OECD progress in preschool enrollment. The region member countries. is home to roughly 119 million children of • Below-Average Performing Systems con- preschool age (3–6 years). In 1980, the gross sistently score at least half a standard enrollment rate for preschool was 13 percent; deviation below the OECD average. by 2014, it had risen to 76 percent. This is a • Emerging Systems do not regularly partici- much faster rate of progress than the global pate in globally comparable standardized gross enrollment rate in preschools, which rose tests, but evidence from other sources sug- from 21 to 48 percent over the same period. gests that learning is very modest. 6 GROWING SMARTER BOX O.3 Nomenclature in this report This report uses several terms with meanings that categorized by their scores. The term Emerging may not be familiar to some readers. The term econ- Systems refers to systems with no globally com- omies refers to nonstate areas and regions. It includes parable standardized test scores. These terms are entities such as Hong Kong SAR, China; Macao capitalized throughout the report to highlight the SAR, China; Taiwan, China; and the four regions of specific designations of systems with regard to test China that participated in the Organisation for Eco- scores. When references to performance are not cap- nomic Co-operation and Development’s (OECD’s) italized, they do not refer to this grouping by test 2015 Programme for International Student Assess- score. ment (PISA) exams. Countries, by contrast, refers to China has not participated in PISA as a country. recognized World Bank member nations. In 2009 and 2012, only the province of Shanghai The terms Top Performing Systems , Above- participated. In 2015, the more economically Average Performing Systems, and Below-Average advanced regions of China—Beijing, Shanghai, Performing Systems refer to the education systems Jiangsu, and Guangdong (B-S-J-G)—participated. of economies and countries that have participated in For convenience, this group is referred to as B-S-J-G PISA and the Trends in International Mathematics (China). The B-S-J-G provinces have only 15 percent and Science Study (TIMSS) since 2000 and are of China’s pretertiary student population. TABLE O.1 Education systems in East Asia and Pacific can be categorized into four performance groups Number of Number of students Total number students assessed assessed Share of of students by PISA/TIMSSa by EGRA cohort Group Economy (millions) (millions) (millions) tested (%)b Top Performing Systems Hong Kong SAR, China; Japan; 24.1 24.1 — 100 (average score = 556) Korea, Rep.; Macao SAR, China; Singapore; Taiwan, China Above-Average Performing China, Vietnam 198.7 39.7c — 20 Systems (average score = 512) Below-Average Performing Indonesia, Malaysia, Philippines, 92.3 92.3 21.7 100 Systems Thailand (average score = 406) Emerging Systems Brunei Darussalam; Cambodia; Fiji; 16.3 0.5 5.1 35 Kiribati; Lao PDR; Marshall Islands; Micronesia, Fed. Sts.; Mongolia; Myanmar; Palau; Papua New Guinea; Samoa; Solomon Islands; Timor-Leste; Tonga; Tuvalu; Vanuatu Subtotal 156.6 26.8 Total 331.4 161.7d 48 Sources: OECD 2015; TIMSS 2015; data from World Bank EdStats (World Bank, various years). Data are the latest available data for each economy. Note: The average score in Organisation for Economic Co-operation and Development countries was 497. — = not available. EGRA = Early Grade Reading Assessment; PISA = Programme for International Student Assessment; TIMSS = Trends in International Mathematics and Science Study. a. PISA is taken by 15-year-olds (typically grade 10). TIMSS is taken by students in grades 4 and 8. b. When sample-based tests are used, the share refers to the sample base measured. c. China as a country has not participated in PISA. Scores are for the more economically advanced regions of Beijing, Shanghai, Jiangsu, and Guangdong, which are home to 15 percent of China’s pretertiary student population. d. This figure combines the number of students tested by PISA/TIMSS with the 5.1 million students tested by EGRA. Countries are not double counted in totals. OVERVIEW 7 About 40 percent of tested students in with an average score above 550 points— the region are enrolled in systems that equivalent to 1.6 more years of learning have high learning outcomes than the average OECD member country. These systems enroll 24 million students, or High-quality test data suggest that roughly 7 percent of the region’s students. 64 million students in East Asia and Pacific All of the highest scorers are middle- or are learning at high levels, but that 98 million high-income countries. But some low- and are in systems in crisis (figure O.2). These middle-income countries perform well, too. data are based on a composite constructed Average performance in Vietnam and in average of PISA and TIMSS performance on B-S-J-G (China) surpassed OECD member the nine iterations of these assessments since countries (table O.2). These systems enroll 2000 (for PISA) and 2003 (for TIMSS).1 As about 40 million students, or 12 percent of of these dates, the two tests used a common the region’s students.2 Their performance scoring system (an average of 500 points, is proof of concept that a low- or middle- with a 100-point standard deviation). For income country can produce students who PISA, 30 points is equivalent to one year’s learn as much as or more than students from worth of learning. high-income countries. Figure O.3 shows the distribution of test The average score of students in Below- scores. It reveals that students in both devel- Average Performing Systems was 106 points oped and developing systems perform well on lower than that of their low- and middle- PISA and TIMSS. income peers in the Above-Average Performing East Asia and Pacific dominates the ranks group—a difference equivalent to more than of top scorers, with 6 of the top 10 and 8 three years of learning. Indonesia, Malaysia, of the top 20 scores since 2000. The Top the Philippines, and Thailand form this group. Performing Systems include seven economies Their 92 million students represent 27 percent of all students in East Asia and Pacific. FIGURE O.2 Sixty percent of students in East Asia and Pacific are in education systems that are in crisis A disproportionate share of students who perform at the highest PISA levels are from East Asia and Pacific Only 1 of 20 test takers attains the two high- est proficiency levels on PISA. Students from East Asia and Pacific represent 34 percent of 40% test takers, but 48 percent of students who (63.8 million) reach the two highest levels of proficiency in science and 40 percent of students who do so 60% (97.9 million) in math. Vietnam and B-S-J-G (China) com- bined have slightly fewer students than the United States, but twice as many top math performers (figure O.4). Students in systems with learning crises Learning outcomes in East Asia and Students in Top Performing and Above-Average Pacific are distributed across income Performing Systems quintiles more evenly than they are in the OECD Source: Calculations based on a composite constructed average of PISA and TIMSS performance on the nine iterations of these assessments since Students in every income quintile in Top 2000 (for PISA) and 2003 (for TIMSS). Performing Systems and Above-Average Note: PISA = Programme for International Student Assessment; TIMSS = Trends in International Mathematics and Science Study. Performing Systems score better than their 8 GROWING SMARTER FIGURE O.3 Students in China and Vietnam are among the top performers in developing East Asia and Pacific B-S-J-G (China) Vietnam Macao SAR, China Japan Taiwan, China Thailand Korea, Rep. Malaysia Hong Kong SAR, China Singapore Shanghai, China Indonesia Philippines 106 points; 3.5 years of schooling 350 400 406 450 500 512 550 556 600 650 Composite constructed average of PISA and TIMSS performance over all available iterations Below-Average Performing Systems Average score, Below-Average Performing Systems Non-East Asia and Above-Average Performing Systems Average score, Above-Average Performing Systems Pacific countries Top Performing Systems Average score, Top Performing Systems Sources: Calculations based on PISA and TIMSS scores on nine assessments since 2000 (for PISA) and 2003 (for TIMSS). Note: Figure shows composite constructed average performance score with mean of 500 points and standard deviation of 100 points. The Philippines has only participated in TIMSS. B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); PISA = Programme for International Student Assessment; TIMSS = Trends in International Mathematics and Science Study. TABLE O.2 PISA scores on science in East Asia and Pacific are higher than predicted based on per capita income Mean PISA science score in 2015 Difference between actual and predicted GDP per capita in 2015 or latest (2011 Prediction based Equivalent years Economy international dollars) Actual on income Score of schooling Top Performing Systems Japan 35,804 538 479 59 2.0 Korea, Rep. 34,387 516 477 39 1.3 Singapore 80,192 556 516 40 1.3 Above-Average Performing Systems B-S-J-G (China) 22,037 518 57 61 2.0 Vietnam 5,668 525 394 131 4.4 Below-Average Performing Systems Indonesia 10,385 403 422 −19 −0.6 Malaysia 25,308 443 463 −20 −0.7 Thailand 15,345 421 440 −19 −0.6 Source: OECD 2016–17. Note: The Philippines is not included because the country has not participated in PISA. B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); GDP = gross domestic product; PISA = Programme for International Student Assessment. counterparts in the OECD in math and sci- Students from the second-lowest income quin- ence. For Vietnam and B-S-J-G (China), this tile scored above 500 on the 2015 math and achievement is remarkable, given their lower science assessments (figure O.5). These results incomes and scarcer resources for education. indicate that, for many students in East Asia OVERVIEW 9 FIGURE O.4 Vietnam and Beijing, Shanghai, Jiangsu, and Guangdong (China) have more top performers than the United States on PISA mathematics assessments 450 400 350 300 Number 250 200 150 100 50 0 Population Students Top performers in Top performers in (millions) (millions) science (thousands) mathematics (thousands) United States Vietnam and B-S-J-G (China) Source: OECD 2016–17. Note: B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); PISA = Programme for International Student Assessment. FIGURE O.5 In Vietnam and Beijing, Shanghai, Jiangsu, and Guangdong (China), even poor students learn more than their OECD counterparts a. Reading b. Mathematics c. Science 600 600 600 581 576 558 548 575 549 574 534 555 533 533 537 519 512 514 525 516 539 534 535 522 503 503 520 506 500 494 500 500 507 485 488 508 476 PISA score PISA score PISA score 497 474 486 470 476 486 471 475 450 470 454 445 451 460 436 444 442 419 400 414 400 410 400 406 399 384 398 373 389 381 390 366 300 300 300 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 (bottom) (top) (bottom) (top) (bottom) (top) Wealth quintile Wealth quintile Wealth quintile OECD Top Performing Systems Above-Average Performing Systems Below-Average Performing Systems Performance of bottom 40 percent of students in Vietnam and B-S-J-G (China) Source: OECD 2016–17. Note: The economies included in Top Performing, Above-Average Performing, and Below-Average Performing Systems are listed in table O.1. The horizontal line in each panel indicates the performance of the bottom 40 percent of students in Vietnam and B-S-J-G (China), which exceeded or matched the OECD average in mathematics and science. The top and bottom wealth quintiles refer to children who come from families with the top 20 percent and the bottom 20 percent of household income, respectively. B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); OECD = Organisation for Economic Co-operation and Development; PISA = Programme for International Student Assessment. and Pacific, poverty is not educational destiny. High PISA subscores belie the myth of The quality of policies and practices and what rote learning happens in school—rather than spending or students’ socioeconomic background— Casual observers and stylized facts have some- determine how much students learn. times conspired to attribute the region’s high 10 GROWING SMARTER scores to overreliance on rote learning and a Differences in performance emerge lack of deep understanding of what is learned. after students enter primary school The empirical evidence shows these assertions PISA performance at age 15 represents the to be false. The three PISA subscores in math accumulation of high-quality schooling measure the ability to recognize and set up over many years rather than rote learn- problems, perform mathematical operations, ing or test-taking skills. Data from the and interpret the meaning and significance Young Lives initiative, which closely fol- of results. Vietnam scored above the OECD lows cohorts from birth through second- average on all three subscores in 2012. These ary school, show that when they start scores indicate mastery of a full range of primary school, Vietnamese children have superior math abilities for complex problems. cognitive skills and abilities that are simi- They are incompatible with rote learning lar to peers in three comparator coun- without conceptual understanding. tries. By third grade, however, Vietnamese students are way ahead of their low- and Roughly 60 percent of students in East middle-income peers in math. At ages 10 Asia and Pacific are in systems facing a and 12, the average Vietnamese student per- learning crisis forms better than all but the top students in Ethiopia, India, and Peru. The 92 million students in Below-Average Performing Systems have both low scores and low measured levels of learning. Students in the 90th percentile in these countries struggle Continuous improvement of to score as high as students in the 10th per- performance has accompanied centile in China and Vietnam. The distribu- “progressive universalism” tion of scores does not overlap with that of A recurring theme among top performers countries in the Top Performing Systems. In is continuous improvement, a trend that the worst cases, scores are little better than is evident from internationally comparable would be obtained by random guessing. standardized tests. Altinok, Diebolt, and Demeulemeester (2014) calculate long- term trends on quality of schooling for 24 Early Grade Reading Assessments mostly high-income economies with suffi- indicate serious learning challenges in cient data. The three highest average annual Emerging Systems growth rates of achievement belong to East Information on countries that do not partici- Asian economies: Singapore (0.98 percent); pate in international tests can be gleaned from the Republic of Korea (0.90 percent); and the results of reading assessments conducted Hong Kong SAR, China (0.55 percent). All in early grades, usually first to third grade. of these rates are three to six times the aver- EGRA scores are not comparable across coun- age rate of improvement (0.165 percent). tries, but the number of students who cannot Japan improved at about the average rate. read a single word at a given age provides a Thailand’s scores declined at an aver- general picture of educational performance in age annual rate of 0.26 percent. Box O.4 the early primary years. In Cambodia, Timor- details the region’s successes in sustaining Leste, and Vanuatu, more than 30 percent of economic growth and in improving educa- second graders cannot read a single word. tional outcomes. OVERVIEW 11 BOX O.4 Education and the extraordinary record of growth in East Asia and Pacific Continuous robust economic growth has made East Asia and Pacific. This stellar growth raised per cap- Asia and Pacific a high- and middle-income region. ita income by a factor of at least 10. The East Asia Since 1960, East Asia and Pacific has grown faster and Pacific region’s economy in 2015 was 10 times and sustained high growth longer than any other larger than it would have been if it had grown at aver- world region (figure BO.4.1). Progress has been age world rates since 1960. Today, these economies remarkable, especially among the region’s low- and account for 30 percent of global output (up from just middle-income economies, which grew at more than 7 percent in 1990). twice the world average in 1960–2015 (7.2 vs. 3.5 As recently as 1991, two-thirds of East Asians percent). Even excluding China’s spectacular growth, worked in agriculture, most as low-income small- low- and middle-income countries in East Asia and holders; by 2012, that figure had dropped to one- Pacific grew more than 2 percentage points faster third. Rising formal employment, wages, and than the world average for nearly half a century. In productivity have made the typical East Asian an 1970–2010, growth among low- and middle-income educated urban dweller rather than a farmer with countries in East Asia and Pacific was almost twice little schooling. the world average (5.9 vs. 3.1 percent). No other low- Countries pursued a broad set of complementary or middle-income region comes close to matching this policies to accelerate growth, with education at the record of steady and rapid long-term growth. forefront. To sustain high growth rates, governments In some cases, growth transformed countries insulated technocratic policy makers from politics from poor agricultural societies to modern knowl- and allowed a set of policies to be consistently pur- edge economies. Their success deeply shaped the core sued. Policy makers tried to reduce inequality, first advice from economists and policy makers on how by boosting rural incomes and then by promot- to achieve prosperity. Nine of the 13 economies stud- ing educational opportunity and outcomes. Policies ied by the Commission on Equitable and Sustainable also improved labor force abilities and skills, mostly Growth (the Growth Commission) were in East through increased schooling, and made education FIGURE BO.4.1 Growth in East Asia and Pacific has exceeded global averages for decades, 1961–2015 10 9 8.9 Annual economic growth (% of real GDP) 8.4 8 7.8 8.0 7.7 7 7.2 7.2 6 5.3 6.1 5.3 5.6 5 5.0 5.1 5.0 5.1 4.6 4.7 4.6 4 3.9 3.6 3.5 3 3.0 2.9 2.7 2.6 2.4 2 1.8 1.5 1 0 1961–70 1971–80 1981–90 1991–2000 2001–10 2011–15 World Low- and middle-income East Asia and Pacific East Asia and Pacific Low- and middle-income East Asia and Pacific (without China) Pacific Island countries Source: Data from World Development Indicators (World Bank, various years). Note: Figures are based on real gross domestic product (GDP) in U.S. dollars. Base year is 2010. Low- and middle-income East Asia and Pacific includes all countries and economies in the region except Japan, the Republic of Korea, and Singapore. Pacific Island countries include Fiji, Kiribati, the Marshall Islands, the Federated States of Micronesia, Nauru, Palau, Samoa, the Solomon Islands, Tonga, Tuvalu, and Vanuatu. box continues next page 12 GROWING SMARTER BOX O.4 Education and the extraordinary record of growth in East Asia and Pacific (continued) broadly relevant to current and expected future years. This increase in average schooling occurred as economic challenges. Governments also sequenced the population more than doubled. reforms and took action for the next phase before the Trends in attainment continue to climb, with more benefits of the current phase were exhausted. and more students completing secondary school and Some countries went from rural agricultural soci- proceeding to tertiary studies. Schools today provide eties to high-tech knowledge economies. Japan, the twice as many students with more than six times as Republic of Korea, and Singapore set their education much instruction. For 40 percent of the region’s stu- policy goals within a larger framework that sought dents, this expansion was accompanied by high levels to eliminate technology gaps with the world’s most of learning. For the remaining 60 percent, the systems advanced countries. Their goal was to create domestic in which they study still struggle to ensure that more capacity to produce knowledge and technology that schooling equals more learning. was new to the world. Long-term increases in produc- Demand for educated workers has increased more tivity depend on continuously improving and apply- rapidly than supply. Income and the ability to earn ing new technologies, which in turn increases demand it are the keys to poverty reduction. Nowhere has for more highly skilled workers. Once countries this been truer than in East Asia and Pacific over the imparted basic skills to one cohort of workers and past several decades. Much of the value of education put them successfully to work, they raised the skills lies in its ability to make workers more productive, bar for the next cohort. raising their incomes. As educated workers become Education raised productivity among farmers and more numerous and common, their wages will fall promoted structural transformation. Investments in if all other things remain equal. But all other things education paid off at all education and income levels, have not remained equal. Amid the massive increases not just for people who worked in high-tech jobs and in educational attainment, the demand for educated industries. Countries pursued policies in agriculture to workers has risen more rapidly than the supply. Wage create environments for the poorest and most vulner- premiums have remained stable or risen, even though able to lift their income by raising their productivity. many more educated workers are seeking work. Rural dwellers with education—even when limited to The legacy of equitable growth is now under a few years of primary school—consistently outpro- threat. Since 1990, Gini coefficients and other duced and outearned their less educated neighbors. measures of inequality have shown a growing gap Poverty rates dropped substantially as jobs and between rich and poor. In countries such as China income-earning opportunities grew. Growth has been and Indonesia, distorted access to high-quality educa- accompanied by unprecedented drops in poverty tion is driving inequality. Additional efforts to make rates and the near elimination of extreme poverty high-quality education available to everyone will help in many countries. In 1981, as many as four of five to stem and reverse rising inequality. people in East Asia and Pacific lived on less than the Policies of “progressive universalism” were key extreme poverty threshold of US$1.90 a day purchas- to ensuring equitable distribution of educational ing power parity (PPP). In 2017, fewer than 2 percent opportunities at the outset. If investments in educa- of people lived in extreme poverty, and three of five tion do not keep pace with demand, technological were economically secure. Still, gains are fragile, and change promotes inequality, as income accrues to a further progress is needed to consolidate them. Even small group of highly skilled workers. Economies though national poverty rates have declined sharply, were able to reduce inequality while they grew rap- many people risk slipping back into poverty. idly partly because of the equitable distribution of Educational attainment increased dramatically to basic educational opportunities. Policies of progres- converge with the global average. In 1950, the aver- sive universalism—focusing on primary and lower- age adult in East Asia and Pacific had only 1.3 years secondary education for all—were a key means of of schooling—less than half the prevailing world aver- ensuring that skills grew in response to increased age of 2.9 years. By 2010, average attainment was demand. more than six times higher than it had been and con- Sources: Barro and Lee 2013; Dollar and Kraay 2002; World Bank 2014; World verged on the world average, which had risen to eight Bank 2018a. OVERVIEW 13 Align institutions to ensure basic were able to accelerate the structural trans- conditions for learning formation of their economies, but they also had few social safety nets. Secure employ- ment in the modern industrial sector served as both a ladder for social mobility and a cush- UT I ONS INS T I T ion against the lack of government-provided P UB LIC social safety nets. Policies offered a vision in NT ME SP which parents saw their children securing jobs S ES EN after completing increasing years of school. ASS D IN G Initial successes in employment of graduates in these newly created industries reinforced LEARNING both the demand for schooling and the value READ parents placed on achievement for their children. INE RS HE SS TO AC LEA RN TE Sound administrative systems start by putting in place the basic conditions for learning Research shows a positive and statistically significant relationship between basic inputs The experience of high-performing education such as blackboards, libraries, and school systems in East Asia and Pacific underscores infrastructure (including walls, ceilings, and the critical role of institutional alignment and roofs) and learning outcomes. sound administrative systems at a variety of Experience in the region indicates that a levels to implement policies and reforms that single national curriculum was critical to the ultimately raise learning outcomes in schools. success of the Top Performing Systems. These Institutional alignment is a critical component curricula generally focused on a clear and of ensuring that policies across domains are unambiguous set of learning goals. Unified synchronized and then implemented, adjusted, curricula were part of the trend of simplify- evaluated, and revised to foster continuous ing the educational endeavor, especially when improvement. From ensuring safe, adequate capacity was low, allowing these systems to physical space for students to developing cur- focus on a narrower set of goals. ricula with a framework for learning, insti- Just as governments exercised firm control tutional alignment can be a key determinant over the curriculum and dictated qualifica- of how much the ideas that underpin policies tion standards for teachers, many mandated are translated into reality for teachers and a single set of textbooks, which some experts students in classrooms. Students who lack believe helps to ensure that the prescribed desks or textbooks or teachers whose training curriculum is implemented across the system. is unrelated to the demands of delivering the This approach better fulfills the promise of curriculum cannot reasonably be expected to equity and quality in education when teacher engage in meaningful classroom interactions competency is low and the capacity to train that produce learning. teachers is limited. Political support for investment in Strong institutions allowed systems to education should be anchored on jobs expand access and improve quality and social mobility Korea and Singapore established goals for High-performing economies in East Asia and compulsory education in the 1950s and 1960s, Pacific had great success in creating industries respectively. They took no more than five that offered employment at scale and thus years to achieve universal primary education. 14 GROWING SMARTER For Korea, evidence of progressive universal- emerge among the four groups of coun- ism is apparent in its sequential expansion tries, although historically the region’s high- of first primary and then junior-secondary performing countries spent a large share of school. China and Vietnam launched similar government resources on basic education. goals almost a decade apart and fulfilled them Public spending per student continued to in 2000. Public spending signified the state’s grow in real terms, even as it moderated as a desire to ensure that basic education was pro- share of GDP and government spending. vided broadly and was high quality. Solid initial public investment among high- performing economies ensured strong foun- dations for education systems later. Singapore Concentrate effective, spent almost a third of its national budget equity-minded public on education in 1952. This share declined spending on basic education steadily as income rose. It now stands at just over a fifth. In Korea, education accounted for 14.3 percent of the total budget in 1963; spending grew to 20.4 percent by 2000 PUB before falling to 12.8 percent in 2013 (OECD LIC SP 2016b; Wong 2017). In Japan, 14.5 percent of government expenditure went to educa- EN NT ME DI tion in 1955. Spending stayed at that level for S NG ES much of the next 30 years, before declining to ASS 8.1–9.3 percent in 2009–13, one of the lowest rates among OECD countries (OECD 2016a; LEARNING Wong 2017). READ High-performing economies prioritized INE RS public spending on basic education HE SS TO AC LEA RN TE High-performing economies in East Asia INST S and Pacific sequenced their investment ITUTION focus from basic to tertiary education over time. Jimenez, Nguyen, and Patrinos (2012) Across the region, countries that spent edu- argue that countries that aim to build strong cational resources effectively concentrated on human capital for economic growth should three key tasks: prioritizing basic education, prioritize spending public resources on basic managing essential inputs, and spending to education to deliver good-quality and uni- promote equity. They also recognized that the versally available education at that level quality of spending, rather than the quantity, before devoting more spending to higher lev- has the greatest impact on learning. They els of education. therefore avoided setting artificial or arbitrary As the economies of the Top Performing targets for allocating a certain share of GDP Systems grew and demand for highly skilled or public expenditure to education. workers rose, they directed increasing shares of education spending to higher levels of edu- Public spending on education does not cation. Their central control of the educa- correlate with learning tion budget enabled them to ensure sustained investments and often provided direct input Public spending on education as a share of into how resources were spent. This influence GDP varies widely, worldwide and within helped to keep schools and districts account- East Asia and Pacific. No clear patterns able for results. OVERVIEW 15 Historically, most wealthier countries and Pacific—less than 3 percent on average, used to allocate more to lower levels of edu- against 6 percent in most Western European cation. However, Vietnam still prioritizes countries and 8 percent in the United States public investment in primary and second- (Wong 2017). ary education more than the Top Performing Teachers’ salaries correlate with student Systems do. China also prioritizes invest- performance in economies with per capita ment in primary, vocational, and preschool GDP above US$20,000 a year. Vietnam has education (OECD 2016c). been a much better performer in PISA than Even as they increased spending on Thailand, where teachers are better paid than higher levels of education, high-performing in Vietnam. In Indonesia, to meet the 2002 economies continued raising per student constitutional mandate to allocate 20 percent spending on primary and secondary educa- of the government budget to education, tion in absolute terms to enhance the qual- teacher salaries increased sharply over the ity of education at those levels. Korea and last decade, but without observable gains in Singapore doubled real spending per student learning outcomes (World Bank 2013). In on basic education, and absolute spending Malaysia, teachers earn more than twice as per student rose in Japan between 2000 and much as GDP per capita, but student per- 2013. In Japan and Korea—where tertiary formance is worse than in Thailand, where education is largely privately financed— teachers earn 25 percent more than GDP per public per student spending on tertiary edu- capita. cation has never exceeded spending for basic Class sizes in high-performing systems education. tend to be larger than the global average, but student-teacher ratios are not higher than the corresponding OECD average. Countries that High performers managed essential lowered student-teacher ratios well below inputs efficiently OECD averages did not enjoy improved High-performing East Asia and Pacific sys- student performance. Since 2000, both tems manage two essential financial inputs Indonesia and Malaysia have reduced their efficiently: spending on teachers and spend- ratios by more than 50 percent—to below the ing on school infrastructure. They allocate OECD average of 16 students per teacher in enough resources to attract and retain the secondary school—without improving learn- best staff, with salaries and benefits that ing outcomes. appropriately reward experienced teachers A lack of basic school facilities remains a with proven classroom performance. challenge throughout East Asia and Pacific, Singapore adjusts salaries for teachers except among its top performers. Reasons frequently, offers other compensation, and for poor school conditions may include links bonuses to performance appraisals. In insufficient public spending on school infra- Korea, teachers with more than 15 years of structure, limited access to water and elec- experience outearn their peers in many pri- tricity in rural areas, and difficult and costly vate sector jobs. In both Japan and Korea, construction conditions. Many schools in teachers with more than 15 years of expe- Indonesia and the Philippines do not meet rience (and whose performance has been basic standards for sanitation facilities, desks, routinely assessed) enjoy salaries that are, chairs, or sufficient space per student.3 In respectively, 125 and 140 percent of per the Lao People’s Democratic Republic, only capita GDP—far higher than the OECD 32 percent of schools have handwashing facil- average of 107 percent. The high reward ities and only 29 percent have working elec- for experience is a likely reason for the tricity (light) in classrooms (Demas, Khan, extremely low annual teacher attrition rates and Arcia, forthcoming). Rural schools in in high-performing economies in East Asia Thailand that serve the most disadvantaged 16 GROWING SMARTER students fall dramatically short of having the Select and support teachers adequate facilities and conditions that urban throughout their careers to allow schools have (World Bank 2015). them to focus on the classroom Top Performing Systems spent to TUT I ONS promote equity INST I PUB N T LIC In the East Asia and Pacific region’s Top ME SP S Performing Systems, the central government EN ES ASS D IN plays a key role in equalizing education fund- G ing across the country. In Japan, the central government subsidizes prefectures (equiva- LEARNING lent to states or provinces) to equalize public READ resources. For nine-year compulsory educa- INE tion, prefectures fund two-thirds of the cost SS of teachers’ salaries, and the central govern- TO RS LEA HE ment subsidizes the remaining third, to help RN AC to equalize the quality of teachers across TE municipalities and schools. Disadvantaged schools have the same share of qualified teachers as advantaged schools and more teachers per student. At the upper-secondary level, students from low-income families are At the heart of high-performing education exempt from tuition fees for public schools; systems is coherence in the recruitment, devel- they receive financial support to pay tuition opment, and support of teachers. Policies and fees for private schools and scholarships practices start from the premise that teaching to cover financial obligations other than is a difficult but learnable skill. Recruitment tuition costs, such as school trips and text- and selection of talented individuals are con- books. In Singapore, the government pro- sidered the beginning of a process in which vides merit-based scholarships and other new teachers learn their craft. Observation, financial assistance for all students as well as collaboration, and feedback are integral tuition subsidies for students from low- and parts of career-long professional development middle-income families to attend independent centered on acquiring and refining pedagog- schools (National Center on Education and ical and content knowledge to improve con- the Economy, n.d.). tinuously the quality of instruction. Career Of the Above-Average Performing Systems, advancement depends, among other things, Vietnam allocates more spending per capita on evaluation of teaching performance. to geographically disadvantaged provinces Career paths allow teachers to be promoted and districts and pays teachers serving in dis- and increase their salaries while remaining in advantaged areas higher salaries than teachers the classroom. Curricula and textbooks align in cities, through various types of allowances. in ways that enhance a teacher’s ability to In China, reducing inequalities in education deliver high-quality instruction. is a government priority. The government has gradually integrated the compulsory Top Performing Systems are more education funding guarantee in rural areas. selective in recruiting and retaining By 2010, 97 percent of the total educational teachers investment in rural compulsory education came from the government budget (OECD As universal access to basic education has 2016c). become the norm, low- and middle-income OVERVIEW 17 countries have had to increase massively the Teachers regularly collaborate with size of their school systems and the number others and receive feedback on their of teachers. Despite the need for more teach- performance ers, effective systems raised selectivity, making salaries and working conditions attractive, Teaching is a “closed-door” profession in so that talented individuals would apply. many OECD countries, where 40 percent Candidates are usually screened and filtered of teachers never teach alongside another both when selected into preservice teacher teacher, observe another teacher, or receive training programs and when hired. feedback. Top Performing Systems—and In Japan, only 14 percent of all appli- increasingly Above-Average Performing cants to education programs are accepted, Systems—treat the classroom as a public and only about 30–40 percent of graduates space and make teacher observation and are hired (Center on International Education feedback routine quality-promotion activi- Benchmarking, n.d.). As a result, newly hired ties. Special attention to observation is part teachers represent only 5 percent of the appli- of teachers’ induction into the profession— cant pool. In Singapore, the government the time when it is most critical to refine, recruits the top third of graduates of universi- improve, or correct teaching practices. ties and polytechnic schools to become teach- Japan’s induction period is designed ers (Tan and Wong 2007). In Korea, teacher around observation, with many demonstra- education programs admit only the top tion lessons conducted in front of panels for 10 percent of high school graduates, and only evaluation and feedback. Shanghai schools 1 in 20 passes the arduous exams to become have lesson observation rooms where lessons a teacher (Ferreras, Kessel, and Kim 2015). can be videotaped and demonstrations con- In Taiwan, China, teacher education pro- ducted with an audience. grams are highly competitive. Typically, only Collaboration and teamwork are required the top third of applicants ranked by perfor- of teachers from induction onward. In mance on high school and university entrance Shanghai, teachers are not promoted unless exams are selected. they can prove that they work collabora- Low pay and delayed or irregular payments tively; mentors are not promoted unless to teachers make teaching less attractive else- they can show that their mentees improve. where in East Asia and Pacific, discourag- Teachers are given ample time for these col- ing talented applicants. In the Philippines, laborative activities. They teach only 10–12 monthly pay for a secondary teacher is less hours a week, less than half the U.S. aver- than US$400 (Ager 2014). In Lao PDR, pre- age of 27 hours (Liang, Kidwai, and Zhang liminary results of a World Bank survey find 2016). that 53 percent of teachers report delays Surveys of teacher professionalism indi- in receiving their salary at least once a year cate that East Asia and Pacific economies (Demas, Khan, and Arcia, forthcoming). score high on collaborative peer networks Indonesia took steps to raise the quality (OECD 2014). On the Teaching and Learning of teachers by doubling salaries as part of International Survey (TALIS) teacher pro- its 2005 teacher reform. The higher salaries fessionalism index of 37 education systems, led to a fourfold increase in enrollment in 4 of 5 Asian participants scored near the top teacher education programs and increased in 2013. The peer network index is based on the average national exam scores of entrants opportunities for the exchange of information (de Ree and others 2017). Indonesia did not and support needed to maintain high stan- systematically link increased pay to observed dards of teaching. It includes participating in teaching performance, however, or put in induction, mentoring, networking with other place other elements of successful teacher teachers, and receiving feedback from direct development. observations. 18 GROWING SMARTER Top Performing Systems establish clear activities to make in-class time much more learning goals and provide uncluttered effective. texts Countries with fewer content standards and Teachers have adequate time for topics in their textbooks tend to have higher professional development, which international assessment scores. The United centers on improving instructional States covers all 79 of the TIMSS science top- practice ics in its content standards. In contrast, Korea Case study research on successful education covers only 8, Japan covers 19, and Hong systems (such as Ontario, Canada; Finland; Kong SAR, China, covers 22. Textbooks cover Japan; Korea; and Singapore) suggests that 78 topics in the United States, 38 in Korea, high-performing systems devote consider- and 17 in Japan (Liang, Kidwai, and Zhang able time to activities related to instruc- 2016). Having fewer topics suggests a nar- tional improvement, especially analysis of rower focus and deeper study of topics, which instructional practice (Darling-Hammond could lead to much deeper understanding. 2010; Darling-Hammond and Rothman A streamlined curriculum allows for 2011; Levin 2008). These systems also tend uncluttered, focused textbooks. Chinese text- to devote a smaller share of teachers’ time to books are typical for the region. They tend to contact with students and more time to on- be thin, narrowly focused on specific topics, site (in-service) professional development and and significantly more demanding than text- research on the effectiveness of various teach- books in the United Kingdom (Qin 2017). ing strategies. In Japan, for example, teach- Normal practice is for students to cover all ers devote about 40 percent of their working textbook content, making study more effi- time to these types of activities; in Ontario, cient and allowing students to master topics. they spend 30 percent of their time on them The mastery approach is believed to have (Darling-Hammond and Rothman 2011). been important in propelling students in Hong Centering teacher professional develop- Kong SAR, China; Shanghai; and Singapore ment on the analysis of instructional practice to the top of the PISA rankings (Qin 2017). is most effective when embedded in a support system that also incorporates active learn- Teachers have adequate time for lesson ing and on-the-job collaboration, uses mod- preparation els of effective practice, provides coaching and expert support, offers opportunities for Time spent in class is only one part of a teach- feedback and reflection, and is of sustained er’s job. Teachers also prepare lessons, grade duration. homework, write tests, and provide after- These principles inform teacher profes- hours support to students. A large proportion sional development in Top and Above- of time spent in class provides less opportu- Average Performing Systems in East Asia and nity for other activities. Pacific more than they do in other countries In Top Performing Systems, a surprisingly in the region or the rest of the world. This small proportion of total working hours is type of integrated support is at its apex in spent in class. In Japan, for example, teach- Shanghai, where “teaching-research groups” ers spend only 18 hours a week teaching on promote continuous improvement of instruc- average, although they have the highest total tional practice. These groups form a profes- working hours (54 hours a week). With nearly sional development network consisting of two-thirds of their working time spent outside same-subject teachers at the school, district, of class, they spend much more time on les- and provincial levels. In larger schools, the son preparation and other quality-enhancing groups are often divided by grade. Each group OVERVIEW 19 has a leader, who is responsible for organiz- this system, schools regularly evaluate teach- ing activities and introducing novice teachers ers for promotion to a higher rank, accompa- to the learning community. The “teacher-as- nied by a salary increase, based on their years researcher” model builds on Japan’s “les- of service and teaching performance. This son study” approach as a vehicle to improve policy helps to ensure maximum benefit from instructional practice. the investment in teachers’ career-long profes- sional development. Career paths allow experienced teachers to stay in the classroom Coherent policies and practices make teachers’ jobs easier Efforts to develop experienced, effective teach- ers pay dividends if those teachers remain in Policies and practices that center on promot- the profession, delivering instruction and ing improved instruction help to develop mentoring peers. High-performing East Asia effective teachers (table O.3). Selectivity in and Pacific economies have developed career recruitment is followed by induction peri- paths that allow teachers to advance in their ods in which observation, collaboration, and careers and remain as classroom teachers. In feedback are routine. Systems promote pro- other countries, promotion, advancement, fessional collaboration around continuous and higher pay are likely to come through improvement and require observation and moving to administrative positions and leav- positive evaluation of teaching practice for ing the classroom. Japan and Singapore have career advancement. Curricular alignment, separate career tracks for teachers, so that the uncluttered teaching materials, and adequate best do not leave the classroom. In Shanghai, preparation time allow teachers to focus on teachers have opportunities to advance pro- execution in the classroom. Specific career fessionally throughout their teaching career paths recognize excellence in teaching and through a five-level ranking system. Under reward it with increased pay and prestige, TABLE O.3 A coherent system covers all aspects of the teacher career cycle Goal Instrument Attraction and selectivity • Good pay • Effective filtering • Mechanisms to increase the attractiveness of teaching as a profession Good preservice • Government control and quality assurance • Filtering at various stages Smooth induction • Open-door culture • Mentoring and extensive support • Time and space for learning Continuous improvement • Teacher support networks • Teamwork and collaboration • Lesson study • Culture of continuous improvement Career development • Promotion policy • Multiple career pathways, including pathways that allow promotion while leaving good teachers in the classroom Making teaching easier • System coherence • Aligned, streamlined curriculum and textbooks • Adequate nonclass time 20 GROWING SMARTER while encouraging the best and most experi- offer, and coordinated across actors to deliver enced teachers to remain in the profession. A needed services. Their efforts to universal- professional, supported teacher can work well ize preschool progressively appear to have when she or he knows precisely what parts of borne fruit. Throughout the region, children the curriculum students know and where they who had access to early childhood education are struggling. and development services posted higher PISA test scores than children who had no such access—even after controlling for socioeco- Ensure that children are ready to nomic differences (OECD 2013–14). learn in school INST I TUT I ONS Nutritional challenges will impede PUB efforts to improve readiness to learn N T LI C ME SP Efforts to improve readiness in East Asia and S EN ES Pacific are beset by nutritional challenges in ASS DIN several countries. In a third of the region’s G countries, stunting remains highly prevalent, LEARNING despite decades of improvement (figure O.6). Evidence from many countries shows that stimulation of a child is consistently and sig- REA RS nificantly beneficial to child development and DIN HE AC school readiness. Inadequate nutrition under- TE ES TO mines efforts to provide stimulation. S LE AR N Gaps in readiness to learn manifest early and can linger if unaddressed Gaps in children’s readiness to learn manifest Intellectual, social, and emotional development themselves early. If unaddressed, they can early in life all affect how well children perform affect children’s cognitive and noncognitive academically in primary school and get along skills over the long term. with their peers and teachers. Governments in There are large cross-country differences in high-performing school systems help to sup- young children’s ability to read, as measured port children’s readiness to learn. by EGRA. In every country in the Emerging Parents are also critical to learning readi- Systems group, the majority of students do ness. They are their children’s first teachers not meet national standards—and many can- and supporters. They provide proper nutri- not read any words at all. Even in countries tion, health care, and a supportive, nurturing where “zero-word” rates are relatively low, environment. Parents can support their chil- reading fluency is not very high and a large dren’s readiness for primary school by send- portion of students are still struggling with ing them to preprimary programs, providing basic subtasks. emotional support, and creating a stimulating EGRA data cover different languages, learning environment at home. making comparability a challenge, and they do not cover all students in some countries. But the overall message is still clear. In many Investments in readiness to learn systems throughout the region, most children appear to generate lasting returns arrive at school unready to learn. By second High-performing systems in the region grade, too many of these children are still appear to have focused on children’s physi- unable to read a single word. cal and cognitive development, assessed and If students are not ready to “read to learn” improved the quality of the services they going into early grades of primary school, OVERVIEW 21 FIGURE O.6 Stunting remains prevalent in many countries in East Asia and Pacific, despite decades of improvement, 1986–2015 60 % of children under five who are stunted 50 40 30 20 10 0 te ea DR sia ilip ia on es M ds Va r Vi u m M u Th a d lia a lu a oa i Fij a i in ng at ur an d ys va nm es na n an go in m ne oP Ch bo nu Na pi ala To ail -L Tu Gu Sa et Isl on do ya m La or M ew Ca In Tim Ph m aN lo pu So Pa 1986–95 1996–2005 2006–15 Sources: Estimates by the World Health Organization, United Nations Children’s Fund, and the World Bank. Note: Some data are unavailable for each of the three time frames. there is little chance that they will attain a Most countries provide broad service cover- high level of functional literacy by the time age during pregnancy and birth, but there is they complete primary school. When exam- a large drop-off in coverage rates for services ined alongside EGRA data, PISA results sug- for families and children before preschool gest that the countries with low early reading age. Coverage of services for family support ability also have low functional literacy (con- and for child health and development tends to ceptualized as the inability to comprehend the be low even in countries where preschool cov- main message in grade-appropriate texts in erage rates are high. In contrast, coverage of late primary school). If students do not learn all of these service packages is high in Korea. to read fluently in the early grades, there is Just how important are interventions little hope that they will develop the skills to aimed at parents? Research from across succeed on tests like PISA or, more important, the globe, including East Asia and Pacific, in a professional workplace. suggests that both parenting practices and children’s participation in preschool services are very important for children’s develop- Low- and middle-income countries in ment. Both have the potential to increase the region lack key packages of services young children’s exposure to developmental Despite growing evidence of the efficacy essentials such as opportunities for stimu- of early childhood education and develop- lating play, rich language experiences, and ment programs, some education systems practice in developing executive function still do not deliver key packages of services. skills. Governments in parts of the region with The stimulation young children receive Below-Average Performing or Emerging at home is often a foundation for the for- Systems are supporting readiness to learn in mal stimulation they receive in preschools, a variety of ways, but disparities in coverage yet many new parents lack the information across five key packages are wide (figure O.7). and tools they need to enrich their children. 22 GROWING SMARTER FIGURE O.7 Families do not have consistent service coverage between pregnancy and preschool 100 90 80 70 % of families 60 50 40 30 20 10 0 Cambodia Indonesia Myanmar Philippines Timor-Leste Korea, Rep. Pregnancy Birth Family support Child health and development Preschool Sources: Data from Demographic and Health Surveys. Preschool attendance data were supplemented by the World Bank EdStats database (World Bank, various years). Note: As described in Denboba and others (2014), pregnancy includes at least four care visits, iron supplementation, and diet counseling during pregnancy. Birth includes doctor or nurse present at delivery and breastfeeding. Family support includes mother having completed at least primary education, last birth interval greater than two years and last pregnancy desired, at least three types of stimulating activities at home, health care facility not too far and visited within 12 months, vitamin A and iron supplementation in last six months, and safe water source and improved sanitation. Child health and development includes at least three types of food besides breast milk from six months on; zinc supplement in case of diarrhea; children with weight, height, and height- weight ratio less than 2 standard deviations below the mean; and access to deworming medicine. Preschool measures enrollment in preprimary education. No data were available on non-breast milk nutrition for the Philippines. Only Cambodia had information on the provision of diet counseling during pregnancy and availability of deworming medicine. Given proper support, parents can help to in achievement between children from the improve their children’s basic literacy. top and bottom wealth quintiles (figure O.8). The lack of key service packages leads to Closing wealth gaps in access to preschools gaps in outcomes. For instance, substantial would cost just a small fraction of total edu- gaps exist between the ability of children cation spending; in most countries, it would from poor and wealthy families to perform cost only a small fraction of spending on basic functions, such as counting from 1 to preprimary education. Estimates for a few 10. These gaps are also apparent in the use countries—particularly Indonesia, Malaysia, of preschool services, with a gap of 65 per- and Thailand—are larger. centage points in Lao PDR and 54 percentage points in Mongolia. Gaps are similar in access Tested and proven solutions exist to high-quality care at home. In Cambodia, the gap in access to preschools by the richest Merely increasing the supply of services and poorest quintiles is 31 percentage points, focused on children’s physical and cognitive and the gap in access to high-quality care is development is not the answer. Countries 24 percentage points. need to assess and improve the quality of these services and coordinate their deliv- ery across actors, if they are to reap the full The costs of inaction during the early benefits. A host of proven solutions exists, years are high—yet, action is affordable including center-based community-managed The social and economic costs of inaction activities that focus on play as learning, during the early years are high. Most govern- home-based programs, reading interventions, ments in the region can afford to close gaps and programs that engage parents. OVERVIEW 23 FIGURE O.8 Closing the achievement gap between socioeconomic groups is affordable 7.0 6.3 6.1 6.0 Government spending (% of GDP) 5.0 4.6 4.1 4.0 3.8 3.9 3.3 3.0 3.0 2.0 0.9 1.0 0.2 0.1 0.4 0.2 0.5 0.1 0.1 0.1 0.1 0.03 0.1 0.1 0.1 0.1 0 Japan Korea, Rep. Singapore B-S-J-G Vietnam Indonesia Malaysia Thailand (China) Top Performing Systems Above-Average Below-Average Performing Systems Performing Systems Total government spending on education Government spending on preprimary education Additional spending required to close learning gap between children from richest and poorest households Sources: Authors’ calculations based on data from PISA 2015 (OECD 2015), UIS, and World Bank and Government of Vietnam 2017. Note: Data on total government spending are the latest available between 2011 and 2014, except 2016 for Indonesia and 2015 for China and Singapore. Cost of inaction is for closing the socioeconomic gap in the percentage of students falling below basic proficiency in science. Data on percentage of government spending on preprimary are from UNESCO Institute for Statistics, Vietnam Ministry of Economy, and Vietnam Ministry of Finance for 2013, except 2012 for the Republic of Korea and Vietnam and 2014 for Japan. No data available for preprimary spending by B-S-J-G (China). B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); PISA = Programme for International Student Assessment; UNESCO = United Nations Educational, Scientific, and Cultural Organization. Assess students to diagnose been integrated with specific ways of tak- issues and inform instruction ing action—linked closely with policies and practices relating to teachers, students, and curricula. The mix of assessments var- ies across countries, but all Top Performing E NT Systems have well-defined ways of feeding SM the information on student learning out- ES PUB comes gained from such assessments back AS S LI C SP into the system to drive quality. At the class- EN room level, good practice involves train- D IN ing teachers to use such assessments and G to incorporate classroom assessment into LEARNING curricula. At the school level, it involves informing principals’ decisions and educa- READ tional strategies. And at the system level, RS INE it involves using assessment data to create HE SS TO AC LEA RN TE a broad commitment to quality and spur policy decisions. INS TITUTIONS A systemic approach to assessment drives High-stakes exams can be useful— high-quality learning outcomes in the but they can have undesirable effects classroom. Efforts to assess student learn- Experience in high-performing East Asia ing in the Top Performing Systems have and Pacific economies shows the value of 24 GROWING SMARTER meritocratic, standardized selection exams fear of failing by providing alternate routes when coupled with good teachers, a strategic to different levels and types of education vision, and labor markets that value produc- (Liang, Kidwai, and Zhang 2016). Although tive workers. For decades, these exams were these efforts are at early stages and their full an integral part of education strategies and impacts are not yet known, they indicate pol- were used to allocate limited learning oppor- icy makers’ attempts at balancing the weight tunities in Hong Kong SAR, China; Japan; of examinations in a country’s assessment Korea; Singapore; and Taiwan, China (Wong system. 2017). Exams played a central role in the push for quality, through mechanisms that Classroom assessments increasingly indirectly gauged teacher effectiveness and inform instruction influenced how teachers were trained. Such exams worked well, especially as education Practices in the region reveal the importance resources were developing. At the time, the of using ongoing measurements of learn- focus was on cognitive development in basic ing in the classroom to guide instruction. A education and technical and vocational regional study by the World Bank in 2012— education and training, with considerable using a version of its Systems Approach for absorption into lower-skill production-based Better Education Results (SABER) bench- factory jobs. marking tool—finds that almost all countries Most countries in the region still use exams surveyed have a framework for large-scale, for entrance decisions at the secondary level. system-level assessment and exam activities, But many high performers have removed or and more than half have such a framework adapted high-stakes exams at lower levels of for classroom assessment activities (Jimenez, education, partly because excessive focus on Nguyen, and Patrinos 2012). 4 The study exams can lead to stress, a shifted or narrowed also finds that teachers’ preservice training focus of education, and reduced equity. Korea increasingly includes techniques for produc- removed middle school entrance exams in the tive use of classroom assessment. The SABER 1960s, and high school entrance exams in analysis also shows areas where improve- the 1970s, as part of its high school equaliza- ments are needed, including monitoring and tion policy. In 2001, Hong Kong SAR, China, ensuring the quality of classroom assessment removed public assessments after primary and raising the priority of its use among school completion. In 2013, Malaysia replaced teachers. the high-stakes exam at the end of lower- Top Performing Systems include assess- secondary school with a mix of school-based ments in teacher training programs and pro- exams and a centralized exam that included vide guidance and monitoring on their use. In more critical thinking skills. Singapore, Singapore, educational reform included efforts however, maintains the primary school– to support assessment in the classroom, includ- leaving exam at the end of sixth grade. ing studying teachers’ practices and designing In response to concerns from parents, a two-year professional development program Korea introduced test-free semesters as a to support assessment (Ho 2012). pilot program at the lower-secondary level in 2013, before expanding the program across International large-scale assessments the system (OECD 2016b). Taiwan, China, have spurred learning-focused reforms introduced exam-free pathways to secondary school in 2014 (Wong 2017). Shanghai has International benchmarking has most value adopted strategies to reduce the dominance of when it leads to the identification of specific the zhongkao lower-secondary school exam areas for improving education quality. In (taken at the end of ninth grade) by increas- Top Performing Systems and Above-Average ing the frequency of testing, broadening the Performing Systems, international assessment domains tested, and reducing the stress and data have spurred changes. Taiwan, China’s OVERVIEW 25 Happy Reading program, launched in 2008, and standardized tests to guide student learn- was a response to low performance on PISA ing (OECD 2016–17). 2006. It used PISA microdata as a bench- mark to align teaching methods, increase the amount of time allocated for reading instruc- Early Grade Reading Assessments are tion, increase resources, and revise teacher critical for Emerging Systems development (Driskell 2014). In response International large-scale assessments such as to PISA results, Vietnam changed the legal PISA and TIMSS provide helpful information framework for large-scale exams to diversify when a national school system has reached a testing methods, improve item quality, and level of performance compatible with the mea- pave the way for competency-based assess- sured outcomes of the test. Where education ment. Broadening sample-based national systems are still emerging, targeted assess- diagnostic assessment of reading, math, and ments of foundational skills provide more rel- Vietnamese was also a key part of curriculum evant information. EGRAs and Early Grade reform. In Japan, PISA has been important in Mathematics Assessments (EGMAs) gauge tandem with the national assessment to drive student progress in early primary school. and monitor education reform. EGRAs provide a snapshot of—and in some cases a wake-up call about—what stu- dents are learning. They have spurred sys- Use of national assessments varies temic changes in teaching methodologies across the region and curricula. In Tonga and Vanuatu, EGRA Top Performing Systems Japan and Korea analysis in 2009 revealed low reading and reintroduced regular large-scale national comprehension. These results informed the assessments about a decade ago. Japan rein- Pacific Early Age Readiness and Learning stituted a national assessment in 2007, cover- program, to address both school readiness ing three subject areas (Japanese, math, and and early grade literacy through a variety of science) for students in sixth and ninth grades. channels, including community-based groups, Korea reintroduced its national assessment in public awareness, teacher training, and a 2008. It is administered in sixth, ninth, and roadmap for early years to guide implemen- tenth grades. China rolled out a national tation of country priorities. Evaluation of the assessment in 2015. It assessed fourth and intervention and tracking of literacy gains eighth graders in six subjects. Uniquely showed improvements in reading of a half to among the high performers, Singapore uses a full year (Patrinos 2016). its system of national examinations as the Below-Average Performing and Emerging primary means of assessing the education Systems should ensure that all students mas- system. ter basic reading and math skills. Widespread Other countries use national assessments use of EGRAs and EGMAs to measure learn- more sporadically. In Vietnam, fifth graders ing should be a cornerstone of their efforts are assessed intermittently. In Mongolia, the to do so. National Assessment of Primary Education, begun in 2004, is administered to fifth-grade A map of countries’ current systems students. Lao PDR conducts the National Assessment of Learning Outcomes (in 2006 The ability of the East Asia and Pacific and 2009 testing fifth-grade students and in region’s Top Performing Systems to improve 2012 testing third-grade students). both schooling and learning provides valuable Many school systems use test results to lessons for all countries struggling to address implement targeted programs aimed at rais- the learning crisis. Although these systems ing learning outcomes. Results from PISA followed different paths, all of them aligned 2015 show that the majority of school sys- and prioritized common elements across five tems in the region use both teacher-developed policy domains to promote learning. 26 GROWING SMARTER Table O.4 maps the current state of the the long term. As these policies are devel- region’s economies for these domains and ele- oped, Emerging Systems are well served to ments. It provides a starting point for coun- review the extent to which teacher train- tries to take stock of the current situation in ing and professional development focus each domain and to envision a path forward. on improving the quality of instruction. Top Performing Systems distinguish them- • Below-Average Performing Systems selves not only by achieving alignment across should review teacher development poli- elements but also by sustaining it. Above- cies to ensure that training is prioritiz- Average Performing Systems attempt align- ing improved instructional quality while ment, but it is not consistent in all domains. building the institutional capacity for Below-Average Performing Systems strive to deeper and more comprehensive reforms. emulate the success of the rest of the region. Introducing selectivity, observation, col- They have devised plans for alignment, but laboration, and feedback while creating implementation is lacking or plans never incentives and career paths for teachers materialize. Emerging Systems face the great- that reward teaching ability are likely to est challenges. Resources are scarcest, few pay off substantially. Ensuring readiness measures of learning exist, and just getting to learn and broadening early childhood all students in school has been difficult. education and development services are also critical. Developing and implement- ing systemwide national assessments of Charting the course ahead cohort progress should complement pro- grams for classroom assessment; assess- Providing learning opportunities is imperative ment systems should also include regular for the millions of children who are out of use of internationally comparable assess- school—or in school but not learning. Lessons ments for benchmarking and system from high performers suggest that countries accountability. Consideration should also can improve learning outcomes by focusing be given to making teachers’ jobs easier, on improving their performance in the five through curriculum and other reforms. policy domains and 15 elements within them Simultaneously raising the attractive- shown in figure O.1. These efforts require ness of teaching as a profession and the detailed, ongoing reforms. accountability of teachers for good class- No one size fits all for reform agendas, but room performance is key. all systems share some priorities (table O.5): • Above-Average Performing Systems • Emerging Systems should concentrate should not rest on their accomplishments. efforts on ensuring that basic conditions Deepening the quality of the teaching are in place for learning in all schools and force and continuing to monitor equity on reviewing spending to ensure that basic are in order. But these systems should education is appropriately prioritized. also endeavor to tie learning to new and Emerging Systems should also commit to emerging needs, which includes introduc- diagnosing cohort progress, especially of ing teaching and measurement of noncog- early learners using early grade assess- nitive and 21st century skills. Redoubling ments, and using test results to inform and the commitment to internationally compa- improve basic reading and math instruc- rable assessment can keep public support tion. They should continue to explore for excellence in education high, along the use of regional and international with public awareness of achievements. assessments for benchmarking. Second- Deepening the availability of high-quality order challenges include channeling early childhood education and develop- resources for equity and considering how ment programs, especially for families to approach the range of teacher support who cannot afford them, is another step. policies that can improve capacities for Also essential is to ensure that good OVERVIEW 27 TABLE O.4 The status of policies and practices that promote learning varies within the region Above- Average Below-Average Performing Performing Top Performing Systems Systems Systems Emerging Systems Pacific Island countries Hong Kong SAR, China Papua New Guinea Macao SAR, China B-S-J-G (China) Taiwan, China Timor-Leste Korea, Rep. Philippines Singapore Cambodia Indonesia Mongolia Myanmar Thailand Malaysia Vietnam Lao PDR Japan Policy Practice Align institutions Ensure that the basic conditions to ensure basic for learning are in place in all conditions for schools. learning. Concentrate Spend effectively. effective, equity- Concentrate public spending on minded public basic education. spending on basic education. Channel resources to schools and districts that are falling behind. Select and Raise the selectiveness of who support teachers becomes a teacher. throughout their Support new teachers by careers to allow observing classroom practices them to focus on and providing feedback. the classroom. Make teachers’ jobs easier by providing clear learning goals and uncluttered texts. Keep experienced teachers in the classroom and leading as peers and researchers. Center teacher training on classroom practice and the ability to teach the curriculum. Ensure that Focus on children’s physical and children are cognitive development from birth. ready to learn in Assess and improve the quality school. of early childhood education and development services. Coordinate across actors to deliver needed services. Assess students Benchmark learning through to diagnose issues participation in international and inform large-scale assessments. instruction. Diagnose cohort progress at every educational sublevel. Inform instruction with data from formative classroom assessment. Alignment is successful Alignment is attempted, Alignment plans exist in writing, but there is little or no No data. and sustained. but success is inconsistent. evidence of implementation, or no such plan exists. Note: B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China). 28 GROWING SMARTER TABLE O.5 Concerted policy action and continuity of implementation drive systems improvement Action Below-Average Above-Average Policy Emerging Systems Performing Systems Performing Systems Top Performing Systems Public Prioritize public spending on basic Continue investing in basic education Continue investing in basic and upper- spending education and ensure high completion to ensure high completion while secondary education while diversifying of primary and lower-secondary school. expanding access to upper-secondary funding for vocational, technical, and school and above. tertiary education. Increase starting teacher salaries to attract qualified teachers. Strengthen teacher compensation Strengthen teacher career development policies to encourage good policies and nonfinancial benefits for Build simple but functional school performance and retain qualified teachers to continue improving their facilities to increase access to basic teachers. quality. education. Build simple but functional school Provide remedial measures such as Provide funding to support facilities to ensure access to basic extracurricular activities to enhance disadvantaged students in accessing education and increase access to learning of disadvantaged students. basic education. preprimary and upper-secondary education. Help disadvantaged students to access basic education; consider mobilizing the private sector to increase access to preprimary and upper-secondary education. Teachers Start the positive foundational cycle Take incremental steps to increase Establish rigorous criteria and multiple of attracting and ensuring reasonable- qualifications and quality, establishing filtering mechanisms, while ensuring quality teachers, while being realistic greater quality assurance and filtering that teaching is attractive in terms of about the system’s capacity to measures. salary, professionalism, and prestige. implement quality assurance measures balanced with supply of candidates. Deepen teacher networks in which Leverage high-performing teachers for more-advanced teachers provide professional development and support. Establish teacher networks (with a focus greater support through mentoring and on supporting new, weaker teachers coaching. Provide more autonomy for high- with mentoring and other support), performing teachers. lesson study, and an “open-door” Provide more advanced professional culture as key elements of professional development that develops deeper Create a highly developed professional development. critical thinking techniques. development framework with tailored, individualized approaches. Establish a basic framework of Create tracks of progression and professional development and how promotion for expert teachers that Deepen critical thinking, elaboration, teachers can reasonably allow them to remain in the classroom. and cognitive activation techniques. be reached. Evolve system- and teacher-related Establish a mentality of teaching to policies in ways that reform and move learn with clear learning goals with all elements forward with alignment. curriculum and textbooks. Decrease classroom hours and increase Begin creating a cohesive, aligned supplementary activities. system that supports teachers and promotes focused, streamlined learning. table continues next page OVERVIEW 29 TABLE O.5 Concerted policy action and continuity of implementation drive systems improvement (continued) Action Below-Average Above-Average Policy Emerging Systems Performing Systems Performing Systems Top Performing Systems Readiness to Develop policies and programs to Ensure that clear quality standards exist, Engage multiple actors in the learn ensure that both cognitive development are disseminated, and are addressed in development and delivery of integrated and physical health are the focus of early an evaluative process. early childhood services. childhood school readiness strategies. Assessments Periodically use early foundational skills Implement national assessments for Participate in regular international assessments to spur policy change and diagnostic purposes. benchmarking (PISA, TIMSS). target reforms every five years. Strengthen systems of assessment Engage in a national dialogue and Integrate assessment policies into (enabling environment, quality control, secure commitment to quality. reforms of curricula and teacher alignment). development systems. Disseminate assessment data to Integrate the use of assessment data stakeholders for accountability. into teacher development, alongside adequate support and quality Reflect changing curricular priorities monitoring. (such as noncognitive skills) in assessment design and methodology Link assessment use to equity, and provide support for change or examining current practices for tracking research, guidance, training, quality, and accountability. monitoring, and so forth. Use internationally benchmarked assessments to check system performance. Note: PISA = Programme for International Student Assessment; TIMSS = Trends in International Mathematics and Science Study. options exist in the labor market for capi- efforts demonstrate that conserving and talizing on knowledge and skills acquired extending existing capacity and expertise in basic and postbasic education. Amid all are preconditions for staying on top. of these activities, building institutional capacity further cannot be neglected. Because learning is critical to East Asia • Top Performing Systems provide compel- and Pacific’s strategy for productivity-driven ling examples of how the work of produc- growth, policy makers have continually ing high learning outcomes in schools is kept their eyes on the next stage of educa- never completed. Looking at how these tion development. In today’s rapidly chang- systems are evolving reveals that promot- ing economies, education systems will ing creativity and new forms of assess- need to prepare students for lifelong learn- ment, ensuring that teachers remain ing. To draw in all countries and all stu- motivated and grow in competency, and dents across the region, the path ahead will learning from other top performers world- involve keeping pace with rapidly changing wide are core tasks. Their continued circumstances. 30 GROWING SMARTER Notes National Center on Education and the Economy, Washington, DC. 1. The constructed averages provide the best Darling-Hammond, L. 2010. “Steady Work: How information on the performance of systems as a Countries Build Successful Systems.” In The whole, but they do not provide any information Flat World and Education: How America’s on trends, and the information on all countries Commitment to Equity Will Determine Our is not equally robust. Future, edited by L. Darling-Hammond, ch. 6. 2. Data from PISA on learning outcomes in New York: Teachers College. China are from only four provinces. However, Darling-Hammond, L., and R. Rothman. 2011. learning outcomes in the rest of China can Teacher and Leader Effectiveness in High- be inferred by looking at the distribution of Performing Education Systems. Washington, proficiency levels in urban and rural areas of DC: Alliance for Excellent Education and B-S-J-G (China). We use the ratios of urban to Stanford Center for Opportunity Policy in rural students at various levels of proficiency Education (SCOPE). in the PISA data to estimate how many urban de Ree, J., K. Muralidharan, M. Pradhan, and and rural students across China are at various H. Rogers. 2017. “Double for Nothing? levels of proficiency. This assumption produces Experimental Evidence on an Unconditional an upper bound on the number of children Teacher Salary Increase in Indonesia” (English). who might be in a learning crisis. At the lowest Policy Research Working Paper 8264, World levels of proficiency, students are considered Bank, Washington, DC. functionally illiterate. Using the score threshold Demas, A., M. M. Khan, and G. Arcia. Forthcoming. for the lowest levels of proficiency to estimate Delivery of Education Services in Lao PDR: the learning levels outside the four tested Results of the SABER Service Delivery Survey, Chinese provinces produces estimates of the 2017. Washington, DC: World Bank. learning crisis in the region that are unchanged, Denboba, A. D., R. K. Sayre, Q. T. Wodon, L. K. implying that these estimates are plausible Elder, L. B. Rawlings, and J. Lombardi. 2014. upper bounds of the learning crisis. “Stepping Up Early Childhood Development: 3. For Indonesia, World Bank (2009); for the Investing in Young Children for High Returns.” Philippines, World Bank (2016). World Bank, Washington, DC. 4. SABER is a set of tools that enables countries Dollar, D., and A. Kraay. 2002. “Growth Is to evaluate and benchmark education policies Good for the Poor.” Journal of Economic across 13 areas, including teachers, early Growth 7 (3): 195–225. http://www.jstor.org childhood development, school autonomy and /stable/40216063. accountability, and student assessments (see Driskell, N. 2014. “Global Perspectives: http://saber.worldbank.org/). Explaining Taiwan’s Dramatic Improvement in PISA Reading.” National Center on Referencesa Education and the Economy, Washington, DC. http://ncee.org/2014/10/global-perspectives Ager, M. 2014. “Poe Wants P25,000 Monthly Pay -explaining-taiwans-dramatic-improvement-in for Public School Teachers.” Inquirer Online, -pisa-reading/. August 8. Ferreras, A., C. Kessel, and M.-H. Kim. Altinok, N., C. Diebolt, and J.-L. Demeulemeester. 2015. Mathematics Curriculum, Teacher 2014. “A New International Database on Professionalism, and Supporting Policies Education Quality: 1965–2010.” Applied in Korea and the United States: Summary Economics 46 (11): 1212–47. of a Workshop. Washington, DC: National Barro, R., and J.-W. Lee. 2013. “A New Data Academies Press. Set of Educational Attainment in the World, Ho, E. S.-C. 2012. “Student Learning Assessment.” 1950–2010.” Journal of Development UNESCO, Paris. http://www.unescobkk.org Economics 104 (C): 184–98. /fileadmin/user_upload/epr/Quality/SLA_8 Center on International Education Benchmarking. August-final.pdf. n.d. “Japan: Teacher and Principal Quality.” Jimenez, E., V. Nguyen, and H. A. Patrinos. 2012. “Stuck in the Middle? Human Capital a. Titles of publications that include “Korea” and “Taiwan” Development and Economic Growth in refer to the Republic of Korea and Taiwan, China, respectively. Malaysia and Thailand.” Policy Research OVERVIEW 31 Working Paper 6283, World Bank, Washington, Tan, S. K. S., and A. F. L. Wong. 2007. “The DC. http://dx.doi.org/10.1596/1813-9450 Qualifications of the Teaching Force: Data -6283. from Singapore.” In A Comparative Study Levin, B. 2008. How to Change 5,000 Schools: A of Teacher Preparation and Qualification Practical and Positive Approach to Leading in Six Nations , edited by R. E. Ingersoll, Change at Every Level. Cambridge, MA: 71–82. Philadelphia: Consortium for Policy Harvard Education Press. Research in Education. http://www.cpre.org Liang, X., H. Kidwai, and M. Zhang. 2016. How /images/stories/cpre_pdfs/sixnations_final.pdf. Shanghai Does It: Insights and Lessons from TIMSS (Trends in International Mathematics the Highest-Ranking Education System in the and Science Study). 2015. TIMSS 2015 World. Washington, DC: World Bank. International Database. Boston: TIMSS and National Bureau of Statistics of China. 2016. PIRLS International Study Center, Lynch School China Statistical Yearbook: Education. Beijing: of Education, Boston College and International China Statistics Press. http://www.stats.gov.cn Association for the Evaluation of Educational /tjsj/ndsj/2016/indexeh.htm. Achievement. https://timssandpirls.bc.edu National Center on Education and the Economy. /timss2015/international-database/. n.d. “Assistance Schemes for Individuals UIS (UNESCO Institute for Statistics). Available at: and Families in Social and Financial Need.” http://uis.unesco.org/. National Center on Education and the Wong, A. 2017. “Insights from East Asia’s High- Economy, Washington, DC. https://www.ncss Performing Education Systems: Leadership, .gov.sg/NCSS/media/NCSS-Documents-and Pragmatism, and Continuous Improvement.” -Forms/NCSS%20Internal%20Documents Background paper for flagship report, /AssistanceSchemes_1.pdf. Growing Smarter: Learning and Equitable OECD (Organisation for Economic Co-operation Development in East Asia and Pacific, World and Development). 2013–14. PISA 2012 Bank, Washington, DC. Results. 6 vols. Paris: OECD Publishing. World Bank. 2009. Investing in Indonesia’s ———. 2 0 1 4 . TA L I S 2 0 1 3 R e s u l t s : A n Education at the District Level: An Analysis International Perspective on Teaching and of Regional Public Expenditure and Financial Learning. Paris: OECD Publishing. http:// Management. Washington, DC: World Bank. dx.doi.org/10.1787/9789264196261-en. ———. 2013. Spending More or Spending ———. 2015. PISA 2015 Results. Vol. I: Better: Improving Education Financing in Excellence and Equity in Education. Paris: Indonesia. Jakarta, Indonesia: World Bank. OECD Publishing. http://www.oecd.org https://openknowledge.worldbank.org/handle /publications/pisa-2015-results-volume-i /10986/13210. -9789264266490-en.htm. ———. 2014. East Asia Pacific at Work: ———. 2016a. Education at a Glance 2016: Employment, Enterprise, and Well-Being . OECD Indicators . Paris: OECD. http:// Washington, DC: World Bank. doi:10.1596 www.oecd .org/edu/education-at-a-glance /978-1-4648-0004-7. -19991487.htm. ———. 2015. Thailand: Wanted, A Quality ———. 2016b. Education at a Glance 2016: Education for All. Washington, DC: World OECD Indicators. Korea Country Note. Paris: Bank. https://openknowledge.worldbank.org OECD. /handle/10986/22355. ———. 2016c. Education in China: A Snapshot. ———. 2016. “Building Better Learning Paris: OECD. Environments.” Policy Note 3, World Bank, ———. 2016–17. PISA 2015 Results. 5 vols. Paris: Washington, DC. OECD Publishing. ———. 2018a. Riding the Wave: An East Asian Patrinos, H. A. 2016. “Investing in Early Miracle for the 21st Century. Washington, DC: Years Learning: It Can Be Done!” World World Bank. Bank blog , November 2. http://blogs ———. 2018b. World Development Report 2018: .worldbank.org/education/investing-early Learning to Realize Education’s Promise . -years-learning-it-can-be-done. Washington, DC: World Bank. Qin, A. 2017. “Britain Turns to Chinese Textbooks ———. Various years. Education Statistics (EdStats) to Improve Its Math Scores.” New York Times, [database]. Washington, DC: World Bank. August 5. http://datatopics.worldbank.org/education/. 32 GROWING SMARTER ———. Various years. World Development Policies towards Sustainability, Efficiency, Indicators [database]. Washington, DC: and Equity. Washington, DC: World Bank. World Bank. https://openknowledge.worldbank.org World Bank and Government of Vietnam. 2017. /handle/10986/28610. Vietnam Public Expenditure Review: Fiscal High-Quality Schooling and Economic Growth in 1 East Asia and Pacific Many economies in East Asia and Pacific Such sustained, high growth rates are the grew very rapidly over the past few decades, path to high-income country status. “Of the with rates of growth in East Asia consistently 71 countries classified by the World Bank in twice the global average. Sound macroeco- 2010 as high-income countries, only 15 were nomic environments allowed human capital not already in this category by 1987,” note to drive growth. Jimenez and Dobson (2013, 2). Five of these countries were in East Asia and Pacific. Extraordinary growth performance The region generated allowed many East Asia and Pacific econo- unprecedented and mies to transform themselves within two transformative economic growth generations or less. The size of the regional The economies of East Asia and Pacific have economy in 2015 was 10 times larger grown faster and sustained high growth lon- than it would have been if it had grown at ger than any other region in the world since world average rates since 1960 (figure 1.2). 1960 (figure 1.1). Progress has been remark- In 1990, economic output in the region able, especially among the region’s low- and accounted for less than 7 percent of global middle-income economies, which grew at output, despite the region having one-third more than twice the world average in 1961– of the world’s labor force. Today, the econ- 2015 (7.2 vs. 3.5 percent). Even excluding omies of East Asia and Pacific account for China’s spectacular growth, the region’s low- 30 percent of global output. and middle-income countries grew more than 2 percentage points faster than the world average for nearly half a century. In 1970– How has sustained high growth 2010, their rate of growth was almost twice changed the region? the world average (5.9 vs. 3.1 percent). No A large and growing share of East Asians other low-income region of the world comes today live prosperous, productive, close to matching this record of steady and middle-class lives. More than 1.4 billion high long-term growth. The Pacific Island East Asians are considered middle class. countries did not enjoy the same strong This emerging group has access to con- growth (box 1.1; spotlight 1). sumer goods and services as well as career 33 34 GROWING SMARTER FIGURE 1.1 Growth in East Asia and Pacific has exceeded global averages for decades, 1961–2015 10 9 8.9 Annual economic growth (% of real GDP) 8.4 8 7.8 8.0 7.7 7 7.2 7.2 6 5.3 6.1 5.3 5.6 5 5.0 5.1 5.0 5.1 4.6 4.7 4.6 4 3.9 3.6 3.5 3 3.0 2.9 2.7 2.6 2.4 2 1.8 1.5 1 0 1961–70 1971–80 1981–90 1991–2000 2001–10 2011–15 World Low- and middle-income East Asia and Pacific East Asia and Pacific Low- and middle-income East Asia and Pacific (without China) Pacific Island countries Source: Data from World Development Indicators (World Bank, various years). Note: Figures are based on real gross domestic product (GDP) in U.S. dollars. Base year is 2010. Low- and middle-income East Asia and Pacific includes all countries and economies in the region except Japan, the Republic of Korea, and Singapore. Pacific Island countries include Fiji, Kiribati, the Marshall Islands, the Federated States of Micronesia, Nauru, Palau, Samoa, the Solomon Islands, Tonga, Tuvalu, and Vanuatu. BOX 1.1 Growth has been slower in the Pacific Island countries The lack of growth in the Pacific Island countries livelihoods of the poor. Tourism (in Fiji, Palau, reflects both their geography and the limited size of Samoa, Tonga, and Vanuatu); mining and min- their populations and economies. Their remoteness erals (in Papua New Guinea); logging (in the from large markets and the dispersion of their popu- Solomon Islands); fisheries and commercial agricul- lations over vast areas of ocean have impeded broad- tural products (coffee in Papua New Guinea and based growth through labor-intensive and export- Vanuatu, copra in several of the smaller countries, oriented manufacturing, as high transport costs make and sugar in Fiji) bring in foreign exchange. it difficult for them to compete with countries closer Limited employment opportunities outside subsis- to major international markets. The high costs of ser- tence farming and the smaller government, tourism, vice delivery to a widely dispersed population have and fishing sectors have made international migra- made it difficult to invest in human capital, expand tion an important economic option for many Pacific economic opportunities for the population, support Islanders. In countries like Samoa and Tonga, where a more diversified economy, and provide basic needs, remittances account for more than 20 percent of such as water, sanitation, and electricity. gross domestic product (GDP), migration helps many The economies of most of these countries have households to avoid slipping into poverty. a narrow production base, with semi- subsistence agriculture playing a major role, especially in the Sources: World Bank 2016a, 2016b, 2017, 2018a. H I G H  Q UA L I T Y S C H O O L I N G A N D E CO N O M I C G R O W T H I N E A S T A S I A A N D PAC I F I C 35 and work opportunities that are increas- This transformation did not happen by ingly comparable with those in high-income accident, nor was it the result of economies countries. in the region merely being carried along For the past two generations, opportuni- by global trends. From the 1960s through ties for economic and social advancement the 1990s and beyond, while the rest of and mobility have been maximized. Many parents have seen their children move up the economic ladder to better employment and FIGURE 1.2 Without sustained high growth since 1960, economic standards of living. Structural transforma- output in East Asia and Pacific would have been less than 10 percent tion of the region’s economies has allowed of its current level hundreds of millions of people whose child- hoods were marked by deprivation and 12.0 11.1 hardship to move out of poverty. Extreme Economic output (US$, trillions) 10.0 poverty rates in East Asia and Pacific have fallen dramatically, from 56 percent in 1990 8.0 to under 8 percent today.1 6.0 As recently as 1991, two-thirds of East 4.8 Asians worked in agriculture, most as low- 4.0 income smallholders. By 2012, that fig- 2.0 1.5 ure had dropped by half, to one-third (figure 1.3). Rising formal employment, wages, and 0 productivity have made the typical East Asian East Asia East Asia and Pacific East Asia and Pacific an educated urban dweller rather than a and Pacific GDP GDP at East Asia GDP at East Asia at world growth and Pacific growth and Pacific growth farmer with little schooling. As overall popu- rate rate without China rate with China lation increased, megacities and educational and formal work opportunities grew faster Source: Data from World Development Indicators (World Bank, various years). (World Bank 2014). Note: Data cover the years 1960–2015. GDP = gross domestic product. FIGURE 1.3 The share of people in East Asia and Pacific working in agriculture has fallen, and the share working in services has risen, 1991–2011 a. Agriculture b. Industry c. Services 70 40 80 60 35 70 Share of employment (%) Share of employment (%) Share of employment (%) 30 60 50 25 50 40 20 40 30 15 30 20 10 20 10 5 10 0 0 0 19 1 19 3 19 5 19 7 20 9 20 1 03 20 5 20 7 20 9 11 19 1 19 3 19 5 19 7 20 9 20 1 03 20 5 20 7 20 9 11 19 1 19 3 19 5 19 7 20 9 20 1 03 20 5 20 7 20 9 11 9 9 9 9 9 0 0 0 0 9 9 9 9 9 0 0 0 0 9 9 9 9 9 0 0 0 0 19 20 19 20 19 20 East Asia and Pacific East Asia and Pacific (high-income) East Asia and Pacific (low- and middle-income) World Source: Data from the International Labour Organization. 36 GROWING SMARTER TABLE 1.1 Success stories of sustained high growth Per capita income (constant 2000 US$) Period of annual growth of Economy 7% a year or more Beginning of period 2005 Botswana 1960–2005 210 3,800 Brazil 1950–80 960 4,000 China 1961–2005 105 1,400 Hong Kong SAR, Chinaa 1960–97 3,100 29,900 Indonesia 1966–97 200 900 Japana 1950–83 3,500 39,600 Korea, Rep.a 1960–2001 1,100 13,200 Malaysia 1967–97 790 4,400 Maltaa 1963–94 1,100 9,600 Oman 1960–99 950 9,000 Singaporea 1967–2002 2,200 25,400 Taiwan, Chinaa 1965–2002 1,500 16,400 Thailand 1960–97 330 2,400 Source: Commission on Equitable and Sustainable Growth 2010. a. Economy has reached high-income countries’ level of income per capita. the low-income world was stuck in pat- cohort are expected to bring more knowledge terns of low growth and losing ground to and skills to their jobs. The massive increases rich countries, the economies of East Asia in schooling in East Asia have vastly increased were converging on rich-country levels the supply of educated workers, but returns (Pritchett 1997). to investment in ever higher levels of human East Asia’s economic success was never capital continue to rise. assured and was largely unforeseen. Indeed, The Commission on Equitable and development economists and other observ- Sustainable Growth (the Growth Commission), ers in the 1960s despaired over the region’s convened under the auspices of the World prospects and saw Latin America as the Bank, studied 13 cases of sustained high region most likely to prosper (Jimenez and growth after World War II (table 1.1). Nine Dobson 2013). of the 13 economies were in East Asia and By the 1980s, East Asia’s formula for suc- Pacific, which grew, on average, at more than cess was an object of increased recognition 7 percent a year for 37 years, raising income and admiration. The recent uptick in growth per capita by an order of magnitude. Although rates throughout the low-income regions rural agricultural work remains the primary of the world can arguably be attributed to occupation for many people in the region average low- and middle-income countries today, economies with decades of high growth taking more steps and implementing more are increasingly centered on large modern, policies that had been first proven in East urban knowledge bases. Asia (Jimenez and Dobson 2013). For more and more people, growth is bringing improved consumption opportu- A sound macroeconomic nities in areas of basic needs and beyond. environment allowed human Perhaps more important, growing numbers capital to drive growth of people have left behind low-productivity Education makes workers more productive smallholder agricultural work for formal when it imparts skills they can use to increase employment and modern careers. Demand their output. Globally, each additional year of for educated workers has risen as rapidly as schooling raises individual earnings by 8–10 supply can respond, and members of every percent (World Bank 2018b). In East Asia H I G H  Q UA L I T Y S C H O O L I N G A N D E CO N O M I C G R O W T H I N E A S T A S I A A N D PAC I F I C 37 and Pacific, the average rate of return to an TABLE 1.2 Rate of return to an additional year of schooling in additional year of schooling is 10.9 percent selected economies in East Asia and Pacific (table 1.2). Economy Year Rate of return (%) Growth-promoting policies increased pro- China 2002 16.6 ductivity through education by committing Japan 2007 14.0 to good plans over decades (more than half Tuvalu 2010 13.4 a century in the case of some economies). Korea, Rep. 2010 13.2 The most successful countries set basic policy Singapore 1998 12.5 directions decades ago. Japan, the Republic Palau 2000 12.4 of Korea, and Singapore began their transi- Malaysia 2010 12.0 tions to high-income status early and have Thailand 2011 9.4 seen their economies rise furthest. Their out- Mongolia 2011 9.1 put per capita is among the highest in the Philippines 2011 8.6 world; their technological capacities make Papua New Guinea 2010 7.7 them full innovation-based economies. They Timor-Leste 2007 7.3 Lao PDR 2008 5.1 are among a small group of economies— Unweighted average for all 10.9 most of them in East Asia—that have gone Unweighted average for men 10.9 from low- to high-income status. Other Unweighted average for women 9.9 countries, like China and Vietnam, set policy Source: Montenegro and Patrinos 2014. directions for education in the late 1980s and 1990s; strong evidence of their educa- tional success emerged only recently. Benefits Promoting positive returns to human abound if governments sustain good policies capital investment through stable, over the long run. market-friendly macroeconomic East Asia and Pacific economies promoted environments fast and sustained growth in a variety of Individuals can use their economically valu- ways: able skills best in predictable conditions that • They created stable, market-friendly mac- favor investment and provide incentives for roeconomic environments. productive work. In rapidly growing East Asia • They improved labor force abilities and and Pacific economies, governments practiced skills, mostly through increased schooling. macroeconomic management and created price • They made education broadly relevant stability and clear rules of the game. These to current and future expected economic policies favored longer-term investments, needs. especially for physical investments such as • They prepared the next phase of reform manufacturing facilities and equipment. They before they exhausted the benefits of the created the conditions under which a better- current phase. skilled labor force can apply its knowledge and • They ensured benefits across the socioeco- skills productively. Favorable business environ- nomic spectrum and “progressively uni- ments and sound basic infrastructure further versalized” opportunities for human capi- promoted the ability to develop domestic and tal investments. export markets and industry. • They avoided inequality of income and By the early 1990s, East Asia’s experi- educational opportunity and outcomes. ence was being held up as the best path to prosperity. In a major World Bank policy These policies—implemented across a report, The East Asian Miracle, Birdsall and variety of sectors—set the stage for increases others (1993, 234) note, in the technological sophistication of national industries. With that sophistica- Macroeconomic management was unusu- tion came increased demand for educated ally good, providing the stable environment workers. essential for private investment. Policies to 38 GROWING SMARTER increase the integrity of the banking system, Measures of both quantity (figure 1.4) and and to make it more accessible to nontradi- quality (figure 1.5) show that rapidly growing tional savers, increased the levels of financial East Asia and Pacific economies are among the savings. Education policies that focused on world’s best performers—and they are continu- primary and secondary schooling generated ing to improve. In 2010, school systems in East rapid increases in labor force skills. Asia and Pacific provided twice as many stu- The report notes that, although the envi- dents with about 10 times as much instruction ronments themselves favored wealth creation, as they did in 1950, in many cases providing adequate labor force skills were still required high levels of learning (Barro and Lee 2013). to realize the potential gains. Although many economies in the region trans- formed schools and the school experience, adjusting attainment years for learning years Improving labor force skills and shows varying levels of success (box 1.2). abilities, mostly through increased This massive increase in supply and qual- schooling ity has been met with an even larger increase In 1950, adults over 25 years of age in East in demand. Returns to tertiary education Asia had less than half the world average for have overtaken returns to primary and sec- educational attainment (1.3 vs. 2.9 years). By ondary education as growing industries 2010, attainment had increased fivefold, con- constantly demand more and better skills. verging with the steadily rising world average Rapidly growing economies in East Asia were of 8.0 years of schooling per adult. The feared the first to realize that modern economies are drop in quality that might have accompanied in a race between education and technology. expansion did not occur. Coverage expanded, If workers lack the skills, abilities, and knowl- and quality was maintained or raised. edge to use leading technologies—or if the FIGURE 1.4 Educational attainment in East Asia is many times higher than it was two generations ago 8.0 8 7.6 7.9 7.5 Average years of schooling (population age 25+) 7.1 7 7.0 6.5 6.3 6 5.9 5.5 5.6 5.0 5.1 5 4.5 4.6 4.1 3.9 4 3.7 3.4 3.3 3.2 3 2.9 2.8 2.2 2 1.8 1.3 1 0 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 World East Asia Source: Data from Barro and Lee 2013. Note: Results are weighted by population. Timor-Leste and Pacific Island countries are missing. H I G H  Q UA L I T Y S C H O O L I N G A N D E CO N O M I C G R O W T H I N E A S T A S I A A N D PAC I F I C 39 FIGURE 1.5 Fast-growing East Asia and Pacific economies generally score well above the OECD average on PISA 650 600 2015 PISA mathematics score 550 500 450 400 350 300 OECD Singapore Hong Taiwan, Macao SAR, Japan Korea, B-S-J-G Vietnam Malaysia Thailand Indonesia average Kong SAR, China China Rep. (China) China OECD Top Performing Systems Above-Average Below-Average Performing Systems Performing Systems 25th percentile 50th percentile 75th percentile Interquartile range Source: Data from PISA 2015 (OECD 2016). Note: B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); OECD = Organisation for Economic Co-operation and Development; PISA = Programme for International Student Assessment. BOX 1.2 Learning-adjusted years of schooling can reveal how effective and efficient school systems are Average years of schooling are an important indi- on students of the same grade (grades 4 and 8) cator that can be used to compare educational throughout all participating economies, allowing for attainment across economies. However, one year of direct international comparisons of learning out- schooling can lead to more learning in some econo- comes after the same number of years in school. mies than in others. Adjusting years of schooling by The World Bank’s Human Capital Project has the amount of learning that takes place within those compiled globally comparable achievement out- years can provide valuable insight into how efficient comes in more than 160 countries and regions over and effective school systems are. the period 1965–2015, allowing for meaningful Using data from large-scale international assess- adjustments of learning internationally and over ments is a way to standardize the measure of student time. What do these adjustments reveal? An illustra- learning across economies. International assessments tion using TIMSS math scores for 2015 shows that such as the Programme for International Student years of schooling can often be very different from Assessment (PISA) and the Trends in International learning-adjusted years. Singapore, the highest- Mathematics and Science Study (TIMSS) can provide scoring economy in TIMSS 2015, served as the basis such measures. These assessments are administered for comparison in the adjustment. box continues next page 40 GROWING SMARTER BOX 1.2 Learning-adjusted years of schooling can reveal how effective and efficient school systems are (continued) While young people in Hong Kong SAR, China have than 7 years of schooling. While Hong Kong SAR, similar average years of schooling to young people in China; Japan; and the Republic of Korea have learning- the United States (14 and 13.7 years, respectively), stu- adjusted years of schooling within 0.5 year of actual dents in the United States have almost 2 fewer learning- years of schooling, youth in Malaysia and Thailand adjusted years of schooling. And whereas young people have learning-adjusted years that are more than 3 years in Singapore have close to 4 more years of schooling below the actual years of schooling. on average than young people in Thailand (14.4 and 10.5 years, respectively), the learning-adjusted mea- Source: World Bank 2018b. Note: For detailed discussion of learning-adjusted years of schooling and data sure shows that Singapore outpaces Thailand by more for more economies, see World Bank 2018b, 48. institutions that produce, sustain, and facili- From the beginning of the 20th century to tate them are ineffective—economies cannot the end of World War II, most jobs required build or sustain a comparative advantage. the skills of a high school graduate. After Technological know-how—broadly defined— World War II, most jobs required the skills raises the value of, demand for, and rewards to of a college graduate. From the mid-1970s, skills, abilities, and knowledge. Governments too few people obtained college-level skills. need to ensure that they have the institutions Wage inequality increased as education levels and infrastructure that allow their economies stagnated. to remain close to the technological frontier. Policies in the high-income East Asian In this race between education and technol- economies caused demand for human capi- ogy in the rapidly growing areas of East tal to outstrip supply, driving long-term real Asia, technology has sprinted ahead—but increases in wages and incomes (table 1.3). education and skills are largely keeping pace. These economies needed more and more diverse high-quality human capital. Their schools, universities, research bodies, and Making education broadly relevant other knowledge institutions now routinely to current and future expected produce this capital, keeping up in the race economic needs between education and technology. Increased education leads to increased pro- Debate continues regarding the size ductivity and income only if graduates find and specifics of the impact of education well-paying jobs. For education to drive eco- on growth and the mechanisms through nomic growth, it must provide graduates with which it acts. A survey of the literature by relevant skills—the kind they need to con- Permani (2009, 1) concludes that, in East front the challenges in the industries in which Asia in 1965–2000, “the complementar- they work. ity between education and other factors is Examining 100 years of data from the commonly seen as the driving force of eco- United States, Goldin and Katz (2008) show nomic growth.” Using data from house- how graduates obtained and used broadly hold surveys from 1990 to 2005 for several defined sets of skills. They document how and large East Asian economies, Di Gropello when the skill sets for high school and col- and Sakellariou (2010) find rising returns lege graduates were relevant to general trends to higher levels of education, despite in production that prevailed in a given era. large increases in the share of the labor H I G H  Q UA L I T Y S C H O O L I N G A N D E CO N O M I C G R O W T H I N E A S T A S I A A N D PAC I F I C 41 TABLE 1.3 Demand for human capital in East Asia has risen even faster than supply Cambodia, China, Indonesia, Philippines, Thailand, Vietnam, Indicator 1997–2007 1999–2005 1994–2007 1988–2006 1990–2004 1992–2006 % change in share of workers with Upper-secondary education −22.1 19.3 59.3 45.0 100.0 20.2 Tertiary education 65.0 15.2 190.3 37.1 98.4 94.0 % change in wage premiums to Upper-secondary education 286.4 75.3 −1.9 18.5 15.4 — Tertiary education 394.2 34.1 17.2 5.0 17.2 273.2 Elasticity of demand Positive Positive Positive Positive Positive Positive (sublevel with highest returns) (tertiary) (upper- secondary) (tertiary) (upper-secondary) (tertiary) (tertiary) Source: Di Gropello and Sakellariou 2010. Note: — = no data available. force educated to these levels (table 1.3). for skills was raised; growth continued Complementarity and two-way causality only where governments had taken earlier explain the virtuous circles in which growth actions to ensure that such skills could be raises demand for education, and higher produced in large quantities. In this policy numbers of skilled workers contribute to of “progressive universalism,” the initial growth. Policy makers want to understand focus on basic education and basic literacy the steps they can take to initiate or acceler- and numeracy shifted to longer, more com- ate this process. plex, and more sophisticated learning pro- The types and sophistication of technolo- grams and educational credentials. gies used in the workplace define what type of skills employers find relevant. Changing or increasing challenges in production deter- Successful economies mine the size of skills premiums. During the systematically decreased the initial phases of growth—from low-income to distance to the technological middle-income status—efficient production frontier usually requires only basic literacy, numeracy, A fundamental element of success has been and reasoning skills. In later phases, it requires the desire to improve technology. Long more advanced, specialized knowledge. before the arrival of the Internet, the high- Specialized institutions, such as technical and performing economies of East Asia were vocational education and training institutes, embracing a proto version of the knowledge provide industry-specific skills. The compat- economy as the long-term focus of national ibility between school-acquired skills and the economic development. Policy makers real- demands of modern work makes the broader ized the value of technology and the serious abilities conferred by rising general education lack of domestic technological capacity and levels increasingly relevant (see spotlight 2). designed policies to address the problem. Early in the East Asian growth miracle, National development strategies prioritized basic literacy and numeracy were compat- the acquisition of technology because of its ible with the demands of newly emerging ability to set in motion a series of positive jobs in manufacturing. Uneducated rural changes. The virtuous circle rested on five dwellers did not make good industrial work- convictions: ers, but lower-secondary-school graduates did. The levels of knowledge and skills of • Economic growth comes from new ideas, the latter were broadly compatible with the which come from the advance of technol- need to read and follow directions in assem- ogy through increased firm capability in the bly and manufacturing jobs. Later, the bar short term and research in the long term. 42 GROWING SMARTER • The use of technology and other products of credit, extensive collaboration with lead- of science, technology, engineering, and ing firms, technology licensing, and reverse mathematics has changed the world econ- engineering. omy (and daily life) and enriched coun- Japan, for instance, took several policy tries that are technology leaders. actions in the postwar reconstruction period • Creating world-class firms systematically to promote firms with increasing technologi- generates many high-paying jobs, reducing cal capacity, including the following: the number of people competing for low- • Subsidizing the importation of machinery skill, low-productivity work. and capital goods • Countries need larger numbers of highly • Aggressively licensing technology and educated workers to keep up with, adapt acquiring new inventions and industrial to, and use new knowledge. designs • Expansion of equitable educational oppor- • Promoting the use of consultants from tunities begins at basic levels but ultimately technologically advanced countries to aims to expand knowledge- generating adapt know-how capacities formed through widespread • Sending engineers abroad to seek out tech- tertiary education and research. nologies, especially through overseas grad- uate education Japan, Korea, Singapore, and Taiwan, • Restricting foreign investment to firms China took advantage of rising global with valuable technologies. demand for inexpensive manufactured goods and used trade and industrial policies Singapore’s Economic Development to create national manufacturing industries. Board granted tax advantages that led to Firms in these industries created the jobs more than 300 multinational firms estab- that drove national economic development. lishing a presence in the newly independent Industrial policy brought firms like LG, country. It also built the infrastructure to Samsung, and Sony into existence by pro- facilitate the opening of factories, initially in viding privileges while not protecting them the Jurong Industrial Park. China’s growth from the need to compete. The aim was to as the “world’s manufacturer” owes much to create competitive firms capable of gain- the role of industrial growth targets in five- ing and defending market share in their year national and regional economic plans. industries, by both acquiring increasingly Vietnam has sought to attract foreign invest- sophisticated technologies and linking ment by firms like Intel and Samsung, but it firm success to continuous technological has struggled to take advantage of the knowl- improvement. This competitive pressure edge spillovers that these firms can catalyze. drove the acquisition of technology—and Increasing national technological capa- with it a need for ever better knowledge and bilities brought myriad benefits. In the short human capital. Policies created firms that term, it raised the quantity and quality of were sources of revenues for the govern- good jobs, creating modern manufacturing ment and profits for the firm, but they also sectors. It also raised demand for more and relied on companies’ ability to narrow the higher-skilled labor. distance to the technological frontier. Such increased demand is an important To keep this process moving forward, at rationale for broader investments in basic some point all rapidly growing East Asian education. Throughout the region, initial suc- economies used virtually every weapon in cesses in manufacturing industries created the arsenal of industrial development pol- jobs that spurred students and families to icy, including tariff protection and import demand more and better education. Pressure restrictions, direct government investment in on firms—even state-owned firms—to remain or ownership of key firms in selected indus- competitive created the continuous need to tries, tax incentives, preferential allocation upgrade technology, skills, and knowledge. H I G H  Q UA L I T Y S C H O O L I N G A N D E CO N O M I C G R O W T H I N E A S T A S I A A N D PAC I F I C 43 Successful economies prepared and use the full range of scientific and tech- early for the next phase in nological knowledge, which is gained by becoming a knowledge economy spending on R&D and advanced human capital. Measures of success include inputs Successful East Asia and Pacific economies (such as the share of GDP spent on R&D and made great efforts to gain access to and even- the number of researchers per 1,000 people tually achieve the capacity to produce lead- employed), but also, more critically, outputs ing technologies. Firms, research institutes, (such as the number of scientific articles pub- and universities all invested in improving lished, citations, patents, and new products their technological and research capacities. and processes). As the distance to the technological frontier Countries created various kinds of institu- in an industry decreases, the dynamism with tions to accelerate the process of technology which new products and technologies can be upgrading. Japan’s Ministry of International produced rises. But firms can remain com- Trade and Industry allocated the scarce for- petitive only if workers can also raise their eign currency needed to license technology ability to adapt and use technology in new to firms it believed could internalize the tech- ways—so demand for education and skills nology as they grew. Korea created dozens rises as well. of government-owned technological research The region’s rapidly growing economies bodies, such as the Korea Institute of Science did more than sustain successful implemen- and Technology and the Korea Institute of tation of policies; they also took early pol- Industrial Technology, which focused on icy actions to prepare for the next phase of applied research, engineering, and technologi- development before the gains from the cur- cal adaptation. rent phase were exhausted. For national eco- Korea transitioned from a low-income nomic development policies, doing so meant economy competing for foreign direct taking early action on two fronts: building investment on the basis of a high skill-to- knowledge institutions and progressively uni- wage ratio to a world leader in innovation versalizing their national education systems. and human capital. Its long-term planning Building knowledge institutions entailed raised investment in R&D from virtually creating trade missions, reverse engineer- zero before 1980 to 2 percent of GDP (just ing, and technological institutes, which grew below the average for the Organisation for into world-class research and development Economic Co-operation and Development) (R&D) and science, technology, and inno- by 1999. By 2014 the figure had risen to vation systems. Progressively universalizing 4.29 percent, the largest share of GDP spent their national education systems meant that, on R&D in the world. Korea has 50 percent after building broad skills through primary more researchers per 1,000 people employed and lower-secondary education, policy mak- than the United States (12.8 vs. 8.8) and ers began preparing for more sophisticated three times as many patent applications. general education systems and large-scale, Seventy-five percent of R&D spending comes high-quality tertiary education. from the private sector, and 80 percent of To achieve mature high-income status, R&D is for applied research or experimental economies continually need to create new development. Korea also leads the world in globally competitive firms, establish condi- the share of population with tertiary educa- tions in which leading firms can adapt and tion (Zastrow 2016). remain on top, or both, in a raft of sectors. In the aftermath of the Cultural Revolution, For this to happen, strong institutions dedi- China embraced the same policy goals and cated to creating and adapting knowledge policies that underpinned the success of Hong must take root. Reaching the technological Kong SAR, China; Japan; Korea; Singapore; frontier requires creating a critical mass of and Taiwan, China. Through policies to domestic capacity to produce, select, absorb, develop manufacturing, it raised the skills of 44 GROWING SMARTER TABLE 1.4 China is quickly becoming the world’s largest producer a high-income country; in 2012, it produced of PhDs in sciences and engineering more PhDs than any other country (table 1.4). Compound This progress is the result of deliberate, long- Degree 1986 2012 growth rate (%) term policy planning and implementation. Bachelor’s 227,764 3,038,473 11 S&E bachelor’s 109,101 1,258,643 10 Human capital protects people Master’s 15,221 434,742 14 from falling back into poverty S&E master’s 9,704 191,048 13 PhD 284 51,717 23 Progress in poverty reduction has been S&E PhD 228 27,652 21 unprecedented in East Asia and Pacific. As Source: Freeman and Huang 2015. recently as 2002, almost 1 billion people Note: S&E = science and engineering; PhD = doctor of philosophy degree. in the region were living in moderate or extreme poverty. Hundreds of millions of them escaped it—but many are vulnerable to its workforce and the demand for skills. As falling back into poverty. Close to 74 million a result of these policies, China became the people in East Asia and Pacific are living “world’s manufacturer” and achieved unprec- on less than US$1.90 a day, and 350 million edented economic growth. are living in moderate poverty (defined as At the same time, China implemented a US$3.10 a day) (World Bank 2013). long-term strategy to become a high-tech Estimates from household surveys show knowledge economy. For more than 25 years, that 50–80 percent of workers in the bottom it sent more students to the United States for socioeconomic quintile in most East Asia and PhD (doctoral) studies than any country in Pacific economies work in agriculture, often the world (Freeman and Huang 2015). At with very little capital or technology (World home, it encouraged firms to upgrade their Bank 2014). Many people who escaped pov- technology, and it built deep technological erty live close to the line and are at risk of and knowledge infrastructure. Freeman and falling back into poverty if conditions do not Huang (2015, 2) highlight China’s tremen- continue to improve (box 1.3). Education’s dous success at this endeavor: value goes beyond its ability to drive techno- logical change at the upper end of the labor In this short span of time [1990–2010] market. It yields benefits for people near or China leaped from bit player in global sci- below the poverty line, too. ence and engineering to become the world’s By definition, the poor have few assets largest source of [science and engineer- ing] graduates, second largest spender on and, in most cases, are completely dependent R&D and second largest producer of sci- on their human capital to earn income. They entific papers, in both cases behind [only] usually have the least social capital and face the United States. The number of patents in obstacles to accessing and benefiting from China increased so rapidly as to make China higher levels of education and social networks the number one country in patents. that decrease vulnerability. To what extent have people in the bot- China is now the world’s second-largest tom 40 percent of the income distribution funder of R&D after the United States, derived concrete benefits from improved spending more than US$200 billion a year on educational opportunities? The short answer R&D (Freeman and Huang 2015). Between is that they have derived substantial benefits 1990 and 2012, China’s output of scientific and continue to do so. Evidence strongly publications grew twentyfold, from just over suggests that the same dynamics that gen- 6,000 to close to 120,000. China also quickly erate returns to education at or near the increased both the number and the quality technological frontier and among high-tech of its researchers. In 1986, China produced firms and industries create strong positive fewer PhDs than a single large university in returns to education for the poor. Seminal H I G H  Q UA L I T Y S C H O O L I N G A N D E CO N O M I C G R O W T H I N E A S T A S I A A N D PAC I F I C 45 BOX 1.3 Gains in poverty reduction are fragile, particularly for people with little human capital In 2015, the extreme poor accounted for 2.5 Not surprisingly, households in higher eco- percent of the East Asia and Pacific region’s popu- nomic classes score higher on several indica- lation, the moderate poor accounted for almost 10 tors that go beyond income and consumption, percent, and the economically vulnerable accounted including access to services, asset ownership, and for about a quarter. The economically secure were employment. They are less likely to be employed the largest group in the region, accounting for in agriculture and more likely to be employed almost half of the population. The middle class accounted for 17 percent (figure B1.3.1). in formal sector salaried jobs, highlighting the importance of structural transformation and bet- ter jobs in supporting higher living standards. FIGURE B1.3.1 Both extreme and moderate Some patterns are less obvious, because they poverty in East Asia and Pacific are decreasing go beyond a simple association with greater con- sumption as economic class improves. Three 100 stand out: 90 • The extreme poor are a heterogeneous group 80 along several dimensions, including location 70 and ethnicity. % of total population • Even better-off groups, including the eco- 60 nomically secure and the middle class, have 50 low access to some basic services, including good-quality housing. 40 • Female labor force participation rates vary 30 widely across economic classes, according to 20 household surveys.a These rates are less than 40 percent for the extreme poor and about 10 43 percent for the moderate poor. In contrast, 0 the rates are 55 percent for the economically vulnerable and more than 68 percent for each 02 20 3 04 20 5 06 20 7 08 20 9 10 20 1 12 20 3 14 15 0 0 0 0 1 1 20 20 20 20 20 20 20 20 of the higher economic classes. Extreme poor (less than PPP US$1.90 a day) Moderate poor (PPP US$1.90–US$3.10 a day) Economically vulnerable (PPP US$3.10–US$5.50 a day) Source: World Bank 2018a. a. Respondents appear to underestimate women’s unpaid work in both Economically secure (PPP US$5.50–US$15.00 a day) the care economy and family businesses, with neither viewed as jobs, Global middle class (PPP US$15.00 and higher a day) as captured by time-use surveys. Although a shift to paid employment often does not lead to a parallel decline in time spent in unpaid activi- Source: World Bank 2018b. ties, participation in jobs that are defined as such (usually paid jobs) sets Note: PPP = purchasing power parity. the nonpoor apart from the rest of the distribution. work by Foster and Rosenzweig (1996) and manage them better, producing higher shows that poor farmers face technology profits. Households led by members with selection decisions that sharply influence completed primary education may earn their income levels. Natural experiments in returns on investment in agricultural inputs a range of low- and middle-income coun- that are 40–70 percent higher than house- try settings show that more-educated poor holds in which no members completed farmers adopt new technologies earlier primary school. The natural experiments 46 GROWING SMARTER allowed the authors to isolate the source of The region’s legacy of equitable the returns as the greater “ability to decode growth is under threat information” among the more educated. “Faced with new information,” they con- The massive increases in educational attain- clude, “educated individuals are better able ment and quality came from policies that to take advantage of technical change” sought to provide everyone with good basic (Foster and Rosenzweig 1996, 941). education, initially through at least lower- Higher returns induce higher rates of secondary school. These policies were imple- household investments in education. School mented both because of the inherent value enrollment rates rise with the increase in of education to the individual and because education’s profitability. “Policies resulting of the advantages accruing to countries with in greater technical change are complemen- well-educated workforces across the socio- tary with those increasing investment in economic spectrum. The growth of good jobs schooling: The returns to technical change demanding improved skills drove prosper- will generally be higher where primary ity at the top, but investments in educating schooling is more accessible and the returns everyone provided equally large and impor- to investment in schooling will be higher tant returns among the poor and vulnerable. where technical change is more rapid” Rising incomes among rural smallholders (Foster and Rosenzweig 1996, 951). helped the region to establish its reputation Duflo (2001) analyzes Indonesia’s mas- for high growth and relatively high equality. sive investments in school infrastructure in East Asia’s “catch-up” phase in educational 1973–78, when it built more than 61,000 attainment lasted roughly two generations. primary schools. She finds that the effort Between 1960 and 1990, increased educational increased both education and wages. attainment was a prominent feature of the E v i d e n c e f r o m t h e L a o P e o p l e ’s “growth with equity miracle” (Jain-Chandra Democratic Republic (Pimhidzai 2017) and others 2016). Economies were able to reveals that this dynamic is at work among reduce inequality while they grew rapidly partly smallholders in other parts of the region because of their equitable distribution of basic (table 1.5). The trend toward positive tech- educational opportunities. “Progressive uni- nical and technological change marks the versalism”—focusing on primary and lower- start of a virtuous circle in which education secondary education for all—was key. becomes more valuable, produces higher Since 1990, some countries have returns, and prompts increased investments struggled to maintain the inclusivity of by the poor in their own human capital. growth (figure 1.6). For the region as a whole, the Gini coefficient is now above TABLE 1.5 Education of the household head is related to selected agriculture indicators in the Lao People’s Democratic Republic, 2008 and 2013 Output per hectare Output per person (kilograms) per year (kilograms) % of produce sold on market Highest level of completed education of household head 2008 2013 2008 2013 2008 2013 No formal education 2,119 2,016 474 532 40.5 39.6 Some primary education 2,437 2,265 529 607 48.7 49.0 Complete primary education 2,644 2,552 660 695 57.7 59.1 Complete lower-secondary or more 2,743 2,700 707 719 59.5 56.7 Weighted average 2,541 2,454 610 659 53.0 53.3 Source: Pimhidzai 2017. H I G H  Q UA L I T Y S C H O O L I N G A N D E CO N O M I C G R O W T H I N E A S T A S I A A N D PAC I F I C 47 (worse than) the global average, and FIGURE 1.6 Inequality in East Asia has been rising since 1990 inequality seems to be rising most among some of the region’s largest countries. Increasing inequality in China and Indonesia China are of particular concern. Indonesia Among the explanations for the post- 1990 divergence is the failure to distribute Taiwan, China the opportunities for tertiary education as Japan equitably as those for basic education. Skill Korea, Rep. premiums to tertiary education are rising, Hong Kong SAR, China and students from poor families are much Singapore less likely than others to attend tertiary Philippines institutions. Analysis by the International Malaysia Monetary Fund shows that the contribution Thailand of skill premiums to inequality is three times as large in Asia as elsewhere (Jain-Chandra –5 0 5 10 15 20 and others 2016). Policy attention to a fair Net change in income inequality, 1990–2014 (Gini points) distribution of the now increasingly impor- tant tertiary level of education is needed if the Source: Data from the Standardized World Income Inequality Database (SWIID) Version 6.0. Note: Data for the Philippines are through 2012. Data for Thailand are through 2013. Data for all other region is to continue growing equitably and economies are through 2014. sustainably. employment, informal work and low-quality More remains to be done if jobs are still the norm. Two generations of countries are to avoid the progress have lifted millions out of poverty, “middle-income trap” but gains are fragile. Further progress is The world’s most successful examples of sus- needed to consolidate them. tained growth in East Asia exist alongside The success of Japan, Korea, Singapore, cases of prolonged economic stagnation. and other economies in the region is the Not all economies in the region have expe- result of sequential implementation of rienced the same economic success, and increasingly sophisticated policies over poverty remains even in the most successful almost four decades. The challenges are economies. Some economies have enjoyed formidable, and success is by no means stretches of high growth punctuated by crisis; guaranteed. Institutional weaknesses or others have tied their fortunes to the fluctuat- unfavorable political developments may ing value of commodities and primary prod- derail good intentions. Strength in one ucts or struggled to raise productivity over the area may be undermined by weaknesses long term. elsewhere. As the World Bank’s Vietnam National poverty rates have declined 2035 study attests, weak institutions or sharply, but many people are at risk of slip- the inability to adopt and enforce com- ping back into poverty. Some economies plementary policies can undercut strong are struggling with or recovering from peri- progress in education coverage and qual- ods of fragility, violence, or conflict. Others ity through secondary levels (World Bank are still in the early phases of structural and Ministry of Planning and Investment transformation. Despite plunging poverty of Vietnam 2016). These deficiencies can rates, East Asia and Pacific is still home to slow growth—and a two-point difference more than 74 million people living on less in growth can easily translate into another than US$1.90 a day. And while strides have generation missing the benefits of high- been made in creating modern industries and income status (figure 1.7). 48 GROWING SMARTER FIGURE 1.7 To achieve the level of GDP that Taiwan, China, reached in the early 2000s, Vietnam needs to grow at an average rate of 7 percent a year through 2035 25,000 GDP per capita (US$, 2005 PPP) 20,000 15,000 10,000 5,000 0 ar +2 +4 +6 +8 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 +1 +1 +1 +1 +1 +2 +2 +2 +2 +2 +3 +3 +3 +3 +3 +4 +4 +4 +4 +4 ye se Number of additional years Ba Taiwan, China Vietnam Vietnam (5% future growth) Vietnam (7% future growth) Source: Adapted from World Bank and Ministry of Planning and Investment of Vietnam 2016. Note: The base year is 1951 for Taiwan, China, and 1990 for Vietnam. GDP = gross domestic product; PPP = purchasing power parity. Conclusions macroeconomic environments. Countries implemented a variety of industrial policies, The record of East Asian economic growth including protection, subsidies, and directed since 1960 is unmatched. Economies that finance, but on the whole firms were required are home to more than two-thirds of the to compete in export markets where techno- region’s inhabitants have grown or are grow- logical prowess and continuous technological ing in transformative ways. The typical East upgrading were rewarded. Growth of firms— Asian worker today is less likely to be an especially in manufacturing—set in motion agricultural smallholder and more likely continually increasing demand for better- to be an urban dweller in formal, modern educated workers. Attention was paid to the employment. Massive investments have been relevance of education; policies promoted tech- made in the quantity of education obtained, nical and vocational education and training to and the quality of education has risen during match skills to industry needs, while preparing this expansion. In economies where growth is for a future in which the capacity to innovate dynamic, demand for educated workers has would be of fundamental importance. outstripped fast-rising supply; the highest Successful economies provided good- returns to schooling are going to those with quality basic education to the broadest pos- the most education. As in the world’s high- sible range of school-age children before income countries, the demand for more and investing in expansion at tertiary and pre- better skills is not exhausted or saturated. On primary levels. Initially, virtually all children the contrary, a main risk for countries is that were educated through the lower-secondary failure to produce adequate numbers of com- level. Later, targets were set for universal sec- petently educated and trained workers will ondary completion. stifle growth. Education’s value was not limited to for- Success relied first on sets of policies mal employment. Education also improved that ensured stable and market-friendly the lives of people in the bottom 40 percent, H I G H  Q UA L I T Y S C H O O L I N G A N D E CO N O M I C G R O W T H I N E A S T A S I A A N D PAC I F I C 49 including people in the most precarious Accumulation.” In The East Asian Miracle: economic circumstances. More and bet- Economic Growth and Public Policy, 189–200. ter schooling raised the productivity and New York: Oxford University Press. income of the poorest, providing both the Commission on Equitable and Sustainable Growth. 2010. Leadership and Growth. Washington, resources and the opportunity for a struc- DC: World Bank. tural transformation toward widespread Di Gropello, E., and C. Sakellariou. 2010. middle-class incomes and jobs. “Industry and Skill Wage Premiums in East Progress has been remarkable—but it has Asia.” Policy Research Working Paper 5379, been neither equitable nor sufficient. Some World Bank, Washington, DC. countries have stagnated, fluctuated, or failed Duflo, E. 2001. “Schooling and Labor Market to convert periods of high growth into sus- Consequences of School Construction in tained changes and prosperity. Poverty rates Indonesia: Evidence from an Unusual Policy have dropped dramatically, but far too many Experiment.” American Economic Review 91 people in the region still live in danger of fall- (4): 795–813. ing back into poverty. Education outcomes Foster, A. D., and M. R. Rosenzweig. 1996. “Technical Change and Human-Capital for 40 percent of the region’s students are Returns and Investments: Evidence from the unambiguously as good as in any country Green Revolution.” American Economic or region—at any income level. But up to 60 Review 86 (4): 931–53. percent of the region’s students are in schools Freeman, R., and W. Huang. 2015. “China’s ‘Great in which internationally comparable test Leap Forward’ in Science and Engineering.” scores point to serious deficiencies in learning NBER Working Paper 21081, National Bureau or in which no such information is available. of Economic Research, Cambridge, MA. The rise in inequality after more than Goldin, C., and L. Katz. 2008. The Race between three decades of growth with equity attests Education and Technology. Cambridge, MA: to the failure of investments in education to Harvard University Press-Belknap Press. keep pace with demand. Rapid technologi- Jain-Chandra, S., T. Kinda, K. Kochar, S. Piao, and J. Schauer. 2016. “Sharing the Growth cal change exerts a strong pull for more and Dividend: Analysis of Inequality in Asia.” better-educated workers, and rates of return Working Paper WP/16/48, International rise with years of education, despite rising Monetary Fund, Washington, DC. supply. Finishing the work of providing all Jimenez, E., and W. Dobson. 2013. “The Skills young people in the region with good-quality of Tigers.” In Human Capital Formation and basic education is as important as ever. Economic Growth in Asia and the Pacific, edited by W. Dobson. London: Routledge. Note Montenegro, C. E., and H. A. Patrinos. 2014. “Comparable Estimates of Returns to 1. See World Bank Poverty and Equity Data Schooling around the World.” Policy Portal: East Asia (http://povertydata Research Working Paper 7020, World Bank, .worldbank.org/poverty/region/EAS). Washington, DC. OECD (Organisation for Economic Co-operation Referencesa and Development). 2016. PISA 2015 Results. Paris: OECD Publishing. http://www.oecd.org Barro, R., and J.-W. Lee. 2013. “A New Data Set /pisa/. of Educational Attainment in the World, 1950– Permani, R. 2009. “The Role of Education in 2010.” Journal of Development Economics Economic Growth in East Asia: A Survey.” 104 (C): 184–98. Asia-Pacific Economic Literature 2 (1): 1–20. Birdsall, N., J. E. Campos, W. M. Corden, Pimhidzai, O. 2017. Unpublished manuscript for C.-S. Kim, H. Pack, R. Sabot, J. E. Stiglitz, Lao PDR Strategic Country Diagnostic, World and M. Uy. 1993. “Strategies for Rapid Bank, Washington, DC. Pritchett, L. 1997. “Divergence, Big Time.” a. Titles of publications that include “South Korea” refer to Journal of Economic Perspectives 11 (3, the Republic of Korea. Summer): 3–17. 50 GROWING SMARTER World Bank. 2013. “PovcalNet: The On-Line of the Marshall Islands, Federated States of Tool for Poverty Measurement Developed Micronesia, Republic of Palau, Independent by the Development Research Group of the State of Samoa, Kingdom of Tonga, Tuvalu, and World Bank.” World Bank, Washington, DC. Vanuatu, FY17–FY21. Suva, Fiji: World Bank. http://iresearch.worldbank.org/PovcalNet ———. 2018a. Riding the Wave: An East Asian /povDuplicateWB.aspx. Miracle for the 21st Century. Washington, DC: ———. 2014. East Asia Pacific at Work: World Bank. Employment, Enterprise, and Well-Being. ———. 2018b. World Development Report 2018: Washington, DC: World Bank. Learning to Realize Education’s Promise. ———. 2016a. “World Bank Concept Notes for Washington, DC: World Bank. the Systematic Country Diagnostics ———. Various years. World Development for Fiji, for Papua New Guinea, and Indicators [database]. Washington, DC: World for the Solomon Islands.” World Bank, Bank. https://data.worldbank.org/data-catalog Washington, DC. /world-development-indicators. ———. 2016b. World Bank Systematic Country World Bank and Ministry of Planning and Diagnostics for the Eight Small Pacific Island Investment of Vietnam. 2016. Vietnam 2035: Countries: Priorities for Ending Poverty and Toward Prosperity, Creativity, Equity, and Boosting Shared Prosperity. Washington, DC: Democracy. Washington, DC: World Bank. World Bank. Zastrow, M. 2016. “Why South Korea Is the ———. 2017. Regional Partnership Framework: World’s Biggest Investor in Research.” Nature For Kiribati, Republic of Nauru, Republic 534 (7605): 20–23. The State of Education in East Asia and Pacific 2 East Asia and Pacific is home to Policy makers throughout the region are a quarter of the world’s school- rightfully gratified with the large amount of age children, most of whom are high-quality schooling, while still recognizing enrolled in school how much more needs to be done. The world’s most successful education sys- tems are in East Asia. In small and large The size of national education systems economies, and at high and moderate levels in the region varies of gross domestic product (GDP) alike, school systems in the region achieve high average As noted, a quarter of the world’s school- performance, boast large shares of top per- age children live in East Asia and Pacific formers, and yield largely equitable learning (figure 2.1). More than 331 million stu- outcomes. Students in many of the region’s dents are enrolled in basic education in its low- and middle-income economies score 29 economies—more than in any other region better than their peers in high-income coun- except South Asia. tries, especially when scores are adjusted for National systems employ more than socioeconomic status. 20 million teachers, according to the Yet up to sixty percent of students in East World Development Indicators for 2016 Asia and Pacific are in poorly performing (World Bank, various years). China has the school systems where performance in key largest system, with 182 million students in subjects is either low or unknown. Students basic education (primary plus lower- and in rural Cambodia, Indonesia, or the Lao upper-secondary); Tuvalu has the smallest, People’s Democratic Republic may perform with fewer than 3,000 students (table 2.1). farm duties before having to undertake a long walk to an ill-equipped rural school and sit Most economies in the region have in a classroom staffed by an underprepared achieved universal primary enrollment teacher. In contrast, a student in Shanghai, China, may take a city bus to a modern Access to education in the region—especially school where a teacher has prepared a les- primary education—is widespread. Rates of son integrating computer-aided instruction coverage improved over the past two decades, that draws on the latest education research. and most economies have achieved universal 51 52 GROWING SMARTER FIGURE 2.1 A quarter of the world’s 1.3 billion school-age children primary enrollment. East Asia and Pacific has live in East Asia and Pacific 26 percent of the world’s school-age children but only 13 percent of the world’s out-of- Middle East and school children (table 2.2). Some 180 million North Africa North America of the region’s 187 million primary-school-age (83 million, 6%) (54 million, 4%) children are enrolled in school. Among countries with Emerging Systems— the categorization into four groups is discussed Europe and South Asia Central Asia further in the pages ahead—the two countries (344 million, 27%) (127 million, 10%) with the largest share of primary-school- age children not enrolled are the Solomon Latin America Islands (21 percent) and the Marshall Islands and the Caribbean (19 percent) (table 2.3). Eighteen countries (129 million, 10%) have greater than 95 percent enrollment in primary. Pacific Island countries are the only subregion with rates of out-of-school pri- Sub-Saharan East Asia mary children above 10 percent, with a total Africa and Pacific (214 million, 17%) (331 million, 26%) out-of-school population amounting to about 200,000 students. Indonesia and Thailand, each with roughly 8 percent of students Source: World Bank EdStats (World Bank, various years). Note: Figures show number of students enrolled in primary and secondary school and percentage out of school at primary level, account for of global enrollment. 3.1 million unenrolled students. TABLE 2.1 Education systems in East Asia and Pacific vary widely in size Number of students in system Economies Fewer than 100,000 Brunei Darussalam; Kiribati; Macao SAR, China; Marshall Islands; Micronesia, Fed. Sts.; Palau; Samoa; Tonga; Tuvalu; Vanuatu 100,000–1 million Fiji; Hong Kong SAR, China; Mongolia; Singapore; Solomon Islands; Timor-Leste 1–10 million Cambodia; Korea, Rep.; Lao PDR; Malaysia; Myanmar; Papua New Guinea; Taiwan, China More than 10 million China, Indonesia, Japan, Philippines, Thailand, Vietnam Source: World Bank EdStats (World Bank, various years). TABLE 2.2 East Asia and Pacific has 26 percent of global enrollment but only 13 percent of the world’s out-of-school children % of world’s % of age Number of Number of % of world’s out-of-school group out-of-school students Region students children not enrolled children (millions) (millions) South Asia 27 38 22 99.7 344.3 East Asia and Pacific 26 13 9 34.8 331.4 Sub-Saharan Africa 17 35 30 93.1 214.4 Latin America and the Caribbean 10 5 10 14.0 129.5 Europe and Central Asia 10 2 5 6.4 127.5 Middle East and North Africa 6 5 13 12.3 83.2 North America 4 1 5 2.7 53.7 World 17 263 1,284 Source: World Bank EdStats (World Bank, various years). T H E S TAT E O F E D U C AT I O N I N E A S T A S I A A N D PAC I F I C 53 TABLE 2.3 Access and coverage are significantly lower in countries with Below-Average Performing and Emerging Systems Number of % of primary Share of region’s primary out-of- Number of primary school-age children out-of-school Type of system school children school students not in school children Top Performing 38,559 11,369,954 0.3 0.6 Above-Average Performing 2,244,573 102,541,968 2.1 34.6 Below-Average Performing 3,559,394 52,659,532 6.3 54.9 Emerging 646,140 10,471,391 5.8 10.0 Source: World Bank EdStats (World Bank, various years). Note: The economies included in Top Performing, Above-Average Performing, Below-Average Performing, and Emerging Systems are listed in table 2.6. Figures for China are estimated from regional totals. Data were not available for Brunei Darussalam; Hong Kong SAR, China; Singapore; or Taiwan, China. FIGURE 2.2 Upper-secondary enrollment rates are high in Top Performing and Above-Average Performing Systems 130 120 110 100 90 Enrollment rate (%) 80 70 60 50 40 30 20 10 0 ea ru na re m ao C . R, a M ina Sa a oa do iji M M sia all sia or ds te Tu a La lu M DR So V mar on tu s ew ines d au un AR n p Pa Ph land SA hin li ng Br g S Japa an F Re va es in Ko sala m ua Tim slan go m l Da hi ne sh ay oP Pa Ch To ail Gu n a N pp -L ei , C lo an a, on ar al ya Is s Th I pu ili In n ac Ko M ng Ho Top Performing Systems Above-Average Below-Average Emerging Systems Performing Systems Performing Systems Source: World Bank EdStats (World Bank, various years). Note: The enrollment rate for some economies exceeds 100 percent because the statistic is calculated from the number of enrolled students versus size of population of a grade’s age cohort. Thus, for example, if there are enrolled students who repeated a year, or who are from regions not calculated in the census, the rate will exceed 100 percent. Progress is also notable in secondary (figure 2.2). Enrollment and completion enrollment, although more remains to be rates are high in Top Performing as well as done to ensure that all children complete Above-Average Performing Systems, but both lower- and upper-secondary school mediocre or low in others. These rates in and have the chance to develop their the Pacific Island countries are especially skills through postsecondary education troubling. 54 GROWING SMARTER The region has made tremendous schooling compulsory and start as early as progress in improving preschool access age three.1 East Asia and Pacific is home to roughly 119 million children of preschool age (3–6 Coverage is unequal across and within years). In 1980, the gross enrollment rate for countries, but it is improving preschool was 13 percent. By 2014, it had In multiple countries, including countries risen to 76 percent (figure 2.3). This prog- as diverse as the Philippines and Palau, ress is particularly notable if one considers more than half of children of preschool age that globally the gross enrollment rate in are enrolled. Yet in too many countries— preschools rose from 21 to 48 percent in the overwhelmingly in countries with Emerging same period. Systems—as few as one in three children is The achievement is striking given that only enrolled (figure 2.4). 4 of East Asia and Pacific’s 29 economies have compulsory preprimary schooling (all of Educational attainment among women them make one year of preprimary schooling has increased compulsory starting at age five). In contrast, 14 of the 30 countries in Latin America and In 1950, women in East Asia had com- the Caribbean have such laws; their current pleted, on average, less than a year of school- gross enrollment rate is just below that of ing, compared with the world average of East Asia and Pacific. In Latin America and 2.5 years. Six decades later, as the population the Caribbean, most countries’ laws cover more than doubled, average attainment of children starting at ages four to five and make women increased to 7.4 years of schooling— one year of preprimary schooling compulsory. catching up to the global average for women Some make up to three years of preprimary (figure 2.5). FIGURE 2.3 The gross enrollment rate in preschools in East Asia and Pacific has surged since 1980 90 80 70 Gross enrollment rate (%) 60 50 40 30 20 10 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 Europe and Central Asia Latin America and the Caribbean South Asia North America Middle East and North Africa Sub-Saharan Africa World East Asia and Pacific Source: World Bank EdStats (World Bank, various years). Note: Figures include high-income countries in each region. T H E S TAT E O F E D U C AT I O N I N E A S T A S I A A N D PAC I F I C 55 FIGURE 2.4 The increase in preschool enrollment has not been even everywhere for everyone 100 90 80 Preschool enrollment rate (%) 70 60 50 40 30 20 10 0 or ia te ar all DR Ph nds do s sia d lau m M na lia re n p. ia alu R, s a st ew G u ea c ld ne d cifi in t a an Tim od ys Re or nm es na ua lan go in i p ne ar ao P Pa Ch uv Ch pi a ala ail Ja Pa W -L u b et pu Van Isl a, on ya Ko n Is ilip T m Th Vi M d L M Ca In SA an Ko ng mo sh aN ng ia lo As So M Pa Ea Ho 1990 2000 2015 Source: World Bank EdStats (World Bank, various years). Note: The starting data for Vanuatu are 1992. The midpoint is 1999 for the Marshall Islands and Myanmar; 2001 for Tuvalu and Vanuatu; and 2002 for Hong Kong SAR, China; Palau; Papua New Guinea; and Timor-Leste. The ending data are 2008 for Papua New Guinea; 2009 for the Philippines; 2011 for Hong Kong SAR, China; 2012 for Mongolia; and 2014 for Indonesia, Japan, and Palau. Some starting and midpoint data are missing. FIGURE 2.5 Average years of schooling among women in East Asia rose sharply between 1950 and 2010 8 8.0 7.6 7.9 Average years of schooling (population age 25+) 7.5 7.4 7.1 7.4 7 7.0 6.9 6.5 6.5 6.9 6.3 6.3 6 5.9 5.9 5.5 5.6 5.7 5.1 5.4 5 5.0 4.9 5.1 4.5 4.6 4.4 4.4 4 4.1 4.0 3.7 3.6 3.9 3.7 3.2 3.4 3.3 3.3 3 3.0 3.1 2.9 2.8 2.8 2.5 2.5 2.2 2.0 2 1.8 1.5 1.3 1.2 1 0.8 0 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 World World female East Asia East Asia female Source: Data from Barro and Lee 2013. Note: Results are weighted by population. Timor-Leste and Pacific Island countries are missing. 56 GROWING SMARTER Returns to schooling are rising, and Singapore; Taiwan, China; and Thailand have labor markets for young people are repeatedly benchmarked learning outcomes changing against those of other countries. China and Vietnam have lately joined the group of East Increased schooling makes a concrete differ- Asia and Pacific economies that regularly par- ence for job prospects and earning among ticipate in PISA. Lower-income countries such youth in East Asia and Pacific. Many young as Cambodia, Lao PDR, Myanmar, Papua people in the region now complete secondary New Guinea, and Timor-Leste as well as school and progress to various types of post- Pacific Island countries have not participated secondary studies. Job prospects for gradu- in either PISA or TIMSS. Some are mak- ates are vastly better than they used to be: the ing plans to start benchmarking. Cambodia, population 20–24 years old in Vietnam has, for example, plans to take part in the first on average, 8.6 years of schooling compared iteration of the Organisation for Economic with 7.5 years of schooling in the working- Co-operation and Development (OECD) age population in general. This age group also PISA for Development exam in 2018. has by far the largest probability of working in the wage sector and obtaining high returns to additional years of schooling (Barro and Education systems fall into Lee 2013). four groups In East Asia and Pacific, economies across the income spectrum have participated in inter- Some economies in the region nationally comparable exams. Samples of benchmark learning outcomes using students representing about 55 percent of the international assessments region’s enrollment have taken internationally Internationally comparable standardized tests comparable exams.2 The picture that emerges are both the best and, from some perspec- is positive: an average composite constructed tives, the only technically sound way to mea- PISA score based on all instances of PISA sure education quality across economies and from 2000 to 2015 for the region comes in school systems. These test scores only assess above the OECD average (table 2.5).3 reading, math, and science knowledge and Figure 2.6 aggregates scores from the nine ability; debates about their value as proxies iterations of PISA and TIMSS administered are vigorous. Although no measure is per- since 2000 and 2003, respectively. Averaging fect, scores on these tests reveal more about the two scores has some drawbacks. It elimi- the relative quantity and quality of learning nates trend information, combines scores of across countries than any other measure or varying robustness (based on how many times combination of measures. a country participated), and may change the Many countries in the region have bench- meaning of a score because the two exams do marked their systems internationally since not test exactly the same content or abilities. the tests appeared (table 2.4). Nine econo- For gauging overall system quality, however, mies have records of frequent participation, the average scores give the clearest picture of dating back to 1995, on the two main inter- a system’s performance. national large-scale assessments: Trends in A 100-point standard deviation is the International Mathematics and Science Study equivalent of a difference of three years of (TIMSS) and Programme for International learning. Figure 2.6 shows that, on average, Student Assessment (PISA). 15-year-old students in China and Vietnam Japan participated in all nine iterations have the equivalent of three to four more since the two tests adopted a common scoring quality-adjusted years of schooling than their system (500 points as the average, with a 100- counterparts in Indonesia and the Philippines. point standard deviation). Hong Kong SAR, Students in Shanghai and Singapore are as China; Indonesia; the Republic of Korea; many as seven to eight years ahead of students Macao SAR, China; Malaysia; the Philippines; from lagging countries. T H E S TAT E O F E D U C AT I O N I N E A S T A S I A A N D PAC I F I C 57 TABLE 2.4 East Asia and Pacific economies have participated in various international assessments since 1995 PISA PILNA (mathematics, science, TIMSS (mathematics 2000 and reading) (mathematics and science) EGRA/EGMA and reading) 2006 2009 2009 2003 2003 2007 1999 1995 2010 2015 2015 2014 2015 2013 2012 2012 2012 2011 2011 Type of system Top Performing Hong Kong SAR, China Japan Korea, Rep. Macao SAR, China Shanghai, China ^ Singapore Taiwan, China Above-Average Performing B-S-J-G (China)a Vietnam Below-Average Performing Indonesia + Malaysia Philippines + Thailand > Emerging Brunei Darussalam Cambodia Fiji Kiribati Lao PDR Marshall Islands Micronesia, Fed. Sts. Mongolia Myanmar Palau Papua New Guinea Samoa Solomon Islands Timor-Leste Tonga Tuvalu Vanuatu Sources: EGRA/EGMA data from Graham and Kelly 2017. PILNA data from PILNA 2016. PISA data sets and results reports are from http://www.oecd.org/pisa/. TIMSS data sets and results reports are from https://timssandpirls.bc.edu. Note: B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); EGMA = Early Grade Mathematics Assessment; EGRA = Early Grade Reading Assessment; PILNA = Pacific Islands Literacy and Numeracy Assessment (see spotlight 1); PISA = Programme for International Student Assessment; TIMSS = Trends in International Mathematics and Science Study; ^ = Shanghai is counted in B-S-J-G (China) in 2015; + = Indonesia and the Philippines participated in TIMSS in 1995 (math) but did not complete the steps necessary for the 1995 data to be reported comparably with those of other countries (see Mullis and others 2000); > = Thailand participated in the fourth-grade TIMSS in 1995 but did not satisfy guidelines for sampling procedures at the classroom level. a. See box 2.1 for definition and explanation of B-S-J-G (China). 58 GROWING SMARTER TABLE 2.5 Economies in East Asia and Pacific routinely score Grouping economies by their scores on above 500 on PISA PISA and TIMSS sheds light on the roots of Average composite high performance. Although test scores never constructed PISA tell the whole story, they are the best available score of test takers Total population comparative indicator. Economy (2000–15) of 15-year-olds Scores from internationally compa- Shanghai, China 582 108,056 rable tests suggest a division of countries Singapore 550 48,218 into four categories: three performance Hong Kong SAR, China 541 65,100 levels plus a group of countries desig- Korea, Rep. 537 620,687 Japan 531 1,201,615 nated as having emerging education sys- Taiwan, China 526 295,056 tems because they do not participate in Macao SAR, China 517 5,100 standardized international assessments B-S-J-G (China) 514 2,084,958 (table 2.6). Table 2.7 shows the number of Vietnam 509 1,803,552 students enrolled at various levels in each Thailand 439 895,513 Indonesia 386 4,534,216 type of system. East Asia and Pacific average 505 Top Performing Systems consistently score OECD average 497 more than half a standard deviation above Sources: PISA 2012 (OECD 2013). Total population of 15-year-olds is from OECD 2016. Total population the OECD average score of the tests they have of 15-year-olds in Shanghai is from PISA 2012 (OECD 2013). taken. Comparator countries participating in Note: Excludes Malaysia’s 2015 score. Includes all economies participating in PISA 2015. Shanghai’s score is from PISA 2012. B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); these tests have almost always included the PISA = Programme for International Student Assessment. world’s top education systems and wealthiest FIGURE 2.6 Students in China and Vietnam are among the top performers in developing East Asia and Pacific B-S-J-G (China) Vietnam Macao SAR, China Japan Taiwan, China Thailand Korea, Rep. Malaysia Hong Kong SAR, China Singapore Shanghai, China Indonesia Philippines 106 points; 3.5 years of schooling 350 400 406 450 500 512 550 556 600 650 Composite constructed average of PISA and TIMSS performance over all available iterations Below-Average Performing Systems Average score, Below-Average Performing Systems Non-East Asia and Above-Average Performing Systems Average score, Above-Average Performing Systems Pacific countries Top Performing Systems Average score, Top Performing Systems Sources: Calculations based on PISA and TIMSS scores on nine assessments since 2000 (for PISA) and 2003 (for TIMSS). Note: Figure shows composite constructed average performance score with mean of 500 points and standard deviation of 100 points. The Philippines has only participated in TIMSS. B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); PISA = Programme for International Student Assessment; TIMSS = Trends in International Mathematics and Science Study. T H E S TAT E O F E D U C AT I O N I N E A S T A S I A A N D PAC I F I C 59 TABLE 2.6 Education systems in East Asia and Pacific can be categorized into four performance groups Number of Number of Share of Total number students assessed students cohort of students by PISA/TIMSSa assessed by tested Group Economy (millions) (millions) EGRA (millions) (%)b Top Performing Systems Hong Kong SAR, China; Japan; 24.1 24.1 — 100 (average score = 556) Korea, Rep.; Macao SAR, China; Singapore; Taiwan, China Above-Average Performing China, Vietnam 198.7 39.7c — 20 Systems (average score = 512) Below-Average Performing Indonesia, Malaysia, Philippines, 92.3 92.3 21.7 100 Systems Thailand (average score = 406) Emerging Systems Brunei Darussalam; Cambodia; Fiji; 16.3 0.5 5.1 35 Kiribati; Lao PDR; Marshall Islands; Micronesia, Fed. Sts.; Mongolia; Myanmar; Palau; Papua New Guinea; Samoa; Solomon Islands; Timor- Leste; Tonga; Tuvalu; Vanuatu Subtotal 156.6 26.8 Total 331.4 161.7d 48 Sources: OECD 2016; TIMSS 2015; data from World Bank EdStats (World Bank, various years). Data are the latest available data for each economy. Note: The average score in Organisation for Economic Co-operation and Development countries was 497. — = not available. EGRA = Early Grade Reading Assessment; PISA = Programme for International Student Assessment; TIMSS = Trends in International Mathematics and Science Study. a. PISA is taken by 15-year-olds (typically grade 10). TIMSS is taken by students in grades 4 and 8. b. When sample-based tests are used, the share refers to the sample base measured. c. China as a country has not participated in PISA. Scores are for the more economically advanced regions of Beijing, Shanghai, Jiangsu, and Guangdong, which are home to 15 percent of China’s pretertiary student population. d. This figure combines the number of students tested by PISA/TIMSS with the 5.1 million students tested by EGRA. Countries are not double counted in totals. TABLE 2.7 Enrollments in East Asia and Pacific, by level of education and type of system Millions Total enrollment, kindergarten Item Preprimary Primary Secondary through grade 12 East Asia and Pacific 58.8 177.1 154.4 331.5 Top Performing Systems 4.3 11.3 12.8 24.1 Above-Average Performing Systems 42.6 102.5 96.2 198.7 Below-Average Performing Systems 8.0 52.7 39.6 92.3 Emerging Systems 0.9 10.5 5.8 16.3 Sources: World Bank EdStats (World Bank, various years); Singapore Department of Statistics 2015; Taiwan Ministry of Education 2016; National Bureau of Statistics of China 2016. Note: Group totals do not equal East Asia and Pacific total because of rounding and the exclusion of Australia and New Zealand from system groups but not from East Asia and Pacific. The group of Beijing, Shanghai, Jiangsu, and Guangdong (China) represents 16 percent of China’s primary enrollment and 11 percent of its secondary enrollment (see box 2.1). economies, so performing consistently above GDP per capita that predicts much lower this average is a notable accomplishment. performance. This group includes the Beijing, This group includes Hong Kong SAR, China; Shanghai, Jiangsu, and Guangdong (B-S-J-G) Japan; Korea; Macao SAR, China; Singapore; regions of China, and Vietnam. (Together, and Taiwan, China. China and Vietnam account for 70 percent of Above-Average Performing Systems con- enrolled students in the region.) Their success sistently score up to half a standard devia- is a testament to how school-related policies tion above the OECD average despite having and practices can lead to high learning 60 GROWING SMARTER BOX 2.1 Educational attainment, achievement, and learning in provinces in China that do not participate in international assessments China as a whole has not yet participated in the Pro- This box characterizes the state of attainment, gramme for International Student Assessment (PISA). achievement, and learning in provinces for which no In 2009 and 2012, only the municipality of Shanghai international assessment data exist. In addition to participated. In 2015, four provinces and municipalities comparing B-S-J-G with the rest of China, it high- of China participated and published the results: Beijing, lights the educational gaps in attainment, access, Shanghai, Jiangsu, and Guangdong. On average, these and learning between rural and urban parts of the four provinces performed 1.3 years ahead of the Organ- country. isation for Economic Co-operation and Development Data on years of attainment are available for (OECD) average in math, 0.8 year ahead of the OECD all of China as well as for a selection of provinces. average in science, and 0.1 year behind the OECD aver- Figure B2.1.1 shows mean years of formal educa- age in reading in 2015. tion for the population 25 and over using data from These four provinces account for 17 percent of Barro and Lee (2013) and the China Household China’s total population and 15 percent of the coun- Income Project (CHIP, various years) for 2002, try’s pretertiary student enrollment. Within China, 2007, 2013. It shows that attainment is more than the provinces and municipalities of B-S-J-G are a year higher in these four provinces than in other among the wealthiest and most urban. They are not provinces surveyed by CHIP.a It also shows that, representative of all of China. among the provinces surveyed by CHIP, mean years of formal schooling are close to, or above, world and FIGURE B2.1.1 Estimated mean years of formal regional levels. Given that the Barro and Lee data schooling in China among adults 25 and older are on mean years of formal schooling for the whole converging to world and regional averages of China are below the mean presented from CHIP 10 data, it is possible to infer that, for the remaining provinces, mean years of formal schooling are below 9 those of the East Asia and Pacific region and the rest 8 of the world. Further disaggregating CHIP 2013 data by region shows that, on average, rural peo- Years of formal schooling 7 ple have 3.4 years less formal schooling than their 6 urban counterparts. Another way to capture the state of education 5 is by looking at enrollment and completion rates. 4 China has made strides in improving access and closing the education gap in the past decade. The 3 net enrollment rate in primary schools reached 99.9 2 percent in 2015, and the gross enrollment rate in junior-secondary school stood at 104 percent. Gaps 1 in educational achievement emerge after the compul- 0 sory education cycle. In the 2013 CHIP survey data, 1950 1960 1970 1980 1990 2000 2010 2020 dropout rates after completing the nine-year compul- Barro and Lee, East Asia and Pacific sory education cycle are higher for students outside Barro and Lee, China of Beijing, Jiangsu, and Guangdong (Shanghai was Barro and Lee, world not surveyed) than for students in these provinces. CHIP, Beijing, Jiangsu, and Guangdong (at start point)/ In Beijing, Jiangsu, and Guangdong, 70 percent of Shanghai, Jiangsu, and Guangdong (at end point) adults age 20–36 who completed junior-secondary CHIP, other provinces in China school go on to complete senior-secondary school; Sources: Data from Barro and Lee 2013; CHIP, various years. for their counterparts in other provinces, the figure Note: CHIP = China Household Income Project. is just 55 percent. box continues next page T H E S TAT E O F E D U C AT I O N I N E A S T A S I A A N D PAC I F I C 61 BOX 2.1 Educational attainment, achievement, and learning in provinces in China that do not participate in international assessments (continued) Differences in educational achievement between children in measures of educational readiness. By the rural and urban students are even more pro- end of the nine-year compulsory education cycle, nounced. As rural children from poor provinces gaps in learning have magnified. On the PISA 2015 progress throughout the education system, an in science, the difference in scores of students in the estimated 20 percent will transition to college, a top and bottom socioeconomic quintile in B-S-J-G rate 11–15 percentage points below the national equaled 4.4 years of schooling. The difference average and 30 percentage points below Beijing between urban and rural students equaled 2.4 years and Shanghai. Upon completing senior-secondary of schooling. school, rural youth from poor counties are 8 times In the absence of PISA data, one way of assess- less likely to attend four-year colleges, 11 times less ing the state of learning in the rest of China is to likely to attend elite “key” colleges, and 43 times extrapolate the urban-rural distribution of students less likely to attend the top two national universities who perform below basic proficiency to the rest of (Peking University and Tsinghua University) than China. This exercise suggests that some students in urban youth. both urban and rural areas face a learning crisis. Educational access is a function of wealth and Rural students constitute a disproportionate share of urbanicity. There is also evidence of fundamental students performing below basic proficiency. gaps in the level of learning between rural and urban children. For the 45 percent of China’s population Sources: CHIP household data for 2002, 2007, and 2013 (CHIP, various years); living in rural areas, gaps in learning manifest early. Barro and Lee 2013; Luo and others 2012; National Bureau of Statistics of China 2016; Wang and others 2011; Yang, Sicular, and Lai 2014. Starting in early childhood, China’s rural children a. CHIP has data on the following provinces or municipalities: Anhui (2002, receive fewer and lower-quality services than their 2007, 2013); Beijing (2002, 2013); Chongqing (2002, 2007, 2013); Gansu urban peers. Results from educational readiness test- (2002, 2013); Guangdong (2002, 2007, 2013); Hebei (2007); Henan (2002, 2007, 2013); Hubei (2002, 2007, 2013); Hunan (2013); Jiangsu (2002, 2007, 2013); ing show that, among children four to five years old, Liaoning (2002, 2012); Shandong (2013); Shanghai (2007); Shanxi (2002, 2013); the average urban child surpasses 94 percent of rural Sichuan (2002, 2007, 2013); Yunnan (2002, 2013); and Zhejiang (2007). outcomes even when overall resources for edu- Darussalam, Cambodia, Lao PDR, Mongolia, cation are far behind those of rich countries. Myanmar, Papua New Guinea, Timor-Leste, Below-Average Performing Systems score and the Pacific Island countries. There is half a standard deviation or more below the reason to believe that levels of educational OECD average and below their predicted achievement would be low if benchmarked score based on the level of GDP per capita. internationally, but many promising edu- Many of these countries have put significant cational initiatives are taking place in these resources and policy efforts into improving countries. their education systems, but results have been mixed. Indonesia, Malaysia, the Philippines, and Thailand constitute this group. Parts of the region are facing a Emerging Systems do not participate reg- learning crisis ularly in globally comparable standardized Reliable test data exist for about half of the tests, although many are preparing to begin region’s 331 million students. They show that benchmarking themselves against comparator roughly 50 million students are performing countries. These countries tend to be smaller below basic proficiency in tests (figure 2.7). and lower-income countries, some of which These poorly performing education systems have recent experience with fragility, violence, are contributing to the global learning crisis or conflict. The group comprises Brunei (World Bank 2018). 62 GROWING SMARTER FIGURE 2.7 Some economies in East Asia and Pacific outperform the OECD, but others face a learning crisis 650 600 2015 PISA mathematics score 550 500 450 400 350 300 OECD Singapore Hong Taiwan, Macao SAR, Japan Korea, B-S-J-G Vietnam Malaysia Thailand Indonesia average Kong SAR, China China Rep. (China) China OECD Top Performing Systems Above-Average Below-Average Performing Systems Performing Systems 25th percentile 50th percentile 75th percentile Interquartile range Source: Data from PISA 2015 (OECD 2016). Note: B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); OECD = Organisation for Economic Co-operation and Development; PISA = Programme for International Student Assessment. About 40 percent of students tested in top 20 scores since 2000 are from East Asia the region are enrolled in systems that and Pacific. The Top Performing Systems rep- produce high learning outcomes resent seven economies with an average score High-quality test data suggest that roughly of more than 550 points or 1.6 more years 63.8 million students in the region are learn- of learning than the average OECD mem- ing at high levels (figure 2.8). The data used ber country. Top Performing Systems enroll are a composite average of PISA and TIMSS some 24 million students, or 7 percent of the performance, constructed based on the results region’s students. of the nine iterations since 2000 (for PISA) The Above-Average Performing Systems— and 2003 (for TIMSS). After these dates, the B-S-J-G (China) and Vietnam—also exceeded two tests used a common scoring system: OECD average performance. Together they 500 points as the average, with a 100-point account for 39.7 million students. Their per- standard deviation. For PISA, 30 points is formance is significantly above what is pre- equivalent to a year’s worth of learning. The dicted based on their GDP per capita. They constructed averages have the advantage of represent 12 percent of the region’s students. providing the best information on the perfor- Together, Top Performing and Above- mance of systems as a whole. As mentioned Average Performing Systems have about 63.8 earlier, combining scores in this manner million students. Below-Average Performing means that the composite scores do not pro- and Emerging Systems combined have about vide trend information and that the scores 97.9 million students. Students in these sys- by country are not equally robust (figure 2.6 tems are performing at levels below pro- shows the distribution of test scores in East ficiency or whose learning outcomes are Asia and Pacific). unknown. Up to 60 percent of students in East Asia and Pacific dominates the ranks East Asia and Pacific are thus enrolled in sys- of top scorers: 6 of the top 10 and 8 of the tems in a learning crisis. T H E S TAT E O F E D U C AT I O N I N E A S T A S I A A N D PAC I F I C 63 FIGURE 2.8 Sixty percent of students in Figure 2.9 assesses the scale of the learn- East Asia and Pacific are in education systems ing crisis in East Asia and Pacific by showing that are in crisis the share of students performing below basic proficiency on the most common assessment available—reading. The results reflect all reading assessments conducted as part of PISA, the Early Grade Reading Assessments (EGRAs), or the Pacific Islands Literacy and 40% Numeracy Assessment (PILNA). Including (63.8 million) all of these results increases the number of countries that can be analyzed, but it has the 60% drawback of putting together tests that are (97.9 million) not strictly comparable. Levels are defined here as similarly as possible. For EGRAs, figure 2.9 reports the percentage of students in second grade who are unable to read a word. PISA students scoring below level 2 Students in systems with learning crises on the reading assessment are considered Students in Top Performing and Above-Average below basic proficiency.4 PILNA scores rep- Performing Systems resent the percentage of fourth-grade stu- dents scoring below level 3 on the reading Sources: Calculations based on a composite constructed average of PISA assessment.5 and TIMSS performance on the nine iterations of these assessments since 2000 (for PISA) and 2003 (for TIMSS). Figure 2.9 shows that all East Asia and Note: PISA = Programme for International Student Assessment; Pacific economies struggle to varying degrees TIMSS = Trends in International Mathematics and Science Study. FIGURE 2.9 East Asia and Pacific economies struggle to varying degrees to ensure that students reach basic proficiency levels % of students performing below basic proficiency on reading assessment 60 50 40 30 20 10 0 Th ia a m J-G hina M na) SA ore a an Ta Viet . In and In sia sia M ga r Gu s La a Ca DR Va dia or tu cifi ste s p Sin hina Pa Ph ma ew ine nd in e Re iw na Tim ua in ys p n ne ne oP Pa -Le Ch bo hi ao gap sla To ail Ja n a N pp B- n, C ala n a, ,C (C do do ya m R, cI re pu ili AR a Ko gS S- n ac Ko M ng Ho PISA (2015) EGRAs (various years) PILNA (2015) Sources: PISA 2015 (OECD 2016); national EGRAs (EGRA, various years); PILNA 2016. Note: For PISA (Programme for International Student Assessment), percentage of students scoring below level 2 on the reading assessment. For EGRA (Early Grade Reading Assessment), percentage of students in second grade unable to read a word. EGRAs were given in the following years: Cambodia (2012), Indonesia (2014/15), Lao PDR (2012), Myanmar (2015), Papua New Guinea (2012–14), the Philippines (2014), Timor-Leste (2011), Tonga (2009), and Vanuatu (2010). For PILNA (Pacific Islands Literacy and Numeracy Assessment), percentage of fourth-grade students scoring below level 3 on the reading assessment. PILNA data are not broken down by country. B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China). a. Weighted response rate in Malaysia fell short of standard PISA response rate. Coverage in Malaysia is too small to ensure comparability. 64 GROWING SMARTER to ensure that students are learning enough Average years of learning of students to demonstrate basic proficiency. These levels in Top Performing and Above-Average are far higher in the Below-Average group school systems are significantly than in the Top Performing and Above- above predictions based on GDP Average groups. In countries with Emerging per capita Systems, a substantial share of students are East Asia and Pacific economies in Top unable to demonstrate basic proficiency (the Performing and Above-Average school sys- share is roughly as large as in the group with tems score significantly above the regres- below-average performance on PISA). sion line for scores by per capita income. Newcomers like B-S-J-G (China) and Developing and high-income countries Vietnam are performing much higher than from the region rank among the world’s their predicted levels. High-income coun- top performers tries have one or two years more learning Japan, Korea, and Singapore are perenni- than predicted by their GDP per capita. The ally top scorers on PISA and TIMSS. And provinces of B-S-J-G (China) and Vietnam the outstanding recent performance of have two to more than four years more B-S-J-G (China) and Vietnam shows that top (table 2.8). performance is not limited to high-income countries. When PISA and TIMSS scores are The region has more than its aggregated—combining reading, math, and share of top performers science for all age groups and years—6 of the The PISA scoring system consists of seven lev- top 10 and 8 of the top 20 countries are from els (from 0 to 6). Students who score at levels East Asia and Pacific. East Asia and Pacific 5 and 6 are considered “top performers.”6 economies also dominate the ranks of coun- Of all test takers, only 4.7 percent score at tries that routinely score above 500 on PISA. level 5, and only 0.7 percent score at level 6. Although only 35 percent of test takers come Students from East Asia and Pacific repre- from East Asia and Pacific, 96 percent of stu- sent 34 percent of all PISA test takers but dents routinely scoring above 500 are from account for 48 percent of level 5 and 6 per- the region. formers in science and 40 percent in math. TABLE 2.8 PISA scores on science in East Asia and Pacific are higher than predicted based on per capita income Mean PISA science score in 2015 Difference between actual and predicted GDP per capita in 2015 or latest (2011 Prediction based Equivalent years Economy international dollars) Actual on income Score of schooling Top Performing Systems Japan 35,804 538 479 59 2.0 Korea, Rep. 34,387 516 477 39 1.3 Singapore 80,192 556 516 40 1.3 Above-Average Performing Systems B-S-J-G (China) 22,037 518 57 61 2.0 Vietnam 5,668 525 394 131 4.4 Below-Average Performing Systems Indonesia 10,385 403 422 −19 −0.6 Malaysia 25,308 443 463 −20 −0.7 Thailand 15,345 421 440 −19 −0.6 Source: OECD 2016. Note: The Philippines is not included because the country has not participated in PISA. B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); GDP = gross domestic product; PISA = Programme for International Student Assessment. T H E S TAT E O F E D U C AT I O N I N E A S T A S I A A N D PAC I F I C 65 FIGURE 2.10 Vietnam and Beijing, Shanghai, Jiangsu, and Guangdong (China) have more top performers than the United States on PISA mathematics assessments 450 400 350 300 Number 250 200 150 100 50 0 Population Students Top performers in Top performers in (millions) (millions) science (thousands) mathematics (thousands) United States Vietnam and B-S-J-G (China) Source: PISA 2015 (OECD 2016). Note: B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); PISA = Programme for International Student Assessment. Although B-S-J-G (China) and Vietnam have the corresponding average for the OECD about the same size population and pool of as a whole, and the spread of learning out- students as the United States, they have more comes was comparable. Only in reading in than twice as many top math performers. The the bottom quintile did performance fail to United States has 19 percent more top per- equal the score for the corresponding OECD formers in science (figure 2.10). income group—and it was only three points lower. China’s concentration of high per- formers tilts up the scores of students from The equity of learning outcomes the richest quintile in this group. China and is greater in East Asia than in the Vietnam are two large middle-income coun- OECD tries whose education systems produce high The region has a low spread of learning out- and equitable learning outcomes for all stu- comes. In B-S-J-G (China) and Vietnam, for dents, outcomes that equal the superior per- example, average performance is high, and formance of the region’s high-income top many poor students do well (figure 2.11). performers (box 2.2). The region also has a disproportionately Girls consistently outperform boys in East large number of top scorers, well above what Asia. In reading, they are one year of learn- national income would predict. ing ahead of boys—a trend consistent with In 2015, the scores of the region’s high- that of the OECD on average. In science and income Top Performing Systems were higher math, with a few exceptions, girls do bet- and less dispersed than the OECD average, ter than boys, although the gaps are smaller. indicating high average learning with equity. In Macao SAR, China and in Thailand, boys In both math and science, Above-Average perform better than girls in science, but the Performing Systems had scores that were difference is only a quarter of a year of learn- higher at every socioeconomic quintile than ing (table 2.9). 66 GROWING SMARTER FIGURE 2.11 In Vietnam and Beijing, Shanghai, Jiangsu, and Guangdong (China), even poor students learn more than their OECD counterparts a. Reading b. Mathematics c. Science 600 600 600 581 576 558 548 575 549 574 534 555 533 533 537 519 512 514 525 516 539 534 535 522 503 503 520 506 500 494 500 500 507 485 488 508 476 PISA score PISA score PISA score 497 474 486 470 476 486 471 475 450 470 454 445 451 460 436 444 442 419 400 414 400 410 400 406 399 384 398 373 389 381 390 366 300 300 300 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 (bottom) (top) (bottom) (top) (bottom) (top) Wealth quintile Wealth quintile Wealth quintile OECD Top Performing Systems Above-Average Performing Systems Below-Average Performing Systems Performance of bottom 40 percent of students in Vietnam and B-S-J-G (China) Source: PISA 2015 (OECD 2016). Note: The economies included in Top Performing, Above-Average Performing, and Below-Average Performing Systems are listed in table 2.6. The horizontal line in each panel indicates the performance of the bottom 40 percent of students in Vietnam and B-S-J-G (China), which exceeded or matched the OECD average in mathematics and science. The top and bottom wealth quintiles refer to children who come from families with the top 20 percent and the bottom 20 percent of household income, respectively. B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); OECD = Organisation for Economic Co-operation and Development; PISA = Programme for International Student Assessment. BOX 2.2 Vietnam’s 2015 PISA performance and the Escuela Nueva initiative Vietnam ranks among the top 10 scorers on the become by 2035, primary school students must not Programme for International Student Assessment only master traditional knowledge and facts, they (PISA) 2015 (OECD 2016). Its performance continues must also develop competencies for independent to be significantly above average for its income group and innovative thinking. Research shows that non- and above that of many high-income countries. Its sci- cognitive and cognitive skills are causally linked and ence score remains 1.1 year of schooling ahead of the related. Children who develop strong socioemotional Organisation for Economic Co-operation and Devel- skills are more likely to persist with effort in the face opment (OECD) and 0.7 year ahead of the regional of hardship and to achieve high test scores. average. These scores were 21 points lower in reading To impart these skills, the Vietnam Escuela Nueva and 17 points lower in math than in 2012; students (VNEN) project was implemented in 2013–16, in this round were 0.67 year of schooling behind in with financial support of US$84.3 million from the reading and 0.5 year of schooling behind in math Global Partnership for Education. The project ini- relative to students who participated in PISA 2012 tially focused on 1,447 schools in disadvantaged (OECD 2013). The difference between the top and areas with ethnic minority populations. By the time bottom socioeconomic quintiles in science equates to the project closed, in May 2016, 2,671 additional 2.1 years of schooling. The difference between urban primary schools had implemented the VNEN model and rural students in science is equivalent to 1.3 years without financial support from the project. Almost of schooling. Girls do as well as boys. Students with twice as many additional schools applied the VNEN some early childhood education and development per- model than was initially envisaged. form 0.7 year ahead of those with no such experience. An impact evaluation was conducted over three To be ready to be productive citizens of a pros- school years (2013–16), with funding from Dubai perous industrial economy, which Vietnam seeks to Cares. The evaluation followed a randomized cohort box continues next page T H E S TAT E O F E D U C AT I O N I N E A S T A S I A A N D PAC I F I C 67 BOX 2.2 Vietnam’s 2015 PISA performance and the Escuela Nueva initiative (continued) of VNEN and regular students in 651 schools. schools already had a higher mean score at the base- Comprehensive sets of data were collected from stu- line of the study, which was 18 months after ini- dents, parents, teachers, principals, and schools. A tiation of the project. This difference was retained qualitative research part of the study implemented throughout the years, although it narrowed over the video recording of lessons and conducted stimulated evaluation period. Regression analyses show statisti- recall interviews in 15 schools. cally significant effects of the VNEN program, with The test scores of students from VNEN as well preferred specifications showing effect sizes of about as control group schools moved up as they moved one-fifth of a standard deviation, about 15 points in through the grades. However, students from VNEN Vietnamese and 18 points in math. TABLE 2.9 Girls in East Asia generally outperform boys Score differences (boys-girls) Assessment and economy Mathematics Science Reading PISA 2015 Japan 14* –14* –13* Korea, Rep. −7 10 –41* B-S-J-G (China) 6 –9* −16* Hong Kong SAR, China 2 1 –28* Indonesia −3 4 –23* Macao SAR, China −8* 8* –32* Singapore 0 −6* –20* Taiwan, China 6 −4 –25* Thailand −3 9* –31* Vietnam −3 3 −25* Malaysiaa −7* 4 –31* OECD average 8* –4* –27* TIMSS 2003, 2007 2003 Philippines grade 8 13* 7 n.a. International average grade 8 1 –6* n.a. 2007 Mongoliab grade 8 –9 1 n.a. International average grade 8 5 6 n.a. Mongoliab grade 4 1 5 n.a. International average grade 4 0 3* n.a. Sources: PISA 2015 (OECD 2016); TIMSS 2003, 2007. Note: For PISA 2015, on average, across countries, the difference between adjacent grades is about 40 score points. OECD average for reading excludes Austria. Score differences for reading are not given for TIMSS because it assesses math and science only. B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); OECD = Organisation for Economic Co-operation and Development; PISA = Programme for International Student Assessment; TIMSS = Trends in International Mathematics and Science Study; n.a. = not applicable. a. Coverage in Malaysia is too small to ensure comparability. b. Mongolia did not provide the necessary documentation for sampling, data collection, and scoring. Its achievement data are not reported in the main tables. * Statistically significant (see annex A3 of OECD 2016). The equity of performance in East Asia other gaps in countries that participate in and Pacific is not as clear when one looks PISA. Box 2.3 looks at educational attain- at gaps in learning outcomes along other ment gaps among ethnic minorities. Box 2.4 dimensions (figure 2.12). But one thing is looks at disability and inclusion in Vietnam’s evident: gender gaps are far smaller than education system. 68 GROWING SMARTER FIGURE 2.12 On the 2015 PISA science assessment, gender gaps are smaller than other kinds of gaps Difference in PISA science performance (PISA points) 140 120 100 80 60 40 20 0 –20 –40 do a sia re n ala . ga ia ail e Vi and m do na sia re n ala . ga ia ai e Vi land m do a sia re an ala . Th ysia Vi land m do a sia re an M ep. ga sia Th ore Vi nd m M Rep M Rep M Rep In Chin In in In in Th por Th por Ko apa Ko apa Sin ys Sin ys na na na na In hi Ko ap Ko ap a ne ne ne ne Sin y Ch Ch R p ala ail C ai et et et et a, a, a, a, J J J J Rich-poor Early childhood development- Urban-rural Female-male no early childhood development Source: PISA 2015 (OECD 2016). Note: PISA = Programme for International Student Assessment. BOX 2.3 Educational attainment gaps between majority and minority ethnic groups in China, the Lao People’s Democratic Republic, and Vietnam Large gaps in educational attainment show up early are a year behind. In the same year (2006), 8 percent between majority and minority ethnic groups in the of the ethnic minority population had completed Lao People’s Democratic Republic and Vietnam. In upper-secondary, and less than 1 percent had com- Vietnam, men who are not from the Hoa and Kinh pleted tertiary. These numbers are around 50 and ethnic groups are most likely to have completed only 20 percent, respectively, of the completion rates for primary school, whereas Hoa and Kinh men are most people of ethnic majority groups in Vietnam. likely to have completed lower-secondary school. In Lao PDR, Lao-Tai (ethnic majority) male Ethnic minority women are most likely to have received students living in rural areas are four times more no schooling at all, whereas Hoa and Kinh women are likely to have attended school than their male non- most likely to have completed primary school. Lao-Tai counterparts. Lao-Tai women are six times Gaps in educational attainment among minority more likely to have attended schools than non-Lao- groups in Vietnam have persisted over time. Data Tai women. from 2006 show that age-grade distortion is more Educational attainment gaps between ethnic severe among ethnic minority students than among groups exist in China, but they have narrowed. ethnic majority students. In first grade, 4.5 percent Between 1990 and 2005, the gap in enrollment of ethnic minority students are more than one year rates in basic education between the most and behind their age-appropriate achievement level. By the least enrolled ethnic group decreased from 18 third grade, 36.3 percent of minority students but to 9 percentage points. Over the same period, the just 15.3 percent among ethnic majority students minority groups among whom college attainment box continues next page T H E S TAT E O F E D U C AT I O N I N E A S T A S I A A N D PAC I F I C 69 BOX 2.3 Educational attainment gaps between majority and minority ethnic groups in China, the Lao People’s Democratic Republic, and Vietnam (continued) rates grew most were the Miao, Uyghur, and women have the same average educational attain- Zhuang. Disparities in tertiary enrollment remain, ment as Lao-Tai men in urban areas. In contrast, but they are declining. there is no sign of gender convergence among non- Across the region, ethnic minority women have Lao-Tai women and men, in either rural or urban lower average levels of educational attainment than areas. Among all ethnolinguistic groups, non-Lao-Tai ethnic minority men. In 2005, in both urban and women in rural areas have the fewest years of school- rural areas of China, ethnic minority women between ing on average, lagging urban Lao-Tai men by 6.6 the ages of 16 and 21 were only a third as likely to years in 2003. be enrolled in school and had nine years less educa- tion than ethnic minority men. In Lao PDR, Lao-Tai Source: Hall and Patrinos 2012. BOX 2.4 Disability and inclusion in Vietnam’s education system In Vietnam, roughly 15,500 children under the age use of sign language. More than 50 hearing people of six are deaf or have a hearing impairment. The were trained as communication facilitators or sign Vietnam Intergenerational Deaf Education Outreach language interpreters. In addition, almost 50 deaf (IDEO) project seeks to build a system of support for children in first grade received sign language–based these children. The project is funded by the Japan education by both deaf and hearing teachers in the Social Development Fund, administered by the 2015/16 school year through a pilot scheme. World Bank, and implemented by the World Concern The project also launched an interactive website Development Organization. (http://ideo.org.vn) to provide online sign language IDEO adopted an innovative model of family learning videos, sign language vocabulary, games, support teams, including a deaf mentor, a sign lan- and other materials on deaf education for deaf chil- guage interpreter, and a hearing teacher, to teach dren and their families, educators, and the public. sign language to deaf children and their families in A series of short sign language videos is expected to their own homes. In five years, the project provided be broadcast on the national education channel to home-based sign language lessons for 255 deaf chil- reach a wider audience. dren under six in Hanoi, Thai Nguyen, Quang Binh, Initial project evaluations show that using sign and Ho Chi Minh City. It also trained more than 50 language has helped to improve deaf children’s deaf adults to become mentors to deaf children and language, cognitive development, and communica- helped to train about 200 hearing teachers in the tion ability. Educational systems in the later in life. Enriched home environments, region’s low- and middle-income social capital, and higher levels of parental countries serve learners from the education are thought to be the main factors bottom 40 percent that contribute to better learning among stu- dents from wealthier families. The socioeconomic gradient for learning out- National education systems seek to ensure comes is a fundamental characteristic of mea- that all students get a chance to make the sured learning outcomes in all countries—and most of their potential. Evidence from a serious threat to equity and opportunity PISA shows how high- and middle-income 70 GROWING SMARTER economies in the region succeed in doing rote learning in their classrooms, along so, often much more so than their OECD with concerns about the purported lack peers (figure 2.11). Students from China and of deep understanding of what is learned. Vietnam in the second income quintile (21st– The empirical evidence indicates that stu- 40th percentiles of family income) have dents are indeed learning. PISA subscores scores in math, reading, and science that are in math measure the ability to recognize as high as or higher than OECD students’ and set up problems, perform mathematical average scores in these subjects regardless of operations, and interpret the meaning and income levels. significance of results. In 2012, Vietnam’s subscores for performing operations were slightly higher than those for structur- Subscores belie the myth of rote learning ing problems and interpreting results. But East Asia and Pacific economies have all three subscores were above the OECD heard concerns about the overreliance on average, pointing to Vietnamese students’ superior abilities on all aspects of mathe- matical processes and undercutting the idea FIGURE 2.13 In several economies in East Asia and Pacific, more that rote memorization accounts for their than a third of second graders cannot read a single word high scores. Further evidence that sound fundamen- tals— rather than rote learning or test- Indonesia taking skills—are being developed comes from Vietnam. Data from the Young Lives Tonga initiative, which closely follows cohorts Myanmar (Yangon) from birth through secondary school, show Vietnamese students pulling away from Philippines (Maguindanaoan) their peers in low-income countries in math Philippines (Hiligaynon) by third grade. When tested at ages 10 and 12, average Vietnamese students per- Cambodia form better than all but the top students Vanuatu in Ethiopia, India, and Peru (Young Lives 2013). Philippines (Ilokano) Timor-Leste Measured performance levels begin to Philippines (Cebuano) emerge early in basic education 0 10 20 30 40 50 60 Whether students learn a little or a lot in % of second graders who cannot school begins to be apparent in primary read a single word school, if not earlier. Countries whose sys- tems have sound fundamentals see results in Source: Graham and Kelly 2017. measures of early grade reading and math— Note: All data are from nationally or regionally representative samples completed in the second and systems on the path to failure often see half of the school year with the exception of the Philippines. The Philippines conducted its only nationally representative Early Grade Reading Assessment (EGRA) in 2013 for third and first graders poor results early. High prevalence of “zero only, in both English and Filipino. The results were similar to those in Indonesia: the rates of zero- scores” (students unable to read a single word readers for third graders were low—at only 1 percent for both languages—and reading comprehension was at 73 percent for Filipino. On the other hand, English reading comprehension word) in Indonesia and the Philippines are was only 32 percent. The 2014 EGRA study targeted regions and schools that taught in mother- a harbinger of the five to six years of dif- tongue languages. Zero-word scores are based on the oral reading fluency subcomponent for all countries except Cambodia and Timor-Leste, for which the score is based on the familiar word ference in years of learning between these reading subcomponent. In the Philippines, tests were conducted in the indicated mother tongues. countries and Vietnam that are brought to Scores for Timor-Leste are for combined tests of Portuguese and Tetum. Scores for Vanuatu are for tests in English. Data are from the following years: Cambodia 2012, Indonesia 2014, Myanmar 2015, light when 15-year-olds are tested on PISA the Philippines 2014, Timor-Leste 2009, Tonga 2009, and Vanuatu 2010. (figure 2.13). T H E S TAT E O F E D U C AT I O N I N E A S T A S I A A N D PAC I F I C 71 Continuous improvement of place in most countries. Few teachers belong performance has accompanied to unions. Teaching is generally a respected progressive universalism profession, although teachers are not uni- versally well paid. Families throughout the A recurring theme among high-performing region are known to value education highly economies is constant improvement—a trend and to place high expectations on their chil- evident in test scores on internationally com- dren to do well in school. Tutoring, extra parable standardized tests. Altinok, Diebolt, classes, and long hours of homework and and Demeulemeester (2014) calculate self-study are common. long-term trends in schooling quality for 24 mostly high-income economies for which sufficient data were available. The top three An emerging literature on the annual average growth rates of achievement roots of high performance belong to East Asian economies: Singapore highlights elements that (0.98 percent), Korea (0.90 percent), and promote learning Hong Kong SAR, China (0.55 percent). These For more than a decade, the education policy rates were all three to six times the average community has emphasized the need to focus rate of improvement of 0.165 percent. Japan policies on learning rather than merely school- improved at about that average rate. Thailand ing. Partially as a result, a body of literature is regressed by 0.26 percent. emerging that seeks to establish the determi- nants of learning outcomes and the effective- ness of interventions aiming to raise learning. The region’s education systems share Several meta-analyses of interventions have some common features been conducted (Evans and Popova 2015; National education systems in the region are McEwan 2015), along with controlled micro- diverse in size, organization, governance, lev- level analyses. Evans and Popova (2015, 12) els of autonomy, and funding systems. Some find that, across reviews, “the intervention school districts in China have more students category which most commonly produces the than the national systems of the smaller largest improvements in student learning is Pacific Island countries. Countries such as pedagogical interventions that match teaching Indonesia and Vietnam rely on contract teach- to students’ learning.” Whether called “adap- ers; in Japan and Korea, virtually all teachers tive instruction,” “structured pedagogy,” are career civil servants. or something else, they involve providing Still, some shared features give education resources for teachers and strategies that they in East Asia and Pacific its regional flavor. use to interact with students. Centralized decision making and control The findings of Evans and Popova (2015) of basic education is common. Most stu- are consistent with the highest-quality micro- dents in the region attend public primary level studies of determinants of learning. and secondary schools. Many countries Dobbie and Fryer (2011) followed charter stu- have a national curriculum and allow only dents in New York City for three years. They authorized textbooks to be used in public found that the ability of schools to produce schools. Through a variety of means, gov- consistently high learning outcomes among ernments often play a strong role in setting randomly assigned students had nothing national standards for teacher quality and to do with traditional resource-based input educational practice. Absenteeism of teach- measures—class size, per pupil expenditure, ers from classrooms is comparatively rare in teacher certification, and teacher credentials. most countries. Basic educational infrastruc- Instead, five factors—all pertaining to class- ture, including reasonably well-functioning room practices—explained half of the varia- administrative and financial systems, is in tion in learning outcomes: high expectations 72 GROWING SMARTER of students, increased time on task, high- The rest of this report shows that the roots dosage tutoring, data-informed instruction, of high performance in East Asia and Pacific and frequent feedback to teachers. economies are grounded in policies and prac- tices that systematically increase learning out- comes for students. Conclusions East Asia and Pacific has enjoyed tremendous, Notes albeit still uneven, success in raising education 1. In East Asia and Pacific, only Brunei quality. Since 1950, as the total population of Darussalam; Macao SAR, China; Myanmar; the region doubled, average per person lev- and the Philippines have such laws (UNESCO els of school attainment rose by a factor of 2015). five. This massive increase in educational 2. Student population numbers represent primary attainment did not lower quality. Economies and secondary school enrollment. Population with some of the highest gains in attainment numbers are from World Development also have the highest measured increases in Indicators for 2016 (World Bank, various achievement. years); enrollment numbers are from World Across income levels, many students in the Bank EdStats (World Bank, various years). region are at or above OECD average levels All figures are for 2016, except the figures of performance on PISA and TIMSS, and for Lao PDR, Malaysia, Myanmar, Thailand, Vanuatu, and Vietnam, which are for 2015. many of its economies—both middle- and Data are also from Singapore Department high-income—are among the world’s top of Statistics (2015); Taiwan Ministry of performers. Perhaps most important, in most Education (2016); and UIS (various years). economies in the region that participate in Figures include economies participating in at these tests, learning outcomes are more equita- least one round of the following assessments: ble than the average for the OECD, with mil- PISA, 2000–15; TIMSS, 1995–2015; the lions of students from the bottom 40 percent Progress in International Reading Literacy of the income distribution scoring as well Study (PIRLS), 2001–11; the Early Grade as average students in the OECD. East Asia Reading Assessment (EGRA), 2009–14; and Pacific’s successes provide clear proof of the Early Grade Mathematics Assessment concept that education systems in low- and (EGMA), 2009–14; Literacy Boost, 2009– 12; the Programme d’Analyse des Systèmes middle-income countries can promote high Éducatifs de la CONFEMEN (PASEC), 2011; learning outcomes among all students. and the Southeast Asia Primary Learning At the same time, a large share of students Metrics (SEA-PLM), 2015. in the region are in school but not learning 3. The OECD uses arithmetic average to nearly enough. Roughly 60 percent of stu- calculate the average OECD score. dents in East Asia and Pacific are in school 4. At level 2, “Some tasks require the reader systems that are struggling to escape from to locate one or more pieces of information, the global learning crisis (or have no data which may need to be inferred and may on which to determine their status). For a need to meet several conditions. Others variety of reasons, these systems have been require recognizing the main idea in a text, unable to impart basic proficiency. Some are understanding relationships, or construing meaning within a limited part of the text in low-income, largely rural, economies at the when the information is not prominent and beginning of their structural transformation. the reader must make low-level inferences. Others—representing the majority of sys- Tasks at this level may involve comparisons tems in crisis—are in large, middle-income or contrasts based on a single feature in countries. New ideas and new approaches the text. Typical reflective tasks at this level are needed to bring learning up to the level of require readers to make a comparison or their high-performing regional peers. several connections between the text and T H E S TAT E O F E D U C AT I O N I N E A S T A S I A A N D PAC I F I C 73 outside knowledge, by drawing on personal Countries?” Policy Research Working Paper experience and attitudes” (OECD 2016). 7203, World Bank, Washington, DC. 5. Level 3 is below basic proficiency in PILNA. Graham, J., and S. Kelly. 2017. “Early Grade Level 3 is described as the ability to “Locate Reading in East Asia and the Pacific.” the main events in a variety of texts. Identify Background paper for the World Bank flagship common language conventions in the use of report, Growing Smarter: Learning and text connectives and synonyms. Spell diagraphs; Equitable Development in East Asia and Pacific, identify and correct errors in some frequently World Bank, Washington, DC. used one syllable words” (PILNA 2016). Hall, G., and H. A. Patrinos, eds. 2012. Indigenous 6. At level 6, students can conceptualize, Peoples, Poverty, and Development. New generalize, and use information based on their York: Cambridge University Press. investigations and can adapt their knowledge Luo, R., L. Zhang, C. Liu, Q. Zhao, Y. Shi, S. Rozelle, to relatively nonstandard contexts. Students and B. Sharbono. 2012. “Behind before They at this level have a mastery of symbolic Begin: The Challenge of Early Childhood and formal mathematical operations and Education in Rural China.” Australasian Journal relationships and can use this understanding of Early Childhood 37 (1): 55–64. to develop new approaches and strategies in McEwan, P. J. 2015. “Improving Learning unfamiliar situations. Students at this level in Primary Schools of Developing can reflect on their actions and precisely Countries: A Meta-Analysis of Randomized communicate their actions and reflections Experiments.” Review of Education Research regarding their findings, interpretations, 85 (3): 353–94. arguments, and adaptations to original Mullis, I. V. S., M. O. Martin, E. J. Gonzalez, situations. At level 5, students can develop K. D. Gregory, R. A. Garden, K. M. O’Connor, and work with models for complex S. J. Chrostowski, and T. A. Smith. 2000. situations, identifying constraints and TIMSS 1999 International Mathematics specifying assumptions. Students at this level Report: Findings from IEA’s Repeat of the Third can work strategically using broad and well- International Mathematics and Science Study developed reasoning skills and are beginning at the Eighth Grade. Boston: International to reflect on their work, formulating and Study Center, Boston College, Lynch School of communicating their interpretations and Education and International Association for the reasoning. Evaluation of Educational Achievement. National Bureau of Statistics of China. 2016. China Statistical Yearbook. Beijing: National References Bureau of Statistics of China. http://www.stats Altinok, N., C. Diebolt, and J.-L. Demeulemeester. .gov.cn/enGliSH/Statisticaldata/AnnualData/. 2014. “A New International Database on OECD (Organisation for Economic Co-operation Education Quality: 1965–2010.” Applied and Development). 2013. PISA 2012 Results. Economics 46 (11): 1212–47. Paris: OECD. http://www.oecd.org/pisa/. Barro, R., and J.-W. Lee. 2013. “A New Data Set ———. 2016. PISA 2015 Results. Paris: OECD. of Educational Attainment in the World, 1950– http://www.oecd.org/pisa/. 2010.” Journal of Development Economics PILNA (Pacific Islands Literacy and Numeracy 104 (C): 184–98. Assessment). 2016. 2015 Pacific Islands CHIP (China Household Income Project). Various Literacy and Numeracy Assessment Regional years. “China Household Income Project Data.” Report. Wellington: New Zealand Foreign China Institution for Income Distribution, Affairs and Trade Aid Programme. Beijing Normal University. Singapore Department of Statistics. 2015. Dobbie, W., and R. G. Fryer Jr. 2011. “Getting “Education and Literacy.” Department Beneath the Veil of Effective Schools: Evidence of Statistics, Singapore. http://www from New York City.” NBER Working Paper .singstat.gov.sg/statistics/browse-by-theme 17632, National Bureau of Economic Research, /education-and-literacy-tables. Cambridge, MA. Taiwan Ministry of Education. 2016. “Education Evans, D. K., and A. Popova. 2015. “What Really Statistical Indicators.” Ministry of Education, Works to Improve Learning in Developing Taipei City. http://english.moe.gov.tw/. 74 GROWING SMARTER TIMSS (Trends in International Mathematics For All (EFA) Global Monitoring Report and Science Study). 2003. TIMSS 2003 Regional Overview: East Asia and Pacific. International Database. Boston: TIMSS and Paris: UNESCO. PIRLS International Study Center, Lynch Wang, X., C. Liu, L. Zhang, R. Luo, T. Glauben, School of Education, Boston College and Y. Shi, S. Rozelle, and B. Sharbono. 2011. International Associations for the Evaluation “What Is Keeping the Poor Out of College? of Educational Achievement. https://timss.bc Enrollment Rates, Educational Barriers, and .edu/timss2003i/mathD.html. College Matriculation in China.” China ———. 2007. TIMSS 2007 International Database. Agricultural Economic Review 3 (2): 131–49. Boston: TIMSS and PIRLS International Study https://doi.org/10.1108/17561371111131281. Center, Lynch School of Education, Boston World Bank. 2018. World Development Report College and International Associations for 2018: Learning to Realize Education’s the Evaluation of Educational Achievement. Promise. Washington, DC: World Bank. https://timss.bc.edu/TIMSS2007/intl_reports ———. Various years. Education Statistics (EdStats) .html. [database]. Washington, DC: World Bank. ———. 2015. TIMSS 2015 International Database. http://datatopics.worldbank.org/education/. Boston: TIMSS and PIRLS International Study ———. Various years. World Development Center, Lynch School of Education, Boston Indicators [database]. Washington, DC: World College and International Association for Bank. https://data.worldbank.org/data-catalog the Evaluation of Educational Achievement. /world-development-indicators. https://timssandpirls.bc.edu/timss2015 Yang, J., T. Sicular, and D. Lai. 2014. “The /international-database/. Changing Determinants of High School UIS (UNESCO Institute for Statistics). Various Attainment in Rural China.” China Economic years. Data for the Sustainable Development Review 30 (2014): 551–66. Goals [database]. Paris: UIS. http://uis.unesco Young Lives. 2013. “Making Progress: Report .org/. of Young Lives School Survey in Vietnam.” UNESCO (United Nations Educational, Scientific, Oxford Department of International and Cultural Organization). 2015. Education Development, University of Oxford. SPOTLIGHT 1 Education in the Pacific Island Countries: Achievements and Challenges The Pacific Island countries are a heteroge- additional challenges, which highlight the neous group of countries with wide differences need to make efficient and effective use of in their geography, institutional structures, resources. historical development, economic drivers, and many other aspects. They can be divided into three subgroups: Melanesia, Micronesia, and Role of the church Polynesia. Melanesia includes Fiji, Papua New The church plays an important and at times Guinea, the Solomon Islands, and Vanuatu. central role in the provision of education ser- Micronesia includes Kiribati, the Marshall vices in most Pacific Island countries. It is offi- Islands, the Federated States of Micronesia, cially recognized as a critical nonstate actor. Nauru, and Palau. Polynesia includes Samoa, Education was introduced in Tonga in 1826 Tonga, and Tuvalu. by the Free Wesleyan Church; primary educa- This subregion—sometimes referred to as tion has been compulsory there since 1876. Oceania—covers an area larger than that of the Russian Federation. Its most populous country is Papua New Guinea, with about Attainment 8.22 million people. The least populous coun- The Pacific Island countries have a mixed tries are Nauru and Tuvalu, with 10,000 and record of attainment in education. Between 11,000 people, respectively. 2005 and 2015, primary and secondary Some countries, like Samoa, are not very enrollments increased substantially. Fiji, dispersed. In contrast, Kiribati is spread over Kiribati, Samoa, Tonga, and Vanuatu have an area the size of India (despite having a land sustained net primary enrollment rates well mass of just 811 square kilometers, about the above 90 percent since the 1990s. Indeed, size of New York City). the Pacific Island countries outperform many The subregion is also home to hundreds other countries at similar income levels in get- of languages, dialects, and sociocultural ting children into school (figure S1.1). groups, making it one of the most diverse Some Pacific Island countries are even suc- regions in the world. Geographic barriers to ceeding in the more difficult task of keeping access—between islands or over difficult ter- children in school throughout primary school rain not adequately serviced by roads—pose (figure S1.2). Enrollment of girls is low in 75 76 GROWING SMARTER FIGURE S1.1 Primary enrollment rates vary in Pacific Island countries 100 90 80 Primary enrollment rate (%) 70 60 50 40 30 20 SPOTLIGHT 1 10 0 Fiji Samoa Kiribati Tonga Papua Vanuatu Tuvalu Micronesia, Marshall Solomon New Guinea Fed. Sts. Islands Islands Baseline (2000 or nearest year) Latest data Sources: World Bank EdStats (World Bank, various years); Pacific Islands Forum Secretariat 2015. FIGURE S1.2 Some Pacific Island countries are succeeding in keeping children in school throughout primary school 100 Rate of persistence to last year of primary school (%) 90 80 70 60 50 40 30 20 10 0 Fiji Tuvalu Tonga Samoa Kiribati Micronesia, Marshall Solomon Vanuatu Papua Fed. Sts. Islands Islands New Guinea Baseline (2000 or nearest year) Latest data Sources: World Bank EdStats (World Bank, various years); Pacific Islands Forum Secretariat 2015. Note: No data on baseline for the Solomon Islands. E D U C AT I O N I N T H E PAC I F I C I S L A N D CO U N T R I E S 77 some countries, but in many Pacific Island the 2012 assessment (43 percent), and the countries their enrollment, participation, proportion of students in the two lowest attendance, and achievement in school are proficiency levels for year 6 fell from 12 higher than those of boys. to 16 percent. At the highest proficiency levels, however, there was a slight decline, from 9 to 7 percent for year 4 and from 20 Achievement to 13 percent for year 6. Time spent in school does not necessarily • Girls demonstrated higher levels of literacy translate into learning outcomes, however. than boys. In 2015 the mean scores were Policy makers in Pacific Island countries are 462 for girls and 444 for boys in year 4 concerned about poor learning outcomes and 487 for girls and 469 for boys in from basic education. They have identified year 6. The trend holds for all three strands improvements in literacy and numeracy as of literacy (reading, language features, and an important shared goal. They view access writing). The share of students in the top SPOTLIGHT 1 to good-quality education not only as an three proficiency levels was also higher inherent right but also as a means of prepar- among girls than boys (21 vs. 13 percent ing young Pacific Islanders for the future. In in year 4 and 33 vs. 23 percent in year 6). 2006, the heads of states agreed that it was The proportion of girls in the bottom three essential to have all children in school and proficiency levels was lower among girls learning. They recognized that, to improve than boys (32 vs. 43 percent for year 4 and learning outcomes, they needed to have tools 17 vs. 27 percent for year 6). in place to measure and benchmark perfor- • In literacy, students in nongovernment mance accurately. schools outperformed students in govern- Fifteen countries worked together to ment schools in both year levels in 2015. develop Pacific-wide benchmark standards Students in nongovernment schools had for literacy and numeracy at grades 2, 4, 6, higher mean scores than students in gov- and 8. These benchmarks were developed ernment schools in both year 4 (459 vs. from curriculum skill components and learn- 453) and year 6 (488 vs. 476). ing outcomes considered to be common • Numeracy improved as children pro- across the national curricula of the 15 mem- gressed from year 4 to year 6. The propor- ber states. tions of students at or above the expected The results of the most recent Pacific proficiency levels for year 4 and year 6 Islands Literacy and Numeracy Assessment were 86 and 67 percent, respectively. The (PILNA) illustrate that the focus on learn- share of year 4 students in the three high- ing has started to make a difference, partic- est proficiency levels rose from 24 percent ularly in numeracy. The key findings from in 2012 to 34 percent in 2015. the 2015 PILNA regional report include the • Girls demonstrated higher levels of perfor- following: mance in numeracy than boys across the region in 2015. In year 4, 37 percent of girls • Literacy improved between 2012 and achieved the three highest proficiency lev- 2015 (PILNA 2016). Students seemed to els compared with 32 percent of the boys, improve during the transition from year and 12 percent of girls were in the bottom 4 to year 6 (equivalent to grade 3 and three proficiency levels compared with grade 5). However, in stark contrast to 16 percent of boys. In year 6, 52 percent numeracy scores, only 46 percent of stu- of girls and 47 percent of boys achieved the dents were at or above expected profi- top three proficiency levels, and 7 percent ciency levels in both year 4 and year 6. The of girls and 10 percent of boys scored in the proportion of students in the three lowest lowest three proficiency levels. proficiency levels for year 4 was lower in • The performance in numeracy in govern- the 2015 assessment (38 percent) than in ment and nongovernment schools varied 78 GROWING SMARTER between year levels in 2015. For year 4, Pacific Island countries struggle to get stu- students from nongovernment schools had dents to the next stage of basic reading skills. a higher mean score (512) than students After three years of instruction, only a frac- from government schools (504). This tion of students can read a simple short story result was evident in all three strands of and understand most or all of it. Although numeracy (numbers, operations, and mea- part of the problem is that students are learn- surement and data). For year 6, the result ing to read in an unfamiliar language, EGRA was reversed, with the mean scores of stu- results suggest that the low proportion of dents from government and nongovern- students meeting the literacy expectations as ment schools at 524 and 520, respectively. measured by PILNA is strongly related to This result was evident in all three strands very weak foundational reading skills in the of numeracy. early years of primary education. The challenge of education systems in the Pacific Islands is the quality of instruction SPOTLIGHT 1 and efficiency in the use of instructional time. Results from Early Grade Reading Teachers and schools struggle to help students Assessments to become readers even after three years of In addition to testing with PILNA, countries instruction (not counting time spent in pre- in the region have been undertaking Early schools). Substantial improvements in the Grade Reading Assessments (EGRAs), with quality of instruction in the early years are support of the World Bank. The results from needed so that all students are able to read EGRAs conducted between 2009 and 2013 and write well no later than grade 3. Box S1.1 are provided in table S1.1. groups countries by their level of performance. TABLE S1.1 Early reading assessments in Pacific Island countries % of students who, after 3 years of schooling, cannot read a single: % of students Word from the who, after 3 years Language of first sentence in a of schooling, are Country or province test in grade 3 Letter Familiar word short story considered readersa Vanuatu Francophone French 12 5 7 24 Anglophone English 8 13 10 24 Tonga Tongan 0 1 3 30 Tuvalu Tuvaluan 8 18 14 21 Kiribati Kiribati 2 4 3 27 Samoa Samoan 2 20 20 19 Solomon Islands English 1 18 22 33 Papua New Guinea Madangb English 2 6 6 22 East New Britainc English 1 3 12 39 Western Highlandsd English 0 8 11 6 National Capital Districte English 2 2 16 11 Sources: World Bank 2010, 2011a, 2011b, 2011c, 2012a, 2012b, 2013. a. Average comprehension at or above 80 percent. b. Average comprehension at 47 percent. c. Average comprehension at 55 percent. d. Average comprehension at 46 percent. e. Average comprehension at 76 percent. E D U C AT I O N I N T H E PAC I F I C I S L A N D CO U N T R I E S 79 BOX S1.1 Grouping Pacific Island countries by level of performance Pacific Island countries can be classified into three modest performers in lower-secondary school performance groups: completion: the Marshall Islands, the Federated States of Micronesia, Nauru, Tonga, and Tuvalu • High performers in access, high performers in • Modest performers in access, low perform- primary school completion, and modest perform- ers in primary school completion, and poor ers in lower-secondary school completion: Fiji, performers in lower-secondary school comple- Kiribati, Palau, and Samoa tion: Papua New Guinea, Solomon Islands, • Modest performers in access, modest perform- and Vanuatu. ers in primary school completion, and very SPOTLIGHT 1 Public expenditure on education education policies have been largely success- ful in expanding access to education, they Spending on education in many Pacific Island come at the expense of good-quality inputs, countries is high, with education consistently such as teacher training, curriculum develop- receiving the largest share of government ment, and classroom materials. budgets from year to year. In percentage terms, Pacific governments have made sizable allocations to human development budgets. References Pacific Island countries spend nearly twice Pacific Islands Forum Secretariat. 2015. 2015 as much per capita on education and health Pacific Regional MDGs Tracking Report. as other low-income small island states Suva, Fiji: Pacific Islands Forum Secretariat. spend. Their per capita expenditures on pri- PILNA (Pacific Islands Literacy and Numeracy mary education are an order of magnitude Assessment). 2016. 2015 Pacific Islands greater as those of many low-income African Literacy and Numeracy Assessment Regional countries. Report. Wellington: New Zealand Foreign But funding for health and education is Affairs and Trade Aid Programme. World Bank. 2010. How Well Are Ni-Vanuatu dependent on bilateral aid, which has declined Children Learning to Read in French? Vol. 2: significantly in recent years or has skewed Vanuatu Report [English]. Pacific early reading away from areas of education where returns assessments series. Washington, DC: World are likely to be greatest. More than half of Bank. the education assistance from Australia’s ———. 2011a. Madang Early Grade Reading Department of Foreign Affairs and Trade Assessment (EGRA) Survey: 2011 Diagnostic goes to scholarships, for example. Results Report. Washington, DC: World Bank. The quality of spending and service deliv- ———. 2011b. National Capital District Early ery is a key factor in poor education out- Grade Reading Assessment (EGRA) Survey: comes. Service delivery is hampered by the 2011 Diagnostic Results Report. Washington, fragmentation of financing and administra- DC: World Bank. ———. 2011c. New East Britain Early Grade tion, inefficient financing flows, and weak- Reading Assessment (EGRA) Survey: 2011 nesses in how education services are managed Diagnostic Results Report. Washington, DC: at the subnational level. World Bank. Salary spending comprises more than 75 ———. 2012a. How Well Are Ni-Vanuatu percent of education budgets in many Pacific Children Learning to Read in English? Vol. 2: Island countries. And although tuition-free Vanuatu Report [English]. Pacific early 80 GROWING SMARTER reading assessments series. Washington, DC: ———. 2013. Western Highlands Province Early World Bank. http://documents.worldbank.org Grade Reading Assessment (EGRA) Survey: /curated/en/690361468133520894/Vanuatu 2011 Diagnostic Results Report. Washington, -report. DC: World Bank. ———. 2012b. How Well Are Tongan Children ———. Various years. Education Statistics Learning to Read? Vol. 2: Tonga Report (EdStats) [database]. Washington, DC: [English]. Pacific early reading assessments World Bank. http://datatopics.worldbank.org series. Washington, DC: World Bank. /education/. SPOTLIGHT 1 Institutional Alignment for Policy Implementation 3 The experience of high-performing education be expected to engage in meaningful classroom systems in East Asia underscores the critical interactions that produce learning. role played by institutional alignment and sound administrative systems at a variety of levels in implementing policies and reforms Institutional alignment is critical that ultimately raise learning outcomes in to ensure that students learn schools. Institutional alignment is a critical Education systems require effective institu- component of ensuring that policies across tional alignment at a variety of levels and key domains are aligned and then imple- among myriad actors (World Bank 2018). mented, adjusted, evaluated, and revised Capacity levels matter for national, provin- to foster continuous improvement. Top cial, and municipal governments; national Performing Systems fostered alignment by ministries of education and subnational edu- setting targets and demanding results, advo- cation authorities; and district- and school- cating for education in national spending, and level bodies. The strength of institutions can providing the impetus for alignment across strongly affect the quality of interactions sectors. Systemic reform was made possible between education officials and providers, by investing in improving the capacity of the on the one hand, and stakeholders from civil education system itself. society, especially parents and employers, on From the very basics of ensuring safe, the other. adequate physical space to the intricacies of These interactions take place within con- decentralizing curricula with a framework for texts shaped by political influences and politi- learning, institutional alignment can be a key cal culture. Politics can drive misalignments determinant of the extent to which the ideas when the vested interests of different stake- that underpin policies are translated into real- holders collide. Misalignment can occur along ity for teachers and students in classrooms. every step of the policy process, from defining Education requires getting millions of details goals to designing and implementing policies right in hundreds of thousands of schools. to evaluating their effectiveness. Together and Students who lack desks or textbooks, or separately, misalignment threatens to under- teachers whose training is unrelated to the mine the efforts of education systems to pro- demands of delivering the curriculum, cannot duce learning.1 81 82 GROWING SMARTER The experiences of Top Performing Systems shipbuilding and jobs related to petrochemi- shed some light on the influences that spurred cals and consumer electronics. their institutional alignment early on and helped them to achieve strong attainment and Advocating for education in national achievement results. Top Performing Systems spending in East Asia and Pacific began to lay the foun- dation for their current school systems in Top Performing Systems in East Asia and the 1960s, when resources were scarce and Pacific advocated for education in national institutional capacity was severely limited. spending. By ensuring that education was Subsequent efforts to upgrade quality and considered a critical input for economic devel- expand access through the 1990s reflected a opment, policy makers made public spending consistent government response to fundamen- on education a national investment in eco- tal changes in industrial structure, the global nomic growth. economy, and parental expectations. Progress Allocations for education among East was made possible by a series of deliberate Asia’s Top Performing Systems were his- policy choices that fostered alignment (Wong torically among the top categories in their 2017). These policies included setting targets national budgets. In Japan, 14.5 percent of all and demanding results, advocating for educa- public spending in 1955 went toward educa- tion in national spending, and providing the tion and stayed at that level for much of the impetus for cross-sectoral alignment. next 30 years. In Korea, education received 14.3 percent of the budget in 1963; the fig- ure grew to 16.1 percent in 1984 and 20.4 Setting targets and demanding results percent in 2000. In Singapore, on average, Top Performing Systems used available 23.3 percent of the national budget went to resources to achieve realistic targets. Universal education between 1959 and 1967. It stayed basic education was often a first target. at roughly this level in subsequent decades. Achieving it was no small feat given wide- spread illiteracy. In 1950, nearly 40 percent Providing the impetus for cross-sectoral of the population of Hong Kong SAR, China alignment had no formal schooling. The figure was 61 percent in Singapore; 46 percent in Taiwan, National leaders in Top Performing Systems China; and 28 percent in the Republic of saw that the full potential of education—or Korea. Under colonial rule, education was any other sectoral policy—could be under- mainly for elites. Only in Japan, which began cut if allowed to function within silos. They its education modernization in the late 19th therefore persistently and proactively under- century, did more than 95 percent of the pop- scored the need to put knowledge and skills ulation go to school in 1950. to work to power economic growth. They By the 1960s, with universal basic educa- put education on a par with physical infra- tion achieved, national leaders recognized that structure, the legal system, and banking and higher productivity and skills were needed to financial institutions, which are perceived as raise wages and sustain economic competi- essential to attracting investors and support tiveness. Consequently, access to secondary industry. education began to expand in the 1970s. National leaders pushed for cross-sectoral Singapore built many new public high schools. policy alignment by bringing into government Other economies allowed private institutions people who shared their vision of human to help to meet demand. Significant upgrades capital–driven economic growth. They used to curricula and teacher quality began to pre- the authority and power of their offices to pare students for the higher skills demanded build support and processes for cross-sectoral by employers as industrial development policy alignment. Education and the quality shifted to higher value-added work, including of the workforce have been constant themes INSTITUTIONAL ALIGNMENT FOR POLIC Y IMPLEMENTATION 83 in National Day speeches of the prime minis- promote success. Near unanimity exists about ter of Singapore since the first celebration in the need to change or improve organizational 1966. In Korea, numerous presidential com- cultures along with or as a means of improv- missions on education have advised presi- ing organizational capacity. Elmore (2004, 11) dents and their cabinets on major education finds that “prevailing cultures of administra- and industry policy reforms. In Vietnam and tion and teaching in schools … do not change Shanghai, China, leadership has been vis- by mandate: they change by the specific dis- ible and proactive in promoting alignment placement of existing norms, processes, and between education and other sectors. Their structures by others … modelling the new val- emphasis on human capital–driven growth ues and behaviors [to] the existing ones.” provides a unifying theme and impetus for Other analysts find that successful reform policy alignment across education, industry, of education systems relies on centering labor, and other relevant departments. reforms on clear targets and results, ensuring that the goals are related directly to improv- ing teachers’ professional practices in order to Systemic reform of education for improve student learning, and focusing on a improved learning reasonable number of achievable goals (Fulan Far too often, policies and investments in 2009; Levin and Marcus 2010). education are not well aligned with national Both Top Performing and Above-Average economic development needs, as noted in the Performing Systems have had policy experi- World Development Report 2018 (World ences that are broadly consistent with the Bank 2018). This lack of alignment has the assessment of the best approaches to systemic potential to undermine system reform of edu- reform. Explicit goals were set, and policies cation for improved learning. Overall govern- were largely sequenced, with emphasis first ment capacity may affect actions that affect placed on getting all children into school student learning, but capacity in education through the lower-secondary level and ensur- systems is likely to have a greater impact. ing the basic adequacy of infrastructure. The literature on promoting systemic Curricula were initially more centrally con- change in education systems highlights inher- trolled; autonomy was ceded gradually and ent obstacles and difficulties. Four obstacles selectively, often only after benchmarks were are ubiquitous: achieved. Top Performing Systems began address- • Political pressure for immediate or short- ing educational challenges amid capacity and term results resource scarcity during post–World War II • Lack of implementation continuity or reconstruction in conditions similar to the abrupt changes of direction ones many low-income countries faced or • Overemphasis on accountability of teach- face decades later. In 1950, the share of the ers and use of punitive measures to force population with no schooling was 34 percent desired behavioral change in Korea; 43 percent in Hong Kong SAR, • Inability or unwillingness to diagnose per- China; 54 percent in Taiwan, China; formance accurately and clearly (Harris and 68 percent in Singapore (figure 3.1). 2011). Expansion of schooling began from low No country is immune to political pressure levels, with the emphasis on quantity, not for fast results, but several high performers in quality. Hong Kong SAR, China; Korea; and East Asia and Pacific inclined toward incre- Singapore relied heavily on exams and testing mentalism and continuity in reforms, through to allocate access to education (Wong 2017). the use of specific, reasonable, and periodi- Despite scarce resources, enormous effort cally revised targets. was made to bring schools to every com- Another important strand of the litera- munity and ensure ease of attendance. In ture emphasizes actions and orientations that Hong Kong SAR, China and in Singapore, 84 GROWING SMARTER FIGURE 3.1 In economies with Top Performing Systems, the for learning—were in place in all schools, share of the population with no schooling plummeted between through implementation of the Fundamental 1950 and 2010 School Quality Audit program. 80 Aligning political support for investment in education with % of population age 25+ with no schooling 70 jobs and social mobility 60 High performers in East Asia and Pacific 50 had great success in building industries that created employment at scale to accel- 40 erate the structural transformation of their 30 economies. They were also widely recog- nized for having few social safety nets. The 20 ability to secure employment in the modern industrial sector served as both a ladder for 10 social mobility and a cushion against the 0 lack of government-provided social safety 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 nets. Policies offered a vision in which par- Singapore Taiwan, Hong Kong SAR, Korea, Japan ents saw their children securing jobs after China China Rep. completing increasing amounts of school. Initial successes in employment of graduates Sources: Barro and Lee 2013. in these newly created industries reinforced both the demand for schooling and the value that parents placed on their children’s for example, new schools were built in newly achievement. constructed public housing estates so that A recurrent theme in Top Performing children could safely walk to school. This and Above-Average Performing Systems has focus on achieving basic universal primary been the tension between government and enrollment explains why the state allowed civil society to ensure access to education as private providers of secondary education a means to more secure and better income to emerge and even dominate the system and the calls from civil society for greater (see spotlight 2). access to initially restricted opportunities for By and large, only after 2000 did these improved job opportunities, which depended Top Performing Systems adopt policies that on education. This dynamic helped to pro- emphasized academic and behavioral skills pel exam-based allocation of opportunity beyond the basics, such as the ability to solve to a distorted state in several East Asia and problems, innovate in complex systems, and Pacific economies. Although many years of collaborate in diverse environments. Progress reforms have broadened opportunities at the toward these goals has been possible because very top of the educational ladder, tension of strong, localized institutional capacity to between widespread access to basic educa- ensure that critical elements are aligned in tion and restricted access to higher levels of their common aim of crowding in learning. education remains part of the political econ- Starting in the 1980s and 1990s, respec- omy of education reform in the region. tively, China and Vietnam pursued poli- The call for greater opportunity at the top cies that sought to imitate the success of came only after success at basic levels was Japan, Korea, and Singapore. In Vietnam, firmly in place. Ensuring success was a func- initial efforts focused on ensuring that tion of simplifying, focusing, and sequencing appropriate physical infrastructure and reforms so that literacy and numeracy skills teaching capacity—the basic conditions for all were secured before new challenges INSTITUTIONAL ALIGNMENT FOR POLIC Y IMPLEMENTATION 85 were attempted. Institutional capacity was the audit was conducted annually between built in the drive for these goals, rather than 2004 and 2010 and cost roughly US$4–US$5 as a precursor to them. per school campus per year. Consistent use for decision making allowed the instrument to be refined and correlated with learning Sound administrative systems outcomes. The data in figure 3.2 underscore start by ensuring that basic the importance of ensuring not only that basic conditions for learning are in inputs are available, but also that they are the place right inputs and that they lead to increased A robust body of literature supports the learning. importance of basic physical inputs on stu- An assessment system is effective only if it dent achievement in school. Glewwe and oth- is updated regularly. In recent years, the use ers (2011) summarize the results of more than of the FSQL has stagnated; the list of basic 40 high-quality research studies that examine inputs is in need of updating. If Vietnam’s the relationship between school inputs and strong performance on the Programme for educational outcomes in low- and middle- International Student Assessment (PISA) in income countries. Their review finds strong recent years is any indication of the usefulness evidence that the availability of basic inputs of having a system such as the FSQL, continu- such as desks, tables, and chairs helps to ing to use the FSQL to set the basic conditions raise student achievement. They find that all for learning is likely to be fruitful. research estimates show a positive relation- Administrative systems in Shanghai also ship, with almost half of them statistically seek to go beyond ensuring that inputs have significant. Other statistically positive inputs reached schools, to measuring their use. were blackboards, libraries, and school infra- The provincial government has established structure (such as walls, ceilings, and roofs). a mechanism for monitoring resource dis- Vietnam’s experience with developing tribution in schools, tracking how funds Fundamental School Quality Level (FSQL) are spent, and linking inputs to use in class- standards illustrates how a centralized admin- rooms. Municipal and district offices carry istrative system can be used to ensure that out annual external audits to verify the fidu- schools have the basic inputs needed for learn- ciary aspects of spending as well as the use of ing. The FSQL began as an effort to collect resources by schools during the year (Liang, data on classroom infrastructure and sanita- Kidwai, and Zhang 2016). tion, furniture, textbooks, learning materials, The Philippines is making similar efforts, teacher training, and community engagement although these efforts do not yet seem to have (box 3.1). Over time, it was used to create an catalyzed learning gains. Like Shanghai and input index, which is used not only to moni- Vietnam, the Philippines has a centralized sys- tor progress but also to assess whether inputs tem for monitoring and managing resources. lead to learning. The Department of Education identifies The first step in developing FSQL stan- school infrastructure needs annually using dards involved the launch, in 2004, of an the Basic Education Information System. It annual school census that became known as uses the data to develop a list of school-level the District FSQL Audit. The 10-page data projects to be carried out using the year’s instrument was distributed to all primary infrastructure budget appropriations. The school campuses nationwide. It enabled the Department of Education also compiles a pri- collection of detailed information from each ority list of schools, identifying first schools campus and provided for both project-specific with high student-classroom ratios and a lack and national monitoring of FSQL status. The of water and sanitation facilities. In the sec- audit produced a detailed national database of ond step, through site visits, the department information on some 40,000 primary schools. verifies the condition of facilities and the Compiled and released within six months, feasibility of the work needed and formally 86 GROWING SMARTER BOX 3.1 The original elements of Vietnam’s Fundamental School Quality Level standards, as designed in 2001 Physical infrastructure Educational activities and quality • Schools and campuses have solid classroom • One set of teaching aids or instructional construction (walls, floors, and roofs), and materials per grade is available to each school adequate natural lighting. and satellite campuses. • Schools and campuses have one blackboard per • One set of teacher supplies (ruler, scissors, classroom, sufficient tables and benches for chalk, paper, pen) is available to each teacher students, one teacher desk-chair per classroom, in all schools and satellite campuses. and sufficient durable and transportable storage • One full set of textbooks, a teaching manual, boxes or lockers for instructional materials. and other guides as required per grade taught is available to each teacher. Teaching staff • One set of supplementary reading materials • Teachers are minimally qualified. appropriate to all grades taught is available • Teachers receive annual professional training to each school and satellite campuses. on relevant classroom management and • All students have math and Vietnamese pedagogical topics (such as crafting teaching textbooks. aids, multigrade teaching, remediation • All students are equipped with a sufficient support, Vietnamese language instruction, (minimum) number of notebooks and pencils. and inclusive education). • All teachers in ethnic minority areas are School organization and management trained in teaching Vietnamese to children whose native language is not Vietnamese. • Schools ensure that each satellite campus • Special Vietnamese language materials are offers the same quality of instruction, services, available in all schools and campuses with and resources as the main school site. ethnic minority populations. • Schools and campuses offer grades based on a full 175-week curriculum. Expected outcomes • Head teachers are trained to manage and • All school-age children are enrolled in school. support satellite campuses. • Ninety-six percent of 14-year-olds complete Education socialization primary education. • Schools and campuses have parent committees. • Gender equity is achieved in primary school • Parent committees at all schools and satellite enrollment. campuses are trained in school and student support strategies. Source: Attfield and Vu 2012. prepares a list of potential projects aligned as unfit for teaching and learning). The study, with the budget available. designed by the World Bank in consulta- The Philippines Public Education tion with the Philippines Department of Expenditure Tracking and Quantitative Education, aimed to track, manage, and Service Delivery Study (PETS-QSDS) was evaluate the presence, use, and efficacy of commissioned after information from the resources in the education sector. It has Basic Education Information System revealed tracked more than 80 percent of the national the need for investment in school facilities government education budget. and infrastructure (one in seven elementary The initial picture revealed by PETS-QSDS and high school classrooms was characterized was discouraging. Widespread deficiencies INSTITUTIONAL ALIGNMENT FOR POLIC Y IMPLEMENTATION 87 FIGURE 3.2 Basic inputs can help to enhance learning outcomes: an example from Vietnam a. Mathematics b. Vietnamese 90 90 R² = 0.3406 R² = 0.2742 80 80 % of students scoring “excellent” in tests % of students scoring “excellent” in tests 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 45 55 65 75 85 95 45 55 65 75 85 95 FSQL input index score FSQL input index score Source: Attfield and Vu 2012. Note: FSQL = Fundamental School Quality Level. FSQL standards were launched in Vietnam in an effort to collect data on classroom infrastructure and sanitation, furniture, textbooks, learning materials, teacher training, and community engagement. were found in the nearly 7,000 classes about it, and even fewer engage in active and classrooms that were observed. They reporting through the system. included unclean classrooms, with poor ventilation; inadequate sound insulation; Providing clear guidance to lack of electricity; and inadequate seating teachers through national (Al-Samarrai 2016). The data were used for curricula and textbooks school renovation and infrastructure budget Curricula appropriations. The Philippines also has a decentralized, Initially, all of the systems in the region that community-based system to track resource are now Top Performing Systems had a single inputs in schools. Check My School is a par- national curriculum. Opting for a unified, ticipatory monitoring initiative that mobilizes compulsory national curriculum served mul- community volunteers to track the provision tiple purposes. of resources in public schools (Shkabatur First, when qualified teachers were in 2012). Volunteers (usually parents with chil- short supply, it was safer to rely on central- dren in the public school system) visit pub- ized experts to develop curricula (Chung and lic schools, verifying the data collected and Ngan 2002; Goh and Gopinathan 2008; Lee released by the Department of Education by 2006; Sweeting 1990). reviewing the existence and quality of school Second, a prescribed curriculum promoted resource inputs and infrastructure such as a relatively uniform education experience for toilets, textbooks, and classrooms and send- students regardless of their school, location, ing pictures and text messages documenting and socioeconomic background (Ashton and what they see. The aim of Check My School others 2002; Chou and Ho 2007; Kim 2002; is to have systems-level impact on account- Lai 2009; Lee 2006). ability and transparency. Its effectiveness is Third, using the same mandatory cur- limited, however, because few schools know riculum throughout the system was seen as 88 GROWING SMARTER giving every student the same fair chance to curriculum updates occur every 10 years. do well on the government exams for selec- In Singapore, the process is continual. In tion to higher levels of education. Some Korea, major revisions typically follow high- observers have credited this last concern to a level government reviews of national human belief in meritocracy; the focus on hard work resources policy and economic development (that is, effort or hard work versus innate strategy needs; and minor amendments are intelligence or natural talent) is strongly held continual (Chang Chien, Lin, and Chen 2013; across these societies (Chou and Ho 2007; Lai 2009; OECD 2010; Seth 2002). Goh and Gopinathan 2008; OECD 2010; Having unified curricula was seen as part Tu 2007). of the goal of simplifying the educational Fourth, and perhaps most important, a endeavor while capacity was scarce by focus- single prescribed curriculum provided guid- ing on a smaller set of goals. ance for curricular development for teacher Strong central leadership and a unified sys- education and training. Curricula were tem are not without critics, some of whom focused and set clear and unambiguous decry a one-size-fits-all approach to prob- learning goals. Teachers were expected to lems and needs. Top-down mandates are not possess the same basic knowledge and skills always sensitive to local conditions and pref- to deliver the prescribed curriculum, in gov- erences and may not allow bottom-up innova- ernment or private schools and regardless tion. Pent-up frustration can undercut morale of location (Chung and Ngan 2002; Fujita among people working in the system, and 2007; Goh and Gopinathan 2008). In Japan, public grievances can undermine trust in it. having a single prescribed curriculum fit especially well into its practice of regularly Textbooks rotating teachers among schools across the country to “level up” the performance of Just as governments exercised firm control of teachers, principals, and students (Fujita the curriculum and set qualification standards 2007; OECD 2010). for teachers, many mandated a single set of For similar reasons, having a single set of textbooks. Singapore follows this practice approved or government-issued textbooks to this day. Advocates of a single set of text- was also the norm in these systems. It had a books see Singapore’s approach as beneficial similar positive effect on teacher education to faithful implementation of the single pre- and training. Teacher education and train- scribed curriculum across the system. They ing programs are expected to adhere to the assert that this approach better fulfills the prescribed curriculum or to follow alterna- promise of equity and quality in education tive guidance from the central government. when teacher competency is low and capacity This decision is partially related to the need to to train them is limited. prepare students for standardized exams that As teacher capacity grows and systems these governments use to select students for respond to new demands on student perfor- secondary and postsecondary school place- mance (such as the need for creativity and ments, therefore requiring teachers to devote problem-solving skills), many systems now much attention to content in the national cur- allow teachers and schools to choose from riculum and its revisions. several government-approved textbooks pro- Review and revision of the curriculum duced by private publishers and encourage occur under central government oversight, them to develop their own syllabi to pro- driven by the push for constant improvement mote student learning. In Hong Kong SAR, and excellence. International trends and mod- China, textbook development, production, els of success are closely studied. Education and dissemination have always been left to experts from universities are brought into the private sector, although private publish- special commissions, and input from teach- ers are careful to reference the government- ers and principals is solicited. In Japan, mandated curriculum. INSTITUTIONAL ALIGNMENT FOR POLIC Y IMPLEMENTATION 89 Sequenced reforms allowed pathways to secondary school were adopted more complex and ambitious in 2014, and more assistance was given to learning goals to be achieved disadvantaged students and students from low-income households (Chen and Fan After universal basic education was achieved, 2014; Clark 2010; National Center on the emphasis of policy reform shifted to Education and the Economy, n.d.). diversifying goals as a means of improving In Japan, reforms introduced in the new quality in schools, classrooms, and facili- Fundamental Education Law in 2006 (the ties (although schools remain fairly basic by first revision in 60 years) devolved some Western standards, and class sizes remain functions to lower levels of government and large, at 30-plus students) (Wong 2017). allowed schools to exercise greater discre- Sports, music, and the arts— previously tion over their budgets and personnel (OECD deemed largely nonessential—found space 2010). A school self-evaluation scheme that in curricula and schools. Educational goals began in 2002 allows schools to set their moved beyond acquiring foundational skills own goals and criteria for evaluation within to seek to impart the full range of skills a framework defined by the central govern- needed for the knowledge economy. ment. Goals can include academic objectives The school experience should fundamen- (such as adding learning activities outside the tally prepare children for the challenges they curriculum with input and support from par- will face as adults in the world of work. The ents, local community, and businesses) as well more straightforward and basic learning goals as nonacademic goals (such as reducing inci- that might allow a person to transition from dents of bullying and absenteeism—two seri- farm to factory would likely not be sufficient ous problems in Japan). to allow her children to move from working In Korea, decentralization and deregula- on an assembly line to working for Facebook. tion of the education system began in the mid- Mastery of factual knowledge and declarative 1990s. The pace picked up when the Asian content needs to be increasingly coupled with financial crisis of 1997–98 forced the central mastery of noncognitive skills and abilities government to downsize. More decision- such as creativity, the ability to communicate, making authority was devolved to provincial, resilience, and grit. Top Performing Systems municipal, and local education offices. Central moved to ensure that these 21st century skills authorities still exercise enormous influence are taught in school. They did so largely because they provide the bulk of education after securing the basics—and often only in spending (75–80 percent) (Ferreras, Kessel, response to strong demand from civil society. and Kim 2015; Kim 2007). The insistence In Taiwan, China, the policy of a single on national standards may be easing, but the national curriculum ended when massive demand for performance remains. demonstrations by middle-class families in Korea continues to revise its vision of what 1994 called for fundamental reforms. Parents its school system should aim to produce. King were concerned about the harmful effects of and Rogers (2014) document how skills val- school stress and the ability of the education ued by the modern innovation economy, such system to prepare their children for work in as creativity and emotional intelligence, pose the 21st century. A “national educational next-generation challenges for education framework” setting out national goals and policy makers in Korea. The foundational ele- priorities in education replaced the national ments of success are still an important part of curriculum. Schools can develop their own the equation—but they are no longer suffi- student-centered curriculum based on this cient for producing students with the cutting- framework, and teachers are encouraged to edge skills and abilities needed to compete in develop innovative pedagogical techniques the 21st century. to connect better with students and condi- High-performing economies continue to tions in local communities. New exam-free evolve their institutions and seek new ways FIGURE 3.3 Timeline of selected educational developments in East Asia and Pacific 1950 1960 1970 1980 Goal or policy 1954: 1957: 1961: Singapore 1965: Hong 1971: Hong 1974: Korea, 1978: Hong 1980: Thailand Korea, Rep. Singapore Plan to provide Kong SAR, China Kong SAR, Rep. Kong SAR, National Six-year Double- universal free Blueprint for China HSEP China Primary compulsory session primary universalization Free and introduced to Free and Education Act education plan schooling education began of primary compulsory equalize compulsory announced, begins begins implementation education six-year school-level nine-year stipulating published education inputs education that all began began villages should have a school Outcome 1959: Korea, Rep. 1965: Singapore 1970: Singapore 1972: Hong Kong 1980: Korea, Rep. Universal primary Universal primary Universal SAR, China Universal achieved achieved lower-secondary Universal primary lower-secondary achieved achieved achieved Institutions 1955: Singapore 1960s–early 1970s: Hong Kong SAR, China Ministry of Education Advisory and school inspection services greatly expanded, Curriculum Development Committee formed, established and regionalized administrative system introduced to secure closer liaison with schools Spending 1952: Singapore 1955: Japan 1958: Korea, Rep. 1960: Korea, Rep. 1963: Korea, Rep. 1971: Korea, Rep. 1974: Indonesia 31.7% of 14.5% of Education Tax 81% of education 14.3% of total Local Education Grant Presidential Instruction government government enacted, tripling budget went to government enacted, earmarking program for school expenditure expenditure education primary years budget went to portion of internal tax construction set aside went to went to spending in one education revenue for teacher portions of oil revenue for education education year salaries and recurrent building new primary spending schools 1959–67: Singapore An average of 23.3% of government expenditures went to education 1973–78: Vietnam Large-scale school construction program 1950s: Hong Kong SAR, China began to boost primary enrollment; 61,000 Extensive school-rebuilding program began, adding 45,000 school schools built in five years places a year at its peak 1950 1960 1970 1980 Teachers 1949: Japan 1950: Singapore 1959: Singapore 1960s: Hong Kong SAR, China 1980: Cambodia All normal TTC established as Unit at TTC set Programs at colleges of education extended from one year to two years; Short-term schools the first full-time up to research special education training began course for transitioned into training college pedagogical preprimary, teacher colleges for teachers; it best practices 1960s: Korea, Rep. primary, and and universities had 1,300 All normal schools transitioned into teacher colleges and universities; secondary students temporary teacher training centers formed to provide preservice training teachers created Readiness to learn 1949: Japan 1969: Korea, Rep. 1970: Vietnam 1971: Japan 1978: The Philippines 1979: Thailand Kindergarten Official Government Plan to promote Law requiring First long-term included in School kindergarten circular announces kindergarten for establishment of plan for child Education Law curriculum establishment of four- to five-year- day-care center in all and youth established crèches and olds announced barangays announced development nursery schools established 1975: Lao PDR Preschools introduced Assessments 1964: FIMS 1970–72: FISS conducted; Japan and conducted; Japan Thailand participate participates Top Performing Systems Above-Average Performing Systems Below-Average Performing Systems Emerging Systems Multiple economies figure continues next page FIGURE 3.3 Timeline of selected educational developments in East Asia and Pacific (continued) 1980 1990 2000 2010 2015 Goal or policy 1981: Vietnam 1984: Indonesia 1990: China 1991: Vietnam 1999: Thailand 2005: Papua 2009: Cambodia 2013: The Education reform Six-year free and “Two Basics” Six-year free and compulsory National New Guinea Five-year Education Philippines defined objectives compulsory goal announced: education established Education Act National Strategic Plan Basic Education as universalization education universal announces Education Plan announced; target Act extended and quality established primary 1994: Indonesia nine-year free announces of 80% completion basic education improvement of education and Nine-year free and compulsory and compulsory prime objective rate in primary cycle from 10 to education elimination of education established education of universal education 13 years illiteracy basic education Outcome 1985: Shanghai, China 2000: Vietnam First city in China to achieve universal Universal primary education achieved primary and secondary education 2000: China “Two Basics” goal achieved Institutions 2002: Japan 2004: Vietnam 2009: Malaysia School FSQL census PEMANDU formed to bring self-evaluation launched together stakeholders in scheme began designing policy Spending 1997: 2000: Korea, 2002: 2005: 2009: 2010: Vietnam Vietnam Rep. Indonesia Vietnam Indonesia Government expenditure Government Government Government Government 21.9% of on education stood at 21% expenditure expenditure mandated expenditure budget spent on education on education 20% of on education on education 2010: Lao PDR stood at stood at budget to be stood at Government expenditure 14% 20.4% spent on 18.6% on education stood at 7.3% education 1980 1990 2000 2010 2015 Teachers 1982: Cambodia 1990: Cambodia 2005: Indonesia 2007: China 2012: Myanmar A year-long preservice Preservice Teacher Law defined Six nationally affiliated Begins training began at the Royal training for competencies required for teachers colleges offer free comprehensive University of Phnom Penh teachers teachers, laws concerning tuition and living stipends teacher reform for lower- and upper- extended to teacher training and for teachers who agree to secondary teachers two-year course accreditation, and role of stay in the teaching various agencies in profession for 10 years supporting teachers Readiness to learn 1981: China 1986: 1987: Vietnam 1991: Vietnam 1995: China 1999: Vietnam 2000: The Philippines 2003: Indonesia 2005: Mongolia 2007: Thailand 2012: China Guidelines Singapore Services for Department Law on Law regarding Early Childhood Care ECED included in National Policy Long-Term Plan and Learning and for Low-fee children 0–3 of Preschool maternal and preschool and Education Law is National on Integrated Strategy for Early Development kindergarten kindergar- and 3–5 years Education unit infant health education issued Education Law Early Childhood Childhood Care and Guidelines for issued tens began merged under created under care enacted Development Development begins 3- to 6-year-olds to be offered the Ministry of new national established 2000: Cambodia adopted established Education administration Policy on ECED 2009: Indonesia articulated National Standards for ECED issued Assessments 1980–82: SIMS 1983–84: SISS 1990–91: SIRS 1995: TIMSS first 2000: PISA first 2009: Malaysia; 2012: Vietnam 2015: Beijing, conducted. Hong Kong conducted. Hong Kong conducted. Hong Kong conducted. Japan; conducted. Hong Kong Shanghai, China; and First year Shanghai, Jiangsu, SAR, China; Japan; and SAR, China; Japan; SAR, China; Indonesia; Korea, Rep.; the SAR, China; Indonesia; Singapore participating and Guangdong Thailand participate Korea, Rep.; Papua New the Philippines; Philippines; Singapore; Japan; Korea, Rep.; and First year participating in PISA represent China Guinea; Singapore; and Singapore; and and Thailand Thailand participate in PISA for the first time in participate 2012: PILNA first PISA 2015 Thailand participate Thailand participate conducted Top Performing Systems Above-Average Performing Systems Below-Average Performing Systems Emerging Systems Multiple economies Sources: Chang and others 2014; EDB 1981; Goh and Gopinathan 2008; Hanushek and Woessmann 2015; Kim 2000; Tandon and Fukao 2015; UNESCO 2015; Wong 2017. Note: ECED = early childhood education and development; FIMS = First International Mathematics Study; FISS = First International Science Study; FSQL = Fundamental School Quality Level; HSEP = High School Equalization Policy; PEMANDU = Performance Management and Delivery Unit; PILNA = Pacific Islands Literacy and Numeracy Assessment; PISA = Programme for International Student Assessment; SIMS = Second International Mathematics Study; SIRS = Second International Reading Study; SISS = Second International Science Study; TIMSS = Trends in International Mathematics and Science Study; TTC = Teachers Training College. 92 GROWING SMARTER to promote alignment between what is taught own learning through an increasing array of and the skills required in a changing economy. digital information sources. In Singapore, for example, in 2016, govern- ment ministers convened the Committee on the Future Economy of Singapore, to develop Institutional alignment and a report to outline Singapore’s economic sequenced reforms helped high- strategies for the next decade (Committee on performing economies to reach the Future Economy, n.d.). Co-chaired by the critical milestones in expanding minister of education, the minister of com- access and improving quality munications and information education, and Figure 3.3 displays a timeline of selected chief executive officers and managing direc- educational developments across the region. tors of the country’s largest private enter- Although each economy’s actions were deter- prises, the Subcommittee on Future Jobs and mined principally by domestic concerns, Skills focused on how to ensure that the skills reforms consistently sought a focused and of the workforce are updated in accordance sequenced set of goals. Universal basic educa- with future jobs and technologies. Education tion and improved attainment were pursued systems that promote creativity, love of learn- first by the Top Performing Systems and then ing, the ability to work in teams, and commu- by others. Concern for having all students nication skills are fundamental to this vision, acquire foundational skills in literacy and as are resilience and the ability to guide one’s numeracy followed. Part of the initial focus BOX 3.2 Elements of policies and practices that promote learning The success of some education systems in East Asia • Make teachers’ jobs easier by providing clear and Pacific shows that students learn most when learning goals and uncluttered texts. efforts focus on five policy domains and align 15 • Keep experienced teachers in the classroom elements. These domains and elements are as follows: and leading as peers and researchers. • Center teacher training on classroom practice Align institutions to ensure basic conditions for and the ability to teach the curriculum. learning: • Ensure that the basic conditions for learning Ensure that children are ready to learn in school: are in place in all schools. • Focus on physical and cognitive development from birth. Concentrate effective, equity-minded public spending • Assess and improve the quality of early on basic education: childhood education and development • Spend effectively. services. • Concentrate public spending on basic • Coordinate across actors to deliver needed education. services. • Channel resources to schools and districts that are falling behind. Assess students to diagnose issues and inform instruction: Select and support teachers throughout their careers • Benchmark learning through participation in to allow them to focus on the classroom: international large-scale assessments. • Raise the selectiveness of who becomes a • Diagnose cohort progress at every educational teacher. sublevel. • Support new teachers by observing classroom • Inform instruction with data from formative practices and providing feedback. classroom assessment. INSTITUTIONAL ALIGNMENT FOR POLIC Y IMPLEMENTATION 93 was an ambitious agenda for school construc- Education Expenditure Tracking and tion and capacity building to ensure that the Quantitative Service Delivery Study . basics were in place. Washington, DC: World Bank. Korea and Singapore established goals Ashton, D., F. Green, J. Sung, and D. James. 2002. “The Evolution of Education and Training for compulsory education in the 1950s and Strategies in Singapore, Taiwan, and S. Korea: 1960s, respectively. In both cases, they took A Development Model of Skill Formation.” no more than five years to achieve univer- Journal of Education and Work 15 (1): 5–30. sal primary enrollment. For Korea, evidence Attfield, I., and B. T. Vu. 2012. “A Rising Tide of the progressive universalism approach is of Primary School Standards—The Role apparent in the sequential expansion of first of Data Systems in Improving Equitable primary and then junior-secondary education. Access for All to Quality Education in China and Vietnam launched similar goals Vietnam.” International Journal of Educational almost a decade apart, meeting them in 2000. Development 33 (1): 74–87. Public spending on basic education testified Barro, R., and J.-W. Lee. 2013. “A New Data Set to the importance that these economies placed of Educational Attainment in the World, 1950– 2010.” Journal of Development Economics on maintaining a degree of quality control 104 (C): 184–98. over education performance, as did the use of Chang, M. C., S. Shaeffer, S. Al-Samarrai, A. B. unified national curricula and textbooks. Ragatz, J. de Ree, and R. Stevenson. 2014. Success in reaching initial goals created Teacher Reform in Indonesia: The Role of tensions and led to demands for more edu- Politics and Evidence in Policy Making. cational opportunity, higher quality, more Directions in Development. Washington, DC: diversified educational experience, and more World Bank. complex goals. Japan embraced support for Chang Chien, C.-K., L.-C. Lin, and C.-F. Chen. readiness to learn very early, but most other 2013. “The Main Features and Key Challenges economies began this expansion of ser- of the Education System in Taiwan.” Higher vice delivery only in the late 1980s or later. Education Studies 3 (6): 1–14. Chen, H., and H.-H. Fan. 2014. Education in Assessment remained principally a device for Taiwan: Vision and Goals of the 12-Year allocating scarce opportunities for advanced Curriculum . Washington, DC: Brookings education. Although exams still serve this Institution Press. function, assessment policies now inform Chou, C. P., and A.-H. Ho. 2007. “Schooling in instruction in basic education. Taiwan.” In Going to School in East Asia, Subsequent chapters examine the details edited by G. Postinglione, 344–75. New York: of policies and practices with regard to pub- Greenwood. lic spending, teachers, readiness to learn, and Chung, C., and M.-Y. Ngan. 2002. “From ‘Rooftop’ assessment (box 3.2) and provide information to ‘Millennium’: The Development of Primary related to the milestones included in figure 3.3. Schools in Hong Kong since 1945.” Hong Kong SAR, China: Hong Kong Institute of Education. Clark, N. 2010. “Taiwan.” World Education Note News and Reviews, May 1. http://wenr.wes .org/2010/05/wenr-may-2010-feature. 1. See, for instance, the examples highlighted in Committee on the Future Economy. n.d. “Future chapter 10 of the World Development Report Economy: The Engagement Process.” 2018 (World Bank 2018). Government of Singapore. https://www.gov .sg/microsites/future-economy/the-cfe-report Referencesa /engagement-process. EDB (Education Bureau Hong Kong). 1981. Al-Samarrai, S. 2016. Assessing Basic Education “Overall Review of the Hong Kong Education Service Delivery in the Philippines: Public System.” EDB, Hong Kong, June. Elmore, R. 2004. School Reform from the Inside a. Titles of publications that include "Hong Kong," "Korea"/"S. Korea"/"South Korea," and "Taiwan" refer to Hong Kong SAR, Out. Cambridge, MA: Harvard Education China; the Republic of Korea; and Taiwan, China, respectively. Press. 94 GROWING SMARTER Ferreras, A., C. Kessel, and M.-H. Kim. Creative and Innovative Economy.” Background 2015. Mathematics Curriculum, Teacher paper for discussion at the symposium on Professionalism, and Supporting Policies “Achieving HOPE: Happiness of People through in Korea and the United States: Summary Education; Innovation in Korean Education for a of a Workshop. Washington, DC: National Creative Economy,” World Bank Group, Seoul, Academies Press. November. http://documents.worldbank.org Fujita, H. 2007. “The Qualifications of the /curated/en/319691468340199371/Intelligence Teaching Force in Japan.” In A Comparative -personality-and-creativity-unleashing-the Study of Teacher Preparation and Qualification -power-of-intelligence-and-personality-traits-to in Six Nations, edited by R. E. Ingersoll, 41–54. -build-a-creative-and-innovative-economy. Philadelphia: Consortium for Policy Research Korea Times. 2016. “One in Three Unemployed in Education. People Have Bachelor’s Degree.” October Fulan, M. 2009. The Challenge of Change: School 17. http://www.koreatimes.co.kr /www/news/ Improvement Now! Thousand Oaks, CA: nation/2016/10/116_216218.html. Corwin. Lai, K. C. 2009. “Qualifications of the Teaching Glewwe, P., E. Hanushek, S. Humpage, and Force in Hong Kong Special Administrative R. Ravina. 2011. “School Resources and Region, China.” In A Comparative Study of Educational Outcomes in Developing Teacher Preparation and Qualification in Six Countries: A Review of the Literature from Nations , edited by R. E. Ingersoll, 29–40. 1990 to 2010.” NBER Working Paper 17554, Philadelphia: Consortium for Policy Research National Bureau of Economic Research, in Education. Cambridge, MA. Lee, C. J. 2006. “The Development of Education Goh, C. B., and S. Gopinathan. 2008. “The in Korea: Past Achievement and Current Development of Education in Singapore since Challenges.” Background paper for the East 1965.” In Toward a Better Future: Education Asia Tour for Senior African Education Policy and Training for Economic Development Makers, World Bank, Washington, DC. in Singapore since 1965, edited by S. K. Lee, Levin, T., and A. Marcus. 2010. “How the C. B. Goh, B. Fredriksen, and J. P. Tan, 12–38. Structure and Focus of Teachers’ Collaborative Washington, DC: World Bank. Activities Facilitate and Constrain Teacher Hanushek, E., and L. Woessmann. 2015. The Learning.” Teaching and Teacher Education Knowledge Capital of Nations: Education and 26 (3, April): 389–98. the Economics of Growth. Cambridge, MA: Liang, X., H. Kidwai, and M. Zhang. 2016. How MIT Press. Shanghai Does It: Insights and Lessons from Harris, A. 2011. “System Improvement through the Highest-Ranking Education System in the Collective Capacity Building.” Journal of World. Washington, DC: World Bank. Educational Administration 49 (6): 624–36. National Center on Education and the Economy. n.d. Kim, E.-G. 2007. “The Quality and “Singapore: System and School Organization.” Qualifications of the Teaching Force in the National Center on Education and the Republic of Korea.” In A Comparative Study Economy, Washington, DC. http://ncee.org of Teacher Preparation and Qualification in /what-we-do/center-on-international-education Six Nations, edited by R. E. Ingersoll, 55–70. -benchmarking/top-performing-countries Philadelphia: Consortium for Policy Research /singapore-overview-2/. in Education. OECD (Organisation for Economic Co-operation Kim, G.-J. 2002. “Education Policies and Reforms and Development). 2010. “Japan: A Story of in South Korea.” In Secondary Education Sustained Excellence.” In Strong Performers in Africa: Strategies for Renewal, 29–40. and Successful Reformers in Education: Washington, DC: World Bank. Lessons from PISA for the United States, ch. 6. Kim, H. S. 2000. “Towards Achieving High Paris: OECD. Quality Pre-Service Teacher Training in Seth, M. J. 2002. Education Fever: Society, Politics, Korea.” Asia-Pacific Journal of Teacher and the Pursuit of Schooling in South Korea. Education and Development 3 (1): 55–77. Honolulu: University of Hawaii Press. King, E., and H. Rogers. 2014. “Intelligence, Shkabatur, Jennifer. 2012. “Check My School: Personality, and Creativity: Unleashing the Power A Case Study on Citizens’ Monitoring of the of Intelligence and Personality Traits to Build a Education Sector in the Philippines.” World Bank, INSTITUTIONAL ALIGNMENT FOR POLIC Y IMPLEMENTATION 95 Washington, DC. https://openknowledge for All (EFA): Korea National Review. Paris: .worldbank.org/handle/10986/23031. OECD. Sweeting, A. 1990. Education in Hong Kong Pre- Wong, A. 2017. “Insights from East Asia’s High 1841 to 1941: Fact and Opinion; Materials Performing Education Systems: Leadership, for a History of Education in Hong Kong. Pragmatism, and Continuous Improvement.” Pokfulam: Hong Kong University Press. Background paper for World Bank flagship Tandon, P., and T. Fukao. 2015. Educating the report, Growing Smarter: Learning and Next Generation: Improving Teacher Quality Equitable Development in East Asia and in Cambodia. Washington, DC: World Bank. Pacific, World Bank, Washington, DC. Tu, C.-S. 2007. “Taiwan’s Educational Reform World Bank. 2013. Spending More or Spending and the Future of Taiwan.” Paper presented at Better: Improving Education Financing in the London School of Economics and Political Indonesia. Public Expenditure Review 73050. Science, London, January 10. http://www.lse Washington, DC: World Bank. .ac.uk/website-archive/publicEvents/pdf/2007 ———. 2018. World Development Report 2018: 0110TaiwanEducationalReform_English.pdf. Learning to Realize Education’s Promise . UNESCO (United Nations Educational, Scientific, Washington, DC: World Bank. and Cultural Organization). 2015. Education SPOTLIGHT 2 What Lessons Can Be Drawn from Top Performing Systems’ Experience with Technical and Vocational Education and Training? Technical and vocational education and their populations to work and lift them and training (TVET) played an important role in the national economy out of poverty. They creating human capital for economic devel- needed low-skill labor that could work in opment in Japan, the Republic of Korea, and labor-intensive light manufacturing indus- Singapore. TVET programs were an impor- tries, such as textiles, clothing, light electron- tant component of education systems in ics, toys, and kitchenware, producing exports all three countries from the 1950s through for these markets. the 1990s (and they remain important in They concentrated their limited resources Singapore). The ministry of education, in and capacity on universal basic education. coordination with the ministry of labor or Large portions of state education funds went industry, oversaw these programs. to building schools, quickly adding teachers TVET produced workers for low-tech, to classrooms, and developing national cur- labor-intensive manufacturing industries that ricula and textbooks. These governments launched high-speed economic growth in the saw TVET programs, rather than academic 1950s and 1960s. TVET subsequently pro- secondary education, as the quickest way to duced workers for heavy industries, such as produce low-skill workers for the factories, shipping, steelmaking, and petrochemicals, construction sites, and shipyards that were which took off in the 1960s and 1970s, and hungry for labor. workers for technology-based manufactur- Demand for workers stayed high for more ing, such as electronics and integrated circuit than two decades of rapid growth. Average chips, in the 1970s and 1980s. unemployment was less than 2 percent in Japan in the 1960s and 1970s. In Korea, it was under 3 percent in the 1960s–80s. In TVET was central to national Singapore, it was 3.6 percent in 1978, after education policy and national a decade of sustained high growth. These low economic development strategy numbers are all the more remarkable given The economies of North America and the countries’ high fertility and rapid popula- Western Europe boomed after World War II. tion growth during these decades (see Benson Leaders in Japan, Korea, and Singapore saw and Zhu 2012; Husna 2018; Singapore in their success a golden opportunity to put Ministry of Trade and Industry 2011). 97 98 GROWING SMARTER In the 1950s through the 1970s, Singapore schools to open. The state refused to invest tracked at least one-fifth of all students into more in secondary academic education. As a TVET programs after primary school, and result, nearly all children in Korea enrolled in more enrolled in TVET programs when they public primary schools, and about half of all failed to pass rigorous exams to advance secondary students attended private institu- from middle to high school (Center on tions (Kim 2001; Sorensen 1994). International Education Benchmarking, Government preference for a combination n.d.; Goh and Gopinathan 2008). Education of primary academic education and TVET was a state monopoly; parents and students programs at the secondary level persisted had no choice but to pursue the path chosen through the 1990s (Center on International for them. Education Benchmarking, n.d.; Chou and Ho In Korea, rapid economic development 2007; Education Commission of Hong Kong took off under President Park Chung-Hee in 1984; Murata and Stern 1993; Seth 2002; the 1960s. His vision was to transform Korea Sorensen 1994; Sweeting 2004). In 1991, SPOTLIGHT 2 into an industrial power. To achieve this, he government policies still tried to push half looked to the German education model of the of all secondary school students into TVET 19th century for guidance. Basic education programs. These programs ended only in was made compulsory for all to build basic 1998, when industry shifted from using low- literacy and numeracy. As much as 80 percent skill labor to adopting labor-saving technolo- of all funds for school construction went to gies to stay competitive and when growth in primary schools through the 1960s. Most pri- insurance, banking, finance, and education mary school graduates enrolled in TVET pro- required workers with higher academic edu- grams at the middle and high school levels, cation (Kim 2001). as only top-performing students could pass In Japan, use of TVET programs to pro- highly competitive national exams and gain duce workers for modern industry dates back entry to the next level of academic education to the end of the 19th century. The founding (Kim 2001). fathers of modern Japan thought the German The view that TVET programs rather than model was best suited for a latecomer to academic education would produce the work- industrial development. In postwar Japan, ers Korea needed to power economic growth TVET programs were revived to produce was articulated in the government’s five-year skilled workers. In the 1950s and 1960s, economic development plans. The first plan, about 40 percent of all senior high school stu- launched in 1962, envisaged that 70 percent dents were enrolled in vocational high schools of secondary school students would be (Abumiya 2012; Center on International enrolled in TVET programs. Education Benchmarking, n.d.; OECD 2010). As incomes and aspirations in Korea rose These students were not tracked at an early in the 1970s, more parents rejected TVET age, as in Singapore, or denied the option of and demanded greater access to academic any academic education, as in Korea. As a secondary education. They saw an academic highly egalitarian society, Japan viewed track- education as promising better jobs, higher ing as taboo. Parents and educators broadly wages, and higher social status for their chil- reject exams as a way to measure student dren. The state acknowledged this parental performance in the kindergarten through preference but refused to abandon its pref- twelfth-grade system. “Exam hell” was (and erence for combining primary schooling with is) only for students seeking to obtain a post- secondary- level TVET programs. As late as secondary education at universities and col- 1979, more than half of all public funds for leges (Fujita 2007). school construction went to the primary level Japan rapidly rebuilt its education sys- (Kim 2001). The government responded to tem on the foundation of knowledge and parental demand for academic secondary edu- experience accumulated in the prewar years. cation by allowing more private secondary Its economy took off earlier and was more W H AT L E S S O N S C A N B E D R AW N F R O M TO P P E R F O R M I N G S Y S T E M S ’ E X P E R I E N C E W I T H T V E T ? 99 advanced than those of Korea and Singapore. invest enormous resources to get their children The government was therefore able to pro- the best academic education they can. In vide greater access to academic education at Korea, more than 70 percent of high school the middle and junior high school levels. graduates go on to university—the highest Enrollment in TVET programs was at rate among Organisation for Economic the senior high school level. It was elective, Co-operation and Development (OECD) and private institutions operated many of countries. However, a university degree does these programs. Students could gain special- not guarantee employment. One out of three ized education in many subjects, including unemployed people in Korea is a university agriculture, plumbing, electrical, accounting, graduate (Korea Times 2016). Technology is mechanics, food processing, and woodwork- central to virtually all industries. Employers in ing. Graduates were regarded as higher-skill Japan, Korea, and Singapore are desperately workers and could easily secure well-paid seeking workers with appropriate technical jobs with good benefits. knowledge and skills. Box S2.1 describes one SPOTLIGHT 2 way in which China is addressing this need. In Singapore, the government knows that TVET 2.0 it has to produce more highly skilled blue- As incomes rose and new job opportunities collar workers if it is to continue to ascend the opened up, families broadly rejected TVET value chain in production activities and make programs, viewing them as inferior in income the island state a regional innovation hub for potential and social status. Families now high-tech manufacturing firms. For Japan, BOX S2.1 World Bank engagement in TVET in China: the Liaoning Urban Construction School’s eco-laboratory The World Bank has a long history of investing The 5,100 square meter eco-lab is a high-tech in China’s technical and vocational education and educational space equipped with state-of-the-art training (TVET), dating back to the first World architectural features, such as solar lighting and Bank project there, the 1981 China Higher Education heating, ground-source heat pumps, energy-saving Project, which supported tertiary TVET programs at technologies, and water recycling. Its construction project universities. Early World Bank TVET proj- contributed significantly to LUCS’s learning activi- ects in China were designed mostly at the national ties. The building became an integral part of the level. Since 2006, demand for TVET directly from school’s architectural education activities, with a provinces has increased. By focusing on competency- large variety of exhibitions and lab sessions for stu- based curricula and pedagogy reform as well as on dents to observe and gain hands-on experience. To implementation, all projects have strived to position foster LUCS’s connection with industry, a commit- TVET in China as a demand-driven, equitable sys- tee on eco-architecture, including academic experts tem within a lifelong learning framework. and leading technical practitioners from industry, One of these provincial-level projects addressed was put in place. The committee has incorporated eco-architecture, a booming field in China. The several additions to LUCS’s architecture curriculum, Liaoning Urban Construction School (LUCS) used including information management technology, funding to construct an “eco-laboratory” building architectural industrialization, use of green building to enhance its architectural education program. materials, energy-efficient building structures, and This nationally recognized, award-winning green construction techniques. The eco-lab provides digi- building expanded the training space on LUCS’s talized equipment to enhance students’ learning campus. The construction process, which lasted experience, helping them to visualize and practice about two years, provided an exemplary case study these concepts. of school-industry collaboration. 100 GROWING SMARTER highly skilled workers are critical to increase Policy Research). http://www.nier.go.jp/English productivity in an aging society. The prospect /educationjapan/pdf/201209SE.pdf. of a shrinking population and workforce is Benson, J., and Y. Zhu, eds. 2012. Unemployment forcing companies to adopt more labor-saving in Asia: Organizational and Institutional Relationships. Singapore: Routledge Studies in technologies, but they still require workers to the Growth Economies of Asia. oversee and maintain them. Center on International Education Benchmarking. Upgrading TVET programs at the second- n.d. “Japan: Teacher and Principal Quality.” ary and postsecondary levels is only part of National Center on Education and the what needs to be done. The bigger challenge Economy, Washington, DC. is changing negative perceptions of TVET. Chou, C. P., and A.-H. Ho. 2007. “Schooling in In Singapore, since the early 1990s, pub- Taiwan.” In Going to School in East Asia, licity campaigns have shown that TVET edited by G. Postinglione, 344–75. New York: and technical jobs are not second-best Greenwood. options in a technology-driven world, and Education Commission of Hong Kong. 1984. “Education Commission Report 1.” SPOTLIGHT 2 enormous state investment has gone into Education Commission of Hong Kong, Tamar, upgrading TVET programs at all levels. October. Today, nearly two-thirds of all postsec- Fujita, H. 2007. “The Qualifications of the ondary enrollments are in TVET programs Teaching Force in Japan.” In A Comparative with state-of-the-art facilities and equip- Study of Teacher Preparation and Qualification ment and top-quality instructors with in Six Nations, edited by R. E. Ingersoll, 41–54. industry experience. Philadelphia: Consortium for Policy Research Although TVET has less of an image in Education. problem in Japan, the government wants Goh, C. B., and S. Gopinathan. 2008. “The to encourage more young people, including Development of Education in Singapore since women, to use postsecondary TVET pro- 1965.” In Toward a Better Future: Education grams to prepare for entry into the work- and Training for Economic Development in Singapore since 1965, edited by S. K. Lee, force. Raising female labor force participation C. B. Goh, B. Fredriksen, and J. P. Tan, 12–38. would help to ease Japan’s severe labor short- Washington, DC: World Bank. age. High-quality TVET classes have been Husna, R. 2018. “Japan Unemployment introduced in many academic junior high Rate.” Trading Economics, January 29. schools to expose students to different indus- http://www.tradingeconomics.com/japan tries and career options. In 2016 the gov- /unemployment-rate. ernment also approved the establishment of Kim, G.-J. 2001. “Education Policies and Reforms four-year technical universities. Programs at in South Korea.” In Secondary Education these institutions are expected to help stu- in Africa: Strategies for Renewal , 29–40. dents to build problem-solving skills relevant Washington, DC: World Bank. to industry needs. Korea Times. 2016. “One in Three Unemployed People Have Bachelor’s Degree.” October 17. http://www.koreatimes.co .kr/www/news /nation/2016/10/116_216218.html. Referencesa Murata, S., and S. Stern. 1993. “Technology Abumiya, M. 2012. “Upper Secondary Education Education in Japan.” Journal of Technology in Japan.” In Education in Japan. Tokyo: Education 5 (1): 29–37. NIER (National Institute for Educational OECD (Organisation for Economic Co-operation and Development). 2010. “Japan: A Story of Sustained Excellence.” In Strong Performers a. Titles of publications that include “Hong Kong,” “South Korea,” and “Taiwan” refer to Hong Kong SAR, China; the and Successful Reformers in Education: Republic of Korea; and Taiwan, China, respectively. W H AT L E S S O N S C A N B E D R AW N F R O M TO P P E R F O R M I N G S Y S T E M S ’ E X P E R I E N C E W I T H T V E T ? 101 Lessons from PISA for the United States, ch. 6. Sorensen, C. 1994. “Success and Education in Paris: OECD. South Korea.” Comparative Education Review Seth, M. J. 2002. Education Fever: Society, Politics, 38 (1, February): 10–35. and the Pursuit of Schooling in South Korea. Sweeting, A. 2004. Education in Hong Kong, 1941 Honolulu: University of Hawaii Press. to 2001: Visions and Revisions. Hong Kong Singapore Ministry of Trade and Industry. 2011. SAR, China: Hong Kong University Press. “MTI Insights 1965–1978.” Singapore Ministry of Trade and Industry. www.mti.gov.sg. SPOTLIGHT 2 Spending Public Resources on Education Effectively 4 Top Performing and Above-Average Trends in public spending on Performing Systems in East Asia and Pacific education have spent their public resources on education effectively by adopting three principles: Wealthier countries do not necessarily spend more on • Prioritize public spending on basic education education There is no overall correlation between a • Manage essential inputs efficiently country’s gross domestic product (GDP) and • Enhance the equitable distribution of government spending on education as a per- resources. centage of GDP. Top Performing Systems are These principles reinforce the con- not among the highest spenders (table 4.1 cept of “progressive universalism” advo- and box 4.1); there are high and low spenders cated by the International Commission on within each group. Among Above-Average Financing Global Education Opportunity Performing Systems, Vietnam has been a (2016), known popularly as the Education consistently high spender, whereas China Commission. Progressive universalism means only recently increased spending. Among that “when balancing spending across dif- Below-Average Performing Systems, spend- ferent levels of education and population ing in Malaysia and Thailand fluctuated over groups, decision makers should prioritize the the years. In Indonesia, spending increased poor and early years where social returns are slowly but steadily, from 1.0 percent in 1995 highest, and minimize household spending on to 3.3 percent in 2014. Spending by the basic education by the poor” (International Philippines remains low. Among Emerging Commission on Financing Global Education Systems, spending is still relatively low, but Opportunity 2016, 20). This chapter distills Cambodia, the Lao People’s Democratic lessons from East Asia and Pacific’s high- Republic, and Myanmar (box 4.2) increased performing systems that other countries—in spending considerably in recent years. In the region and beyond—can adopt to spend contrast, spending declined in Mongolia public resources for education effectively. It and Timor-Leste. This lack of correlation is does not describe how much to spend but not unique to East Asia and Pacific; it can be rather how to spend. observed worldwide. 103 104 GROWING SMARTER TABLE 4.1 Government spending on education as a share of GDP varies widely, even among economies with similar levels of performance Government spending on education as % of GDP GDP per capita (US$) Country 2000 2005 2010 Latesta 2015 Top Performing Systems Japan 3.6 3.5 3.8 3.8 34,524 Korea, Rep. 4.3 3.9 4.9 4.6 27,222 Singapore 3.6 2.9 3.1 3.0 52,889 Above-Average Performing Systems China — 2.8 3.6 3.9 8,069 Vietnam — — 6.3 6.3 3,968 Below-Average Performing Systems Indonesia — 2.9 2.8 3.3 3,346 Malaysia 6.0 — 5.0 6.1 9,768 Philippines 3.3 2.4 2.2 2.6 2,904 Thailand 5.3 3.9 3.5 4.1 5,815 Emerging Systems Cambodia 1.7 — 1.9 2.7 1,159 Lao PDR 1.5 2.4 2.8 4.2 1,818 Mongolia 5.6 — 4.6 4.6 3,968 Myanmar <1.0 — 0.7 2.0 1,161 Papua New Guinea — — — 4.8 2,268b Timor-Leste — — 10.5 7.7 1,217 Sources: Cambodia Ministry of Economy and Finance 2017 (for 2010) and World Bank 2017a (for latest) for Cambodia; National Bureau of Statistics of China 2016 for China; World Development Indicators (last data, updated April 2017) for GDP per capita (World Bank, various years); Indonesia Ministry of Finance for Indonesia (unpublished data; latest only); OECD 2003, 2013, and 2016a for the Republic of Korea; Government of the Republic of the Union of Myanmar 2014 and UNESCO 2016 for Myanmar; World Bank 2014b for Papua New Guinea; World Bank 2016 for the Philippines; Singapore Ministry of Education (http://data.gov.sg/) for Singapore; World Bank EdStats (World Bank, various years) for others. Note: — = not available; GDP = gross domestic product. a. Latest between 2011 and 2014, except 2016 for Indonesia and 2015 for Cambodia, China, Myanmar, the Philippines, and Singapore. b. GDP per capita for 2014. BOX 4.1 Government spending on education as a share of GDP is a better indicator of education spending than is spending as a share of total government spending The two most frequently used indicators for demands, given budget availability and demo- education spending—government spending graphic conditions. on education as a percentage of GDP and as The two indicators are only loosely correlated a percentage of total government spending— (figure B4.1.1). For instance, Timor-Leste spends illustrate different pictures. Spending as a a very small portion of its government budget share of GDP illustrates the magnitude of edu- on education, but its spending still represents a cation spending; spending as a share of total very large share of GDP. Indonesia spends almost government spending illustrates how each 20 percent of its government budget on education, country places education among competing but the figure represents only 3.3 percent of GDP. box continues next page S P E N D I N G P U B L I C R E S O U R C E S O N E D U C AT I O N E F F E C T I V E LY 105 BOX 4.1 Government spending on education as a share of GDP is a better indicator of education spending than is spending as a share of total government spending (continued) FIGURE B4.1.1 Spending on education as a share of government expenditure is largely uncorrelated with spending as a share of GDP 25 Malaysia Vietnam 20 % of government expenditure Indonesia Thailand Singapore 15 Philippines China Lao PDR Papua New Guinea Myanmar Mongolia Korea, Rep. 10 Cambodia Japan Timor-Leste 5 0 0 1 2 3 4 5 6 7 8 9 % of GDP Top Performing Systems Above-Average Performing Systems Below-Average Performing Systems Emerging Systems Sources: World Bank 2017a for Cambodia; National Bureau of Statistics of China 2016 for China; Indonesia Ministry of Finance for Indonesia; OECD 2016a for the Republic of Korea; Government of the Republic of the Union of Myanmar 2014 and UNESCO 2016 for Myanmar; World Bank 2014b for Papua New Guinea; World Bank 2016 for the Philippines; Singapore Ministry of Education (http:// data.gov.sg/) for Singapore; World Bank EdStats (World Bank, various years) for others. Note: Figures are latest between 2011 and 2014, except 2016 for Indonesia and 2015 for Cambodia, China, Myanmar, the Philippines, and Singapore. BOX 4.2 Myanmar is gradually increasing public spending on education For many decades, Myanmar allocated a much approved a substantial increase in the education smaller share of GDP to education than did most budget, as tax revenues began to rise. Estimated other countries. In 1973/74, government expen- actual spending on education increased to diture on education was estimated to be about 2.1 percent of GDP, or about 14 percent of over- 3.3 percent of GDP. The share subsequently fell all government spending in 2013/14, although steadily, reaching just 0.7 percent of GDP by it dropped to 2.0 and 13.0 percent, respectively, 2011/12. Households financed about 70 percent of in 2015. total education spending that same year. Decades of inadequate spending have affected the quality of education. In 2012/13, Parliament Sources: UNESCO 2016; World Bank 2015a. 106 GROWING SMARTER More spending does not promise Prioritizing public spending on higher performance basic education Most participants in the Programme for Top Performing and Above-Average International Student Assessment (PISA) Performing Systems sequenced spend 18–25 percent of GDP per capita per subsectoral allocations student per year (figure 4.1). Among high performers, Japan, the Republic of Korea, Countries that aim to build strong human and Vietnam spend about 24 percent of capital for economic growth should prioritize GDP per capita on each student, but the public spending on basic education to deliver highest performer, Singapore, spends good-quality and universally available educa- only 11 percent. Several countries that tion at that level before devoting more spend- spend 20–25 percent of GDP per capita ing to higher levels of education (Jimenez, per student—and even some that spend Nguyen, and Patrinos 2012). East Asia and more than 30 percent of GDP per capita— Pacific’s Top Performing and Above-Average performed as poorly as those spending less Performing Systems followed this principle, than 15 percent. Thailand, for example, shifting their focus from basic to tertiary edu- with an average PISA score of 409, spends cation over time. During their early period 21 percent of GDP per capita per student, of economic growth, they invested in pri- and Indonesia, with a score only slightly mary education to establish basic literacy and lower (397), spends only 11 percent. High numeracy for the largest number possible per student spending relative to the coun- and to build strong foundations before mov- try’s income level is thus no guarantee of ing gradually to higher levels of education. high performance. Then, as their economies grew and demand for highly skilled workers rose, these systems devoted an increasing share of education expenditure to higher levels of education. Their central control over the education bud- FIGURE 4.1 Per student spending as a share of GDP per capita get enabled them to sustain the investment. does not correlate with 2015 PISA science scores In Korea, the government more than tri- pled the education budget, from 4.2 percent of government spending in 1954 to 14.9 600 Singapore Japan percent in 1959. To achieve free compulsory 550 Hong Kong SAR, China primary education, it allocated 69–80 percent 2015 PISA science score Vietnam 500 of the education budget to primary education Korea, Rep. between 1948 and 1959. Much of this money 450 was spent to build schools in order to reduce 400 Indonesia Thailand double-shifting. The allocation to primary education peaked at 81 percent in 1960 and 350 leveled off at 54 percent by 1979 (Kim 2002). 300 In Singapore, education expenditure 10 15 20 25 30 35 40 45 expanded dramatically in the 1950s, from Spending per student as a % of GDP per capita 1–8 percent of total government expenditure Top Performing Systems in the prewar period to 13.2 percent in 1952 Above-Average Performing Systems and 24.9 percent in 1958 (Blackburn 2017). Below-Average Performing Systems Non-East Asia and Pacific country It peaked at 31.7 percent in 1964 and 28.8 percent in 1965. In 1965, the government Sources: Data from World Bank EdStats (World Bank, various years) and PISA 2015 (OECD 2016d). allocated 59 percent of the education budget Note: GDP = gross domestic product; PISA = Programme for International Student Assessment. to primary education. S P E N D I N G P U B L I C R E S O U R C E S O N E D U C AT I O N E F F E C T I V E LY 107 In Japan, the education system that con- FIGURE 4.2 Japan prioritized basic education in its earlier stituted the foundation of postwar educa- years of economic growth, 1960–2014 tion reforms underwent rapid quantitative expansion beginning in the 1950s (IFIC and 100 JICA 2004). A large share of the educa- 90 tion budget in the early postwar period was spent on extending compulsory education 80 % of total education spending from six (elementary school) to nine years 70 (elementary and lower-secondary school) (figure 4.2). Increases in enrollment in voca- 60 tional high school followed. The stock of 50 mid-level human capital capable of using 40 borrowed technology efficiently was further augmented during the period of rapid eco- 30 nomic growth by allocating a large share of 20 the education budget to secondary schools, at 10 the expense of tertiary education (Godo and Hayami 2009). 0 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2014 Above-Average Performing Systems are still focusing on basic education. Vietnam Higher education Upper-secondary school continues to prioritize public investment in Specialized training college Elementary and lower- Special needs school secondary school primary and secondary education. China also Secondary school Kindergarten prioritizes investment in pretertiary educa- tion, particularly primary, vocational, and preschool education (OECD 2016b). Source: Japanese Ministry of Education, Culture, Sports, Science, and Technology (http://www.mext .go.jp/b_menu/toukei/002/002b/1383990.htm). Note: Historical data for the breakdown by elementary and lower-secondary levels are not available. Secondary schools consist of lower- and upper-secondary levels, which are not common in Japan. Top Performing Systems continue prioritizing investment in basic Relative to poor countries in East Asia and education Pacific, wealthier countries tend to spend a Although increasing spending on higher levels larger share of GDP per primary and second- of education, Top Performing Systems contin- ary education student and a smaller share per ued raising per student spending on primary tertiary education student. Patterns of public and secondary education. In Singapore, for spending vary widely (figure 4.4). example, between 1986 and 2015, per student Both worldwide and in East Asia and spending grew by 122 percent in real terms at Pacific, per student public spending at the the primary level and by 113 percent at the primary level and countries’ wealth are cor- secondary level.1 In Japan and Korea, per stu- related. This correlation is less clear at the dent spending on primary and secondary lev- secondary level worldwide, although it is still els grew steadily in real terms between 2000 evident in East Asia and Pacific. At the ter- and 2013 (figure 4.3). A notable difference tiary level, there is a negative correlation in between Singapore and the other two coun- the rest of the world, a slightly negative corre- tries is that per student spending for tertiary lation in East Asia and Pacific, and a positive education in the other two was always lower correlation in the Organisation for Economic than spending for the other levels, except for Co-operation and Development (OECD). a recent increase in Japan, primarily because These findings reflect the fact that govern- tertiary education is financed largely by pri- ments in East Asia and Pacific play an impor- vate spending in Japan and Korea. tant role at the primary and secondary levels 108 GROWING SMARTER FIGURE 4.3 Annual per student spending in Japan and the Republic of Korea rose at all levels of education between 2000 and 2013 a. Japan b. Republic of Korea 10,000 10,000 Annual spending per student (US$) Annual spending per student (US$) 9,000 9,000 8,000 8,000 7,000 7,000 6,000 6,000 5,000 5,000 4,000 4,000 3,000 3,000 2,000 2,000 1,000 1,000 0 0 20 0 20 1 20 9 20 0 20 1 02 20 3 20 4 20 5 06 20 7 20 8 20 9 20 0 20 1 12 13 20 0 20 1 20 2 03 20 4 20 5 20 6 20 7 20 8 12 13 0 1 1 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 20 20 20 20 20 20 20 Primary Lower-secondary Upper-secondary Secondary Tertiary Source: World Bank EdStats (World Bank, various years). Note: Values are in 2013 purchasing power parity prices. Dotted lines in panel b indicate missing data. and spend more per student as their econo- lower-secondary education are financed pre- mies grow. In East Asia and Pacific’s low- dominantly by public resources. In Singapore, income countries, the unit cost for tertiary private spending on formal education was education is about the same as or slightly never more than 1 percent of total spending higher than that for lower levels of education; on education (Wong 2017). These statistics it is not higher than in many poorer countries illustrate that Top Performing Systems in East outside the region. Asia and Pacific prioritize public funding for Even in Japan and Korea, where private basic education. spending on education is substantial, primary Returns to different levels of education and lower-secondary education (and upper- may help to explain why countries shift their secondary education to a lesser extent) are subsectoral focus yet continue investing in mostly publicly funded. The proportion of basic education as their economies grow. private spending on educational institutions Figure 4.5 presents the returns to educa- within the formal education system is large tion in low- and middle-income countries by in Japan and Korea relative to the OECD country income group and level of schooling average. In 2013, of spending on educational (Psacharopoulos, Montenegro, and Patrinos institutions, only 72 percent in Japan and 2017). It shows that returns are higher in 64 percent in Korea was publicly financed, lower-income countries in general, where compared with an average of 84 percent in the quantity of schooling is scarcer. Private the OECD (table 4.2). However, the large returns to all levels of education are higher share of private spending in these two coun- than social returns in these countries, but tries is driven primarily by spending on ter- the difference is much larger at the tertiary tiary education and, to a lesser degree, on level. Given the high social returns at the pri- upper-secondary education; primary and mary level, it would make sense for low- and S P E N D I N G P U B L I C R E S O U R C E S O N E D U C AT I O N E F F E C T I V E LY 109 FIGURE 4.4 Government spending on education as a share of GDP per capita varies widely a. Primary education a.1. World a.2. East Asia and Pacific 60 30 Government spending per primary Government spending per primary R ² = 0.2378 student as a % of GDP per capita student as a % of GDP per capita 50 25 40 20 30 15 20 10 10 5 0 0 6 7 8 9 10 11 12 6 7 8 9 10 11 12 Log GDP per capita Log GDP per capita b. Secondary education b.1. World b.2. East Asia and Pacific 80 30 R ² = 0.3758 Government spending per primary Government spending per primary student as a % of GDP per capita student as a % of GDP per capita 70 25 60 20 50 40 15 30 10 20 5 10 0 0 6 7 8 9 10 11 12 7 8 9 10 11 12 Log GDP per capita Log GDP per capita c. Tertiary education c.1. World c.2. East Asia and Pacific 400 60 Government spending per primary Government spending per primary student as a % of GDP per capita student as a % of GDP per capita 350 R ² = 0.1774 R ² = 0.0062 50 300 40 250 200 30 150 20 100 10 50 0 0 6 7 8 9 10 11 12 7 8 9 10 11 12 Log GDP per capita Log GDP per capita Top Performing Systems Emerging Systems Above-Average Performing Systems OECD excluding Top Performing Systems Below-Average Performing Systems Rest of world Source: Data from World Bank EdStats (World Bank, various years). Note: GDP = gross domestic product; OECD = Organisation for Economic Co-operation and Development. 110 GROWING SMARTER TABLE 4.2 Share of public expenditure on educational institutions in Japan, the Republic of Korea, and the OECD, by level of education, 2013 Percent Country or country group Primary Lower-secondary Upper-secondary Tertiary All Japan 99 94 82 35 72 Korea, Rep. 91 93 71 32 64 OECD average 93 93 87 70 84 Source: OECD 2016a, table B3.1a. Note: OECD = Organisation for Economic Co-operation and Development. FIGURE 4.5 Returns to education differ by level of schooling and country income group a. Lower-income countries b. Lower-middle-income countries c. Upper-middle-income countries 25 22.8 25 25 Returns to education (%) Returns to education (%) Returns to education (%) 20 20 17.6 20 17.1 16.9 15 14.1 15 13.8 15 12.2 9.6 9.4 9.2 10 10 8.4 10 7.5 6.7 7.0 5 5 5 4.5 2.3 2.9 0.4 0 0 0 Primary Secondary Tertiary Primary Secondary Tertiary Primary Secondary Tertiary Education level Education level Education level Private returns Social returns Source: Psacharopoulos, Montenegro, and Patrinos 2017. lower-middle-income countries to prioritize Performing and Emerging Systems have investment at that level. not sequenced the focus of their spending The low returns to primary and second- from primary to tertiary, and they have not ary education in upper-middle-income coun- sufficiently prioritized investment in lower tries stem from the fact that primary and, levels of education. Indonesia, Malaysia, to a lesser degree, secondary education have and Thailand have increased per student already reached most of the population. spending for primary and secondary educa- Governments need to continue investing to tion in recent years. This spending is now enhance the quality of primary and secondary higher than spending on tertiary education education and to strengthen the human capi- in Thailand, but it is still much lower in tal foundation to meet rapidly changing labor Indonesia and Malaysia. In Malaysia, per market demand. student spending on tertiary education (in purchasing power parity dollars) is even much higher than in Japan and Korea. Below-Average Performing and In Indonesia, public spending on public Emerging Systems have not tertiary institutions (US$5,794 in 2013 prioritized primary education prices) is more than four times greater than Unlike Top Performing and Above-Average spending on public primary institutions P e r f o r m i n g S y s t e m s , B e l o w - Av e r a g e (US$1,319) (OECD 2016a, table B3.3). S P E N D I N G P U B L I C R E S O U R C E S O N E D U C AT I O N E F F E C T I V E LY 111 FIGURE 4.6 Indonesia spends more public 36 percent of the budget (World Bank funds on higher education than on upper- EdStats; World Bank, various years). secondary education Since the late 2000s, Papua New Guinea has prioritized universal access to basic 100 education, rapidly increasing education spending to achieve it. The abolition of 90 school tuition fees for basic education in 80 2012 led to a dramatic increase in enroll- Spending on education (%) 70 ment. As a result, per student spending 60 has decreased in recent years (World Bank 2014b). 50 40 High-performing systems 30 manage teachers and school 20 infrastructure efficiently 10 More spending does not necessarily result 0 in better outcomes. Efficient management of Basic Upper- Higher essential inputs is key. This section discusses secondary education how high-performing systems finance teach- Public spending Household spending ers and school infrastructure. Source: Adapted from World Bank 2013b. Teachers East Asia and Pacific’s Top Performing and Above-Average Performing Systems tend to Public spending represents only two-thirds spend less on recurrent spending than other of total spending on noncompulsory upper- countries, leaving room for capital spend- secondary education, less than spending on ing. Japan and Korea allocate the smallest higher education (figure 4.6). shares to recurrent spending among OECD In the Philippines, total government countries: 85–87 percent for primary com- spending on education and government pared with the OECD average of 92 percent spending on basic education fell in real terms (figure 4.7). The patterns are similar for at the start of the 2000s. The government secondary and tertiary education: 85–89 began increasing spending on basic educa- percent for secondary and 84–86 percent for tion in 2005, but as a share of GDP, spend- tertiary in Japan and Korea compared with ing on basic education reached the rates of the OECD average of 93 and 89 percent, the early 2000s only recently, and it is still respectively.2 much lower than in comparator countries, The proportions of recurrent and capital such as Malaysia, Thailand, and Vietnam expenditures vary over time, depending on (World Bank 2016). investment needs. In Singapore, for instance, Among Emerging Systems, in 2013/14, development spending peaked in the late Lao PDR allocated only 64 percent of edu- 1990s to early 2000s. It represented more cation spending to pretertiary education (3 than 25 percent of total education spending percent to preprimary, 32 percent to primary, over five years during the period; in 2015 the and 30 percent to secondary, including tech- figure was less than 10 percent (figure 4.8). nical and vocational education and train- The composition of spending in Indonesia ing). Tertiary and other education received and Japan was similar. Vietnam allocated only 112 GROWING SMARTER FIGURE 4.7 The composition of spending on primary 80 percent to recurrent spending. By contrast, education varies widely across countries Malaysia, Mongolia, the Philippines, and Thailand spent more on recurrent expendi- 100 ture than the OECD average. 90 East Asia and Pacific’s Top Performing and Above-Average Performing Systems offer suf- % of spending on primary education 80 ficiently attractive starting salaries and bene- 70 fits, as well as progressive salary schemes and 60 other incentives, to attract and retain quali- fied teachers. 50 The Ministry of Education in Singapore 40 regularly adjusts salaries for new teachers 30 so that teaching remains as attractive to new 20 graduates as other occupations. Although teacher salaries do not increase as much as 10 salaries in the private sector, the government 0 offers retention bonuses and nonfinancial m do n re a ilip . Th es M and M lia Es ia a k R via OE Hun lic av ry e incentives, such as career opportunities within Ph Rep Ko nesi ni ag pa ys CD ga na n ub go va Lat to pi ala ail Ja er et ep a, on education. High-performing teachers can earn Vi In significant performance bonuses under the Slo Non-East Asia and comprehensive performance appraisal system East Asia and Pacific countries Pacific countries (OECD 2011). Capital spending New teachers in OECD countries earn an Recurrent spending on other than compensation average of 77 percent of GDP per capita. In Recurrent spending on compensation Japan and Korea, new teachers earn 76 and 92 percent of GDP per capita, respectively. Sources: OECD 2016a for Estonia, Hungary, Indonesia, Japan, the Republic of Korea, Latvia, and the Slovak Republic; World Bank EdStats for other countries (World Bank, various years). They can expect salary increases to reach Note: Data are for 2009 for the Philippines, 2012 for Indonesia, 2013 for OECD countries, and 2012–14 133 and 159 percent of GDP per capita, for Malaysia, Mongolia, Thailand, and Vietnam. OECD = Organisation for Economic Co-operation and Development. respectively, after 15 years (figure 4.9). These increases are much higher than the OECD FIGURE 4.8 In Singapore, the proportion of recurrent and average, where teacher salaries are 4–13 development education spending fluctuated over time, percent lower than those of workers with 1981–2015 similar qualifications (OECD 2016c). The high reward for experience is a likely reason 12 100 for the extremely low average annual teacher 90 attrition rates in the East Asia and Pacific’s 10 Top Performing Systems—less than 3 percent % of education spending 80 Spending (US$, millions) 70 compared with 6 percent in most Western 8 60 European countries and 8 percent in the United 6 50 States (Wong 2017). 40 In China, teachers receive average sala- 4 30 ries for civil servants. Civil service jobs are 20 stable and provide good benefits, which 2 10 makes teaching popular, especially in big cities 0 0 (OECD 2016b). In Shanghai, teacher salaries are based partly on performance evaluations 81 83 85 87 89 91 93 95 97 99 01 03 05 07 09 11 13 15 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 (Liang, Kidwai, and Zhang 2016). Development Recurrent % of recurrent spending spending spending In Vietnam, where teachers are promoted to a higher salary category every three years Source: Singapore Ministry of Education 2017. and receive a “working-year allowance” S P E N D I N G P U B L I C R E S O U R C E S O N E D U C AT I O N E F F E C T I V E LY 113 every year, teachers at the top of the scale FIGURE 4.9 Salaries of primary school teachers in East Asia with more than 30 years of experience can rise markedly with seniority earn 2.77 times as much as beginning teach- ers (World Bank 2017b).3 300 Although teaching is a relatively well-paid Annual salary of primary teachers profession in Malaysia and Thailand, the 250 school system is unable to attract enough as % of GDP per capita qualified teachers. In Malaysia, the salary 200 increment for teachers increases substantially with seniority (UNESCO 2015a). But the 150 very low starting salaries relative to GDP per 100 capita fail to attract qualified teachers in the first place (figure 4.9). In 2010, 93 percent 50 of applicants to the bachelor’s in education program did not meet academic require- 0 ments. Among candidates offered a place, 70 ey ico sia s n Th a d p. d ile ne i pa an lan ys Re rk Ch ne ex pi ala ail Ja percent did not meet academic requirements Tu Fin a, do M ilip re M In Ph Ko (World Bank 2013a). Thailand also faces Non-East Asia chronic teacher shortages. Starting salaries East Asia countries countries for teachers there are very low relative to Starting salary Salary after 15 years Salary at top of scale GDP per capita (OECD and UNESCO 2016). of experience In Cambodia, teaching is not a particularly attractive profession for top graduates, pri- Source: UNESCO 2012. Note: Salaries reflect official pay scales in 2010. They refer to the average scheduled gross annual marily because of low salaries. Almost all salary for a full-time teacher with the minimum training necessary to be fully qualified at the teacher trainees enrolled in teacher training beginning of his or her teaching career. GDP = gross domestic product. centers are bottom-half performers on the grade 12 exam (World Bank 2014a). in Thailand are much better paid than teach- Analysis of PISA results suggests that the ers in Vietnam (in 2011, teachers earned 144 only type of resource that is correlated with percent of comparable professions in Thailand student performance is teacher salaries relative and only 88 percent in Vietnam). To meet the to national income. There is a statistically sig- constitutional mandate to allocate 20 percent nificant positive relationship between teacher of the government budget to education (intro- salaries relative to GDP per capita and PISA duced in 2002), Indonesia increased teacher 2012 math scores in economies with GDP per salaries considerably over the past decade— capita of more than US$20,000 (figure 4.10). without observable improvements in learning All of East Asia and Pacific’s high perform- outcomes (World Bank 2013b). ers are above the trend line, but teachers in Student-teacher ratios declined over the Japan; Shanghai, China; and Singapore are past two decades in most of the region, not particularly highly paid, unlike teachers at both the primary and secondary levels in Korea and Hong Kong SAR, China. (figure 4.11). The decline reflects both the There is no statistically significant relation- fact that governments did not reduce the size ship between teacher salaries and test scores of their teaching workforces in the face of among economies with GDP per capita of falling numbers of school-age children and less than US$20,000. In Malaysia, where the fact that some governments hired more teachers earn more than twice GDP per teachers to respond to increasing enrollments capita on average, student performance was or to reduce student-teacher ratios. Changes worse than in Thailand, where teachers earn in the school-age population (5- to 19-year- only 25 percent more than GDP per capita. olds) since 2000 and projected changes over Vietnam has been a much better performer the next two decades vary considerably on PISA than Thailand, even though teachers across the region. They suggest that changes 114 GROWING SMARTER FIGURE 4.10 Higher teacher salaries relative to GDP per capita are correlated with student performance on the 2012 PISA among wealthier economies but not among less wealthy ones 650 Shanghai, China 600 Singapore Hong Kong SAR, China Macao SAR, China Korea, Rep. 550 Japan 2012 PISA mathematics score 500 R ² = 0.09 450 Thailand Malaysia R ² = 0.05 400 Indonesia 350 300 20 40 60 80 100 120 140 160 180 200 220 Teacher salaries relative to per capita GDP (%) Economies with per capita GDP exceeding US$20,000 Fitted line for economies with per capita GDP exceeding US$20,000 Economies with per capita GDP less than US$20,000 Fitted line for economies with per capita GDP less than US$20,000 Source: Adapted from OECD 2014. Note: The solid line shows a significant relationship ( p < 0.10). The dotted line shows a nonsignificant relationship ( p > 0.10). Teacher salaries are the weighted average of upper- and lower-secondary school teacher salaries. The average is computed by weighting salaries based on the enrollments of 15-year-old students (for economies with valid information for both upper- and lower-secondary education). GDP = gross domestic product; PISA = Programme for International Student Assessment. in the size of the school-age population low ratios do not guarantee high perfor- were a driving factor for the rapidly declin- mance. Evidence from recent impact evalu- ing student-teacher ratios in about half the ations in low- and middle-income countries region. Compared with 2000, there were shows that reducing class sizes in primary fewer people age 19 and younger in 2015 in schools improves learning (MacEwan 2015). about half the economies in the region; com- However, large class sizes in themselves pared with 2015, it is projected that there will may not be detrimental to learning, as long be fewer 5- to 19-year-olds in 2035 in more as student-teacher ratios are not very high, than half of them (figure 4.12). Depending on as observed in East Asia and Pacific econo- the size of the school-age population, govern- mies. Class sizes in PISA-participating East ments need to adjust the size of the teaching Asia and Pacific economies are larger than workforce efficiently. in most other PISA-participating economies, Very high student-teacher ratios are likely but student-teacher ratios are not, meaning to result in poor student performance, but that more teachers are present in classrooms S P E N D I N G P U B L I C R E S O U R C E S O N E D U C AT I O N E F F E C T I V E LY 115 FIGURE 4.11 In most of the region, student-teacher ratios declined at the primary and secondary levels between 1998 and 2014 a. Primary b. Secondary 60 60 (number of students to one teacher) (number of students to one teacher) 50 50 Student-teacher ratio Student-teacher ratio 40 40 30 30 20 20 10 10 0 0 20 0 20 1 02 20 4 20 6 20 7 20 8 20 9 10 20 1 20 2 20 3 14 20 9 20 3 05 20 9 20 0 20 1 02 20 3 20 4 05 20 6 20 7 08 20 9 10 20 1 12 20 3 14 19 8 19 8 0 0 0 0 0 0 0 1 1 1 9 0 9 0 0 0 0 0 0 0 1 1 9 9 20 20 20 20 20 19 20 20 19 20 Brunei Darussalam Kiribati Mongolia Solomon Islands Cambodia Lao PDR Myanmar Thailand Fiji Macao SAR, China Philippines Tonga Hong Kong SAR, China Malaysia Singapore Vietnam Source: World Bank EdStats (World Bank, various years). in East Asia and Pacific. Except for Thailand, and from 16.3 to 13.0 at the secondary level. these ratios are about average or below aver- The numbers in 2013 were significantly lower age (figure 4.13). Student-teacher ratios are than the OECD average of 16.1 in 2011 more closely correlated with learning out- at the primary level and 13.5 at the secondary comes than are class sizes. Except for the level (World Bank 2013a). Netherlands, no country with a student- These lower student-teacher ratios did not teacher ratio higher than 17 performed above improve learning outcomes. In Indonesia, average in math, although a ratio lower than there is no correlation between the student- 17 did not guarantee good performance either teacher ratio and learning for math or Bahasa (based on an analysis of PISA 2015 data). Indonesia for schools with student-teacher Very low student-teacher ratios have not ratios below 32 and only a slight correlation improved student performance in Indonesia for schools with ratios above 32 (World Bank and Malaysia. In Indonesia since the early 2015b). In Malaysia, Trends in International 2000s, the number of teachers for primary Mathematics and Science Study (TIMSS) and secondary education grew faster than the scores continued to decline between 1999 number of students. As a result, the student- and 2011 for eighth-grade math, and they fell teacher ratio continued to decline in primary between 2003 and 2011 for eighth-grade sci- and senior-secondary schools. By 2010, the ence. PISA scores improved for all three sub- student-teacher ratio was 20 for primary jects between 2012 and 2015, but Malaysia and 12 for secondary schools. In Malaysia, still lags countries at a similar level of eco- the student-teacher ratio declined from 17.2 nomic development and the regional and in 2004 to 11.5 in 2013 at the primary level OECD average by 1.5–2.2 years. 116 GROWING SMARTER FIGURE 4.12 In about half the economies in the region, School infrastructure the school-age population has decreased since 2000 and is projected to decrease over the next two decades Evidence from around the world shows that improving school infrastructure leads to bet- ter learning outcomes.4 Reviewing 39 studies a. 2000–15 Timor-Leste published between 1990 and 2012 (of which Solomon Islands 1.5 only 19 are of high quality), Cuesta, Glewwe, Projected relative change in school-age population, ages 5–19 (2000 = 1) Papua New Guinea Vanuatu and Krause (2016) assess the extent to which 1.4 Kiribati school infrastructure affects educational out- Philippines comes and enrollment.5 They find that school Singapore 1.3 libraries and the creation of new schools lead Malaysia Samoa to improved learning and enrollment. Toilets 1.2 Indonesia improve student learning, and laboratories Brunei Darussalam and drinking water facilities increase enroll- Tonga ment. Having roofs, walls, and floors in good 1.1 Lao PDR Myanmar condition improves student learning. No 1.0 Cambodia other classroom-level variables have clear Fiji effects. Micronesia, Fed. Sts. Woolner and others (2007, 60) review 0.9 Thailand Japan school environments in the United Kingdom. 0.8 Mongolia They find that “beyond the necessity of meet- Vietnam ing basic standards, there is not enough evi- Korea, Rep. dence to give clear guidance to policy makers 0.7 Taiwan, China on how to set priorities for funding or to eval- China 0.6 Hong Kong SAR, China uate the relative value for money to different 2000 2005 2010 2015 Macao SAR, China design initiatives.” Classrooms can become more effective if they are designed well. It is not necessary to b. 2015–35 Macao SAR, China Timor-Leste invest massive resources (box 4.3). 1.8 Projected relative change in school-age population, ages 5–19 (2000 = 1) Hong Kong SAR, China East Asia and Pacific’s Top Performing Mongolia Systems generally provide school facilities Vanuatu efficiently by keeping them simple but func- 1.6 Kiribati Papua New Guinea tional. Japanese schools are built to ministry Solomon Islands designs. They are functional but very plain, Philippines without many of the special features of 1.4 Cambodia schools in other high-income countries. There Vietnam is no cafeteria, and students take their meals Malaysia Lao PDR from a central kitchen to their classrooms. 1.2 Indonesia Students are also responsible for cleaning Tonga their classrooms. These practices keep spend- Fiji ing on school construction and maintenance 1.0 Micronesia, Fed. Sts. Samoa low. The savings can be spent on teachers Brunei Darussalam and instruction that matter more for student China learning (OECD 2011). 0.8 Myanmar In Korea and in Taiwan, China, school Singapore Taiwan, China buildings are also basic and built to accom- Korea, Rep. modate large classes (Wong 2017). In 0.6 Japan Singapore, schools built in the 1970s typically 2015 2020 2025 2030 2035 Thailand had simple designs and standard features (Heng 2011); today the government invests Source: UN 2017. heavily to improve schools’ information and Note: The figures for 2000–15 are estimates. The figures for 2015–35 are projections. S P E N D I N G P U B L I C R E S O U R C E S O N E D U C AT I O N E F F E C T I V E LY 117 FIGURE 4.13 Class sizes are larger in East Asia and Pacific than in most PISA-participating economies, but the student-teacher ratios are not higher 30 Student-teacher ratio (number of students to one teacher) 25 Thailand 20 Vietnam Indonesia Macao SAR, China 15 Korea, Rep. Malaysia B-S-J-G (China) Singapore Japan 10 Hong Kong SAR, China 5 15 20 25 30 35 40 45 50 Class size East Asia and Pacific economy Non-East Asia and Pacific OECD economy Source: PISA 2015 (OECD 2016d). Note: B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); OECD = Organisation for Economic Co-operation and Development; PISA = Programme for International Student Assessment. BOX 4.3 Inexpensive, small changes in classrooms can make big differences in student performance A study in the United Kingdom assessed three types facilities. The same school often includes a mix of of physical characteristics of classrooms: naturalness, more and less effective classrooms. individualization, and stimulation. Its results present Teachers can readily improve many design clear evidence that well-designed and decorated pri- elements. Small changes that cost little or nothing— mary schools boost children’s academic performance such as changing the layout of the classroom, the in reading, writing, and math. Differences in the phys- displays, and the color of the walls—can make a ical characteristics of classrooms explain 16 percent real difference. To stimulate learning, the classroom of the variation in learning progress over a year. should be neither chaotic nor boring but somewhere The design of individual classrooms is much in between. more important than whole-school factors such as size, navigation routes, specialist facilities, and play Source: Barrette and others 2015. communication technology infrastructure for school conditions may include insufficient the future (Govtech Singapore 2017). public spending on school infrastructure, Lack of basic school facilities remains limited access to water and electricity in a challenge in the region, except among rural areas, and difficult and costly construc- Top Performing Systems. Reasons for poor tion conditions. Many schools in Indonesia 118 GROWING SMARTER (World Bank 2009) and the Philippines (figure 4.14). This section explores policies (World Bank 2016) are overcrowded and do that Top Performing and Above-Average not meet basic standards for sanitation facili- Performing Systems have adopted to promote ties, desks, and chairs. In Lao PDR, almost the equitable distribution of resources to min- half of schools have leaky roofs. Only slightly imize gaps in learning outcomes. more than half of schools have a water sup- ply, only 20 percent have electricity, and The role of the central government fewer than half have toilets. Only 40 percent in ensuring equity have telephone service (mostly provided through the principal’s cell phone), and fewer In East Asia and Pacific’s Top Performing than 3 percent have access to computers Systems, the central government plays a (Santibañez 2014). The inequality of resource key role in equalizing education funding. In allocation is much worse than in OECD Japan, the government subsidizes prefec- countries. In Papua New Guinea, the rapid tures (equivalent to states or provinces) in increase in capital spending on schools in order to equalize public resources. For nine- recent years has resulted in a dramatic expan- year compulsory education, prefectures fund sion of access to basic education, but it has two-thirds of the cost of teacher salaries, and not led to learning improvement (World Bank the central government subsidizes the rest. 2014b). Disadvantaged schools have the same share When access is still an issue, heavy invest- of qualified teachers as advantaged schools ment in school facilities can improve edu- and more teachers per student. At the upper- cational attainment, performance, and secondary level, students from low-income wages. In 1973, the Indonesian government families are exempt from tuition fees for launched an unprecedented school construc- public schools, receive financial support to tion program to boost primary enrollment, pay tuition fees for private schools, and gain building more than 61,000 primary schools scholarships to cover financial obligations between 1973 and 1978. Duflo (2001) ana- other than tuition, such as school trips and lyzes the impact of the program on edu- textbooks (OECD 2015). In Singapore, the cational attainment and wages. She finds government provides merit-based scholar- that each new school built led to increases ships and other financial assistance for all stu- of 0.12–0.19 year of education, 1.5–2.7 dents as well as tuition subsidies for students percent of earnings, and 6.8–10.6 percent from low- and middle-income families to in economic returns to education. A study attend independent schools (National Center of Indonesia finds that the availability of at on Education and the Economy, n.d.). least one functioning toilet led to better stu- Vietnam allocates more per capita to geo- dent performance among girls (but not boys) graphically disadvantaged provinces and (Suryadarma and others 2006). districts and pays teachers serving in disad- vantaged areas higher salaries than teachers in cities, through various types of allowances.6 Enhancing the equitable In China, reducing inequalities in education distribution of resources is a government priority. The government has Public resources should be distributed equi- gradually integrated the compulsory educa- tably, not only because equity has an intrin- tion funding guarantee mechanism in rural sic value but also because it helps countries areas. By 2010, 97 percent of the total edu- to improve student performance. In Top cational investment in rural compulsory edu- Performing and Above-Average Performing cation came from the government budget Systems such as Korea, Singapore, and (OECD 2016b). Vietnam, learning outcomes are consistently Governments in Below-Average Perform- more equitable than in most other economies ing and Emerging Systems could improve S P E N D I N G P U B L I C R E S O U R C E S O N E D U C AT I O N E F F E C T I V E LY 119 FIGURE 4.14 Learning outcomes in the region’s Top Performing and Above-Average Performing Systems are consistently more equitable than in other economies a. Singapore b. Republic of Korea 100 100 90 90 80 80 70 70 60 60 Percent Percent 50 50 40 40 30 30 20 20 10 10 0 0 2009 2015 2009 2015 2009 2015 2006 2015 2006 2015 2006 2015 Mathematics Reading Science Mathematics Reading Science c. Vietnam d. Thailand 100 100 90 90 80 80 70 70 60 60 Percent Percent 50 50 40 40 30 30 20 20 10 10 0 0 2012 2015 2012 2015 2012 2015 2006 2015 2006 2015 2006 2015 Mathematics Reading Science Mathematics Reading Science Below basic proficiency Basic proficiency High proficiency Sources: PISA 2006, 2009, 2012, and 2015 (OECD 2006, 2009, 2014, and 2016d). Note: Singapore started participating in the Programme for International Student Assessment (PISA) only in 2009; Vietnam began in 2012. their funding measures to support disad- equalize the distribution of teachers across the vantaged areas and students, with a goal of country (World Bank 2013b). In Thailand, equalizing learning outcomes. In Malaysia, the government allocates higher per student the government has taken various steps to subsidies for small village schools, but their address inequities in the system, creating spe- small size makes it difficult for these schools cial programs for the indigenous population, to provide a high-quality education (World offering support programs for poor students, Bank 2015c). In the Philippines, the national upgrading and expanding educational facili- government does not allocate the education ties, and deploying more qualified teachers budget equitably across regions and prov- to rural areas. Performance gaps on national inces. Local government patterns of spending exams are evident across and within states, reinforce these differences, leading to wide dif- however (UNESCO 2015a). The Indonesian ferences in levels of per student spending and central government provides scholarships for the quality of learning environment (World the poor; it does not have central programs to Bank 2016). In Myanmar, the government 120 GROWING SMARTER provides a hardship allowance to incentiv- ESC-awarded schools may charge top-up ize teachers to work in remote areas (World fees to students, raising equity concerns. To Bank 2015a). But remote schools face chal- increase access to senior high school educa- lenges in retaining specialized teachers (Muta tion (grades 11–12), the government also 2015). In 2012, Papua New Guinea abolished introduced a voucher program that provides school fees in basic education and introduced qualified junior high school completers with subsidized fees for postbasic education in government vouchers to enroll in private order to increase access. This policy increased senior high schools.7 The voucher amount enrollment in basic education, particularly for reflects the cost of public provision, so that girls, reducing gender disparities (Papua New whether a student enrolls in a public or a pri- Guinea Department of Education 2016). The vate senior high school, the cost to the govern- tuition-free policy does not take into account ment is the same (Department of Education different needs across the country, however of the Philippines 2016). (UNESCO 2015b). In Indonesia, private schooling continues to account for about 40 percent of secondary school enrollments. Although government- Increasing the role of private dependent private schools are underfunded, spending in increasing access with a high proportion of uncertified, under- The government is the dominant financier paid teachers, demand remains high because of basic education, but many East Asia and of their focus on religious training and educa- Pacific economies promote private funding tion and their ability to increase educational to increase access to secondary education access for low-income families (Stern and and, to a lesser extent, primary education Smith 2015). (see spotlight 3). For instance, the Chinese Governments of East Asia and Pacific’s government promotes privately run schools Top Performing Systems have also mobilized by requiring the government at all levels to private spending to increase the overall pre- establish and improve the government sub- school enrollment and to narrow the gaps sidy system and to specify projects, targets, between the poor and the rich. As discussed standards, and purposes of the subsidies in chapter 6, the economic value of early (National People’s Congress 2012). The share childhood development programs is often of Chinese elementary students enrolled in highest for the most disadvantaged chil- private schools rose from 2.9 percent in 2004 dren. Investment in their early years there- to 6.7 percent in 2015 (Simon 2015). Private fore improves both equity and efficiency enrollment at the secondary level increased to (Heckman and Masterov 2007). However, 10.1 percent by 2015. across the world, except in South Asia, chil- Korea uses a lottery to assign students dren from the richest households are at least randomly to academic (not vocational) high twice as likely—and in some regions three schools. It applies the equalization policy times as likely—as children from the poor- not only to public schools but also to private est households to be enrolled in preprimary schools, which account for about half of all school (Neuman and Devercelli 2013). high schools in the country (Park 2013). Most economies in East Asia and Pacific In the Philippines, the government intro- allocate a small proportion of GDP to pre- duced the Educational Service Contracting primary education. On average, OECD (ESC) scheme in the 1980s, under which the economies spent 0.61 percent of GDP on government provides subsidies to private high preprimary education in 2013. Japan spent schools (grades 7–10) to enroll primary edu- 0.24 percent and Korea spent 0.39 percent. cation completers, in order to ease congestion In East Asia and Pacific, only Mongolia and in public high schools. The subsidies do not Vietnam spent more than the OECD average necessarily cover the full cost of education; (table 4.3). S P E N D I N G P U B L I C R E S O U R C E S O N E D U C AT I O N E F F E C T I V E LY 121 When only government spending is con- low level of public spending and high level sidered, this figure came down to 0.10 of private spending, preprimary education percent for Japan and 0.18 percent for enrollments for Japan and Korea are among Korea compared with the OECD average of the highest in the OECD, at nearly 100 0.54 percent. In these two countries, more percent (OECD 2016a).8 In contrast, enroll- than half of spending on preprimary educa- ment rates vary widely between the poor and tion comes from private sources. Despite the rich in Below-Average Performing Systems and Emerging Systems, where private fund- ing for preschool education is limited. In TABLE 4.3 Most East Asian countries spend only Indonesia, for instance, the enrollment rate a small share of GDP on preprimary education for four- to six-year-olds was 51 percent among the poorest quintile and 71 percent Total spending Government among the richest quintile in 2016. In the Country as % of GDP spending as % of GDP Philippines, increased government spend- Philippines — 0.04 ing on school infrastructure and key inputs Indonesia 0.06 0.06 raised the kindergarten enrollment rate Malaysia — 0.08 among the poorest 20 percent the most— from 33 percent in 2008 to 63 percent by Japan 0.24 0.10 2013 (World Bank 2016). Korea, Rep. 0.39 0.18 Thailand — 0.22 OECD average 0.61 0.54 Pros and cons of private tutoring Vietnam — 0.62 Spending on private tutoring in primary Mongolia — 1.09 and secondary education has been grow- Sources: OECD 2016a for total spending; World Bank EdStats (World Bank, ing rapidly across the world (box 4.4). This various years) for government spending. increase is a concern because the propor- Note: Figures are for 2011–14, except for the Philippines (2009). GDP = gross domestic product; OECD = Organisation for Economic tion of students who use private tutors is Co-operation and Development. — = not available. higher among wealthier students than among BOX 4.4 Spending on private tutoring is on the rise Spending on private tutoring is considerable and TABLE B4.4.1 Use of private tutoring is high across increasing in East Asia and Pacific. More than half the world of the 15-year-olds in East Asia and Pacific and the Middle East and North Africa who participated in % of 15-year-olds the 2012 PISA took at least one hour a week of pri- Region taking private tutoring vate tutoring (table B4.4.1). These shares are more Middle East and North Africa 58.3 than twice the share in the European Union. In some East Asia and Pacific 54.5 countries, spending by households on private tutor- Europe and Central Asia 43.5 ing now rivals public sector education expenditures. Latin America and the Caribbean 41.9 Japan has been a pioneer in private tutoring. In European Union 24.1 2010, families spent US$12 billion (0.2 percent of Others 23.2 GDP) on private tutoring. In the Republic of Korea, Source: Data from PISA 2012 (OECD 2012b). families spent US$20 billion or about 1.5 percent box continues next page 122 GROWING SMARTER BOX 4.4 Spending on private tutoring is on the rise (continued) TABLE B4.4.2 Effect of one additional hour of of GDP on private tutoring in 2012; these expenses private tutoring on PISA math scores in select account for 12 percent of household spending. Some economies in East Asia and Pacific 80 percent of elementary, 69 percent of middle school, and 50 percent of high school students used Economy Effect private tutors in 2014. Indonesia 0.042 Even in Singapore, where the state generously (1.118) funds education, parents spent 1.2 percent of GDP Japan 0.425 on education services in 2013, one-third of which (1.296) was spent on private tuition and other educational China 0.675 courses. The proportion of household spending on (0.986) preprimary and general education relative to public Korea, Rep. 2.064*** spending is very high in Vietnam as well. (0.631) The impact of private tutoring on academic Macao SAR, China 2.236** achievement is mixed. Outcomes depend on the (0.978) grade level; the type and duration of provision and Thailand 3.487*** delivery modes; specific national and local circum- (0.621) stances; definitions of achievement; and probably Singapore 3.644** most important, the motivation, attitudes, and learn- (1.612) ing styles of learners as well as the motivation, atti- Taiwan, China 3.841*** tudes, and teaching styles of tutors. (1.268) An additional tutoring hour at a commercial Vietnam 3.921*** (0.477) company is significantly correlated with higher PISA Hong Kong SAR, China 3.954*** scores in 8 of the 11 economies analyzed (Hong (1.511) Kong SAR, China; Korea; Macao SAR, China; Malaysia 7.319*** Malaysia; Singapore; Taiwan, China; Thailand; and (0.989) Vietnam) (table B4.4.2). The results are less evident for tutoring with a personal tutor. In contrast, there Source: Data from PISA 2012 (OECD 2012b). Note: The following control variables are included in the regressions. Student is no significant impact of an additional hour study- characteristics: age; gender (1 = male); grade; and index of economic, social, ing at a commercial company or with a private tutor and cultural status. School resources: school average of the index of economic, in most Western European countries (one exception social, and cultural status; regime school (proportion of funds from public sources); rural school (1 = rural); proportion of certified teachers; teaching is Spain) or the United States. These outcomes may practices (teacher-directed instruction); disciplinary climate in the classroom; suggest that in East Asia and Pacific economies, pri- student-teacher ratio; school autonomy in curriculum; and school autonomy in resource allocation. Student motivations: index of sense of belonging at school, vate tutoring is increasing because it is effective and students who strongly agree that trying hard at school will help them to get because students and parents tend to value extra- into a good college (1 = strongly agree), and students who strongly agree that curricular classes more than school education. trying hard at school will help them to get a good job (1 = strongly agree). Parental engagement: parental pressure at school for achievement. Standard errors are in parentheses. ***p < 0.01, **p < 0.05. Source: Bray and Lykins 2012. poorer students (figure 4.15). In Vietnam, for In Cambodia, teacher salaries are low and instance, over 60 percent of children from often paid late, pushing teachers to provide poorer households use private tutors, com- private tutoring for additional income. pared to close to 90 percent of children from Prohibiting private tutoring is likely to better-off households. be ineffective; Cambodia, Korea, Myanmar, Tutoring is also a concern because it may and other countries tried and failed to do be an outcome of badly functioning school so (Bray and Lykins 2012). Private tutor- systems and poor teacher governance. ing is likely to continue to expand in S P E N D I N G P U B L I C R E S O U R C E S O N E D U C AT I O N E F F E C T I V E LY 123 many countries. Therefore, governments FIGURE 4.15 In almost all economies, wealthier students are should focus on regulating it better and more likely than poorer students to use private tutors ensuring that it does not displace classroom teaching (UNESCO 2014). 100 Private tutoring tends to have a greater impact on the academic performance of poor 90 Vietnam students than on the performance of rich stu- % of richest students in private tutoring 80 Korea, Rep. Malaysia dents. The difference is statistically significant Singapore only in China and Singapore for personal 70 Shanghai, China Indonesia tutors and in China, Singapore, and Taiwan, Hong Kong SAR, China 60 China for commercial companies, however Taiwan, China (analysis based on data from OECD 2013). 50 As an alternative to private tutoring, Macao SAR, governments of East Asia and Pacific’s 40 China Top Performing Systems provide funding Japan 30 to schools and parents to support extra- curricular activities to mitigate widening 20 gaps in student performance. In Singapore, 10 the Edusave Scheme provides Ministry of Education–funded schools (government and 0 government-aided schools and government- 0 10 20 30 40 50 60 70 80 90 100 funded independent and special education % of poorest students in private tutoring schools) with resources for school enrich- East Asia and Pacific economy Non-East Asia and Pacific OECD economy ment activities to support students’ holis- tic development and encourage students to Source: PISA 2012 (OECD 2012b). Note: OECD = Organisation for Economic Co-operation and Development. excel in both academic and nonacademic areas. The government also provides finan- help to build social capital. Japan can afford cial assistance to needy students in gov- these activities because it keeps overall edu- ernment and government-aided schools cation spending relatively low by maintain- and institutes of higher learning so that all ing comparatively large class sizes (OECD Singaporeans, regardless of their income 2012c). The effectiveness of these practices is level, can benefit from the best opportuni- consistent with the finding that “high-dosage ties in education. It also provides a subsidy tutoring” is one of the key factors contrib- for school fees to students from lower- and uting to high academic performance among middle-income families in independent charter schools in New York (Fryer and schools.9 Dobbie 2013). In Korea, where 77 percent of students in primary and secondary schools have private tutors, the government provides additional Conclusions and public resources to schools to offer extra recommendations instruction after school and financial support East Asia and Pacific’s Top Performing and to poor parents to make private tutors more Above-Average Performing Systems adopted affordable (OECD 2012a). three principles for effective education In Japan, the central and local govern- spending: ments provide subsidies to students with financial needs, students with disabilities, and • Prioritize public spending on basic students living in remote areas. It also pro- education vides teachers with ample time for work other • Manage essential inputs efficiently than teaching. Teachers use nonteaching time • Enhance the equitable distribution of on activities that support weaker students and resources. 124 GROWING SMARTER Policy makers in other economies in the student performance in many East Asia and region and beyond may wish to consider Pacific countries. They keep schools simple adopting these principles to improve the but functional. In countries where access to effectiveness with which they spend public primary and secondary education is still an resources (table 4.4). issue, large-scale investment in school facili- ties has resulted in improving educational attainment and student performance. Principle 1: Prioritize public spending P o l i c y m a k e r s i n B e l o w - Av e r a g e on basic education Performing Systems and Emerging Systems Top Performing and Above-Average should review their teacher compensation Performing Systems in East Asia and Pacific and career development policies holistically sequenced investments from basic to ter- to attract and retain qualified teachers. They tiary education. During their early economic should also prioritize simple but functional growth period, they prioritized government school buildings and basic school facilities investment in lower levels of education to to meet increasing demand for access. Such build a strong human capital foundation. Even capital investment will have to be accom- after they began increasing investment in ter- panied by other reforms, including training tiary education, they continued increasing per and balanced deployment of teachers, timely student spending at the lower levels of educa- provision of textbooks, and curriculum tion to enhance its quality. Even in Japan and development. Korea, where private spending on education is substantial, primary and lower-secondary Principle 3: Enhance the equitable (and to a lesser extent upper-secondary) edu- distribution of resources cation is mainly publicly funded. Below-Average Performing Systems and Top Performing and Above-Average Emerging Systems have not sequenced the Performing Systems not only perform well, focus of their spending from primary to but they also achieve consistently more equi- tertiary education, and they have not pri- table learning outcomes than most other sys- oritized investment in lower levels of educa- tems. The central government plays a key tion. Countries such as Indonesia, Lao PDR, role in equalizing education funding across Malaysia, and the Philippines could consider the country. At the same time, although the reallocating public resources to lower levels government is still the dominant financier for of education; per student spending on tertiary basic education, many countries, including education in these countries seems too high China, Indonesia, Korea, and the Philippines, relative to spending on lower levels of educa- have promoted private funding to increase tion. Other Emerging Systems have focused access and narrow gaps between the poor on investing in basic education. They should and rich. Ever-increasing private tutoring is a continue doing so to establish strong basic concern for widening inequity in student per- education systems. formance, but it can support disadvantaged students. As an alternative to private tutor- ing, governments of Top Performing Systems Principle 2: Manage essential inputs in East Asia and Pacific provide funding to efficiently schools and parents to support extracurricu- Top Performing and Above-Average lar activities and to mitigate widening gaps in Performing Systems in East Asia and Pacific student performance. manage two essential inputs efficiently: teach- Where equity in access to good-quality ers and school infrastructure. They offer education is an issue, governments should competitive salaries and benefits to attract consider not only increasing scholarships and retain qualified teachers, but higher sala- for disadvantaged students and schools ries and more teachers do not lead to higher but also mobilizing the private sector by S P E N D I N G P U B L I C R E S O U R C E S O N E D U C AT I O N E F F E C T I V E LY 125 TABLE 4.4 Steps toward achieving more effective spending of public resources on education Action Below-Average Above-Average Policy Emerging Systems Performing Systems Performing Systems Top Performing Systems Public Prioritize public spending on basic Continue investing in basic education Continue investing in basic and upper- spending education and ensure high completion to ensure high completion while secondary education while diversifying of primary and lower-secondary school. expanding access to upper-secondary funding for vocational, technical, and school and above. tertiary education. Increase starting teacher salaries to attract qualified teachers. Strengthen teacher compensation Strengthen teacher career development policies to encourage good policies and nonfinancial benefits for Build simple but functional school performance and retain qualified teachers to continue improving their facilities to increase access to basic teachers. quality. education. Build simple but functional school Provide remedial measures such as Provide funding to support facilities to ensure access to basic extracurricular activities to enhance disadvantaged students in accessing education and increase access to learning of disadvantaged students. basic education. preprimary and upper-secondary education. Help disadvantaged students to access basic education; consider mobilizing the private sector to increase access to preprimary and upper-secondary education. creating a conducive environment for private 3. The data are not shown in figure 4.9 because investment. the source and definition for Vietnam are This chapter explores the effectiveness of different. education financing among East Asia and 4. School infrastructure includes classroom- level infrastructure and other classroom Pacific’s high-performing systems and identi- characteristics (natural light, temperature, fies the principles for effective spending com- acoustics) as well as school-level monly observed in those countries. Analysis infrastructure, which includes school utilities in other countries requires better student (availability of electricity, potable water, assessment data. and the condition of the building) and other features of the school (library, computer lab, science labs). Notes 5. This study is an update of Glewwe and others 1. Calculations are based on data from the (2011). Singapore government (https://data.gov.sg/). 6. World Bank (2017b) discusses some caveats 2. The shares ranged from 85 percent (Japan) to in implementation of the formula. Chingos 98 percent (Hungary) in primary education; and Blagg (2017) describe implementation from 86 percent (Estonia) to 98 percent challenges of formula funding in the (Austria, Belgium, Hungary, Italy, and the United States, where many states that have United Kingdom) in secondary education; and progressive funding formulas on paper do not from 74 percent (Luxembourg) to 97 percent achieve equity in practice, partly because it is (Argentina, Denmark, Finland, and Sweden) difficult to segregate and target students by in tertiary education. income. 126 GROWING SMARTER 7. Department of Education of the Philippines Teachers’ Salary Subsidy Programs in Junior (2017a, 2017b). The Senior High School High School Effective School Year 2017– Voucher Program subsidizes senior high 2018.” DepEd Orders DO 20, Department of school students enrolled in non–Department Education, Manila. http://www.deped.gov.ph of Education senior high schools, including /orders. private high schools, private colleges and ———. 2017b. “Guidelines on the Implementation universities, local universities and colleges, of the Senior High School Voucher Program state universities and colleges, and technical Effective School Year 2017–2018.” DepEd and vocational institutions. Orders DO 19, Department of Education, 8. See chapter 6 for a more detailed discussion of Manila. http://www.deped.gov.ph/orders. how Japan, Korea, and Singapore mobilized Duflo, E. 2001. “Schooling and Labor Market the private sector for preprimary education. Consequences of School Construction in 9. For the Edusave Scheme, see https://www Indonesia: Evidence from an Unusual Policy .moe.gov.sg/education/edusave; for the Experiment.” American Economic Review Financial Assistance Scheme, see https://www 91 (4): 795–813. http://www.jstor.org/stable .moe.gov.sg/education/financial-assistance. /2677813. Fryer, R. G. Jr., and W. Dobbie. 2013. “Getting Beneath the Veil of Effective Schools: Evidence Referencesa from New York City.” American Economic Barrette, P., Y. Zhang, F. Davies, and L. Barrett. 2015. Journal: Applied Economics 5 (4): 28–60. “Clever Classrooms: Summary Report of the Glewwe, P. W., E. A. Hanushek, S. D. Humpage, HEAD Project (Holistic Evidence and Design).” and R. Ravina. 2011. “School Resources University of Salford, Manchester, U.K. and Educational Outcomes in Developing Blackburn, K. 2017. Education, Industrialization Countries: A Review of the Literature from and the End of Empire in Singapore. New 1990 to 2010.” NBER Working Paper 17554, York: Routledge. National Bureau of Economic Research, Bray, M., and C. Lykins. 2012. Shadow Cambridge, MA. http://www.nber.org/papers Education: Private Supplementary Tutoring /w17554.pdf. and Its Implications for Policy Makers in Godo, Y., and Y. Hayami. 2009. “The Human Asia. Mandaluyong City: Asian Development Capital Basis of the Japanese Miracle: A Bank. https://www.adb.org/sites/default/files Historical Perspective.” Paper presented at /publication/29777/shadow-education.pdf. “Festschrift Workshop for Prof. Hayami,” Cambodia Ministry of Economy and Finance. Tokyo, February 27–28. http://www3.grips 2017. “More and Better Spending for .ac.jp/~otsuka/Workshop2009/paper/S5%20 Education Results.” Unpublished PowerPoint Godo.pdf. presentation. Government of the Republic of the Union of Chingos, M. M., and K. Blagg. 2017. Do Poor Myanmar. 2014. Education for All 2015 Kids Get Their Fair Share of School Funding? National Review: Myanmar. http://unesdoc Washington, DC: Urban Institute. http://www .unesco.org/images/0022/002297/229723E.pdf. .urban.org/sites/default/files/publication/90586 Govtech Singapore. 2017. “Next-Level Upgrades for /school_funding_brief.pdf. School Tech.” Govtech Singapore, July 13. https:// Cuesta, A., P. Glewwe, and B. Krause. 2016. www.tech.gov.sg/TechNews/Upclose/2017/07 “School Infrastructure and Educational /Next-level-upgrades-for-school-tech. Outcomes: A Literature Review, with Special Heckman, J. J., and D. V. Masterov. 2007. “The Reference to Latin America.” Economia Productivity Argument for Investing in Young 17 (1): 95–130. Children.” Review of Agricultural Economics Department of Education of the Philippines. 2016. 29 (3): 446–93. http://jenni.uchicago.edu “Voucher Program.” http://www.deped.gov / papers/Heckman_Masterov_RAE _2007 .ph/k-to-12/faq/voucher-program. _v29_n3.pdf. ———. 2017a. “Guidelines on the Implementation Heng, S. K. 2011. “Singapore Education Reform of the Educational Service Contracting and Journey: Keynote Address.” Address given by the Singapore minister of education at the 14th a. Titles of publications that include “Korea” and “South Korea” OECD Japan Seminar, “Strong Performers, refer to the Republic of Korea. Successful Reformers in Education,” Tokyo, S P E N D I N G P U B L I C R E S O U R C E S O N E D U C AT I O N E F F E C T I V E LY 127 June 28. http://www.google.com/url?sa=t&rct Individual Basic Education School Data.” =j&q=&esrc=s&source=web&cd=8&ved=0ahU Unpublished research paper 20150221, Tokyo KEwjW8ITSltfVAhVGLSYKHef5AEwQFghDM Institute of Technology, Tokyo. Ac&url=http%3A%2F%2Fwww.oecd.org%2F National Bureau of Statistics of China. 2016. pisa%2Fpisaproducts%2F48357652.ppt&usg= China Statistical Yearbook 2016. Beijing: AFQjCNEDkAAPG4Ivh3W_1jrw27icp5BORA. National Bureau of Statistics. http://www.stats IFIC (Institute for International Cooperation) .gov.cn/tjsj/ndsj/2016/indexeh.htm. and JICA (Japan International Cooperation National Center on Education and the Economy. n.d. Agency). 2004. “Part I. Overview of the “Singapore: System and School Organization.” History of Japan’s Education.” In The History National Center on Education and the of Japan’s Educational Development: What Economy, Washington, DC. http://ncee.org Implications Can Be Drawn for Developing /what-we-do/center-on-international-education Countries Today. Tokyo: Japan International -benchmarking/top-performing-countries Cooperation Agency. https://www.jica .go /singapore-overview-2/. .jp/jica-ri/IFIC_and_JBICI-Studies/english National People’s Congress. 2012. “Law of the /publications/reports/study/topical/educational People’s Republic of China on the Promotion /index.html. of Privately-Run Schools.” Adopted at the International Commission on Financing 31st Meeting of the Standing Committee of Global Education Opportunity. 2016. the Ninth National People’s Congress, Beijing, The Learning Generation: Investing in December 28. http://www.npc.gov.cn/englishnpc Education for a Changing World. New /Law/2007-12/06/content_1382110.htm. York: Education Commission. http://report Neuman, M. J., and A. E. Devercelli. 2013. .educationcommission.org/wp-content “What Matters Most for Early Childhood /uploads/2016/09/Learning_Generation_Full Development: A Framework Paper.” World _Report.pdf. Bank, Washington, DC. Jimenez, E., V. Nguyen, and H. A. Patrinos. OECD (Organisation for Economic Co-operation 2012. “Stuck in the Middle? Human Capital and Development). 2003. Education Development and Economic Growth in Malaysia at a Glance 2016: OECD Indicators . and Thailand.” Policy Research Working Paper Paris: OECD. http://www.oecd.org/site 6283, World Bank, Washington, DC. http:// /worldforum/33703760.pdf. dx.doi.org/10.1596/1813-9450-6283. ———. 2006. PISA Science Competencies for Kim, G. 2002. “Education Policies and Reform Tomorrow’s World: Volume I Analyses. http:// in South Korea.” In Secondary Education www.oecd.org/education/school / program in Africa: Strategies for Renewal , 29–40. meforinternationalstudentassessment pisa Africa Region Human Development /pisa2006results.htm. Working Paper, World Bank, Washington, ———. 2009. PISA 2009 Key Findings. Paris: DC. http://siteresources .worldbank OECD Publishing. http://www.oecd.org/pisa . o r g / I N TA F R R E G T O P E D U C AT I O N /pisaproducts/pisa2009keyfindings.htm. /Resources/444659-1220976732806/Secondary ——— . 2 0 1 1 . S t r o n g P e r f o r m e r s a n d _Education_Strategies_renewal.pdf. Successful Reformers in Education: Lessons Liang, X., H. Kidwai, and M. Zhang. 2016. How from PISA for the United States. Paris: Shanghai Does It: Insights and Lessons from OECD Publishing. http://www.oecd.org the Highest-Ranking Education System in the /pisa/46623978.pdf. World. Washington, DC: World Bank. ———. 2012a. Equity and Quality in Education: MacEwan, P. J. 2015. “Improving Learning in Supporting Disadvantaged Students and Primary Schools of Developing Countries: A Schools . Paris: OECD Publishing. http:// Meta-Analysis of Randomized Experiments.” asiasociety.org/files/oecd-0512report.pdf. Review of Educational Research 85 (3): 353– ———. 2012b. PISA 2012 [database]. Paris: 94. http://academics.wellesley.edu/Economics OECD. http://www.oecd.org/pisa/pisaproducts /mcewan/PDF/meta.pdf. /pisa2012database-downloadabledata.htm. Muta, H. 2015. “Regional Disparity of ———. 2012c. Strong Performers and Successful Educational Conditions in Townships and the Reformers in Education: Lessons from PISA Effect of the Policy on Daily Wage Temporary for Japan. Paris: OECD Publishing. http:// Teachers: Statistical Analyses Based on dx.doi.org/10.1787/9789264118539-en. 128 GROWING SMARTER ———. 2013. Education at a Glance 2013: OECD -conditions-and-recommendations-for-the Indicators. Paris: OECD. https://www.oecd -future. .org/edu/eag2013%20(eng)--FINAL%20 Simon, K. 2015. “Private Schools, on the Rise in 20%20June%202013.pdf. China, Are Now in Beijing’s ‘Crosshairs’.” Asia ———. 2014. PISA 2012 Results: What Students Society (blog), March 13. http://asiasociety.org Know and Can Do: Student Performance in /blog/asia/private-schools-rise-china-are-now Mathematics, Reading and Science (Volume I). -beijings-crosshairs. Paris: OECD Publishing. http://www.oecd.org Singapore Ministry of Education. 2017. Education /pisa/keyfindings/pisa-2012-results-volume-i [database]. https://data.gov.sg/group/education. .htm. Stern, J. M. B., and T. M. Smith. 2015. “Private ———. 2015. Education Policy Outlook: Japan. Secondary Schools in Indonesia: What Is http://www.oecd.org/education/policyoutlook Driving the Demand?” International Journal .htm; https://www.oecd.org/edu/Japan of Educational Development 46: 1–11. https:// -country-profile.pdf. doi.org/10.1016/j.ijedudev.2015.11.002. ———. 2016a. Education at a Glance 2016: Suryadarma, D., A. Suryahadi, S. Sumarto, and OECD Indicators. Paris: OECD Publishing. F. H. Rogers. 2006. “Improving Student http://www.oecd.org/edu/education-at-a Performance in Public Primary Schools -glance-19991487.htm. in Developing Countries: Evidence from ———. 2016b. Education in China: A Snapshot. Indonesia.” Education Economics 14 (4): Paris: OECD Publishing. 401–29. http://www.tandfonline.com/doi ———. 2016c. Education Policy Outlook: Korea. /pdf/10.1080/09645290600854110. Paris: OECD Publishing. www.oecd.org U N ( U n i t e d N a t i o n s ) . 2 0 1 7 . Wo r l d /education/policyoutlook.htm. Population Prospects 2017. New York: UN. ———. 2016d. PISA 2015 Results. Paris: OECD https://esa.un.org/unpd/wpp/Download Publishing. https://www.oecd.org/pisa/. /Standard/Population/. OECD and UNESCO (United Nations UNESCO (United Nations Educational, Scientific Educational, Scientific, and Cultural and Cultural Organization). 2012. Global Organization). 2016. Education in Education Digest 2012: Opportunities Thailand: An OECD-UNESCO Perspective, Lost: The Impact of Grade Repetition and Reviews of National Policies for Education, Early School Leaving. Montreal: UNESCO OECD Publishing, Paris. http://dx.doi Institute for Statistics. https://reliefweb.int/sites .org/10.1787/9789264259119-en. /reliefweb.int/files/resources/global-education Papua New Guinea Department of Education. -digest-opportunities-lost-impact-grade 2016. National Education Plan 2015–2019: -repetition-early-school-leaving-2012-en.pdf. Quality Learning for All . Port Moresby: ———. 2014. EFA Global Monitoring Report Department of Education. http://www 2013/4, Teaching and Learning: Achieving .education.gov.pg/documents/NEP_2015 Quality for All. Paris: UNESCO. -2019_DoE.pdf. ———. 2015a. Education for All 2015 Park, H. 2013. Re-evaluating Education in National Review Report: Malaysia . Paris: Japan and Korea: De-mystifying Stereotypes. UNESCO. http://unesdoc.unesco.org London: Routledge. /images/0022/002297/229719E.pdf. Psacharopoulos, G., C. Montenegro, and ———. 2015b. Education for All 2015 National H. A. Patrinos. 2017. “Education Financing Review Report: Papua New Guinea. Paris: Priorities in Developing Countries.” Journal UNESCO. http://unesdoc.unesco.org of Educational Planning and Administration /images/0023/002316/231679e.pdf. 31 (1): 5–16. https://harrypatrinos.files ———. 2016. Global Education Monitoring .wordpress.com/2015/03/psach-monte-patr-ed Report 2016: Education for People and -financing-002.pdf. Planet: Creating Sustainable Futures for All. Santibañez, L. 2014. School-Based Management Paris: UNESCO. http://unesdoc.unesco.org in Lao P.D.R.: Current Conditions and /images/0024/002457/245752e.pdf. Recommendations for the Future. Washington, Wong, A. 2017. “Insights from East Asia’s High- DC: World Bank. http://documents.worldbank Performing Education Systems: Leadership, .org/curated/en/258391468263928480/School Pragmatism, and Continuous Improvement.” -based-management-in-Lao-P-D-R-current Background paper for World Bank flagship S P E N D I N G P U B L I C R E S O U R C E S O N E D U C AT I O N E F F E C T I V E LY 129 report, Growing Smarter: Learning and World Bank. http://documents.worldbank Equitable Development in East Asia and .org/curated / en/504121467987907393 Pacific, World Bank, Washington, DC. /pdf/103993-WP-P132668-PUBLIC-Myanmar Woolner, P., E. Hall, S. Higgins, C. McCaughey, and -PER-Dec-2015.pdf. K. Wall. 2007. “A Sound Foundation? What ———. 2015b. Teacher Certification and We Know about the Impact of Environments Beyond: An Empirical Evaluation of the on Learning and the Implications for Building Teacher Certification Program and Education Schools for the Future.” Oxford Review Quality Improvements in Indonesia. Jakarta: of Education 33 (1): 47–70. http://dx.doi World Bank. http://documents.worldbank .org/10.1080/03054980601094693. .org/curated/en/129551468196175672 World Bank. 2009. Investing in Indonesia’s /pdf/104599-WP-P102259-PUBLIC-Teacher Education at the District Level: An Analysis -Certification-and-beyond-final.pdf. of Regional Public Expenditure and Financial ———. 2015c. Thailand: Wanted, a Quality Management. Washington, DC: World Bank. Education for All. Washington, DC: World ———. 2013a. Malaysia Economic Monitor: Bank. https://openknowledge.worldbank.org High-Performing Education. Bangkok: World /handle/10986/22355. Bank. http://documents.worldbank.org/curated ———. 2016. Assessing Basic Education Service /en/391841468051873989/pdf/832910WP0P1 Delivery in the Philippines: Public Education 4640Box0382083B00PUBLIC0.pdf. Expenditure Tracking and Quantitative ———. 2013b. Spending More or Spending Service Delivery Study . Washington, DC: Better: Improving Education Financing in World Bank. http://documents.worldbank Indonesia. Jakarta, Indonesia: World Bank. .org/curated / en/507531468325807323/pdf https://openknowledge.worldbank.org /AUS6799 -REVISED-PH-PETS-QSDS-Final /handle/10986/13210. -Report.pdf. ———. 2014a. Educating the Next Generation: ———. 2017a . “ Kingdom of Cambodia: Improving Teacher Quality in Cambodia. Secondary Education Improvement Project Washington, DC: World Bank. https://open A p p r a i s a l D o c u m e n t . ” Wo r l d B a n k , knowledge.worldbank.org / handle/109 86 Washington, DC. /21002. ———. 2017b. Vietnam: Public Expenditure ———. 2014b. The Independent State of Papua Review 2016: Spending for Results . New Guinea: From Expanded Access to Washington, DC: World Bank. Better Quality; A Call for Consolidation of ———. Various years. Education Statistics Education Investment; A Public Expenditure (EdStats) [database]. Washington, DC: Review of the Education Sector. Washington, World Bank. http://datatopics.worldbank.org DC: World Bank. /education/. ———. 2015a. Myanmar: Public Expenditure Review; ———. Various years. World Development Realigning the Union Budget to Myanmar’s Indicators [database]. Washington, DC: Development Priorities. Yangon, Myanmar: World Bank. SPOTLIGHT 3 How Have Governments Used Public-Private Partnerships to Improve Education Outcomes? Economies in the East Asia and Pacific region Gross preprimary enrollment rates are have embraced public-private partnerships in relatively high in Malaysia and Vietnam, education and used them to varying degrees despite only modest private enrollment (box S3.1). Private provision has increased in (lower-right quadrant). In contrast, prepri- recent years in many countries in the region, mary enrollment rates remain low in many especially in preprimary education, where 54 Below-Average Performing Systems (upper- percent of preprimary students were enrolled in and lower-left quadrants). In Indonesia, for private institutions in 2014 (World Bank, vari- instance, almost all preprimary students ous years). Although the public sector remains are enrolled in private schools, and the the principal provider of primary and second- gross enrollment rate is below 60 percent. ary education in most of the region, the num- In Cambodia, the Lao People’s Democratic ber of primary and secondary students enrolled Republic, and Timor-Leste, both private in private institutions is growing (table S3.1). enrollment and total enrollment are low. This spotlight provides examples of how These data suggest that the private sector economies—especially those in East Asia and helps to increase preprimary enrollments, but Pacific—have used public-private partnerships that it alone cannot guarantee access for all, to improve their education systems. particularly in lower-income countries. The Korean case illustrates the importance of public investment at least at an initial stage. Does private enrollment The rapid growth of public kindergartens increase access? in 1982 contributed to the rapid growth of The role of private enrollment in increasing enrollment in the 1980s (figure S3.2) (Kim access differs at different levels of education. and Na 2003). East Asia and Pacific’s high-performing sys- Figure S3.3 presents the relationship tems except Vietnam rely heavily on the pri- between the size of private enrollment and vate sector to enroll preprimary education gross enrollment rate at the primary, lower- and keep enrollment rates high (figure S3.1). secondary, and upper-secondary levels. 131 132 GROWING SMARTER BOX S3.1 What are public-private partnerships in education? The concept of a public-private partnership recog- design, and textbook provision (table SB3.1.1). Each nizes the existence of options for providing educa- type of contract has different expected effects on the tion services other than public finance and public four objectives. For instance, vouchers and subsidies delivery. Governments may guide and finance the are expected to have strong effects on increasing private sector to deliver education services through enrollment, whereas private finance initiatives are different types of contracts and with different objec- expected to reduce costs. tives. Public-private partnerships in education have The World Bank’s Systems Approach for Better four main objectives: increasing enrollment (access), Education Results-Engaging the Private Sector improving education outcomes, reducing inequality, (SABER-EPS) aims to help countries to engage the and reducing costs. private sector more effectively in promoting equita- The provision of education services can range ble learning for all. SABER-EPS assesses how well a from the construction, management, or mainte- country’s policies are oriented toward ensuring that SPOTLIGHT 3 nance of infrastructure (often referred to as a private the services of nonstate providers promote learning finance initiative) to the provision of education ser- for all with regard to four policy goals: encouraging vices and operations. Contracts for education-related innovation by providers; holding schools account- services can cover a range of services and inputs, able; empowering all parents, students, and commu- including the private management of public schools; nities; and promoting diversity of supply (see World subsidies and vouchers; private finance initiatives for Bank 2014 for details). school construction and maintenance; and profes- sional services such as teacher training, curriculum Source: Patrinos, Barrera-Osorio, and Guáqueta 2009. TABLE SB3.1.1 Public-private partnerships in education What governments contract for What governments buy Management, professional, support services (input) • School management (financial and human resources management) • Support services (meals and transportation) • Professional services (teacher training, curriculum design, textbook delivery, quality assurance, and supplemental services) Operational services (process) • Education of students, financial and human resources management, professional services, and building maintenance Education services (outputs) • Student places in private schools Facility availability (inputs) • Infrastructure and building maintenance Facility availability and education services • Infrastructure combined with services (operational or educational (both inputs and outputs) outputs) Source: Patrinos, Barrera-Osorio, and Guáqueta 2009. TABLE S3.1 The share of private enrollment has changed over the last two decades, 1990–2014 Percent Preprimary Primary Secondary Economy 1990 2000 2014 1990 2000 2014 2000 2014 Myanmar — — 82.1 — — 0.2 — 1.3 Vietnam — 51.1 11.8 — 0.3 0.6 — — Japan 77.4 65.3 72.7 0.7 0.9 1.2 18.4 19.7 Korea, Rep. 64.6 — 80.5 1.4 1.5 1.6 39.6 31.0 Finland — 9.8 8.5 0.9 1.1 1.7 7.8 14.2 Cambodia — 22.5 13.4 — 1.6 2.6 0.5 — Lao PDR — 16.6 21.1 — 2.0 4.5 0.9 3.0 table continues next page H O W H AV E G O V E R N M E N T S U S E D P U B L I C  P R I VAT E PA R T N E R S H I P S TO I M P R O V E E D U C AT I O N O U TCO M E S ? 133 TABLE S3.1 The share of private enrollment has changed over the last two decades, 1990–2014 (continued) Percent Preprimary Primary Secondary Economy 1990 2000 2014 1990 2000 2014 2000 2014 Poland 0.0 3.9 19.8 — — 4.5 — 11.2 Mongolia — 3.7 7.1 — 0.9 5.7 0.3 — Philippines 53.0 48.5 — 6.6 7.2 8.3 — 19.4 East Asia and Pacific average 17.2 30.4 53.7 3.6 4.4 8.7 — 17.1 Upper-middle-income average 8.0 17.0 36.6 3.1 4.5 9.2 8.5 10.8 OECD average 31.7 31.9 36.8 8.8 9.5 10.0 14.5 17.5 Timor-Leste — — 46.8 — — 12.9 — 23.3 Low-income average 31.5 38.1 34.5 13.3 14.0 13.9 19.7 18.7 Malaysia 59.9 48.1 44.5 0.3 — 13.9 5.8 13.4 SPOTLIGHT 3 United Kingdom 6.0 6.0 45.2 4.8 4.7 14.0 26.0 69.5 Denmark 9.0 — 17.4 9.7 10.8 15.3 11.5 13.3 Indonesia 99.6 99.2 96.8 17.5 15.7 18.4 — 43.4 Colombia — 40.8 28.0 15.2 18.7 18.7 30.1 20.3 Middle-income average 18.2 25.8 40.6 8.3 9.6 18.7 20.1 27.2 Hong Kong SAR, China 100.0 — 99.0 9.6 — 18.9 — 18.0 Thailand 23.7 — 30.4 9.5 13.1 20.0 — 11.0 Brunei Darussalam — 61.5 77.2 — 35.1 39.2 10.7 16.8 Chile 47.7 — 67.2 — — 61.5 — 61.0 Macao SAR, China — 91.7 97.5 — 94.2 97.7 93.4 95.7 Source: World Bank EdStats (World Bank, various years). Note: Countries are ordered according to the percentage of primary school enrollment in private schools in 2014. Data for 2014 include latest data since 2011. — = not available; OECD = Organisation for Economic Co-operation and Development. It suggests that the role of private provision FIGURE S3.1 There is little correlation in East Asia and Pacific in increasing access is not important at the between the share of private enrollment and the gross enrollment primary level but is important at the upper- rate for preprimary education secondary level and, to a less extent, the Vanuatu 100 Tonga Samoa Indonesia Macao lower-secondary level. SAR, China 90 Hong Kong SAR, In Thailand, the National Education Act Korea, Rep. China 80 Myanmar obligated the state to provide general subsi- Brunei Darussalam in private institutions Japan dies for basic 12-year education, whether it 70 % of enrollment East Asia and is provided by the state or by private insti- 60 Pacific average tutions (Australian Education International 50 China Middle-income 2013). Private schools providing basic edu- 40 Timor-Leste average Malaysia cation can qualify for state-subsidized loans 30 Low-income average Thailand Solomon Islands to build new school buildings or renovate 20 Lao PDR Palau Micronesia, old ones. The government also has a revolv- 10 Cambodia Fed. Sts. Vietnam Tuvalu ing fund for private schools, which offers 0 loans at 4 percent interest with a repayment 0 10 20 30 40 50 60 70 80 90 100 110 120 period of 10–15 years to schools that can Gross enrollment rate (%) provide collateral (Patrinos, Barrera-Osorio, Top Performing Systems and Guáqueta 2009). Private education has Above-Average Performing Systems played an important role in delivering basic Below-Average Performing Systems Emerging Systems as well as other levels of education. It also Country group average has contributed to reducing public spend- Median ing on education (Pinyakong, Virasilp, and Source: World Bank EdStats (World Bank, various years). Somboon 2007). Note: Gross enrollment rate can exceed 100 percent due to over- and underage students. 134 GROWING SMARTER FIGURE S3.2 The private sector helped to increase preprimary enrollments, but it alone cannot guarantee access, particularly in lower-income countries a. Indonesia b. Republic of Korea 100 100 90 90 80 80 Enrollment rate (%) Enrollment rate (%) 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 80 82 84 86 88 90 92 94 96 98 00 02 04 06 08 10 12 14 80 82 84 86 88 90 92 94 96 13 15 SPOTLIGHT 3 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 19 19 19 19 19 19 19 19 19 20 20 c. Myanmar d. Samoa 100 100 90 90 80 80 Enrollment rate (%) Enrollment rate (%) 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 92 94 96 98 00 02 04 06 08 10 12 14 06 07 08 09 10 11 12 13 14 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 % of enrollment in private institutions Gross enrollment rate (%) Source: World Bank EdStats (World Bank, various years). Does private provision of private school students in Indonesia signifi- education improve learning cantly underperformed public school stu- outcomes? dents on the Programme for International The evidence is mixed on the effect of private Student Assessment (PISA) 2000, although provision of education on learning outcomes the difference was negligible after control- in East Asia and Pacific. These results con- ling for student and school characteristics trast with the positive correlation observed (OECD 2005).1 across the world (Patrinos, Barrera-Osorio, Mixed outcomes are observed in the and Guáqueta 2009). Republic of Korea. One study finds that stu- A study of Indonesia finds that gradu- dents in private foreign language high schools ates of public schools score higher on outperformed students in general private and the national exit exam than graduates of public high schools on the College Scholastic private schools, controlling for a wide Aptitude Test of Korean, English, and math- variety of student characteristics and fam- ematics (Kim 2012). After controlling for ily background (Hendajany 2016). An prior achievement, the achievement gap Organisation for Economic Co-operation decreased significantly in English and Korean and Development (OECD) study finds that scores, although it held up in math scores. H O W H AV E G O V E R N M E N T S U S E D P U B L I C  P R I VAT E PA R T N E R S H I P S TO I M P R O V E E D U C AT I O N O U TCO M E S ? 135 FIGURE S3.3 Private enrollment is not correlated with access at the primary level, but it is correlated at the upper-secondary and, to a less extent, lower-secondary levels a. Primary b. Lower-secondary c. Upper-secondary 60 60 60 % of enrollment in private institutions % of enrollment in private institutions % of enrollment in private institutions Indonesia 50 50 50 Korea, Rep. 40 Brunei Darussalam 40 40 Samoa Indonesia Hong Kong SAR, Solomon Islands Japan 30 Micronesia, 30 30 Fed. Sts. Solomon Islands China Timor-Leste Solomon Islands Philippines Thailand Middle-income average Philippines Korea, Rep. 20 Thailand Indonesia East Asia 20 Timor-Leste 20 Malaysia Brunei Hong Kong Samoa Pacific Timor-Leste Brunei Malaysia Low-income and average Samoa Darussalam Papua Darussalam SAR, China 10 average 10 China 10 New Guinea Mongolia Mongolia Philippines Malaysia Japan Thailand China China Lao PDR Cambodia Myanmar Korea, Rep. Lao PDR Mongolia SPOTLIGHT 3 Cambodia Myanmar Vietnam Lao PDR 0 0 0 Myanmar Japan Vietnam 50 60 70 80 90 100 110 120 130 0 20 40 60 80 100 120 140 90 100 110 120 130 140 Gross enrollment rate (%) Gross enrollment rate (%) Gross enrollment rate (%) Top Performing Systems Emerging Systems Above-Average Performing Systems Country group Below-Average Performing Systems Source: World Bank EdStats (World Bank, various years). Note: For most countries, data are for 2014 or the closest year. Private provision in Macao SAR, China, is almost 100 percent. It is omitted from the graph because including it would have made it difficult to display the other data. Gross enrollment rate can exceed 100 percent due to over- and underage students. Private high school students scored higher Do public-private partnerships in than public high school students in Korean education reduce inequity? even after controlling for student-school backgrounds and prior achievement; their Public funding for privately managed schools scores were not higher in English or math. may have helped to reduce inequity in East For lower grades, the results of the 2015 Asia and Pacific’s Top Performing Systems. Pacific Islands Literacy and Numeracy Korea has used private schools to address Assessment (PILNA)—which measures lit- between-school differences in overall levels of eracy and numeracy skills of students who student ability and the quality of high school have completed four and six years of for- education by randomly assigning students to mal primary education in 13 countries in general high schools, including both public the Pacific—present mixed results as well and private schools. Private schools are not (figure S3.4). In the year 4 numeracy assess- allowed to charge tuition fees; the govern- ment, students in nongovernment schools ment subsidizes 95 percent of their costs. performed better than students in government Analysis of the effect of random assignments schools (fewer low scores and more high on academic achievement shows that, in areas scores), but this difference was reversed in the where students were randomly assigned, top two highest proficiency levels in year 6. school effects explain only 0.5 percent of the On both the year 4 and year 6 literacy assess- observed variation in academic achievement. ment, students in nongovernment schools had By contrast, in areas where students were not slightly higher mean performance in literacy randomly assigned, school effects account than students in government schools across for more than 30 percent of the variation in all proficiency levels. test scores (Bastos, Cristia, and Kim 2016). 136 GROWING SMARTER FIGURE S3.4 Regional assessments in the Pacific show an unclear impact of the school authority (governmental, nongovernmental) on students’ numeracy and literacy proficiency 25 a. Numeracy, year 4 25 b. Numeracy, year 6 20 20 % of students scoring % of students scoring 15 15 10 10 5 5 0 0 SPOTLIGHT 3 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 Proficiency level Proficiency level 20 c. Literacy, year 4 25 d. Literacy, year 6 20 15 % of students scoring % of students scoring 15 10 10 5 5 0 0 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 Proficiency level Proficiency level Nongovernmental Governmental Source: Adapted from EQAP 2015. Analysis of PISA scores also suggests that (Jimenez and others 2011). The program— countries with higher levels of public fund- the Education Service Contracting Program— ing for privately managed schools tend to aims to increase access to good-quality basic have less socioeconomic stratification.2 Japan education at the secondary level by extending and Korea are among the countries with low financial assistance from the public budget stratification. that allows “poor but deserving” elemen- The Philippines uses targeted vouchers to tary school graduates to attend private high help to increase access for disadvantaged stu- schools that have contracts with the gov- dents. It has one of the largest public-private ernment (Jimenez and others 2011). The partnership programs in education in the program has faced many shortcomings in world. More than 567,500 students—almost relation to equity, as grantees have tended 9 percent of the country’s 6.5 million high to come from relatively well-off households, school students—received vouchers in 2009 because poorer households are unable to pay H O W H AV E G O V E R N M E N T S U S E D P U B L I C  P R I VAT E PA R T N E R S H I P S TO I M P R O V E E D U C AT I O N O U TCO M E S ? 137 even the difference between the subsidy and 2. OECD (2012) defines stratification as the the fees that schools charge. The contract difference in socioeconomic background does not specify performance criteria for par- between students in privately and publicly ticipating private schools, such as targeting managed schools, as measured by the PISA index of economic, social, and cultural status. students most in need or requiring minimum Japan and Korea were among the countries student achievement to ensure the continua- with low stratification, with privately tion of their public funding (Aslam, Rawal, managed schools receiving 38–50 percent of and Saeed 2017; Jimenez and others 2011). their budgets from public resources. What can governments do Referencesa to promote public-private Aslam, M., S. Rawal, and S. Saeed. 2017. partnerships that lead to Public-Private Partnerships in Education in positive outcomes? Developing Countries: A Rigorous Review SPOTLIGHT 3 Evidence on the impact of public-private of the Evidence. London: Ark Education partnerships is still limited in East Asia and Partnerships Group. http://arkonline.org/sites /default/files/Ark_EPG_PPP_report.pdf. Pacific, but country experiences yield some Australian Education International. 2013. lessons: “Thailand Regulatory Fact Sheet 2013.” • Private provision and financing of educa- Australian Education International, Canberra, h t t p s : / / i n t e r n a t i o n a l e d u c a t i o n . g o v. a u tion can help to increase access where public /International-network/thailand/Publications resources are limited. In particular, in Below- /Documents/AEI%20Thailand%20Fact%20 Average Performing Systems and Emerging Sheet%20March%202014%20Updates%20 Systems, where access to preprimary edu- Word%20Version%20(2).pdf. cation is very limited, promoting public- Bastos, P., J. Cristia, and B. Kim. 2016. “Good private partnerships, with or without public Schools or Good Students: Evidence on School subsidies, is likely to increase enrollment at Effects from Universal Random Assignment that level. Private provision and financing of Students to High School.” University of are likely to help to increase enrollment in Korea Discussion Paper 1607, Institute of upper-secondary education. Economic Research, University of Korea, Seoul. • Impacts on learning outcomes in private http://econ.korea.ac.kr/~ri/WorkingPapers /w1607.pdf. schools are mixed. Incentives for improved EQAP (Educational Quality and Assessment learning outcomes, government monitor- Programme). 2015. Pacific Islands Literacy ing of outcomes, and both financial and and Numeracy Assessment (PILNA): Regional technical support for private schools are Report. Suva, Fiji: EQAP. http://www.eqap essential to produce positive outcomes. .org.fj/getattachment /work/Assessment • For vouchers and subsidies to help to /Regional/Final-Regional-Report.pdf.aspx. reduce inequity, targeting mechanisms Hendajany, N. 2016. “The Effectiveness of Public need to be well designed. Poorly targeted vs Private Schools in Indonesia.” Journal of subsidies may actually widen inequity. Indonesian Applied Economics 6 (1): 66–89. http://jiae.ub.ac.id/index.php/jiae/article /view/176. Jimenez, E., B. Hofman, E. Velez, and Notes H. A. Patrinos. 2011. Philippines—Private 1. Private schools include independent and Provision, Public Purpose: A Review of the government-dependent private schools. The Government’s Education Service Contracting OECD defines independent private schools as Program . Washington, DC: World Bank. schools that receive at least half of their funds http://documents.worldbank.org/curated from private sources; government-dependent private schools receive half or more of their a. Titles of publications that include “Korea” and “South core funding from government agencies. Korea” refer to the Republic of Korea. 138 GROWING SMARTER /en/486651468092652040/Philippines-Private Patrinos, H. A., F. Barrera-Osorio, and J. Guáqueta. -provision-public-purpose-a-review-of-the 2009. The Role and Impact of Public-Private -governments-education-service-contracting Partnerships in Education. Washington, -program. DC: World Bank. https://openknowledge Kim, K. J. 2012. The Effectiveness of School Type .worldbank.org/bitstream/handle/10986/261 on Students’ Academic Achievement: Focusing 2/479490PUB0Role101OFFICIAL0USE0O on Private High School of South Korea. PhD NLY1.pdf?sequence=1. dissertation, Pennsylvania State University. Pinyakong, K., P. Virasilp, and U. Samboon. 2007. Kim, M., and J. Na. 2003. “Early Childhood Care “Development of Private Secondary Schools and Education in Korea: National Policies and in Thailand.” United Nations Educational, Practices.” Korean Educational Development Scientific, and Cultural Organization, Institute, Seoul. International Institute for Educational OECD (Organisation for Economic Co-operation Planning, Paris. and Development). 2005. School Factors Related World Bank. 2014. “What Matters Most for to Quality and Equity: Results from PISA 2000. Engaging the Private Sector in Education: Paris: OECD Publishing. https://www.oecd.org A Framework Paper.” SABER Working SPOTLIGHT 3 / edu/school/pro gramme forinternational Paper 8, World Bank, Washington, DC. studentassessmentpisa/34668095.pdf. https://openknowledge.worldbank.org ———. 2012. Public and Private Schools: How /handle/10986/21756. Management and Funding Relate to Their Socio- ———. Various years. Education Statistics (EdStats) Economic Profile. Paris: OECD Publishing. [database]. Washington, DC: World Bank. http://dx.doi.org/10.1787/9789264175006-en. http://datatopics.worldbank.org/education. Developing Skilled Teachers and Supporting Effective Teaching 5 Teachers are a core element of East Asia clear goals, and reform in incremental steps. and Pacific’s Top Performing Systems. These They also provide comprehensive support systems have established competent, quali- that increases equity, enables teachers to fied, and motivated teaching forces that pro- develop professionally, provides the space and mote sustained learning. Elsewhere in the resources they need to do so, and motivates region, teacher reform has produced only them to be effective. mixed results. Lessons can be learned from Beyond direct support, the systems pro- both types of experiences. vide strong auxiliary elements aligned in In East Asia and Pacific’s Top Performing ways that make the teacher’s job easier. Systems, policies focus on the principle that These elements include clear learning good teachers are made, not born. Policy objectives and standards; well-aligned cur- makers consider effective teaching a process ricula that tend to focus on a few topics; that requires extensive training and practice, textbooks that are uncluttered and focused a mentality of continual learning, and a tac- on learning goals; more time for teachers to tic of constant refinement to hone the craft of prepare lessons and engage in professional teaching. These systems tend to be thought- development activities; and efficient assess- fully designed and continually enhanced ment systems. through an evolutionary process that devel- A foundational aspect of the approach ops and supports teachers throughout the to professional development is the mental- teaching life cycle and provides them with the ity of constant improvement, exemplified by tools and environment to be effective. an emphasis on in-service training. In many Teachers in Top Performing Systems ben- Western systems, teacher development often efit from coherent systems of recruitment, stagnates after a burst of improvement in the development, and support that make their initial years. In East Asia and Pacific’s Top jobs easier and promote effective teaching. Performing Systems, the mind-set is one of These systems begin by attracting top candi- continuous improvement. dates into teaching, filtering for quality, and Japan’s “lesson study” is the epitome of providing an excellent induction into the this approach, which has proliferated in profession. They have a history of preser- many East Asian systems. Lesson study brings vice institutions with initial quality control, together teachers of different levels of skills 139 140 GROWING SMARTER and experience to discuss techniques and that lower-performing students do not fall formulate sample teaching plans. The Top behind (Tucker 2016). Performing and Above-Average Performing Many of the low-performing East Asia Systems also foster extensive collaboration, and Pacific education systems, including with networks and learning communities of Cambodia, Indonesia, Myanmar, and the teachers. Policies encourage collaboration, Philippines, have tackled teacher reform with providing space for teachers to work together varying degrees of success. Comprehensive and to support one another. reform requires a sober assessment of capac- Another feature of high-performing ity and constraints, areas of high impact and systems is that teaching is viewed as an “open- with “leapfrog” potential, and the steps to be door” profession. Hong Kong SAR, China; performed. Japan; the Republic of Korea; Shanghai, China; and Singapore have a culture and accompanying policies that encourage obser- Coherent systems of teacher vation and critiquing of classroom prac- recruitment, development, and tices (see, for example, Choy 1996; Doig support and Groves 2011; Economist 2016; Liang, When policies are developed without a Kidwai, and Zhang 2016). blueprint or overall vision, they tend to The practices and aspects of high- be misaligned (box 5.1) or even contradic- performing classrooms offer important les- tory. Policies may be a reaction to pressures sons regarding teaching effectiveness. Good and demands from stakeholders, including teachers take a focused learning outcome– teachers themselves, or budget constraints oriented approach, through a combination may allow for only pieces of a blueprint of meticulous preparation and purposeful to be implemented. These issues may stem execution. Before even entering the class- from constant changes in governments room, good teachers have clear, concise and education ministry officials, which do learning goals and sequential steps that are not allow an overall vision to be concret- heavily focused on ensuring that students ized and to prosper. Such disjointed policy understand by elaborating and directing development can create friction, inefficien- them to underlying principles, which devel- cies, and frustration. ops habits of mind. Lessons are constantly High-performing East Asian systems have refined and improved. They are executed generally developed coherent teacher poli- using practices that stimulate learning cies, driven by clear goals and vision from an (Echazarra and others 2016) and ensure early stage. Although they have by no means BOX 5.1 Policy misalignment hurt Indonesia’s education system and teachers Indonesia’s rollout—and eventual rollback—of also a major issue. The new curriculum required a very the 2013 curriculum presents a cautionary tale of different role for teachers and the approach to teaching misalignment. With pressure to implement the new as well as content. But only one or two weeks of train- curriculum before a change of government, the govern- ing were given, and the curriculum was rolled out in a ment announced and developed the new curriculum cascaded manner that did not reach many teachers. hurriedly, without consulting with key stakeholders. The combination of misalignment of curriculum Rapidly developed and printed textbooks were mis- and textbooks, missing textbooks and materials, and aligned with the curriculum content and of low quality. inadequate training created confusion and frustra- In addition, they were not available for many schools tion among teachers. In 2015, the curriculum was and teachers at the time of the rollout. Training was rolled back. DEVELOPING SKILLED TEACHERS AND SUPPORTING EFFECTIVE TEACHING 141 been free of challenges, most have seen fewer East Asia’s Top Performing Systems have radical shifts or short-term thinking than low- policies that make teaching attractive performing systems did. and allow for selectivity of teachers In the 1970s, Singapore’s educational Attracting top candidates into the teaching outcomes lagged the rest of the world’s. profession depends on a complex combina- Transformation was achieved through sys- tion of cultural and policy factors. Cultural temic change at the national level that encom- aspects such as the value a society places on passed curriculum development, national education and the level of esteem in which textbooks, and preservice and in-service teachers are held play a critical role. But there teacher education. Hong Kong SAR, China are also important policy factors, including provides a strong example of supporting financial incentives, that signal that teachers teachers with integrated curriculum develop- are supported and taken care of. ment. In Korea and Shanghai, China educa- Top Performing Systems in East Asia and tional change and improvement were planned Pacific stand out for how much they value and directed at the national level. All schools teachers. They hire good teachers, pay them use government-approved curriculum and well, and continually invest in their profes- materials, entry qualifications to become a sional development. Curriculum and text- teacher are consistent, and there is much less books provide clear guidelines, but teachers diversity of types of schools than in many are also given considerable autonomy to other systems (Boylan 2016). apply their knowledge and skills in class- A coherent system is important at all stages rooms and are encouraged to experiment and of the teacher career cycle (table 5.1). The innovate. central importance of teachers to the educa- Results from the Organisation for tion enterprise was recognized from the start Economic Cooperation and Development in Hong Kong SAR, China; Japan; Korea; (OECD) Teaching and Learning International Singapore; and Taiwan, China. Their expe- Survey (TALIS) indicate that there is a posi- riences suggest that teacher qualifications, tive correlation between the value society teacher quality, and the goal of education for places on teachers and student learning out- the state and parents can have an enormous comes as measured by the Programme for influence on teacher effectiveness. TABLE 5.1 A coherent system covers all aspects of the teacher career cycle Goal Instrument Attraction and selectivity • Good pay • Effective filtering • Mechanisms to increase the attractiveness of teaching as a profession Good preservice • Government control and quality assurance • Filtering at various stages Smooth induction • Open-door culture • Mentoring and extensive support • Time and space for learning Continuous improvement • Teacher support networks • Teamwork and collaboration • Lesson study • Culture of continuous improvement Career development • Progression and promotion policy • Career ladders with multiple pathways in teaching Making teaching easier • System coherence • Aligned, streamlined curriculum and textbooks • Adequate nonclass time 142 GROWING SMARTER FIGURE 5.1 A positive correlation exists between the value that society places on teachers, and student learning outcomes as measured by PISA 45 40 Singapore 35 % of top PISA performers in mathematics Korea, Rep. 30 25 Japan Netherlands 20 Czech Poland Alberta (Canada) Republic Portugal Australia 15 Estonia Finland France England (United Kingdom) Iceland Norway Israel Flanders (Belgium) 10 Slovak Republic Italy Denmark Latvia United States Spain 5 Sweden Serbia Croatia Brazil Bulgaria Romania Chile Mexico 0 0 10 20 30 40 50 60 70 80 % of teachers who agree teaching is valued in society Source: PISA 2015 (OECD 2016a). Note: PISA = Programme for International Student Assessment. International Student Assessment (PISA). All graduates meeting the minimum qualifi- Korea and Singapore are among the highest cation requirements were hired. In-service performers on PISA. In these countries, 67 training or exemptions to these qualification and 68 percent, respectively, of teachers agree requirements were used liberally to remedy that teaching is valued in society (figure 5.1). weaknesses and meet urgent needs. With the exception of Finland, all other coun- In Japan, only 14 percent of applicants to tries are below 50 percent; the average is only education programs are accepted, and only 28 percent. Although there are challenges about 30–40 percent of graduates are hired in disentangling causation—higher perfor- annually (National Center on Education and mance on tests could lead to a higher value of the Economy, n.d.). In Singapore, the govern- teachers—there is likely a virtuous cycle. ment recruits the top one-third of university Today’s high-performing East Asia and graduates and top graduates of polytechnic Pacific systems tended to attract the best schools to become teachers (Tan and Wong teachers from the beginning. Even in the early 2007). In Korea, teacher education program years, when the pool of potential recruits entrants are among the top 10 percent of high was small and demand was great, only stu- school graduates, and only 1 in 20 passes the dents who took top scores on national uni- arduous exams for employment to become versity entrance exams were selected for a teacher (Ferreras, Kessel, and Kim 2015). admission to teacher education and training In Taiwan, China, education programs are programs in Hong Kong SAR, China; Japan; highly competitive. Typically, only the top Korea; and Singapore (Tan and Wong 2007). third of applicants ranked by their academic DEVELOPING SKILLED TEACHERS AND SUPPORTING EFFECTIVE TEACHING 143 performance in high school and university Philippines, for example, monthly pay for a entrance exams are selected (Chou and Wang secondary teacher is less than US$400, com- 2012). Teaching is a well-regarded profession pared with a minimum wage of US$570 for in Shanghai with established entry require- domestic work in nearby Hong Kong SAR, ments; the central government provides China (plus room and board) (Ager 2014; incentives to increase the pool of teaching Ubalde 2007). In the Lao People’s Democratic candidates and encourage teacher retention; Republic, preliminary results of a World Bank and teacher pay is appealing, varying by per- survey on service delivery in the school system formance and years of service. indicate that 53 percent of teachers surveyed Pay is an important element of attracting reported delays in receiving their salary at good candidates. In Japan, salaries for new least once a year (Demas, Khan, and Arcia, teachers compete well against professions like forthcoming). engineering and the civil service. Singapore has gone the furthest to make teaching a financially East Asia’s Top Performing Systems attractive option. The starting salary of teach- generally filter so that good teacher ers is equal to or higher than that for engineers candidates enter the system and lawyers in the civil sector. Indonesia dou- bled teacher salaries through its 2005 teacher Filtering is important in ensuring teacher can- reform. The resulting inflow of candidates into didate quality. Filtering points vary across the system was huge, with a fourfold increase the region, but all systems have mechanisms in enrollment into teacher education programs to ensure that qualified, prepared, and moti- as well as an increase in the average national vated teachers enter the system (table 5.2). exam scores of entrants, indicating higher- Singapore uses intensive selection at entry quality candidates (Chang and others 2013). into teacher education colleges, recruiting Low pay and delayed or irregular pay- its teachers from the top third of high school ments are problems in some systems. In the graduates. The filtering is front-loaded; nearly TABLE 5.2 Different systems use different methods of selecting and filtering teachers Entry to Exit from Evaluation teacher Evaluation teacher Evaluation Evaluation of of probation education of practical education of induction professional period (for Economy program experience program Certification Hiring period development tenure) East Asia and Pacific Hong Kong SAR, China § § ° ° ° ° § l Japan l l ° ° § § l l Korea, Rep. l § l ° l ° l ° Singapore l § l ° § § ° ° Non-East Asia and Pacific Australia § § l § ° ° ° § Netherlands l l § ° ° ° ° § United Kingdom ° l ° l ° ° ° § United Statesa § ° § l ° ° ° ° Source: Wang and others 2003. Note: l = high stakes; § = medium stakes; ° = low or no stakes. Stakes are based on author-defined rubrics and are indicative of whether there is a high, medium, or low level of filtering and quality control measures for the given category. a. Teacher education and certification are the responsibility of individual states; practices differ among them. 144 GROWING SMARTER all candidates who enter the preservice sys- better-quality education for their children, tem receive a position on graduation, mak- and the state targeted higher value-added jobs ing it extremely efficient. In contrast, Japan and industries. Over time, new teachers were filters at multiple points. It filters at entry, required to complete preservice professional with only 14 percent acceptance of applicants training at teacher high schools, then two- into schools of education, as well as at hir- year teacher colleges, then four-year colleges. ing, with only 30–40 percent finding work in In all these systems, an undergraduate degree public schools, and through a rigorous induc- in education or other fields plus a one- to tion program, followed in later stages by an two-year professional teaching program are evaluation of professional development and now the minimum education requirement for evaluation of a probation period for tenure. teachers. Japan was the first to make this shift, In Japan, 98 percent of classes at the second- in the 1950s. Other economies followed, in ary level are taught by teachers who hold a the 1980s and 1990s. Teachers already in the certificate in the field or subject they teach, system were also required or incentivized to and a majority of teachers remain in the pro- undertake continuing professional education fession until retirement. and training for higher pay and promotion (Lai 2009; Sweeting 2004). The state as a major employer of teachers East Asia’s Top Performing Systems has helped to raise teacher qualifications. State- raised their standards as their mandated education and other qualification economies developed requirements can be uniformly applied. From the start, Hong Kong SAR, China; Nongovernment schools are either mandated Japan; Korea; Singapore; and Taiwan, to use the same teacher education and quali- China recognized the central importance of fication requirements or incentivized to adopt teachers to the education enterprise. Given them via public subsidies that raise teacher the enormous gap between demand and salaries to the levels in government schools. capacity, a practical response was the only option. The initial focus in all five systems East Asia’s Top Performing Systems was on expanding the number of teachers. ensure good preservice teacher Individuals meeting minimum qualifications, training, often through direct provision such as graduates of regular and teacher high schools, were hired. Minimum qualifications Although approaches and experiences of were deemed sufficient when the state priority systems at the different performance lev- was basic literacy and numeracy at the pri- els (Top, Above-Average, Below-Average, mary level. Simply having enough teachers to and Emerging) differ, a few features stand staff classrooms was a measure of success for out. First, state control of teacher training, the education system.1 In-service training was qualifications, and hiring is generally more a pragmatic solution to meet the urgent need far-reaching at the primary level than at the for teachers. In Singapore, for example, teach- secondary level. This dominance aligns with ers in training were recruited from teacher state goals to establish basic literacy and training colleges. Teachers who taught in numeracy and to promote common core morning sessions studied in the afternoon social values or political ideology. and vice versa. Through massive recruit- Second, the state is frequently the sole ment and training, the number of teachers in provider of primary education. Virtually all Singapore nearly doubled in a decade, from children attend government primary schools 10,590 in 1959 to 19,216 in 1968 (Goh and in these systems. Having unitary forms of Gopinathan 2008). government, the central authority can impose Qualification requirements were gradually standards via mandates or tie requirements to raised as capacity in the system grew, par- the allocation of education funds to lower lev- ents demanded higher levels of schooling and els of government. DEVELOPING SKILLED TEACHERS AND SUPPORTING EFFECTIVE TEACHING 145 Third, the state accepted expansion of an important guide to developing content universal access to academic secondary edu- for teacher education and training programs cation only in the later years of high growth, (Chou and Wang 2012; Fujita 2007; Kim by which point industry was adopting more 2000; Kim 2007; Lai 2009; Tan and Wong technology to offset rising labor costs and 2007). The five systems also conducted move up the value chain in production. More benchmarking against the best systems in the parents saw academic secondary and postsec- world and integrated the latest concepts and ondary education as critical to the economic techniques in education with their teacher future of their children. Besides expand- education and training programs (Beauchamp ing access at public and private secondary and Vardaman 2015; OECD 2010). schools, the state raised teacher qualifications by upgrading teacher education to four-year Supporting induction into an “open- undergraduate programs, relinquishing its door” profession monopoly on teacher education to expand the production pipeline, and providing state sub- Support of new teachers is a prominent and sidies for teacher salaries at private secondary well-supported component of many success- schools that meet the higher teacher qualifi- ful systems. Japan and Singapore, in partic- cation standards mandated by the state. By ular, use induction as an integral part of the raising teacher qualifications across the entire process. Japan, for example, sets itself apart secondary school system, the state hoped to through active mentoring and a collabora- minimize gaps in quality between public and tive approach to the continual upgrading and private secondary schools. “leveling up” of teaching skills (OECD 2010). Two points are key features of these All new teachers must complete six weeks of systems’ experience. The first is the quality practical training under senior teachers at the of teacher education and training programs. school to which they are assigned. Mentoring The integrity of the teacher training institu- can come in many forms, from pairing new tions is also important. A corrupt system may hires with senior teachers to sending skilled sell diplomas or push students through just teachers with proven success in the classroom to show high completion rates. The second to assist teachers in struggling schools. On is the cognitive ability of teachers. Continual leveling up, teachers and administrators are upgrades in the curriculum mean that teach- transferred to other schools within the same ers must possess knowledge and skills essen- school district and beyond about once every tial to deliver it and meet other goals of the six or seven years; these moves rotate teachers state and parents in education. Consequently, between urban and rural areas and between attracting top-performing students into the big and small cities (OECD 2010). With teaching profession and keeping them became informal “lesson studies” that are used in core to improving education systems. all Japanese schools, teacher quality is lifted In the postwar reconstruction period, across the entire system. these five systems had at most a handful of In Korea, teachers from high-performing teacher training institutions, and most were schools and education experts from univer- teacher training high schools and two-year sities are sent to low-performing schools to colleges (also known as “normal” schools identify problems, develop solutions, and and colleges). Gradually over the next two build capacity. decades, two-year teacher training colleges were elevated to become full undergraduate Teachers need to be observed and institutions under government control, and critiqued in-service training and continuing education and training all became more rigorous. State- Teaching is a “closed-door” profession in most mandated and standardized curriculum for Western systems. In the OECD, 40 percent of primary and secondary education became teachers have never taught alongside another 146 GROWING SMARTER BOX 5.2 Models of teachers observing teachers Observation can yield its greatest benefits when • Peer coaching. Educators work together to discuss used to share instructional techniques and ideolo- and share teaching practices, observe one another’s gies. This form of practice has a long tradition in classrooms, and provide mutual support. many East Asian systems, as exemplified by three • Cognitive coaching. Teachers are taught specific practices: skills that involve asking questions so that the teacher observed is given the opportunity to • Lesson study. Designed by Japanese educators, process learning associated with teaching the this approach has teachers collaboratively develop lesson. a lesson, observe it being taught to students, and discuss and refine it. Source: Education World, n.d. teacher, been observed, or given feedback. to theoretical constructs and that training is Teaching is still a closed-door profession, and classroom based (Walter and Briggs 2012). teaching unions have made it hard for observ- Specificity means that teachers are taught ers to take notes in classes (Economist 2016). teaching pedagogy that is specific to a subject In contrast, in many East Asian coun- area. Continuity means that teachers receive tries, the classroom is seen as a public space, significant, continual support, not one-off and lesson observation and peer support are workshops (Darling-Hammond and others ingrained characteristics (box 5.2). Japan’s 2009; Yoon and others 2007). induction period involves lesson demonstra- Teacher improvement must be a continu- tions conducted in front of panels for evalu- ous process that does not end at preservice ation and feedback. Many Shanghai schools or induction. Preservice preparation sets the have separate lesson observation rooms, foundation of a teacher’s knowledge and skill where lessons can be videotaped and demon- set. Induction in the initial years is critical in strations conducted with an audience. getting the teacher comfortable and effective in the classroom. But the next 30 years or more of a teacher’s career must also be used Effective professional development to build skills and continually improve. The leads to effective teaching in-service portion typically makes up more One of the three “promising principles” than 80 percent of the teaching cycle. identified in the World Development Report There is evidence that teachers often stall 2018 to make teaching more effective is the in their skills development after an initial delivery of professional development: “For jump early in their careers. A differentiat- effective teacher training, design it to be indi- ing factor in high-performing systems is that vidually targeted and repeated, with follow-up learning and improvement continue through- coaching—often around a specific pedagogi- out the teaching cycle. cal technique” (World Bank 2018, 22). This Traditional approaches of bringing teach- practice lies in stark contrast to much teacher ers to a training center for one-off training professional development in a range of tend not to have longer-term effects on chang- countries. ing practices in the classroom. Professional Experience from high-income countries development is often poorly aligned to what shows that practicality, specificity, and conti- Darling-Hammond and others (2017) iden- nuity are key to effective teacher professional tify as the seven characteristics of effective development (Popova, Evans, and Arancibia professional development, which (a) is con- 2016). Practicality means that teachers are tent focused; (b) incorporates active learn- trained with concrete methods as opposed ing using adult learning theory; (c) supports DEVELOPING SKILLED TEACHERS AND SUPPORTING EFFECTIVE TEACHING 147 collaboration, typically in job-embedded of 10 teachers in the OECD say they are well contexts; (d) uses models and modeling of prepared (Economist 2016). effective practice; (e) provides coaching and In contrast, in many East Asian countries, expert support; (f) offers opportunities for teachers think that they need to improve con- feedback and reflection; and (g) is of sus- stantly. In the 2015 Trends in International tained duration. Mathematics and Science Study (TIMSS) Professional development programs in East survey, teachers were asked to character- Asia and Pacific’s Top Performing Systems ize their confidence in developing students’ tend to embody these characteristics. They higher-order thinking skills. Many of the Top emphasize content, collaboration, modeling, Performing Systems tended to have the lowest coaching, and expert support with opportuni- proportion of teachers saying they had high ties for feedback and reflection. The rest of or very high confidence. This result partly this section explores how these systems imple- reflects a cultural tradition of modesty, but it ment these efforts. also likely indicates a belief that there is room for improvement (figure 5.2). In Shanghai, teachers participate in A mentality of constant improvement teaching-research groups at the school, dis- is the foundation for professional trict, provincial or municipal, and national development levels. At the municipal level, Shanghai’s In most education systems, after the burst Municipal Education Commission Teaching- of improvement at the start of their careers, Research Office, founded in 1949, is the teachers rarely get much better. This stagna- key agency for conducting teaching-related tion often reflects the fact that teachers do research. At the school level, teaching- not think they need to improve. Three out research groups are professional development of five low-performing teachers in the United networks consisting of same-subject teachers. States think they are doing a great job; 9 out In larger schools, the groups are often further FIGURE 5.2 Teacher modesty in self-assessment of skills may facilitate or complement continuous professional development 100 % of teachers self-assessing their skills 90 at indicated confidence level 80 70 60 50 40 30 20 10 0 Fe oro a ng ia Ka thua a Qa el Ba hile ain ai a ,C n Ta ing land , C re G e ina Sw gia en s y No ep. ite Ma way ba s de cco Ko orw n Ira New a, R ay Isl ea . ic d Ire om Ca and ng a M ry Un T alta d ey ts es kh a Eg au Jord n yp di an So rab bia h A p. Ku ica Isr it ab v r i a Om n C n n, Z ep Ar Slo ta Le rate ian M trali Li wan Au Ital Th hin Hu nad za ni Em eni AR apa N tio sta no a am lan wa Ki ys a ut Re an o a Bo Stat ite urk ed fr hr R or h t, A Ara iw ap d d la l ra r gS J re S S on Un ss d gK ite Ru Un n Ho Low confidence in skills Medium confidence in skills High confidence in skills Very high confidence in skills Source: TIMSS 2015. Note: Economies labeled with purple type are East Asia and Pacific Top Performing Systems. 148 GROWING SMARTER divided by grade. Each group has a leader 44 percent host groups once a week and 53 responsible for organizing activities and intro- percent host them every two weeks. ducing novice teachers to the learning com- munity. The groups normally meet for two Collaborative environment with to three hours a week. Activities include the feedback following: Collaboration and teamwork are explicit • Professional development. Teachers share features of many successful East Asian sys- resources, set clear goals for student tems. In Shanghai, for example, teachers are learning, engage in conversations about not promoted unless they can prove they are instruction and subject knowledge, and collaborative. Mentors are not promoted ask questions and seek answers from col- unless they can show that their student teach- leagues or external advisers. ers improve. Teachers are given ample time • Coaching and guidance. Senior teachers for these collaborative activities. They teach provide guidance and coaching to junior only 10–12 hours a week—less than half the teachers in a wide range of activities, U.S. average of 27 hours (Liang, Kidwai, and including exam design and teaching expe- Zhang 2016). rience discussions. The 2013 TALIS developed a peer network • New teacher induction. New teachers par- index that measures opportunities for the ticipate in group activities, receive guid- information exchange and support needed ance from experienced teachers, and gain to maintain high standards of teaching. The expertise on a wide range of teaching top- index includes participation in induction and ics. The leader of the group often coaches mentoring programs, participation in teacher new teachers. Mentoring includes lesson networks, and the receipt of feedback from observations and critiques. direct observations. Of 37 education systems • Research. Teachers conduct research on in the study, the three top-scoring systems subject areas and pedagogical practices were Malaysia, Shanghai, and Singapore, to improve their instructional ability. with Korea coming in 6 and Japan in the Schools provide teachers with various middle at 19. The fact that four of the top six research approaches. Among the schools systems are from East Asia indicates the high surveyed, 84 percent have an overarch- importance the region places on collaboration ing topic for all teachers to research, 81 among teachers. percent encourage teachers to conduct Teachers who meet more frequently for individual-topic research, and 88 percent professional collaboration tend to have higher provide research topic options for teachers levels of self-efficacy (belief in their own abili- on various subjects. ties). Most of the best teachers have a strong • Performance evaluations. Teachers often sense of self-efficacy (Künsting, Neuber, and evaluate one another during group activi- Lipowsky 2016). High levels of self-efficacy ties. The group leader provides feedback are also found among teachers who engage in annual teacher performance evaluations in joint activities—observing other teach- (Liang, Kidwai, and Zhang 2016). ers’ classes, providing feedback, and teach- The groups’ collaborative nature helps ing jointly in the same class. Although even the entire teaching community to grow. Such meeting just once a year in professional col- groups create a constructive work environ- laboration activities can improve self-efficacy ment and allow teachers to forge close bonds. (figure 5.3), there is a clear and steady trend The groups’ tiered network at the school, dis- of increasing self-efficacy with incremental trict, and municipal levels allows for quick increases in frequency. and far-reaching dissemination of curricula, Teacher collaboration is common in nearly best practices, and other ideas on teaching all East Asian systems. Indonesia has a strong and learning. Of the 154 schools surveyed, system of teacher working groups. Networks DEVELOPING SKILLED TEACHERS AND SUPPORTING EFFECTIVE TEACHING 149 FIGURE 5.3 Teachers who meet more frequently for professional collaboration tend to have higher levels of self-efficacy (belief in their own abilities) 13.4 13.2 13.0 Index of teacher self-efficacy 12.8 12.6 12.4 12.2 12.0 11.8 11.6 11.4 Never Once a year 2–4 times 5–10 times 1–3 times Once a week or less a year a year a month or more Frequency Teach jointly as a team Observe other teachers’ Engage in joint Take part in collaborative in the same class classes and provide activities across professional learning feedback different classes Source: OECD 2014, using an index to measure self-efficacy. of general teachers at the primary level and systems shows that most systems offer teach- subject teachers at the secondary level are ers the possibility of horizontal promotions complemented by working groups of prin- to positions that allow them to grow pro- cipals and supervisors to provide teacher fessionally as teachers but remain closely support from different levels. In Myanmar, connected to instruction, instead of mov- full-time mentors provide pedagogical guid- ing “up” to managerial positions (Darling- ance and demonstrations. In Vietnam, teacher Hammond 2010; OECD 2012b). In many clusters play a prominent role in supporting countries, the way to get ahead in a school is to professional development for early childhood move into management. U.S. school districts, teachers and in introducing new teaching for example, “pay people in inverse propor- techniques, such as Vietnam Escuela Nueva tion to the value they add” (Economist 2016). and its student-centered 21st century skills District superintendents make more money approach. In Thailand, a peer review method than teachers, although they generally have monitors the practice of math teachers. less impact on students’ lives. Singapore has a separate career track for teachers, so that the best teachers do not Career progression and promotion that leave the classroom. In Shanghai, teachers attracts, supports, and motivates can apply for academic or administrative Career opportunities are important to attract positions to assume more responsibilities. talented individuals into teaching and pro- Unlike many other systems, in which one is vide incentives for them to stay. In addition either a teacher or not, teachers have oppor- to vertical promotions into leadership posi- tunities to advance professionally through- tions, case study research on high-performing out their teaching career through a five-level 150 GROWING SMARTER ranking system. Shanghai regularly evaluates that is not accompanied by actions to ensure teachers for promotion (Liang, Kidwai, and alignment with all teacher-related aspects Zhang 2016). such as teacher standards, the accompanying textbook modification, and teacher training will weaken teachers’ ability to carry out their Strong supporting elements are tasks and possibly reduce their motivation. An aligned in ways that make the approach to policy that considers the impact teacher’s job easier on teachers and the direct or indirect role that The top systems provide strong auxil- teachers play in policy implementation can iary elements that are aligned in ways that make policies more effective (figure 5.4). make the teacher’s job easier, including clear learning objectives and standards; Curriculum alignment well-aligned curricula, usually focusing on relatively few topics; textbooks that are A well-aligned curriculum with a small number uncluttered, focused, and effective for learn- of topics focuses learning and allows deeper ing goals; more time for teachers to prepare study of the topics covered. Schmidt and oth- lessons; professional development activities; ers (2001) analyze textbooks and standards and assessment systems. in 34 countries. They find that the highest- performing countries tend to have fewer con- tent standards and topics in their textbooks Alignment of system elements than other countries. The United States cov- Because teachers are intimately linked to most ers all 79 of the TIMSS science topics in its aspects of the education system, policies affect content standards. In contrast, Korea cov- them as well as their ability to get students ers only 8, Japan covers 19, and Hong Kong to learn. A major change in the curriculum SAR, China covers 22. Textbooks in the United States cover 78, whereas Korea cov- FIGURE 5.4 Considering the direct or indirect role played by ers 38 and Japan only covers 17. The nar- teachers in implementation can make policies more effective rower focus and deeper study allow teachers to develop better lessons rather than rushing through topics (table 5.3). Curriculum and textbooks Focused, streamlined textbooks Student Teacher A streamlined curriculum allows for unclut- assessment standards tered, focused textbooks. Textbooks in the United Kingdom and the United States tend to be extremely thick and full of content, much of which is not covered. In contrast, Resources TEACHERS AND Preservice Chinese textbooks tend to be thin and nar- and school training environment TEACHER rowly focused on specific topics, allowing POLICY them to go into greater depth on those topics (textbooks in Japan, Korea, and Vietnam also tend to be narrowly focused). Chinese text- Career In-service books are also significantly more demanding progression professional and promotion development than textbooks in the United Kingdom (Qin 2017). All students cover all of the content in Licensing the textbooks, making the study more focused and allowing students to master topics. The mastery approach is believed to have been important in propelling DEVELOPING SKILLED TEACHERS AND SUPPORTING EFFECTIVE TEACHING 151 TABLE 5.3 Top Performing East Asia and Pacific systems tend to focus learning on fewer content standards and topics in textbooks compared with other countries Content Textbook Topics tested in Topics tested in Economy standards topics coverage content standards textbook Korea, Rep. 8 38 6 29 Japan 19 17 18 15 Hong Kong SAR, China 22 37 17 26 Greece 25 49 22 37 Romania 29 53 19 33 Czech Republic 33 49 22 30 Germany 35 32 21 22 Israel 37 32 25 21 Singapore 38 27 27 20 South Africa 39 49 26 35 Cyprus 41 53 31 39 Bulgaria 43 57 28 37 Sweden 47 49 34 32 Netherlands 48 67 32 43 Slovak Republic 48 49 29 30 Denmark 50 9 29 8 France 50 37 30 25 Hungary 53 62 36 43 Iceland 56 46 42 30 Scotland 57 44 42 36 Austria 58 60 36 39 Canada 58 74 38 46 Colombia 58 65 43 46 Ireland 61 58 42 40 Russian Federation 62 42 37 28 Australia 66 65 43 44 Portugal 66 64 40 39 Spain 66 67 43 41 Norway 69 61 41 40 Slovenia 69 62 44 41 Switzerland 69 78 44 48 Latvia 70 35 44 24 New Zealand 79 52 48 37 United States 79 78 48 48 Source: Schmidt and others 2001. Note: Economies labeled with purple type are East Asia and Pacific Top Performing Systems. Green shading is used to indicate systems that cover 30 or fewer topics, yellow indicates systems that cover 31 to 60 topics, and red indicates systems that cover 61 or more topics. Topics are within science curricula. students in Hong Kong SAR, China; Shanghai, Time and space for lesson preparation China; and Singapore to the top of the and professional development PISA rankings (Qin 2017). It also allows Spending time in class is only one part of a teachers to focus and have a clearer sequence teacher’s job. Teachers also prepare lessons, of content. 152 GROWING SMARTER FIGURE 5.5 Teachers in Top Performing East Asia and Pacific In the high-performing Asian countries, systems tend to spend a smaller proportion of total working hours a small proportion of total working hours teaching and more time preparing lessons compared with teachers is spent in class. Japan’s teachers spend only in other countries 18 hours teaching on average, even though they have the most total working hours (54) Korea, Rep. (figure 5.5). With nearly two-thirds of their economies and Pacific East Asia Japan working time spent outside of class, they Singapore Malaysia spend much more time on lesson preparation United States and other quality-enhancing activities that Chile make in-class time much more effective. Alberta (Canada) Brazil Mexico What happens in the classroom? Abu Dhabi (United Arab Emirates) Insights into effective teaching Estonia PISA top performers focus their time and Portugal effort not on the school system but rather Finland Slovak Republic on what goes on in the classroom. PISA England (United Kingdom) results reveal what does not work. Spending Non-East Asia and Pacific economies Croatia more money, for example, is associated Average with higher scores only in poorer countries. Latvia Among systems that already spend more than Flanders (Belgium) US$50,000 per pupil throughout their time in Iceland school, money alone brings no improvement. Denmark But PISA also teaches what does work. Spain Its most important insight is that what mat- Poland France ters most is what happens in the classroom. Australia Successful children are exposed to good Bulgaria teaching more often (OECD 2016a). Serbia Case study research on successful edu- Israel cation systems in Finland; Japan; Korea; Czech Republic Ontario, Canada; and Singapore suggests Sweden that high-performing education systems Italy devote considerable time to activities related Netherlands to instructional improvement, such as col- Cyprus Romania laboration among teachers on the analysis Norway of instructional practice (Darling-Hammond 2010; Darling-Hammond and Rothman 0 10 20 30 40 50 60 Average hours spent teaching 2011; Levin 2008). These systems also devote and working per week a smaller share of teachers’ time to contact Total working hours with students and a larger share to teacher Hours spent teaching collaboration, on-site professional devel- opment, and research on the effectiveness Source: TIMSS 2015. Note: Economies labeled with purple type are East Asia and Pacific Top Performing Systems. of teaching strategies. Japan, for example, The economy labeled with blue type is an East Asia and Pacific Below-Average Performing System. devotes about 40 percent of teachers’ work- ing time to these types of activities; Ontario devotes 30 percent (Darling-Hammond and check homework, develop tests, and provide Rothman 2011). after-hours support to students. If a large pro- According to an extensive meta-analysis, portion of time is spent in class, less time is the 20 most powerful ways to improve learn- available for other activities. ing are related to teaching. Many factors shape DEVELOPING SKILLED TEACHERS AND SUPPORTING EFFECTIVE TEACHING 153 a child’s success, but in schools nothing matters the strongest evidence in improving student out- as much as the quality of teaching. In a meta- comes are content knowledge and the quality study updated in 2015, John Hattie examines of pedagogical practices (box 5.3). Teachers the results of more than 65,000 research papers with strong knowledge and understanding of on the effects of hundreds of interventions and their subject have greater impact on student their effects on student learning (Hattie 2003, learning. Teachers also need to understand how 2015). He finds that aspects of schools that students think about content and be able to parents care about a lot—such as class size, identify common misconceptions. Pedagogical uniforms, and streaming by ability—make lit- practices include effective questioning and the tle or no difference to whether children learn. use of assessment by teachers. Practices such as What matters is “teacher expertise.” All 20 of reviewing previous learning, providing model the most powerful ways to improve school- responses for students, giving adequate time for time learning identified by the study depend on practice to embed skills securely, and progres- what a teacher did in the classroom. sively introducing new learning (“scaffolding”) Many “good practices” have turned out improve attainment. to be ineffective. Coe and others (2014) note In many high-performing East Asian sys- that many common classroom techniques tems, teachers have very good mastery of do not work. Unearned praise, grouping by content. In a comparison of third-grade math ability, and accepting or encouraging chil- teachers in China and the United States, dren’s different “learning styles” are widely researchers find that U.S. teachers were espoused but bad ideas. So, too, is the notion more knowledgeable about general educa- that students can discover complex ideas all tional theories and classroom skills, whereas by themselves. Teachers must impart knowl- Chinese teachers had stronger knowledge edge and critical thinking. of the subject matter they were teaching as Research offers clear pointers to the most well as a better understanding of the over- likely effective approaches. The two factors with all elementary curriculum that their students BOX 5.3 Subject and pedagogical knowledge are critical for effective teaching A video study conducted in Indonesia that measured • Demonstrate more confidence and command in subject and pedagogical knowledge of math teach- the classroom ers finds that 57 percent of teachers scored below • Stimulate more teacher-student interaction, with 50 percent (Ragatz and others 2014). Higher-knowledge interaction involving more higher-order thinking teachers tended to use a greater range of practices, • Use questioning more and encourage students such as investigation, open-ended questioning, the use to think and share ideas rather than just follow of mathematical language and symbols, and the use of instructions nonroutine problems and applications. They also used • Make connections between procedures and concepts. these practices more effectively. These teachers also When teachers do not reach a minimal level • Make fewer mathematical errors of mastery of the subject, they are unable to con- • Use more accurate language and are more concise vey concepts and procedures effectively, even if in their explanations they use good pedagogical practices. Many low- • Come up with new problems that were not origi- and middle-income countries in the region, includ- nally in the lesson plan ing Cambodia and the Lao People’s Democratic • Make greater attempts to probe student thinking Republic, have teachers with low subject knowl- • A re more able to spot and correct student edge, which limits their effectiveness. misconceptions • Involve a wider range of students, including stu- Sources: Benveniste, Marshall, and Santibañez 2007; Ragatz and others 2014; dents who might not provide correct answers Tandon and Fukao 2015. 154 GROWING SMARTER had covered and would cover in later years. encompasses a Japanese business philosophy Chinese teachers spoke in great detail about of continuous improvement of working prac- the content they present to students and dem- tices, personal efficiency, and productivity. onstrated a deep understanding of the sub- Lessons are approached through investigation ject matter as well as knowledge of the entire and research into the teaching-learning pro- elementary math curriculum. When asked cess, with constant refinements for improve- about their teaching methods, few U.S. teach- ment. Lessons are planned with a clear sense ers mentioned content (Zheng, Peverly, and of the goal and how to reach it. Xin 2006). Japan’s lesson study breaks a lesson down in detail and indicates how to teach it effec- What do effective teachers do? tively. In lesson study, the principal typically organizes and oversees activities. A teacher in “I don’t teach physics,” says Charles Chew, a the group tries out the sample teaching plans teacher in Singapore, “I teach my pupils how in the classroom while other teachers (and fre- to learn physics” (Economist 2016). This quently university researchers) observe. The statement encompasses a critical learning group then reviews, critiques, and adjusts the outcome approach to teaching often found in sample teaching plans before adopting revised high-performing East Asian systems that com- versions for use. This collaborative approach bines high-quality instruction with pedagogi- strengthens work culture and gives teachers cal content knowledge—a blend of subject considerable autonomy to determine how best knowledge and pedagogy for effective teach- to implement the curriculum simultaneously to ing. Teachers teach students how to learn by achieve education policy objectives and meet directing them to underlying principles, devel- local needs and conditions (OECD 2010). oping habits of mind. Schools in Korea have been using the les- Effective teaching is a combination of son study approach for more than a decade. preparation (well-planned lessons with a clear In addition, the central government provides goal of learning outcomes) and execution grants of US$5,000–US$10,000 to sup- (practices that stimulate thinking and learn- port research by small groups of teachers to ing) (table 5.4). develop new instructional techniques and materials (Ferreras, Kessel, and Kim 2015). Singapore is also working to adopt lesson Meticulous preparation—before even study (Lu and Yee 2011–12). Besides peer col- entering the classroom—with clear, laboration within schools, the government concise goals and sequence gives teachers time and funding to undertake In East Asia and Pacific’s Top Performing “attachment” (internship or observation) in Systems, lessons are refined through a kai- local and overseas schools as well as in busi- zen (change for better) approach, with a ness and community organizations. The goal focus on detail, quality, and learning. Kaizen is to expand teachers’ perspectives and bring TABLE 5.4 Content and pedagogy needed for lesson preparation and execution Item Content Pedagogy Preparation • Clear goal of learning outcome • Pedagogical approach that will most effectively convey the • Deep thought on how students will learn the content content • Identification of potential misconceptions • Identification of alternative pathways for students to learn • Assessment for learning plan content • Effective ways to address potential misconceptions Execution • Concise, accurate explanations of content • Teacher-centered but interactive pedagogy • Dynamic approach to content based on circumstances • Practices that stimulate thinking and learning • Elaboration and use of cognitive activation strategies DEVELOPING SKILLED TEACHERS AND SUPPORTING EFFECTIVE TEACHING 155 new experiences and insights back to their the use of cognitive activation strategies and classrooms so that they can better prepare high test scores among mostly high-income students for life and work.2 countries. Cognitive workouts such as “turn and talk” (when a teacher asks students to turn around and talk to a surrounding peer Practices that stimulate thinking and about a topic) and cold calling are common in learning Shanghai and Singapore (OECD 2016b). Classes in East Asia tend to be planned, The PISA index of teacher-directed instruc- teacher-led but interactive, with a clear sense tion provides insights on (and breaks some of the goal and how to reach it. Effective stereotypes about) East Asian education sys- teachers ask probing questions of all students tems. The OECD used the PISA survey results and assign short writing tasks that get chil- to create indexes that measure the extent to dren thinking and allow them to check stu- which instruction is teacher directed or stu- dent progress. Teachers anticipate errors, such dent oriented, as well as the extent to which as the tendency in math to mix up remain- learning is memorization versus elaboration. ders and decimals. They space out and vary The Top Performing Systems in East Asia and ways in which children practice, based on the Pacific tend to use more elaboration and less results of cognitive science that doing so aids memorization (figure 5.6). This finding does long-term retention. These techniques work. not fit the stereotype that the region tends to A 2016 OECD report finds a link between use memorization and rote learning. FIGURE 5.6 The Top Performing Systems in East Asia and Pacific tend to use more elaboration and less memorization in teaching math More memorization 3.0 United Kingdom 2.5 New Zealand Australia Ireland 2.0 Canada Singapore France Learning index United Indonesia States Japan OECD Finland Hong Kong SAR, China 1.5 average Thailand Macao Korea, Rep. Shanghai, China SAR, China 1.0 Malaysia Vietnam Taiwan, China 0.5 0 0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 More Teaching index elaboration More student-oriented More teacher-directed instruction instruction Top Performing Systems Non-East Asia and Pacific country Above-Average Performing Systems OECD average Below-Average Performing Systems Source: PISA 2012 (OECD 2012a). Note: OECD = Organisation for Economic Co-operation and Development. 156 GROWING SMARTER FIGURE 5.7 Memorization is less useful, and elaboration Students in classrooms in which elabo- strategies are more useful, as math problems become more difficult ration strategies are more prevalent tend to perform better, particularly on more difficult Greater problems. Memorization can be effective success a. Memorization for easy problems (figure 5.7, panel a). But 1.05 as difficulty increases and critical thinking 1.00 is required, elaboration strategies become Odds ratio of success much more important (figure 5.7, panel b). using memorization 0.95 The high-performing Asian countries include 0.90 some memorization, but they focus on 0.85 elaboration. 0.80 0.75 Directed learning versus student-driven 0.70 learning 300 400 500 600 700 800 900 Less Less Difficulty of mathematics item More A study by Miao and Reynolds com- success difficult on the PISA scale difficult pares math teaching in Nanjing, China Greater b. Elaboration strategies and Southampton, England. It finds that success “whole-class interactive teaching” (when a 1.50 teacher interacts with pupils in the whole using elaboration strategies 1.40 class through two-way communication, typ- Odds ratio of success 1.30 ically in the form of questioning-answering 1.20 or discussion) was used 72 percent of the time in Nanjing and just 24 percent of the 1.10 time in Southampton (Miao and Reynolds 1.00 2018). Stigler and Hiebert (2009) find that 0.90 U.S. classrooms tend to ask “what” ques- tions. In Japan, teachers asked more “why” 0.80 300 400 500 600 700 800 900 and “how” questions that checked whether Less Less Difficulty of mathematics item More students understood what they were learn- success difficult on the PISA scale difficult ing. Box 5.4 describes how the Classroom Assessment Scoring System (CLASS) was Source: PISA 2015 (OECD 2016a). used in China to observe teaching practices. BOX 5.4 Teaching practices in Guangdong, China The Classroom Assessment Scoring System (CLASS) (figure B5.4.1). Scores on student engagement is a classroom observation instrument that has been were also very high, and scores on emotional and applied extensively in many international contexts to other support were similar to averages in other measure teaching practices. Coflan and others (2018) countries. Student engagement also appears to be use it to compare practices in 16 poor locales. linked to cognitive activation techniques, where Teachers in Guangdong, China, scored particu- understanding of content and quality of feedback larly high on the domain of classroom organization were relatively high dimensions of instruction. box continues next page DEVELOPING SKILLED TEACHERS AND SUPPORTING EFFECTIVE TEACHING 157 BOX 5.4 Teaching practices in Guangdong, China (continued) FIGURE B5.4.1 Teachers in Guangdong, China, perform on average at higher levels on CLASS studies than do teachers in other countries a. Emotional support 1 2 3 4 5 6 7 Less effective Moderately effective Highly effective CLASS score b. Classroom organization 1 2 3 4 5 6 7 Less effective Moderately effective Highly effective CLASS score c. Instructional support 1 2 3 4 5 6 7 Less effective Moderately effective Highly effective CLASS score d. Student engagement 1 2 3 4 5 6 7 Less effective Moderately effective Highly effective CLASS score Source: Coflan and others 2018. Note: The bars are Guandong’s pilot values; the grey, smoothed overlays are the averages from other upper-elementary studies collected by Teachstone, the company responsible for the management and certification of CLASS. CLASS = Classroom Assessment Scoring System. 158 GROWING SMARTER Reducing the proportion of low- teachers and rebuilding its system, starting performing students with teachers at very low education levels. Myanmar is embarking on many important Thanks to an approach of equity and learn- teacher reforms that began mainly in 2012, in ing for all, East Asia and Pacific systems have an effort to expand its young teacher work- the lowest percentage of low-performing stu- force and rebuild its system after years of dents on international assessments. Hong military rule. The Philippines introduced a Kong SAR, China; Korea; Shanghai, China; New Teacher Education Curriculum in 2004 and Singapore all rank in the global top 10 for and has experimented with many teacher ini- achievement as measured by PISA, and in each tiatives. It is attempting to support teachers of them, less than 10 percent of students score through major changes to the kindergarten at the lowest levels (Tucker 2016). Improving through twelfth-grade curriculum, but it has the performance of the low performers does struggled to develop a competent teaching not require an education system to sacrifice the force (Tucker 2016). performance of the average or top performers. Starting in 2005, Indonesia undertook one In the United States and many other sys- of the most comprehensive regional teacher tems, as students start to fall behind, they find reforms. It has accomplished a great deal— it harder and harder to comprehend what is establishing teacher rights, developing an going on in class and fall even further behind. array of teacher policies, creating teacher stan- Sometimes, as their morale sinks and embar- dards, establishing new systems, and making rassment rises, they stop coming to school and the profession more attractive through both then drop out. East Asia and Pacific systems higher salaries and a degree of professional- deal with this downward spiral by stopping it ization of teaching. But the reform has gone before it gains momentum. They start from a in fits and starts, with many struggles along commitment to the idea that all students can the way, and has often fallen short of what and will meet high standards as they progress was envisioned. Although the accomplish- through the years. Although it requires extra ments should be recognized, Indonesia also commitment, this approach helps teachers to highlights the challenges of teacher reform. be more effective in the long run (Tucker 2016). Teacher quality was a key focus of Indonesia’s reform, a cornerstone of which was teacher certification. Certification was Challenges in teacher reform intended to ensure that teachers met a mini- Although no two paths are alike, some fun- mum threshold of quality, but it was politi- damental similarities of teacher reform exist cized and became a watered-down and among East Asia’s high performers. In the early toothless process in which nearly all teach- stages, teacher qualifications were not necessar- ers received certification and little was done ily high, but a strong emphasis was placed on to establish clear standards and thresholds attracting high-quality candidates, and preser- of quality. The plan to establish an inte- vice training provided a strong base. Advances grated Teacher Professional Management in teacher policy tended to be made in tandem System brought a strong vision for evaluat- with other system components to ensure a ing teachers through observation and assess- coherent and aligned system. Although reform ment, which would then link to targeted efforts and accomplishments of the high per- professional development. It ran into imple- formers were rapid, they still spanned decades mentation realities, however, such as dif- (see the timelines in figure 3.3 in chapter 3). ficulty conducting objective assessments or Lower-performing East Asia and Pacific making sufficient professional development countries have adopted many teacher reforms, opportunities available. Financial, technical, with varying ambitions and varying degrees and political realities forced an ad hoc imple- of success. Cambodia has made impres- mentation, with some aspects implemented in sive progress since the 1980s in educating a hasty and ineffective manner. DEVELOPING SKILLED TEACHERS AND SUPPORTING EFFECTIVE TEACHING 159 Indonesia’s experience offers the following profession. With this foundation, a virtuous lessons: circle can be created. The system of support is critical. It begins • The political economy of reforms, includ- with induction into an “open-door” profes- ing the political will of the implement- sion, created by policies that assume that ers as well as the recipients, needs to be new teachers need help and that it is the role considered. of all actors in the education community to • Ambitious reform has to take into account provide it. In an open-door profession, peers, the realities of financial and technical con- school leaders, and others provide support, straints and be coherent. and teachers habituate themselves to being • A degree of ad hoc implementation is observed and critiqued. This kind of culture unavoidable, but a coherent vision should is marked by collaboration with feedback. underpin activities and their sequence, and A mentality of constant improvement on the certain activities should be implemented in part of teachers provides a foundation for sync to ensure coherence. continuous professional development activi- • Establishment of teacher quality standards ties, which are closely linked to career pro- requires the political will to enforce them; gression and promotion. standards need to reflect realistic thresh- Strong supporting elements within the sys- olds that teachers can meet or work to tem include an aligned and focused curricu- achieve. lum, streamlined textbooks, time and space Indonesia’s reform is just over a decade in for lesson preparation and professional devel- the making, and many important elements are opment, and a focus on learning (with mini- now in place, albeit imperfectly. The teaching mized administrative activities). profession has become an attractive prospect, Systems that focus on classroom and with large numbers of applicants to preser- teacher-related policies and devote consid- vice programs. Whether a sufficient founda- erable time at the school level to activities tion for the future has been established will related to instructional improvement are the depend on the political will to refine the ele- most effective. ments, fill missing gaps, and push for system High-performing systems often adopt a crit- coherence. ical learning outcome approach—teaching stu- dents to learn rather than just teaching students to know. It combines high-quality instruction Conclusions and with pedagogical content knowledge—a blend recommendations of subject knowledge and pedagogy for effec- Educational systems perform best when tive teaching. Teachers teach students how they have teachers who are respected, pre- to learn by elaborating and directing them to pared, selected on merit, and supported in underlying principles, combining meticulous their work (World Bank 2018). East Asia preparation and solid execution. and Pacific systems provide many lessons for Many systems in East Asia and Pacific now developing such high-performing systems. A use the lesson study approach, which origi- foundational aspect is attracting high-quality nated in Japan. Effective classroom practices candidates into the profession by offering not tend to be teacher-led but interactive, involve only competitive compensation but also an elaboration strategies, and aim to stimulate appealing professional environment founded thinking and learning. They also seek to on respect and support. By attracting good ensure that lower-performing students are candidates, education systems can also be given extra support and do not fall behind. selective. Combined with a process of sup- As the World Development Report 2018 port and effective development throughout a notes, in the short run, systems can improve teacher’s career, such selectivity leads to more the quality of professional development by effective teachers and greater respect for the shifting resources to the kinds of professional 160 GROWING SMARTER TABLE 5.5 Steps toward attracting and developing skilled teachers and supporting effective teaching Action Emerging Below-Average Above-Average Top Performing Policy Systems Performing Systems Performing Systems Systems Teachers Start the positive foundational cycle Take incremental steps to increase Establish rigorous criteria and multiple of attracting and ensuring reasonable- qualifications and quality, establishing filtering mechanisms, while ensuring quality teachers, while being realistic greater quality assurance and filtering that teaching is attractive in terms of about the system’s capacity to measures. salary, professionalism, and prestige. implement quality assurance measures balanced with supply of candidates. Deepen teacher networks in which Leverage high-performing teachers for more-advanced teachers provide professional development and support. Establish teacher networks (with a focus greater support through mentoring and on supporting new, weaker teachers coaching. Provide more autonomy for high- with mentoring and other support), performing teachers. lesson study, and an “open-door” Provide more advanced professional culture as key elements of professional development that develops deeper Create a highly developed professional development. critical thinking techniques. development framework with tailored, individualized approaches. Establish a basic framework of Create tracks of progression and professional development and how promotion for expert teachers that Deepen critical thinking, elaboration, teachers can reasonably allow them to remain in the classroom. and cognitive activation techniques. be reached. Evolve system- and teacher-related Establish a mentality of teaching to policies in ways that reform and move learn with clear learning goals with all elements forward with alignment. curriculum and textbooks. Decrease classroom hours and increase Begin creating a cohesive, aligned supplementary activities. system that supports teachers and promotes focused, streamlined learning. development that will change teacher perfor- Lessons from the region’s successful sys- mance in the classroom. They can support tems suggest a way forward (table 5.5). They teachers in teaching to the level of the child. suggest a step-by-step approach that includes They can provide a professional structure ambitious but realistic goals based on the so that teachers feel motivated to apply what capacity of the system and teachers within they know. the system; alignment with key elements; and Comprehensive reform requires assess- thoughtful implementation that ensures that ment of capacity and constraints, identifica- key supporting elements are in place. tion of areas that are high impact and areas On specific policy areas, it is critical to that provide leapfrog potential, and guidance establish a virtuous circle of attracting good on the steps that must be taken. Challenges candidates through adequate pay, creating to success include ad hoc policies that are not an attractive professional environment, and part of a comprehensive plan, lack of criti- establishing a culture of respect for teach- cal elements necessary for successful imple- ers. Quality control in preservice and filtering mentation, rushed implementation without mechanisms are also important. a long-term vision, unrealistic goals, the set- Professional development in high-performing ting of education policies in other areas with- systems includes collaborative teacher networks out proper consideration of their impact on and a culture of teamwork, knowledge shar- teachers, and policy misalignment. ing, peer observation, feedback, and support. DEVELOPING SKILLED TEACHERS AND SUPPORTING EFFECTIVE TEACHING 161 A complementary area is career progression The Role of Politics and Evidence-Based and promotion linked to professional develop- Policy Making: The Case of Teacher Reform in ment, such as tracks that keep the best teachers Indonesia. Washington, DC: World Bank. in teaching by recognizing and rewarding high Chou, C. P., and L.-T. Wang. 2012. “Who Benefits from the Popularization of Higher Education performers. A key takeaway in classroom prac- in Taiwan?” Chinese Education and Society 45 tices is the importance of meticulous prepara- (5-6): 8–20. tion and refinement of lessons with a deep focus Choy, M. C. 1996. “In Singapore / SOCRATES on student learning, followed by lesson execu- Streamlines Lesson Observations, Improving tion with practices that teach students how to Professional Practice.” Educational Leadership learn by directing them to underlying principles, 53 (6): 76–78. developing habits of mind. Chung, C., and M.-Y. Ngan. 2002. From ‘Rooftop’ to ‘Millennium’: The Development of Primary Schools in Hong Kong since 1945. Notes Hong Kong SAR, China: Hong Kong Institute 1. In Hong Kong SAR, China, for example, of Education. teachers received training at night and taught Coe, R., C. Aloisi, L. E. Major, and S. Higgins. during the day (Chung and Ngan 2002; 2014. What Makes Great Teaching? London: Sweeting 2004). Sutton Trust. 2. See “Enabling Teachers,” Ministry of Coflan, A., A. Ragatz, A. Hasan, and Y. Pan. 2018. Education Singapore (https://www.moe Understanding Effective Teaching Practices in . g o v. s g / e d u c a t i o n / e d u c a t i o n - s y s t e m Chinese Classrooms: Evidence from Primary /enabling-teachers). and Junior Secondary Schools in Guangdong, China. Washington, DC: World Bank. Darling-Hammond, L. 2010. “Steady Work: How Countries Build Successful Systems.” In The Referencesa Flat World and Education: How America’s Ager, M. 2014. “Poe Wants P25,000 Monthly Pay Commitment to Equity Will Determine Our for Public School Teachers.” Inquirer Online, Future , edited by L. Darling-Hammond. August 8. New York: Teachers College. Beauchamp, E. R., and J. M. Vardaman. Darling-Hammond, L., M. E. Hyler, M. Gardner, 2015. Japanese Education since 1945: and D. Espinoza. 2017. Effective Teacher A Documentary Study . Abingdon, U.K.: Development. Palo Alto, CA: Learning Policy Routledge. Institute. Benveniste, L., J. Marshall, and L. Santibañez. Darling-Hammond, L., and R. Rothman. 2011. 2007. “Teaching in Lao PDR.” World Bank, Teacher and Leader Effectiveness in High- Washington, DC. http://documents.worldbank Performing Education Systems. Washington, .org/curated/en/967301468265782991 DC: Alliance for Excellent Education and /Teaching-in-Lao-PDR. Stanford, CA: Stanford Center for Opportunity Boylan, M. 2016. “National Curriculum Sub- Policy in Education (SCOPE). Levels: Assessment Practices as Assemblage.” Darling-Hammond, L., R. C. Wei, A. Andree, Paper presented at the “Third Manchester N. Richardson, and S. Orphanos. 2009. Conference on Mathematics Education and Professional Learning in the Learning Profession: Contemporary Theory,” Manchester, U.K., A Status Report on Teacher Development in the July 18–21. United States and Abroad. Stanford, CA: School Center for International Education Benchmarking. Redesign Network at Stanford University. n.d. “Hong Kong.” National Center on Demas, A., M. M. Khan, and G. Arcia. Education and the Economy, Washington, DC. Forthcoming. Delivery of Education Services http://ncee.org. in Lao PDR: Results of the SABER Service Chang, M. C., S. Shaeffer, S. Al-Samarrai, A. B. Delivery Survey , 2017. Washington, DC: Ragatz, J. de Ree, and R. Stevenson. 2013. World Bank. Doig, B., and S. Groves. 2011. “Japanese Lesson a. Titles of publications that include “Hong Kong,” “Korea,” Study: Teacher Professional Development and “Taiwan” refer to Hong Kong SAR, China; the Republic of t h r o u g h C o m m u n i t i e s o f I n q u i r y. ” Korea; and Taiwan, China, respectively. 162 GROWING SMARTER Mathematics Teacher Education and Classroom.” European Journal of Psychology Development 13 (1): 77–93. of Education (31) 3: 299–322. Echazarra, A., D. Salinas, I. Méndez, V. Denis, Lai, K. C. 2009. “Qualifications of the Teaching and G. Reich. 2016. “How Teachers Teach Force in Hong Kong Special Administrative and Students Learn: Successful Strategies for Region, China.” In A Comparative Study of School.” OECD Education Working Paper Teacher Preparation and Qualification in Six 130, OECD Publishing, Paris. http://dx.doi Nations , edited by R. E. Ingersoll, 29–40. .org/10.1787/5jm29kpt0xxx-en. Philadelphia: Consortium for Policy Research Economist . 2016. “Teaching the Teachers.” in Education. June 11. Levin, B. 2008. How to Change 5,000 Schools: A Education World. n.d. “Teachers Observing Practical and Positive Approach to Leading Teachers: A Professional Development Change at Every Level. Cambridge, MA: Tool for Every School.” Education World. Harvard Education Press. http://www.educationworld.com/a_admin Liang, X., H. Kidwai, and M. Zhang. 2016. How /admin/admin297.shtml. Shanghai Does It: Insights and Lessons from Ferreras, A., C. Kessel, and M.-H. Kim. the Highest-Ranking Education System in the 2015. Mathematics Curriculum, Teacher World. Washington, DC: World Bank. Professionalism, and Supporting Policies Lu, P. C., and L. P. Yee. 2011–12. “A Singapore in Korea and the United States: Summary Case of Lesson Study.” Mathematics Educator of a Workshop. Washington, DC: National 21 (2): 34–57. Academies Press. Miao, Z., and D. Reynolds. 2018. “The Fujita, H. 2007. “The Qualifications of the Effectiveness of Mathematics Teaching in Teaching Force in Japan.” In A Comparative Primary Schools: Lessons from England and Study of Teacher Preparation and Qualification China.” Routledge Research in International in Six Nations, edited by R. E. Ingersoll, 41–54. and Comparative Education. Abingdon-on- Philadelphia: Consortium for Policy Research Thames: Routledge Publishing. in Education. National Center on Education and the Goh, C. B., and S. Gopinathan. 2008. “The E c o n o m y. n . d . “ J a p a n : Te a c h e r a n d Development of Education in Singapore since Principal Quality.” National Center on 1965.” In Toward a Better Future: Education Education and the Economy, Washington, and Training for Economic Development DC. http://ncee.org/what-we-do/center-on in Singapore since 1965, edited by S. K. Lee, -international-education-benchmarking C. B. Goh, B. Fredriksen, and J. P. Tan, 12–38. / top-performing-countries/japan-overview Washington, DC: World Bank. /japan-teacher-and-principal-quality/. Hattie, J. 2003. “Teachers Make a Difference: What OECD (Organisation for Economic Co-operation Is the Research Evidence?” Paper presented and Development). 2010. Japan: A Story of at the Research Conference of the Australian Sustained Excellence. Paris: OECD Publishing. Council for Educational Research, Melbourne. ———. 2012a. PISA 2012 [database]. Paris: ———. 2015. “The Applicability of Visible Learning OECD. http://www.oecd.org/pisa/pisaproducts to Higher Education.” Scholarship of Teaching /pisa2012database-downloadabledata.htm. and Learning in Psychology 1 (1): 79–91. ———. 2012b. Preparing Teachers and Developing Kim, E.-G. 2007. “The Quality and Qualifications School Leaders for the 21st Century. Paris: of the Teaching Force in the Republic of Korea.” OECD Publishing. In A Comparative Study of Teacher Preparation ———. 2014. TALIS 2013 Results: An and Qualification in Six Nations, edited by R. E. International Perspective on Teaching and Ingersoll, 55–70. Philadelphia: Consortium for Learning. Paris: OECD Publishing. Policy Research in Education. ———. 2016a. PISA 2015 Results. Vol. I: Kim, H. S. 2000. “Towards Achieving High Excellence and Equity in Education. Paris: Quality Pre-service Teacher Training in Korea.” OECD Publishing. http://www.oecd.org Asia-Pacific Journal of Teacher Education and / publications/pisa-2015-results-volume-i Development 3 (1): 55–77. -9789264266490-en.htm. Künsting, J., V. Neuber, and F. Lipowsky. 2016. ———. 2016b. Ten Questions for Mathematics “Teacher Self-Efficacy as a Long-Term Teachers... and How PISA Can Help Predictor of Instructional Quality in the Answer Them. Paris: OECD Publishing. DEVELOPING SKILLED TEACHERS AND SUPPORTING EFFECTIVE TEACHING 163 http://www.oecd.org/education/ten-questions Bank. https://openknowledge.worldbank.org -for-mathematics-teachers-and-how-pisa-can /handle/10986/21002. -help-answer-them-9789264265387-en.htm. TIMSS (Trends in International Mathematics Popova, A., D. K. Evans, and V. Arancibia. and Science Study). 2015. TIMSS 2015 2016. “Training Teachers on the Job: What International Database. Boston: TIMSS Works and How to Measure It.” Policy and PIRLS International Study Center, Research Working Paper 7834, World Bank, Lynch School of Education, Boston College Washington, DC. and International Association for the Qin, A. 2017. “Britain Turns to Chinese Textbooks Evaluation of Educational Achievement. to Improve Its Math Scores.” New York Times, https://timssandpirls.bc.edu / timss2015 August 5. /international-database. Ragatz, A., S. Iskandar, R. Kesuma, and S. Sugiarti. Tucker, M. 2016. “Asian Countries Take the U.S. 2014. A Video Study of Teaching Practices in to School.” Atlantic, February 29. TIMSS Eighth Grade Mathematics Classrooms: U b a l d e , M . 2 0 0 7 . “ R P ’s B e s t Te a c h e r s Understanding What Teaching Practices Are Are Leaving in Droves.” GMA News, Used, Why They Are Used and How They December 18. http://www.gmanetwork Relate to Student Learning. Jakarta: World .com/news/news/content/73387/rp-s -best Bank. -teachers-are-leaving-in-droves/story/. Schmidt, W. H., C. C. McKnight, R. T. Houang, Walter, C., and J. Briggs. 2012. “What Professional H. C. Wang, D. E. Wiley, L. S. Cogan, and Development Makes the Most Difference to R. G. Wolfe. 2001. Why Schools Matter: A Teachers? A Report Sponsored by Oxford Cross-National Comparison of Curriculum University Press.” University of Oxford, and Learning. Hoboken, NJ: Wiley, Jossey- Department of Education, Oxford. Bass Education Series. Wang, A., A. Coleman, R. Coley, and R. Phelps. Stigler, J. W., and J. Hiebert. 2009. The Teaching 2003. “Preparing Teachers around the World.” Gap: Best Ideas from the World’s Teachers Policy Information Report, Educational Testing for Improving Education in the Classroom. Service, Princeton, NJ. New York: Simon and Schuster. World Bank. 2018. World Development Report Sweeting, A. 2004. Education in Hong Kong, 1941 2018: Learning to Realize Education’s to 2001: Visions and Revisions. Pokfulam: Promise. Washington, DC: World Bank. Hong Kong University Press. Yoon, K. S., T. Duncan, S. Wen-Yu Lee, B. Scarloss, Tan, S. K. S., and A. F. L. Wong. 2007. “The and K. L. Shapley. 2007. “Reviewing the Qualifications of the Teaching Force: Data Evidence on How Teacher Professional from Singapore.” In A Comparative Study Development Affects Student Achievement.” of Teacher Preparation and Qualification in Issues & Answers Rel 2007-no. 033, Regional Six Nations, edited by R. E. Ingersoll, 71–82. Educational Laboratory Southwest (NJ1), Philadelphia: Consortium for Policy Research U.S. Department of Education, Institute of in Education. http://www.cpre.org / images Education Sciences, Washington, DC. /stories/cpre_pdfs/sixnations_final.pdf. Zheng, Z., S. T. Peverly, and T. Xin. 2006. Tandon, P., and T. Fukao. 2015. Educating the “Knowing and Teaching Fractions: A Cross- Next Generation: Improving Teacher Quality Cultural Study of American and Chinese in Cambodia. Directions in Development— Mathematics Teachers.” Contemporary Human Development. Washington, DC: World Educational Psychology 31 (4): 438–57. Getting Children Ready to Learn Data from international student assessments services. Ensuring quality is a continual chal- suggest that early efforts by Top Performing lenge, especially when seeking to coordinate Systems in East Asia to universalize preschool across actors such as private sector provid- progressively have contributed to their success. ers, parents, and families. The costs of inac- A review of the region’s post–World War II tion are high, but acting to close the access history suggests that these systems focused on gap is affordable. Solutions have been tested children’s physical and cognitive development, and proven in the region, and in some systems assessed and improved the quality of the ser- they are being attempted at scale. vices they offered, and coordinated across Merely increasing the supply of services actors to deliver needed services. for children’s physical and cognitive devel- As other countries in the region seek to opment will not improve school readiness— emulate this success, their strategies will need countries need to improve the quality of these to focus on physical and cognitive well-being services. In most cases, close coordination and emphasize quality as central elements across actors will be required to ensure that of improving children’s readiness to learn these services are catering to the full range of (boxes 6.1 and 6.2). Country strategies also physical, cognitive, and socioemotional needs need to make allowances for their own current of young children. state of preparedness to learn from others. For instance, efforts at improving readiness to learn are beset by large-scale nutritional chal- Top Performing Systems lenges in several East Asia and Pacific econo- invested in readiness to learn mies. Policies have to be grounded in the fact for lasting returns that gaps in children’s readiness to learn mani- Data from the Organisation for Economic fest themselves early and, if unaddressed, can Co-operation and Development (OECD) affect children’s cognitive and noncognitive Programme for International Student skills in the long term. Assessment (PISA) reveal that, among Despite growing evidence of the efficacy of 15-year-olds in school, children who attended early childhood education and development preschool scored a full year ahead of their services, some education systems continue to peers who did not, after controlling for socio- lag in ensuring delivery of key packages of economic differences (OECD 2013, 2016c). 165 166 GROWING SMARTER BOX 6.1 What is readiness to learn? Readiness to learn is a multifaceted construct. It is sounds and letters, they will quickly fall behind, as much about children’s readiness for school as it is leading to life-long gaps in achievement. about schools’ readiness for children. Readiness to Unfamiliarity with numbers can cause primary learn describes children’s intellectual and socioemo- school students to fall behind in their first year of tional development as they prepare to enter primary primary school, with a magnifying effect as they school. Successful transition to school depends on progress through grades. children’s ability to keep up with the academic con- Children should be able to regulate their emo- tent as well as to integrate into the social environ- tions, follow instructions, interact with peers, and ment, including through relationships with peers and pay attention. Educators note the importance of the teachers. In some settings, children who are socio- socioemotional side of readiness to learn. Familiarity emotionally vulnerable lag behind children who are with letters and numbers is not sufficient if the child not vulnerable by about 9–12 months of learning in lacks the ability to sit still and engage in the lesson. primary school. A child’s success in primary school relies on many Readiness to learn also encompasses a family’s social skills, such as cooperation with peers, team- readiness to support their child’s education. Parents work, emotional regulation—all skills that are also are a child’s first teachers and supporters. From birth essential for success in a career and everyday life as to about three years old, they need to provide proper an adult. nutrition, health care, and a supportive, nurtur- These skills begin at home and are strengthened ing environment. From about age four to entrance through preprimary programs. Readiness to learn into primary school, they can support readiness by is a holistic concept, essential to success not just in putting their children into preschool and providing primary school but throughout life. Strong support emotional support and a stimulating learning envi- for families’ efforts to assist in their child’s academic ronment at home. and socioemotional development pay high dividends Academically, children need foundational at low cost. numeracy and literacy skills; schools have to be ready to help children to build on these founda- Sources: Blair 2002; Bowman, Donovan, and Burns 2001; Bradley and Corwyn 2005; Brazelton and Greenspan 2000; Britto, Fuligni, and Brooks-Gunn 2003; tions. If children enter primary school without Dockett and Perry 2007; Ginsburg, Lee, and Boyd 2008; Minujin, Delamonica, the requisite language skills, such as recognizing and Komarecki 2006; Perez and Gauvain 2009; Raver 2004. BOX 6.2 The early years are the cornerstone of readiness to learn The early years are the foundation for a child’s Development starts in the womb. The brain devel- readiness to learn. There is no other time during ops rapidly throughout gestation. Poor maternal which the brain develops so rapidly or is as sensi- nutrition and ill health can hinder child develop- tive to learning. It is during this time that the child ment. However, improved nutrition alone is not suf- begins developing the framework for later cogni- ficient to make sure children are able to reach their tive and noncognitive skills. The skills developed in full potential. To support the developing brain, coun- the early years influence learning and development tries need to bring to bear both health and education not just through primary school but also late into policy tools in support of expectant mothers, their life. A focus on the early years includes both at- families, and their children. home and preprimary education interventions and The early years are the most rapid period of support. neuron growth for the brain. Essential to this box continues next page GETTING CHILDREN READY TO LEARN 167 BOX 6.2 The early years are the cornerstone of readiness to learn (continued) growth is interaction with the child. Lack of inter- early years relates directly to readiness to learn start- action or a stressful environment can inhibit or ing in primary education. reverse growth. Investments yield benefits across a range of out- Early development has many facets, includ- comes, in the short term and the long term. A focus ing cognitive and noncognitive domains. Cognitive on the early years pays dividends over the life domains include thinking and reasoning, while non- span, from primary school through adulthood. In cognitive domains encompass socioemotional and fact, early inputs have a multiplier effect, as skills other related skills. Families and communities need beget skills. Support for development in the early to support development of these domains long before years leads to more than just increased educational children enter compulsory education. outcomes. It also improves employment outcomes There is strong evidence that these facets are and health. linked to readiness to learn. Far from being abstract Sources: Conti, Heckman, and Urzua 2010; Cunha and Heckman 2007; Gunnar concepts with little relation to education, develop- and Fernald 2009; Heckman and Kautz 2012; Heckman, Pinto, and Savelyev ment of cognitive and noncognitive domains in the 2013; Schweinhart 2005; Young and Mustard 2007. Abundant evidence suggests a link between TABLE 6.1 Long-term gains in test scores underscore the importance access to preschool services and long-term of investments in readiness to learn benefits. Among Top Performing Systems, Compound annual growth research from Japan (Nitta and Nagano Economy rate in test scores (%) Period 1975), the Republic of Korea (Rhee and Lee Macao SAR, China 0.63 2005–15 1990), and Singapore (Sim and Kam 1992) Hong Kong SAR, China 0.57 1980–2015 suggests that participating in preschool pre- Korea, Rep. 0.32 1995–2015 pared children for elementary school and Japan 0.18 1965–2015 supported the development of traits such as Taiwan, China 0.15 2000–15 sharing and cooperating. The evidence on Thailand –0.21 1980–2015 the relationship between access to preschool Source: Calculations based on Altinok, Angrist, and Patrinos 2018. and long-run improvements in learning is mixed, but the number of rigorous stud- ies showing a correlation is growing (see, for instance, Berlinski, Galiani, and Gertler 2009; Brinkman and others 2017a). Top Performing Systems strived Long-term improvements in test scores are progressively to universalize early also suggestive of the benefits of school readi- childhood education ness (Altinok, Angrist, and Patrinos 2018). Economies can be differentiated by the degree Hong Kong SAR, China; Korea; Macao of continuity of their early childhood educa- SAR, China; and Taiwan, China experienced tion policies, whether services were targeted or strong improvements in the learning out- universal, and whether multiple sectors were comes of their secondary schools (table 6.1). involved.1 Top Performing Systems such as Since 2015 (the end year of the data reported Japan and Korea had policies and curricula in in table 6.1), various policies have sought to place well before the turn of the 21st century, increase access to and improve the quality of but they continue to refine both curricula and preschool services. standards for services. Singapore has been 168 GROWING SMARTER offering low-fee kindergarten since the mid- beginning of compulsory schooling). The 1980s. In all three countries, different minis- National Guidelines for Care and Education tries have been involved in the provision of at Day Nursery were first devised in 1965. child care and kindergarten services. In each They were revised in 1990, 1999, and 2008, country, policies have ebbed and flowed with based on social transformation and current demographic and socioeconomic trends. As best practices. Major changes in the most more women entered the labor force, there recent guidelines increase support for parents was an increase in child care services. The of preschool-age children and enhance the timing varied across Top Performing Systems. requirements and objectives of nurseries and In Japan, the increase took place in the after- child care centers (Shishido 2008). math of World War II; in Singapore it took In Singapore, three waves of reforms—in place during the 1970s. 2000, 2008, and 2012—sought to improve the quality of teachers, centers, and programs and improve accessibility and affordability. Top Performing Systems assessed and By 2009, 97 percent of children were reported improved the quality of preschools to be in preschool. Top Performing Systems such as Hong Kong SAR, China and Korea are especially Top Performing Systems leveraged the comprehensive in ensuring quality through private sector but insisted on quality school self-assessment and monitoring by and affordability the Department of Education. In Korea, the Nuri Curriculum was introduced in 2012 to The private sector provides a substantial share integrate curriculum standards for nursery of preschool enrollment in East Asia and centers and kindergartens for an emphasis on Pacific. Governments differ in the degree to holistic child development. It lays out com- which they monitor the quality of preschools prehensive lesson content and learning objec- and incentivize the provision of services to all tives. Korea assesses service and staff quality, children. providing funding and training to monitor In Japan, 55 percent of total expendi- results (OECD 2016a). ture on early childhood institutions comes In Hong Kong SAR, China, the govern- from private sources, largely households ment has put in place performance indica- (the OECD average is 19 percent). Less than tors to measure the quality of early childhood 30 percent of children enrolled in preschool education.2 It has also issued a guide to the are in public institutions (the OECD average preprimary curriculum, which stipulates areas is 68 percent); 71 percent of children in pre- and objectives of early childhood learning. school attend independent private institutions Its quality assurance framework comprises (the OECD average is 11 percent) (OECD annual school self-evaluations and external 2014). In 2013, Japan had 24,043 kindergar- quality reviews. Self-evaluations collect data tens, of which more than 58 percent were pri- on the qualifications of teachers and ask vately owned (figure 6.1). teachers to reflect on administrative struc- In Korea, the government has encouraged ture, parent involvement, and goal-setting an independent private market. As a result, for the next year. External quality reviews more than 91 percent of children between the compare results from classroom observations ages of three and five had access to kindergar- with performance indicators. tens and child care centers in 2013 (KICCE In Japan, curricula and standards for 2013). Historically, parents were expected to early childhood education have been refined fund the full cost of provision (OECD 2004), since the turn of the 21st century, with paral- some of which was borne by their employers. lel frameworks for child care (from birth to Where available, government funding from three) and early education (age three to the the Ministry of Education was targeted GETTING CHILDREN READY TO LEARN 169 FIGURE 6.1 There is a trend of increasing private ownership among early childhood institutions in Japan, 1999–2013 25,000 Number of early childhood institutions 20,000 10,132 10,480 10,108 10,851 9,395 11,246 11,582 9,472 9,625 14,005 9,842 13,405 11,916 12,309 12,707 15,000 10,000 12,875 12,723 12,589 12,426 12,358 12,246 12,090 11,848 11,602 11,327 11,009 10,760 10,280 10,242 10,038 5,000 0 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Public ownership Private ownership Source: UNESCO 2016. at providing kindergartens in rural areas As low-income countries in (Kamerman 2006). Financial support was the region strive to emulate available to children whose households were the success of Top Performing below the 70th percentile of the income distri- Systems, they need to consider bution. As of 2012, all households regardless the status quo of income can be subsidized. In Singapore, the government relies on the Outside the Top Performing Systems, private sector to deliver services but maintains countries are at various stages a strong role in defining and assessing the of drafting and refining national quality and affordability of services. Private strategies sector provision was seen as advantageous Above-Average Performing Systems began because it allowed a spectrum of institutions offering early childhood education and develop- to meet the diverse needs of parents. ment services in the 1970s and 1980s but con- After three decades of nondirect gov- tinued to refine service offerings and curricula ernment involvement, in 2012 Singapore well into the 1980s and 1990s. In China, a pol- announced that it would open 15 government icy and legal framework is in place to support kindergartens, a decision driven by the desire early childhood development in all relevant sec- to maintain affordability in the sector (Choo tors. The Law on Maternal and Infant Health 2010). Early childhood care in Singapore falls Care and the Outline for Chinese Children under the auspices of two ministries, which Development (2011–20) are the two core policy functioned independently until recently. They frameworks. Some services for maternal and have now merged on policies and require- infant health care and early childhood educa- ments regarding accreditation, teacher quali- tion services are neither free nor mandatory. fication, and teacher training and collaborate In the 2000s, Below-Average Performing in developing the Kindergarten Curriculum Systems often issued laws detailing their poli- Framework of the Ministry of Education cies and goals. Thailand established its first (Tan 2017). long-term plans for early childhood education 170 GROWING SMARTER and development in 1979. It developed more The major breakthrough was adoption of detailed policies in 2002 and in the context the Law on Preschool Education in 2008. It of the 2007–16 Long-Term Plan and Strategy stipulates that the provision of food, books, for Early Childhood Care and Development manuals, and toys for children attending state- (children from birth through age five) (Shaeffer owned kindergartens, as well as the norm- 2015). Indonesia passed National Education based variable costs, should be financed from System Law No. 20 in 2003, and the Ministry the state budget. Before the law’s implementa- of Education and Culture developed the tion, parents were obligated to pay half of the Rencana Strategis (Village Development food costs and other expenses. In addition, the Strategic Plan) in 2004. In 2008, the govern- government decided to provide its own finan- ment issued an ambitious policy strategy and cial and technical support for nonformal pro- accompanying guidelines. In 2009, it devel- grams for young children, programs that had oped national standards for early childhood earlier been supported by external stakehold- education and development, which established ers only (UNICEF 2013). the first level of the country’s education system. In Emerging Systems, such as Mongolia, Nutritional challenges impede some early attempts to create a favorable envi- countries’ efforts to improve readiness ronment for preschool education (as early to learn as the mid-1990s) faced fiscal difficulties. In 2005, the National Policy of Integrated The youngest children in East Asia and Pacific Early Childhood Development was endorsed still face high rates of moderate to severe stunt- by a joint order of the Ministry of Health, ing. In a third of the region’s countries, stunting Education, Labor, and Social Welfare to pro- remains highly prevalent, despite decades of mote policy coordination among stakeholders. improvement (figure 6.2). Stunting challenges FIGURE 6.2 Stunting remains prevalent in many countries in East Asia and Pacific, despite decades of improvement, 1986–2015 60 % of children under five who are stunted 50 40 30 20 10 0 te La a DR Ca sia Ph odia on es M ds Va r Vi u m M u Th a d lia a lu a oa i Fij a e i in ng at ur an ys va nm a N -Les na lo ppin an go in m ne oP Ch nu Na ala To ail Tu Gu b Sa et Isl on do ya m or li M ew In i Tim m pu So Pa 1986–95 1996–2005 2006–15 Sources: Estimates by the World Health Organization, United Nations Children’s Fund, and the World Bank. Note: Some data are unavailable for each of the three time frames. GETTING CHILDREN READY TO LEARN 171 national efforts to improve readiness to learn. FIGURE 6.3 Large shares of second-grade students in East Asia Evidence from many countries shows that and Pacific cannot read a single word inadequate nutrition undermines brain stimu- lation that is critical to early development. It also suggests that policies and programs to Indonesia improve readiness to learn may be particularly Tonga beneficial for poor and marginalized groups, especially in poorer countries (Gertler and oth- Myanmar (Yangon) ers 2014; Grantham-McGregor and others Philippines (Maguindanaoan) 2014, cited in UNESCO 2015). Philippines (Hiligaynon) Gaps in readiness to learn manifest Cambodia themselves early Vanuatu Across the region, gaps in readiness to learn Philippines (Ilokano) open up early. If unaddressed, they can linger. Timor-Leste Ability—or inability—to read Philippines (Cebuano) Large differences in early development out- comes are evident across and within countries 0 10 20 30 40 50 60 (figure 6.3). The worst-performing coun- % of second graders who cannot tries generally have Emerging Systems. For read a single word every country in this group, too few students are meeting national standards and an unac- Source: Graham and Kelly 2017. Note: All data are from nationally or regionally representative samples completed in the second ceptable number of students cannot read any half of the school year with the exception of the Philippines. The Philippines conducted its only words at all. Even in countries where zero- nationally representative Early Grade Reading Assessment (EGRA) in 2013 for third and first graders only, in both English and Filipino. The results were similar to those in Indonesia: the rates of zero- word rates are relatively low, results from word readers for third graders were low—at only 1 percent for both languages—and reading Early Grade Reading Assessment (EGRA) data comprehension was at 73 percent for Filipino. On the other hand, English reading comprehension was only 32 percent. The 2014 EGRA study targeted regions and schools that taught in mother- suggest that fluency is not very high. A large tongue languages. Zero-word scores are based on the oral reading fluency subcomponent for all portion of students are still struggling with countries except Cambodia and Timor-Leste, for which the score is based on the familiar word reading subcomponent. In the Philippines, tests were conducted in the indicated mother tongues. basic subtasks. Scores for Timor-Leste are for combined tests of Portuguese and Tetum. Scores for Vanuatu are for Top Performing and Above-Average tests in English. Data are from the following years: Cambodia 2012, Indonesia 2014, Myanmar 2015, the Philippines 2014, Timor-Leste 2009, Tonga 2009, and Vanuatu 2010. Performing Systems tend to perform better than Emerging Systems, with a higher propor- tion of students meeting curricular standards, level of comprehension by the end of second smaller shares of students failing on basic or third grade. They are not ready to “read subtasks, and higher levels of reading compre- to learn.” Data from the nationally repre- hension. However, reports from Indonesia, the sentative Dynamic Indicators of Basic Early Philippines, and Vietnam suggest that too few Literacy Skills test in the United States reveal students are able to read for comprehension that, by the end of the school year, the median even by the end of second and third grades. first grader can read with 95 percent compre- National averages mask regional inequalities, hension. Even a student in the 15th percentile especially in Indonesia and the Philippines. can read with 75 percent comprehension, and In all reports in which disaggregated data are a student in the 1st percentile can read with available, inequality in achievement by gen- 15 percent comprehension (Dewey, Kaminski, der and socioeconomic background is evident and Good 2014). In the EGRA-tested East (Graham and Kelly 2017). Asia and Pacific economies, students in In all the countries and provinces for the 1st percentile are usually reading zero which EGRA data are available, the major- words—far from comprehending any written ity of students are not reading with a high text at all. 172 GROWING SMARTER FIGURE 6.4 Poor children score lower than Ability—or inability—to manage time and nonpoor children on tests of executive function, pay attention and, in the absence of any services, the difference An assessment in rural Indonesia shows that can grow over time gaps in young children’s executive function 90 are apparent at as early as age four (figure 6.4). Executive function is a set of skills that % of children passing stage 1 of a 3-stage card-sorting game 80 allows one to manage time and pay atten- tion. These gaps do not close over time (Jung and Hasan 2016), especially in the absence of 70 interventions. 60 If unaddressed, gaps in readiness to learn linger 50 48 49 50 51 52 53 54 55 56 57 58 59 60 If students are not ready to “read to learn” Age of child (months) going into grades 3 or 4, there is only a weak Poor Nonpoor chance that they will attain a high level of functional literacy by the time they com- Source: Jung and Hasan 2016. plete primary school. Figure 6.5 shows that Note: Data are from a study conducted in rural Indonesia. functional illiteracy (defined as the inability FIGURE 6.5 Rates of functional illiteracy among 15-year-old to comprehend the main message in grade- students vary widely in East Asia appropriate texts) is much more preva- lent among 15-year-olds in Below-Average Indonesia Performing Systems than it is among Top Thailand Performing Systems. If students do not Malaysia learn to read fluently in the early grades, there is little hope that they will develop B-S-J-G (China) the skills to succeed on tests like PISA or in Taiwan, China the workplace. Vietnam Japan Macao SAR, China Low-income countries in the region lack certain key packages Korea, Rep. of services Singapore Hong Kong SAR, China Denboba and others (2014) identify five pack- ages of services that are key for ensuring that 0 10 20 30 40 50 60 all young children reach their full potential % of 15-year-old students who are functionally illiterate (figure 6.6). The stimulation young children Below level 1b Level 1b Level 1a receive at home is often a foundation for the formal stimulation they receive in preschool. Source: PISA 2015 (OECD 2016b). Note: At level 1a, tasks require the reader to locate one or more independent pieces of explicitly stated Consequently, some of these packages focus information; recognize the main theme or author’s purpose in a text about a familiar topic; or make a on families. Investments in these interventions simple connection between information in the text and common, everyday knowledge. The required information in the text is prominent and there is little, if any, competing information. The reader is explicitly lay the foundation for building human capi- directed to consider relevant factors in the task and in the text. At level 1b, tasks require the reader to tal (physical health, cognitive development, locate a single piece of explicitly stated information in a prominent position in a short, syntactically simple text with a familiar context and text type, such as a narrative or a simple list. The text provides support to and socioemotional maturity) and preparing the reader, such as repetition of information, pictures, or familiar symbols. There is minimal competing children to learn. Countries are often able to information. In tasks requiring interpretation, the reader may need to make simple connections between adjacent pieces of information. Below level 1b, the reader is unable to locate a single piece of explicitly support families and parents through preg- stated information in a short and syntactically simple text. Economies labeled with purple type are East nancy and birth, but many struggle to provide Asia and Pacific Top Performing Systems. Those labeled with green type are East Asia and Pacific Above- Average Performing Systems, and those labeled with blue type are East Asia and Pacific Below-Average support for families or for child health and Performing Systems. B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China). development. As a result, many new parents GETTING CHILDREN READY TO LEARN 173 FIGURE 6.6 Five packages of services for children under six improve readiness to learn Quality early childhood Preschool and preprimary programs; package continuity to quality primary schools Immunizations; deworming; Child health growth monitoring and and promotion; optimal feeding development practices; therapeutic zinc package supplementation for diarrhea Birth package Attended delivery; exclusive Pregnancy breastfeeding; package birth registration Antenatal visits; iron and folic acid; counseling on adequate diets Parental support for vulnerable families: planning for families; maternal education; parenting and social networks of Family support and community education; parental leave and adequate child care; prevention and treatment of maternal support depression; social assistance transfer programs; child protection regulatory frameworks package Health, nutrition, and sanitation for families: access to health care; access to safe water; adequate sanitation; hygiene and handwashing; micronutrient supplementation and fortification Pregnancy Birth 1 year 2 years 3 years 4 years 5 years 6 years Source: Denboba and others 2014. lack the information and tools needed to How important are interventions aimed enrich their child’s learning. at parents? Research from across the globe, Below-Average Performing and Emerging including from East Asia and Pacific, Systems appear to be supporting readiness to suggests that both parenting practices learn in a variety of ways, but there are large and children’s participation in preschool disparities in access and take-up across the five services have the potential to increase key packages (figure 6.7). Take-up rates for young children’s exposure to developmen- services during pregnancy and birth are high in tal essentials—opportunities for stimulating most of these countries; availability and take- play, rich language experiences, practice in up is lower for services aimed at families and developing executive function skills, and children before their children reach preschool more. age. In particular, availability and take-up of family support services and services for child Home environments matter for health and development tend to be low even in school readiness countries where preschool coverage rates are high. In contrast, in a country such as Korea, The stimulation young children receive availability and take-up of all five packages are at home is often a foundation for the for- higher and more uniform. mal stimulation they receive in preschools. 174 GROWING SMARTER FIGURE 6.7 Families do not have consistent service coverage between pregnancy and preschool 100 90 80 70 % of families 60 50 40 30 20 10 0 Cambodia Indonesia Myanmar Philippines Timor-Leste Korea, Rep. Pregnancy Birth Family support Child health and development Preschool Sources: Data from Demographic and Health Surveys. Preschool attendance data were supplemented by the World Bank EdStats database (World Bank, various years). Note: As described in Denboba and others (2014), pregnancy includes at least four care visits, iron supplementation, and diet counseling during pregnancy. Birth includes having a doctor or nurse present at delivery and breastfeeding. Family support includes mother having completed at least primary education, last birth interval greater than two years and last pregnancy desired, at least three types of stimulating activities at home, health care facility not too far and visited within 12 months, vitamin A and iron supplementation in last six months, safe water source, and improved sanitation. Child health and development includes at least three types of food besides breast milk from six months on; zinc supplement in case of diarrhea; children with weight, height, and height- weight ratio less than 2 standard deviations below the mean; and access to deworming medicine. Preschool measures enrollment in preprimary education. No data were available on non-breast milk nutrition for the Philippines. Only Cambodia had information on the provision of diet counseling during pregnancy and availability of deworming medicine. Yet many new parents lack the information in Mongolia, Thailand, and Vietnam. In the and tools they need to enrich their children. Lao People’s Democratic Republic, where With the proper support, parents can help the gap is extremely wide, preschool atten- to improve their children’s basic literacy. dance accounts for most of the variance There are substantial wealth gaps in children’s (figure 6.10).4 These results are consistent ability to perform basic functions such as with evidence from low-income settings counting from 1 to 10 (figure 6.8). These gaps showing that 10–20 percent of the gap in are also apparent in preschool attendance learning outcomes can be explained by differ- and high-quality care at home (figure 6.9).3 ences in home stimulation, even after control- In Cambodia, the gap in access to preschools ling for maternal education and endowments between the richest and poorest quintiles is (Galasso, Weber, and Fernald 2017). 31 percentage points, and the gap in access to high-quality care at home is 24 percentage Parenting education programs can also points. support noncognitive skills Although both home and school factors help to explain differences in basic skills, The importance of home and preschool home environments (interactions with the environments is not limited to basic literacy child, number of books at home, having at and numeracy. Studies conducted in settings least two playthings at home, leaving the child as diverse as Indonesia and Mongolia that unattended, and mother’s years of school- collected detailed information on socioemo- ing) account for most of the gap observed tional development of children show that GETTING CHILDREN READY TO LEARN 175 FIGURE 6.8 Children from the poorest households are far less likely than children from the richest households to be able to count from 1 to 10 100 90 % of children able to count from 1 to 10 80 80 70 72 65 67 60 61 58 55 56 54 50 52 40 39 41 33 35 30 32 31 23 25 20 21 20 20 10 0 Cambodia Lao PDR Mongolia Mongolia Thailand Vietnam Vietnam (2014) (2012) (2010) (2013) (2012) (2010) (2014) Poorest quintile Richest quintile Mean Source: Authors’ calculations based on data from Multiple Indicator Cluster Surveys (MICSs). Note: The richest and poorest quintiles refer to children who come from families with the top 20 percent and the bottom 20 percent of household income, respectively. FIGURE 6.9 Access to both preschool and high-quality home care shows wide wealth gaps 100 99 95 96 % of children receiving high-quality home care 90 91 88 86 87 88 85 80 84 75 % of children in preschool and 70 70 71 73 69 60 58 57 56 54 50 49 44 40 38 42 37 37 30 32 24 20 21 10 7 5 0 Preschool High-quality Preschool High-quality Preschool High-quality Preschool High-quality Preschool High-quality (DHS 2014) home care (MICS4 home care (MICS5 home care (MICS4 home care (MICS5 home care (DHS 2014) 2012) (MICS4 2012) 2013) (MICS5 2013) 2012) (MICS4 2012) 2014) (MICS5 2014) Cambodia Lao PDR Mongolia Thailand Vietnam Poorest quintile Richest quintile Mean Source: Authors’ calculations based on data from Multiple Indicator Cluster Surveys (MICSs) and Demographic and Health Surveys (DHS). Note: The richest and poorest quintiles refer to children who come from families with the top 20 percent and the bottom 20 percent of household income, respectively. 176 GROWING SMARTER FIGURE 6.10 Home environments explain a holds annual seminars for parents of chil- substantial portion of the wealth gap in ability dren from birth to age six, inviting speak- to count from 1 to 10 ers such as professors, Education Bureau officers, and specialists in childhood devel- 40 opment to present and create written mate- Decomposition of wealth gap (%) 5 35 rial.5 Parenting guides are available in five 30 3 17 languages.6 The number of parents and fam- 25 ilies that take advantage of such government 20 15 24 resources is unclear. 10 18 22 Indonesia’s National Board on Family 12 Planning (BKKBN) provides content on all 5 2 5 0 domains of child development for children –2 0 –5 from birth to age six. Topics covered include Lao PDR, Mongolia, Thailand, Vietnam, 2011–12 2013 2012–13 2013–14 discipline and behavior management, promo- Early childhood development attendance tion of social skills, and parental attitudes Home environment Individual characteristics and self-awareness. In 2012, there were 84,000 BKKBN programs across Indonesia, Source: Authors’ calculations based on data from Multiple Indicator Cluster serving 3.7 million families. The rate of active Surveys (MICSs). participation (defined as having attended a BKKBN program in the past three months) both environments matter, with different was just 16 percent, however (Tomlinson degrees of importance depending on the and Andina 2015), and the quality of the domain of child development (figures 6.11 material covered was less than satisfactory, and 6.12). according to an evaluation. The language Parent support services on topics relat- used in the material was convoluted, and ing to early childhood development vary certain resources were confusing, resulting in widely in content and coverage. In Hong the limited application of new knowledge at Kong SAR, China, the Education Bureau home by parents. FIGURE 6.11 Both parenting practices and participation in early childhood education and FIGURE 6.12 Both home engagement and development services affect the development of time in kindergarten affect the development of four-year-olds in Indonesia five-year-olds in Mongolia Participation in early childhood education and At-home development engagement services Parenting practices Time in kindergarten 0 .05 .10 .15 .20 .25 .30 Influence on development of four-year-olds as 0 .01 .02 .03 .04 .05 .06 .07 .08 .09 .10 measured by Early Development Instrument (EDI) (standard deviation units) Influence on development of five-year-olds (standard deviation units) Physical health and well-being Emotional maturity Numeracy Literacy Communication and general knowledge Language and cognitive skills Socioemotional Executive function Social competence Draw-a-man Source: Hasan, Hyson, and Chang 2013. Source: World Bank 2017b. GETTING CHILDREN READY TO LEARN 177 Across all settings—including high- average center is seldom rated “good,” let income countries—the average preschool alone “excellent,” on this scale. seldom achieves a rating of “good” Evidence from across East Asia and Pacific suggests that the quality of services Three measures of preschool quality appear remains low even when countries have stan- regularly in the literature: observed pre- dards for service delivery. This is particularly school quality, teacher characteristics, true in countries striving to increase the sup- and structural characteristics. The Early ply of services while simultaneously upgrad- Childhood Environment Rating Scale-Revised ing quality. Evidence from Indonesia, for (ECERS-R) (Harms, Clifford, and Cryer instance, suggests that only about 80 percent 2005) collects information on each of them. of services can meet local service standards ECERS-R scores incorporate observer assess- (figure 6.14) and even less meet the mini- ments of space and furnishings, personal care mum score for quality as outlined in the and routine, language and reasoning, activi- ECERS-R. ties, interactions, program structure, and In Thailand, the trade-off between qual- information on parents and staff. Figure 6.13 ity and supply is clear. Of more than 20,000 captures how varied the quality of preschool early childhood development centers in the services can be. It also underscores the diffi- country, 67 percent passed the quality assess- culty of providing high quality. Across all set- ment survey conducted in 2013.7 tings—including high-income countries—the FIGURE 6.13 The quality of preschools varies widely across settings, and average quality is low Canada (326) 4.7 Sweden (27) 4.5 United Kingdom (141) 4.3 Beijing, China (40) 4.3 Kenya, Tanzania, and Uganda (46) 4.1 Rural Indonesia: project playgroup (236) 3.1 Rural Indonesia: kindergarten (221) 2.9 Korea, Rep. (24) 2.9 Bangladesh (22) 2.9 Rural Indonesia: non-project playgroup (70) 2.7 Brazil (138) 2.6 Rural Indonesia: Islamic kindergarten (50) 2.5 Kunming, China (24) 2.2 0 1 2 3 4 5 6 7 Inadequate Minimal Good Excellent Preschool quality (ECERS-R) Sources: Aboud 2006 for Bangladesh; Campos and others 2010 for Brazil; Goelman and others 2006 for Canada; Liang, Zhang, and Fu 2013 for Beijing and Kunming; Malmberg, Mwaura, and Sylva 2011 for East Africa; calculations for Indonesia based on data collected in 2013 from World Bank Indonesia Early Childhood Education and Development Project; Sheridan and others 2009 for the Republic of Korea and Sweden; and Sylva and others 2006 for the United Kingdom. Note: None of the samples is nationally or regionally representative. Numbers in parentheses are number of centers observed. ECERS-R = Early Childhood Environment Rating Scale-Revised. Rural Indonesia project playgroup refers to playgroups opened as part of the World Bank Indonesia Early Childhood Education and Development Project. 178 GROWING SMARTER FIGURE 6.14 About 80 percent of preschools in rural Indonesia meet local standards, and fewer meet the minimum score for quality as outlined in ECERS-R 100 90 87 81 % of preschools meeting standards 80 79 79 70 69 60 50 48 45 43 40 33 30 22 20 10 0 Project Kindergarten Islamic Non-project Total playgroup kindergarten playgroup Meets minimum Indonesia ECED standards (ECERS-R >/= 2) Meets minimum score for quality on ECERS-R (ECERS-R >/= 3) Source: Brinkman, Hasan, and others 2016. Note: Evidence is based on a sample of preschool services in rural Indonesia. Data are not nationally representative. ECED = early childhood education and development; ECERS-R = Early Childhood Environment Rating Scale-Revised. Access to early childhood education is The costs of inaction are enormous. Without linked to positive outcomes for both intervention, children who are at risk of poor children and their mothers development in low- and middle-income countries are expected to earn 26 percent Evidence from 7,520 children and their care- less as adults than they would if they had givers in northern Lao PDR shows that chil- reached their full developmental poten- dren who attended preschool outperformed tial (Richter and others 2017). Globally, a children who did not across a range of devel- 25 percent increase in preschool education is opmental domains (figure 6.15). estimated to yield returns of US$10.6 billion Figure 6.16 documents the impacts of (Lancet 2011). expanding access to preschool services on Beyond earning potential, there are school readiness among a cohort of poor four- long-term economic and social benefits to year-old children in rural Indonesia. It sug- acting. An abundance of research shows the gests that across several domains—including correlation between effective early child- social competence, emotional maturity, and hood intervention and an individual’s qual- language and cognitive development—access ity of life, participation in crime, intelligence to preschool is beneficial if the quality of ser- quotient, and mothers’ labor outcomes vices is high. (see box 6.3). Research suggests a 13 percent return on investment for comprehensive, The costs of inaction are high, high-quality education for children from birth and countries can afford to act to age five (Garcia and others 2016). Left unchecked, the wide socioeconomic Lopez Boo, Behrman, and Vazquez (2016) gaps in early childhood interventions will estimate the costs of inaction for closing the have high costs for individuals and society. socioeconomic gap in two early childhood GETTING CHILDREN READY TO LEARN 179 FIGURE 6.15 Access to preschool is associated with a range of child development outcomes in the Lao People’s Democratic Republic 1.0 0.9 Average child development score 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Overall Perseverance Cultural Socioemotional Approaches Numbers Literacy Communication development knowledge skills to learning and concepts Child attended early childhood education Child had no early childhood education Source: Brinkman, Sincovich, and others 2016. FIGURE 6.16 Access to preschool is associated interventions: preschool and stimulation with a range of child development outcomes in home visits in six countries in Latin America Indonesia, particularly for poor children and the Caribbean. Applying their approach for the subset of East Asia and Pacific econ- omies that took part in PISA 2015 suggests Physical health 0.143 that closing the gaps in learning between and well-being children in the highest and lowest income Social quintiles would cost countries about 0.03– competence 0.267** 0.46 percent of gross domestic product (GDP) (figure 6.17). The cost of inaction is higher for Emotional 0.223** countries with larger proportions of students maturity performing below basic proficiency. Indonesia Language and is an exception only because government cognitive 0.199** expenditure for preprimary education is very development low; if Indonesia had the same level of public Communication expenditure per student as Thailand, the cost and general 0.089 knowledge of inaction would be 0.35 percent of GDP. 0 0.050 0.100 0.150 0.200 0.250 0.300 Impact of preschool Tested and proven solutions exist across (standard deviation units) the region Community-managed play-based activities Source: Brinkman and others 2017a. In rural Indonesia, increasing the sup- Note: Data are for 2013. Data in clear bars show no statistical significance. **p < .05. ply of low-cost, government-sponsored, 180 GROWING SMARTER BOX 6.3 Child care services can improve women’s labor force participation: evidence from Indonesia Increasing female labor force participation is essen- unpaid family worker during the year of giving birth tial for economic growth and prosperity; economic is higher for women without elderly family members. empowerment of women is a prerequisite for sus- Without access to elderly household members to pro- tainable development. Gender equality is a catalyst vide informal child care, it takes an average of six years for change in multiple dimensions of development, for women in rural areas to transition out of unpaid including the health and educational attainment of family work (eight years in urban areas). Women with- future generations. In countries and regions with out support from elders are more likely to move out rapidly aging workforces, raising female labor force of manufacturing jobs and into less lucrative jobs than participation can help to offset the impact of the are women who have such support. shrinking working-age population. Unmet child care needs have an economic Access to child care is an important determi- cost in terms of forgone earnings. In urban areas, nant of labor force participation for urban women. Indonesian women with access to elderly family mem- Elderly members within the household often act as bers return to work an average of two years after giv- informal child care providers, alleviating the need for ing birth; women without access to such care return women to stay home to take care of their children. after four years. It is estimated that women who Data from Indonesia show that, before motherhood, do not have elderly family members at home forgo the presence of elderly family members at home has US$1,300 in earnings in the two extra years it takes no effect on the likelihood of an urban woman work- for them to return to regular work. The transition out ing. But at age 26, when fertility begins to peak, the of manufacturing occupations and into jobs like sales share of urban women who work is 10 percentage and agricultural work is associated with annual for- points higher among women with elderly family gone earnings of US$319 and US$255, respectively. members at home. And by age 27, this margin grows This sector switch is permanent; the likelihood of to 19 percentage points. working in other sectors does not return to prepreg- Without access to child care, women tend not to nancy levels in either rural or urban areas. return to the jobs they had before giving birth. In both urban and rural areas, the probability of working as an Source: Halim, Johnson, and Perova 2017. community-based playgroups helped to raise enrollment in preschools and the provi- enrollment rates and the duration of enroll- sion of lunch subsidies for poor and ethnic ment. Greater exposure to the program led minority children to increase demand. The to modest and sustained impacts on child project also promoted quality through more development, especially for children from and better teacher training, the develop- more disadvantaged backgrounds (Brinkman ment of a quality assurance system, and the and others 2017b). These improvements were establishment of an accreditation process. large enough to close the achievement gap It supported policy measures to determine between richer and poorer children (Jung and standards and authorize quality assurance Hasan 2016). At an annual cost of US$30 per systems that would guide investments in child (in 2014 US$), conservative estimates early childhood education and development suggest that this intervention had a benefit- and strengthen monitoring and reporting. cost ratio of 1.3–4.3 (Brinkman and others The results-based project enabled the gov- 2017a). ernment to increase its financial commit- In Vietnam, the School Readiness ment to early childhood efforts. After four Promotion Project was designed to sup- years, the program achieved the following port the expansion of full-day, full-year (World Bank 2017a): GETTING CHILDREN READY TO LEARN 181 FIGURE 6.17 Closing the achievement gap between socioeconomic groups is affordable 7.0 6.3 6.1 6.0 Government spending (% of GDP) 5.0 4.6 4.1 4.0 3.8 3.9 3.3 3.0 3.0 2.0 0.9 1.0 0.2 0.1 0.4 0.2 0.5 0.1 0.1 0.1 0.1 0.03 0.1 0.1 0.1 0.1 0 Japan Korea, Rep. Singapore B-S-J-G Vietnam Indonesia Malaysia Thailand (China) Top Performing Systems Above-Average Below-Average Performing Systems Performing Systems Total government spending on education Government spending on preprimary education Additional spending required to close learning gap between children from richest and poorest households Sources: Authors’ calculations based on data from PISA 2015 (OECD 2016b), UIS, and World Bank and Government of Vietnam 2017. Note: Data on total government spending are the latest available between 2011 and 2014, except 2016 for Indonesia and 2015 for China and Singapore. Cost of inaction is for closing the socioeconomic gap in the percentage of students falling below basic proficiency in science. Data on percentage of government spending on preprimary are from UNESCO Institute for Statistics, Vietnam Ministry of Economy, and Vietnam Ministry of Finance for 2013, except 2012 for the Republic of Korea and Vietnam and 2014 for Japan. No data available for preprimary spending by B-S-J-G (China). B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); PISA = Programme for International Student Assessment; UNESCO = United Nations Educational, Scientific, and Cultural Organization. • The share of vulnerable five-year-olds per US$100, depending on the reading test declined from 24 percent in 2012 to domain (Macdonald and others 2017). 12 percent in 2016. • The share of five-year-olds in full-day pre- Home-based programs school rose from 74 percent in 2011 to A home-based school preparation program 88 percent in 2017. in Mongolia helps five-year-olds living in • The number of preschools accredited with the most remote rural areas to get ready for quality level 1 or above reached 41 percent school. Children use a home-based school in 2017. No preschools were accredited in preparation learning program and selected 2012. educational toys and materials. Parents are the • More than 90 percent of preschool teachers primary teachers, engaging with their children and managers completed compulsory pro- through reading, singing, and playing. Local fessional development training programs. teachers train them, using guide books pre- In Tonga, researchers examined the impact pared by the program (World Bank 2017b). on early grade reading outcomes of a low- In Vanuatu, the Literacy Education cost community play-based activity. Using Programme (2005–10) included a component an instrumental variables approach, they on home literacy and parent education. The demonstrated positive impacts of about 0.2 program, which focused on children in grades standard deviation in many, but not all, read- 3 and 4, showed positive impacts on student ing domains. The intervention improved literacy of parents’ attitudes toward educat- test scores by 0.21–0.47 standard deviation ing their children and parent-teacher relation- (equivalent to 0.1–0.6 year of schooling) ships (UNESCO, n.d.). 182 GROWING SMARTER Reading interventions two Khmer-language textbooks were pub- In Tonga, researchers examined the impacts lished. In 2010, the Cambodian Ministry of of a reading intervention on early grade read- Education, Youth, and Sport announced that ing outcomes. The average estimated impact the teaching of reading would be a national was about 0.3 standard deviation, although in priority. The government also established a some reading domains impacts were 0.6–0.7 dedicated National Assessment Office within standard deviation (Macdonald and others a new Department of Education Quality 2017). The cost per child was US$183, and Assurance. These investments in early literacy the estimated net present value of benefits was have reaped significant rewards. In 2014, an US$792, yielding a benefit-cost ratio of 4:1. analysis of the Cambodian Grade 8 National Box 6.4 describes the impact of the Pacific Assessment showed that almost all students in Early Age Readiness and Learning (PEARL) that grade were capable of reading and com- program on improving literacy outcomes in prehending text (World Bank 2016b). the Pacific. Papua New Guinea’s READ PNG pro- Cambodia assessed students on their read- gram boosted literacy at low cost. It ing of Khmer in 2006 and 2009. The poor coupled innovative strategies for word recog- results led to a change in reading method from nition with teacher training and monitoring. whole-word recognition to a vowel and con- Program results showed an average increase sonant combination approach. In 2010–12, of 0.54 standard deviation across the differ- Cambodia focused on reading skills in pilot ent domains of literacy among third graders, schools. Thousands of students and their as measured by the EGRA. The improve- teachers received their own copies of Khmer ment was more than six times the difference language books—a unique intervention at a in reading scores between third and fourth time when most children had to share books— grades, underlining the efficacy of teacher and students were encouraged to read at home. training and support at the core of early lit- More than 24,000 teachers were trained, and eracy program efforts. Teachers had access BOX 6.4 Improving literacy and readiness across the Pacific through the PEARL program Early reading assessments in the Pacific reveal that and more effective teaching methods that target a large proportion of children are not functionally reading and comprehension. It also supports the literate and not ready to succeed in school. A 2009 ministries of education of Papua New Guinea, assessment revealed that only 3 of 10 third graders in Samoa, Tuvalu, and Vanuatu in producing road- Tonga and 2 of 10 in Vanuatu were able to read and maps on how to support early literacy develop- comprehend what they read. Poor reading and com- ment in their respective countries. The program prehension skills persist over time, as evidenced by has advocated for increased parental and commu- the fact that more than half of fourth and sixth grad- nity involvement in helping children to prepare for ers in the Pacific are not meeting regional benchmarks school through play-based learning. in literacy. Results from pilot projects in Tonga indicate that The Pacific Early Age Readiness and Learning the program has raised student learning outcomes. (PEARL) program was designed to improve lit- A randomized control trial showed significant eracy and readiness to learn outcomes across the reading gains, ranging from a half to a full school Pacific. It takes a holistic approach to develop- year’s worth of advancement in one academic year. ing foundations for literacy, while working with Community-based play groups have been operat- families and communities to build enabling envi- ing in 49 communities, benefiting more than 2,000 ronments for early childhood development. The children and their parents. program has provided classroom support and training for teachers in Tonga to incorporate new Source: Patrinos 2016. GETTING CHILDREN READY TO LEARN 183 to scripted instructional materials, structured attending preprimary institutions. The gov- lesson plans, feedback through consistent ernment has created and distributed materi- class observations, and remote support using als and videos on topics such as choosing a mobile text messaging. A training model was good kindergarten, appreciating child growth developed to be responsive to teachers’ needs. and development, and understanding the Key elements included relevant and participa- kindergarten curriculum. It promotes par- tory training, opportunities to practice new ent education multisectorally through differ- skills, and monitoring of teachers as they ent departments and bureaus. The Education apply new pedagogy within the classroom Bureau organizes seminars for parents every (World Bank 2016a). year, with the aim of helping them to under- stand age-appropriate expectations for Programs that engage parents cognitive and physical development. The In Mongolia, parents borrow learning materi- Department of Health and the Department of als from a mobile toy and book library estab- Social Welfare also provide parent education. lished in 30 soums (county subdivisions) as These initiatives focus on imparting knowl- part of an early childhood education program. edge and skills on child-rearing, family func- The library includes at least 30 educational tioning, and health.8 kits, with different sets of educational and In Indonesia, the government and non- storybooks, toys, digital tools, and listening governmental organizations run parent resources. Parents and children use each kit education programs that cover many areas for about two or three weeks before receiving of the country. The Ministry of Health dis- another one. The cognitive and noncognitive seminates information on breastfeeding, skills of the children enrolled in the program immunization, health, and safety practices. were significantly higher than those of chil- District health offices offer classes taught dren in alternative programs, underlining the by paraprofessional health workers and potential for a home-based model to improve sometimes invite specialized professionals school readiness among hard-to-reach popu- to speak to parent groups (Tomlinson and lations (World Bank 2017b). Andina 2015). The Ministry of Education and Culture provides grants for funding Programs that simultaneously engage preschool programs that submit successful parents and children proposals to create parent education pro- An initiative in the Philippines that included grams. Parenting programs receiving the a wide range of health, nutrition, early government grant need to require that par- education, and social services programs ents bring their child and interact with him significantly improved the cognitive, social, or her during class. motor, and language development of par- ticipants. By combining center-based services with a program of parent effectiveness ser- Conclusions and vices, the program increased the short-term recommendations nutritional status of children living in pro- Top Performing Systems in East Asia and gram areas, particularly for children under Pacific made substantial investments to sup- age four (Armecin and others 2005). port readiness to learn. These systems con- tinue to do the following: Some economies in the region are • Focus on children’s physical and cognitive attempting to bring these solutions development from birth to scale • Assess and improve the quality of services Hong Kong SAR, China provides compre- • Coordinate across actors to deliver needed hensive support for families with children services. 184 GROWING SMARTER TABLE 6.2 Steps toward supporting children’s readiness to learn Action Below-Average Above-Average Policy Emerging Systems Performing Systems Performing Systems Top Performing Systems Readiness to Develop policies and programs to ensure Ensure that clear quality standards exist, Engage multiple actors in the learn that both cognitive development and are disseminated, and are addressed in development and delivery of integrated physical health are the focus of early an evaluative process. early childhood services. childhood school readiness strategies. Investments in early childhood education issues. For more information about how have been linked to lower crime rates, higher Hong Kong SAR, China measures quality in economic output, and improved quality of life classrooms, see the Education Bureau (http:// (Garcia and others 2016). The costs of inac- www.edb.gov.hk/en/edu-system/preprimary -kindergarten/quality-assurance-framework tion are thus high. /index.html). For others in the region, coverage and 3. If an adult household member engages in quality of early childhood education policies four or more activities that promote learning are mixed. Their governments need to recog- and school readiness during the three days nize the developmental state of their coun- preceding the survey, the quality of home care tries’ early childhood policies and craft plans is considered high; otherwise, it is considered grounded in realities. low. Activities include reading books or Priorities need to be tailored to each econ- looking at picture books; telling stories; omy’s circumstances (table 6.2). For Emerging singing songs; taking children outside the Systems, a primary challenge is to develop pro- home; playing with children; and spending grams that emphasize cognitive and physical time with children in naming, counting, and drawing things. development at a sustained level from before 4. The decompositions focus on the share birth to entry into compulsory education. Once of the gap that can be attributed to these policies are in place, governments need to home environments and the share that can be develop the tools and mechanisms to ensure that attributed to preschool enrollment over and quality standards for readiness to learn services above the influence of parents’ characteristics, are addressed clearly and improved continu- such as education; child characteristics, ously. Above-Average Performing Systems need such as age and gender; and household to engage across actors to deliver services in a characteristics, such as household wealth. manner that leverages multisectoral expertise. The results are qualitatively similar when an index of development is used as the dependent variable. 5. A sample of seminar and pamphlet topics Notes since 2009 includes the following: “Strategies 1. Multisectoral coordination is especially critical, on Development of Young Children’s Reading because programs with multiple components Habit,” “To Nurture or to Pressure?” that target health, cognitive development, and “Helping Your Preschooler Be Resilient,” and emotional development tend to yield greater “Cooperative Parenting with Grandparents.” and longer-term results (Sánchez Puerta, 6. For more information on parenting resources, Valerio, and Gutiérrez Bernal 2016). see the Education Bureau (http://www.edb 2. These indicators cover teaching methods, .gov.hk/en/student-parents/parents-related classroom management, learning atmosphere, /parent-info/index.html). professional attitude, classroom environment 7. For more statistics on early childhood and facilities, home-school relationships, and education in Thailand, see https://www.unicef provision of support to children with family .org/thailand/education_14938.html. GETTING CHILDREN READY TO LEARN 185 8. See the Education Bureau website (http:// Brinkman, S. A., A. Hasan, H. Jung, A. Kinnell, www.edb.gov.hk/en/student-parents/parents N. Nakajima, and M. Pradhan. 2017a. “The -related/parent-info/index.html) for more Role of Preschool Quality in Promoting information on parent education initiatives. Child Development: Evidence from Rural Indonesia.” European Early Childhood Education Research Journal 25 (4): 483–505. Referencesa Brinkman, S. A., A. Hasan, H. Jung, A. Kinnell, Aboud, F. 2006. “Evaluation of an Early Childhood and M. Pradhan. 2017b. “The Impact Preschool Program in Rural Bangladesh.” of Expanding Access to Early Childhood Early Childhood Research Quarterly 21 (1): Education Services in Rural Indonesia.” 46–60. doi:10.1016/j.ecresq.2006.01.008. Journal of Labor Economics 35 (S1): S305–35. Altinok, N., N. Angrist, H. A. Patrinos. 2018. Brinkman, S. A., A. Sincovich, P. Cerdan-Infantes, “Global Data Set on Education Quality (1965– and P. N. Danchev. 2016. “Lao PDR Early 2015).” Policy Research Working Paper 8314, Childhood Education Project: Snapshot World Bank, Washington, DC. Four—Services and Facilities.” World Bank, Armecin, G., J. Behrman, P. Duazo, S. Ghuman, Washington, DC. S. Gultiano, E. King, and N. Lee. 2005. “Early Britto, P. R., A. S. Fuligni, and J. Brooks-Gunn. Childhood Development through an Integrated 2003. “Literacy, Childhood.” In Encyclopedia Program: Evidence from the Philippines.” of Primary Prevention and Health Promotion. Impact Evaluation Series 2, World Bank, Boston, MA: Springer. Washington, DC. Campos, M., Y. Espocito, E. Bhering, N. Gimenes, Berlinski, S., S. Galiani, and P. Gertler. 2009. “The and B. Abuchaim. 2010. “Quality of Early Effect of Pre-primary Education on Primary Childhood Education: A Study in Six Brazilian School Performance.” Journal of Public State Capitals.” Cadernos de Pesquisa Economics 93 (1–2): 219–34. 41 (142). Blair, C. 2002. “School Readiness: Integrating Choo, K. K. 2010. “The Shaping of Childcare Cognition and Emotion in a Neurobiological and Preschool Education in Singapore: From Conceptualization of Children’s Functioning Separatism to Collaboration.” International at School Entry.” American Psychologist Journal of Child Care and Education Policy 57 (2): 111. 4 (1): 23–34. Bowman, B. T., M. S. Donovan, and M. S. Conti, G., J. Heckman, and S. Urzua. 2010. Burns. 2001. Eager to Learn: Educating Our “The Education-Health Gradient.” American Preschoolers [Full Report and Executive Economic Review 100 (2): 234–38. Summary]. Washington, DC: National Cunha, F., and J. Heckman. 2007. “The Research Council, National Academies Press. Technology of Skills Formation.” American https://doi.org/10.17226/9745. Economic Review 97 (2): 31–47. Bradley, R. H., and R. F. Corwyn. 2005. “Caring Denboba, A., R. Sayre, Q. Wodon, L. Elder, for Children around the World: A View from L. Rawlings, and J. Lombardi. 2014. “Stepping Home.” International Journal of Behavioral Up Early Childhood Development: Investing Development 29 (6): 468–78. in Young Children for High Returns.” World Brazelton, T. B., and S. I. Greenspan. 2000. Bank, Washington, DC. The Irreducible Needs of Children: What Dewey, E. N., R. A. Kaminski, and R. H. Good. Every Child Must Have to Grow, Learn, and 2014. DIBELS Next® National Norms Flourish. Cambridge, MA: Da Capo Press. 2012–2013. Technical Report 17. Eugene, OR: Brinkman, S. A., A. Hasan, H. Jung, A. Kinnell, Dynamic Measurement Group. N. Nakajima, and M. Pradhan. 2016. “The Dockett, S., and B. Perry. 2007. Transitions to Role of Preschool Quality in Promoting Child School: Perceptions, Expectations, Experiences. Development: Evidence from Rural Indonesia.” Sydney, Australia: UNSW Press. Policy Research Working Paper 7529, World Galasso, E., A. Weber, and L. Fernald. 2017. Bank, Washington, DC. “Dynamics of Child Development: An Analysis of a Longitudinal Cohort in a Very Low- Income Country.” Policy Research Working a. Titles of publications that include “Korea” and “South Korea” Paper 7973, World Bank, Washington, DC. refer to the Republic of Korea. https://doi.org/10.1093/wber/lhw065. 186 GROWING SMARTER Garcia, J. L., J. Heckman, D. Leaf, and M. Prados. Directions in Development. Washington, DC: 2016. “The Life-Cycle Benefits of an Influential World Bank. Early Childhood Program.” NBER Working Heckman, J. J., and T. Kautz. 2012. “Hard Paper 22993, National Bureau of Economic Evidence on Soft Skills.” Labour Economics Research, Cambridge, MA. 19 (4): 451–64. Gertler, P., J. Heckman, R. Pinto, A. Zanolini, Heckman, J., R. Pinto, and P. Savelyev. 2013. C. Vermeersch, S. Walker, S. M. Chang, and “Understanding the Mechanisms through S. Grantham-McGregor. 2014. “Labor Market Which an Influential Early Childhood Returns to an Early Childhood Stimulation Program Boosted Adult Outcomes.” American Intervention in Jamaica.” Science 344 (998): Economic Review 103 (6): 2052–86. 998–1001. Jung, H., and A. Hasan. 2016. “The Impact Ginsburg, H. P., J. S. Lee, and J. S. Boyd. 2008. of Early Childhood Education on Early “Mathematics Education for Young Children: Achievement Gaps in Indonesia.” Journal of What It Is and How to Promote It.” Social Development Effectiveness 8 (2): 216–233. Policy Report 22 (1), Society for Research in Kamerman, S. B. 2006. “Strong Foundations: Child Development, Washington, DC. Early Childhood Care and Education.” Paper Goelman, H., B. Forer, P. Kershaw, G. Doherty, commissioned for the Education for All Global D. Lero, and A. LaGrange. 2006. “Towards a Monitoring Report 2007, UNESCO, Paris. Predictive Model of Quality in Canadian Child KICCE (Korea Institute of Child Care and Care Centers.” Early Childhood Research Education). 2013. “Policy Brief.” KICCE, Quarterly 21 (3): 280–95. Seoul, Republic of Korea. Graham, J., and S. Kelly. 2017. “Early Grade Lancet. 2011. “Early Childhood Development—A Reading in East Asia and the Pacific.” Winning Combination.” The Lancet 378 Background paper for the East Asia and (9799): 1276. Pacific regional flagship study, Growing Liang, X., Y. Zhang, and Y. Fu. 2013. Challenges Smarter: Learning and Equitable Development and Opportunities: Early Childhood Education in East Asia and Pacific, World Bank, in Yunnan. Washington, DC: World Bank. Washington, DC. Lopez Boo, F., J. R. Behrman, and C. Vazquez. Grantham-McGregor, S., L. C. H. Fernald, R. M. 2016. “The Cost of Inaction in Early C. Kagawa, and S. Walker. 2014. “Effects of Childhood Interventions in Latin America and Integrated Child Development and Nutrition the Caribbean.” Lancet Web Appendix 12. Interventions on Child Development and Macdonald, K., S. Brinkman, W. Jarvie, Nutritional Status.” Annals of the New York M. Machuca-Sierra, K. McDonall, S. Academy of Sciences 1308 (January): 11–32. Messaoud-Galusi, S. Tapueluelu, and B. T. Vu. Gunnar, M., and L. Fernald. 2009. “Effects of a 2017. “Pedagogy versus School Readiness: The Poverty-Alleviation Intervention on Salivary Impact of a Randomized Reading Instruction Cortisol in Very Low-Income Children.” Social Intervention and Community-Based Playgroup Science and Medicine 68 (12): 2180–89. Intervention on Early Grade Reading Halim, D. Z., H. C. Johnson, and E. Perova. 2017. Outcomes in Tonga.” Policy Research Working “Could Childcare Services Improve Women’s Paper 7944, World Bank, Washington, DC. Labor Market Outcomes in Indonesia?” Malmberg, K. A., P. Mwaura, and K. Sylva. East Asia and Pacific Gender Policy Brief 2011. “Effects of a Preschool Intervention on 1, World Bank Group, Washington, DC. Cognitive Development among East-African http://documents.worldbank.org/curated Preschool Children: A Flexibly Time-Coded /en/855851490958133680/Could-childcare Growth Model.” Early Childhood Research -services-improve-women-s-labor-market Quarterly 26 (1): 124–33. doi: 10.1016/j -outcomes-in-Indonesia. .ecresq.2010.04.003. Harms, T., R. Clifford, and D. Cryer. 2005. Minujin, A., E. Delamonica, and M. Komarecki. “Early Childhood Environment Rating Scale.” 2006. Global Child Poverty and Well-Being: Teachers College, Columbia University, Measurement, Concepts, Policy, and Action. New York. Bristol, U.K.: Policy Press. Hasan, A., M. Hyson, and M. C. Chang, eds. Nitta, N., and S. Nagano. 1975. “The Effects of 2013. Early Childhood Education and Preschool Education.” Research Bulletin of the Development in Poor Villages of Indonesia: National Institute for Educational Research Strong Foundations, Later Success. 13: 17–19. GETTING CHILDREN READY TO LEARN 187 OECD (Organization for Economic Co-operation High/Scope Educational Research Foundation, and Development). 2004. “Early Childhood Ypsilanti, MI. Education and Care Policy in the Republic Shaeffer, S. 2015. “The Demand for and the of Korea.” OECD Country Note, OECD Provision of Early Childhood Services since Publishing, Paris. 2000: Policies and Strategies.” Background ———. 2013. PISA 2012 Results: Excellence paper prepared for the Education for All through Equity. Vol II: Giving Every Student Global Monitoring Report 2015, UNESCO, the Chance to Succeed. Paris: OECD Paris. Publishing. Sheridan, S., J. Giota, Y. M. Han, and J. Y. Kwon. ———. 2014. Starting Strong IV: Monitoring 2009. “A Cross-Cultural Study of Preschool Quality in Early Childhood Education and Quality in South Korea and Sweden: ECERS Care. Paris: OECD Publishing. Evaluations.” Early Childhood Research ———. 2016a. Low-Performing Students: Why Quarterly 24 (2): 142–56. They Fall Behind and How to Help Them Shishido, T. 2008. “The Revision of Japan’s Succeed. Paris: OECD Publishing. http:// National Guidelines for Care and Education at dx.doi.org/10.1787/9789264250246-en. Day Nursery and Its Problems.” International ———. 2016b. PISA 2015 Results. Vol. I: Journal of Child Care and Education Policy. Excellence and Equity in Education. Paris: 2 (1): 27–38. OECD Publishing. Sim, K. P., and H. W. Kam. 1992. Growing Up ———. 2016c. PISA 2015 Results. Vol. II: Policies in Singapore: The Preschool Years. Singapore: and Practices for Successful Schools. Paris: Longman Singapore. OECD Publishing. Sylva, K., I. Siraj-Blatchford, B. Taggart, Patrinos, H. A. 2016. “Investing in Early P. Sammons, E. Melhuish, K. Elliot, and Ye a r s L e a r n i n g : I t C a n B e D o n e ! ” V. Totsika. 2006. “Capturing Quality in Early World Bank (blog), Washington, DC. Childhood through Environmental Rating http://blogs.worldbank.org/education Scales.” Early Childhood Research Quarterly /investing-early-years-learning-it-can-be-done. 21 (1): 76–92. Perez, S. M., and M. Gauvain. 2009. “Mother- Tan, C. T. 2017. “Enhancing the Quality of Child Planning, Child Emotional Functioning, Kindergarten Education in Singapore: a n d C h i l d r e n ’s Tr a n s i t i o n t o F i r s t Policies and Strategies in the 21st Century.” Grade.” Child Development 80 (3): 776–91. International Journal of Child Care and Raver, C. C. 2004. “Placing Emotional Self- Education Policy 11 (7): 1–22. Regulation in Sociocultural and Socioeconomic Tomlinson, H. B., and S. Andina. 2015. Contexts.” Child Development 75 (2): 346–53. Parenting Education in Indonesia: Review Rhee, U., and K. Lee. 1990. “The Effectiveness and Recommendations to Strengthen of Four Early Childhood Program Models: Programs and Systems. Washington, DC: Follow-Up at Middle School.” Journal of World Bank. Educational Research 28 (3): 147–62. UIS (UNESCO Institute for Statistics). Available at: Richter, L., B. Daelmans, J. Lombardi, http://uis.unesco.org/. J. Heymann, L. Lopez Boo, J. R. Behrman, UNESCO (United Nations Educational, Scientific, C. Lu, J. E. Lucas, R. Perez-Escamilla, and Cultural Organization). 2015. Education T. Dua, Z. A. Bhutta, K. Stenberg, and G. L. for All 2000–2015: Achievements and Darmstadt. 2017. “Investing in the Foundation Challenges. Paris: OECD Publishing. of Sustainable Development: Pathways to Scale ———. 2016. “Investing in the Foundation Up for Early Childhood Development.” The for Lifelong Learning and Sustainable Lancet 387 (10064): 103–18. Development.” Paper commissioned for the Sánchez Puerta, M. L., A. Valerio, and M. Gutiérrez Financing for Early Childhood Care and Bernal. 2016. Taking Stock of Programs to Education (ECCE), UNESCO, Paris. Develop Socioemotional Skills: A Systematic ———. n.d. “Vanuatu Literacy Education Review of Program Evidence. Directions in P r o g r a m m e ( VA N L E P ) . ” U N E S C O , Development: Human Development Series. Paris. http://litbase.uil.unesco.org/ ?menu Washington, DC: World Bank. =14&programme=34. Schweinhart, L. J. 2005. “The High/Scope Perry UNICEF (United Nations Children’s Fund). 2013. Preschool Study through Age 40: Summary, UNICEF Annual Report 2013—Mongolia. Conclusions, and Frequently Asked Questions.” New York: UNICEF. 188 GROWING SMARTER World Bank. 2016a. Papua New Guinea—Reading ———. Various years. Education Statistics Education Project [English]. Washington, DC: (EdStats) [database]. Washington, DC: World Bank Group. World Bank. ———. 2016b. Raising Reading Scores in World Bank and Government of Vietnam. 2017. Cambodia. Washington, DC: World Bank. Vietnam Public Expenditure Review: Fiscal ———. 2017a. “Implementation Completion Policies towards Sustainability, Efficiency, and Results Report: Vietnam School and Equity. Washington, DC: World Bank. Readiness Promotion Project.” World Bank, https://openknowledge.worldbank.org Washington, DC. /handle/10986/28610. ———. 2017b. Pre-primary Education in Young, M. E., and F. Mustard. 2007. “Brain Mongolia : Access, Quality of Service Development and ECD: A Case for Investment.” D e l i v e r y, a n d C h i l d D e v e l o p m e n t In Africa’s Future, Africa’s Challenge: Early Outcomes . Washington, DC: World Bank. Childhood Care and Development in Sub- https://openknowledge .worldbank.org Saharan Africa, edited by M. Garcia, A. Pence, /handle/10986/26402. and J. L. Evans. Washington, DC: World Bank. Using Student Assessments to Improve Instruction and Learning 7 Because the quality, not just the quantity, of and international large-scale assessments to learning is crucial for growth and develop- spur policy change. A focus on high-stakes ment, countries need to measure learning exams is not the only viable path for other to ensure that the benefits of education countries to follow; the principal lesson from reach all students. It is not enough that chil- the experience of the Top Performing Systems dren are in the classroom—it is imperative that is that the right tool should be used for the they learn. right purpose and that the system should be The literature has established a strong asso- flexible enough to adapt. Using assessments ciation between economic growth and edu- is not enough; systems need to have the right cation quality (see, for example, Hanushek policies and frameworks to support a system and Woessman 2007). Quality is increasingly of assessments. Many countries in the region becoming a global priority. The 2015 Incheon have built systems that include an adequate Declaration’s call for high-quality inclusive enabling environment, mechanisms for qual- education is a testament to the importance ity control, and alignment of assessments of learning outcomes; it also emphasizes the with the rest of the education system. need for systems and practices for assess- Efforts to assess student learning have been ing quality (UNESCO 2015). Its framework closely integrated with ways of taking action emphasizes that improvements to quality and in the education matrix—linking to policies relevance of learning must accompany gains and practices relating to teachers, students, in access.1 Many high-performing countries and curricula. Although the mix of assess- recognize this link and use student assess- ments may vary, high-performing systems ments as a central element in advancing their have well-defined ways of feeding the infor- education systems. mation on student learning outcomes gained Many countries systematically use assess- from such assessments back into the system ments to develop their education systems to drive quality. Assessments can be used (box 7.1). Many Top Performing Systems to feed information back into the system at strategically used high-stakes exams early on the classroom level (training teachers to use to allocate spaces to higher levels of such assessments and incorporating class- education. They have been flexible in using room assessment into curricula), at the school assessments in the classroom and national level (informing principals’ decisions and 189 190 GROWING SMARTER BOX 7.1 Multiple types of assessments can be used Each type of assessment has a different purpose: • L arge-scale assessments (international and national) are generally directed at evaluating • Classroom assessments are used to provide system performance. Examples include inter- results that teachers can use to adapt strategies national assessments such as the Programme and content to students’ needs. Examples include for International Student Assessment (PISA) teacher-designed tests, quizzes, oral questioning, and Trends in International Mathematics and and portfolios. Science Study (TIMSS), regional assessments, • Exams are generally high-stakes assessments used and country-specific assessments, such as the to determine students’ schooling trajectory (often National Assessment of Educational Achievement for transitioning between levels or accessing in the Republic of Korea. specific educational streams). Examples include school-leaving and matriculation exams. Source: Clarke 2011. educational strategies), and at the system level problem solving, well-being, and communica- (using assessment data to enable systemwide tion skills—measurements of learning need to accountability, create a broad commitment to keep pace. This shift presents a range of chal- quality, and spur policy decisions). lenges for assessment. International exams Assessments also have a potentially stron- such as the Programme for International ger link to the public in East Asia and Pacific: Student Assessment (PISA) are already work- they can function as a lever for accountability ing to broaden the scope of skills assessed. to spur quality and equity. With all types of Innovations in assessments in tandem with assessment, adequate communication and use curricular reforms in the Republic of Korea, of test results are important for improving edu- Singapore, and Vietnam offer insights into pri- cation quality and equity. Learning assessment orities for the future. data provide transparency and allow parents to demand quality, make informed choices, and strengthen their role in governance. Within the Multilevel measuring and region, policies and practices for disseminating monitoring of learning outcomes assessment data could be developed further to are critical strengthen the basis for such accountability. Regular information on student learning out- Analysis of practices in the region reveals that comes provides policy makers and educators the poorest students are less likely to attend with a critical tool for improving their edu- schools that publicly share achievement data. cation systems. Various types of assessments Assessments are changing and must con- offer information on students’ knowledge, tinue to do so to keep pace with education skills, and attitudes. Such information can trends, which are looking ahead to the fourth inform key decisions by teachers, including industrial revolution. Many assessments mea- how to tailor instruction to learners’ needs, sure content knowledge and cognitive skills. provide real-time feedback to learners, deter- But education reforms in the region and mine placement and progression of students beyond show that the scope of expected learn- in the education system, and certify levels ing is expanding to meet changing economic of education or skills achieved (Kellaghan realities. As curricula increasingly emphasize 2004). For policy makers, information on noncognitive skills—such as collaborative learning outcomes serves as a window into USING STUDENT ASSESSMENTS TO IMPROVE INSTRUCTION AND LEARNING 191 whether the education system is meeting the into support for the principle of assessment country’s needs. This information provides in education and a strong culture for the use critical input for policy reforms such as cur- of exams (OECD 2012). Such exams comple- ricula and teacher training and for monitor- ment classroom-based assessments, which ing trends and implementation of reforms offer teachers an opportunity to fine-tune (Kellaghan 2004). For the public, assessment instruction. International assessments offer data can increase transparency about the opportunities to benchmark and drive quality. quality of services and provide parents with information on their children’s progress. Exams At each level of the system, data from learning assessments are central in informing High-stakes exams have dominated assessments efforts to improve the quality of the educa- in the region’s Top Performing Systems. In Hong tion and learning experience. At the country Kong SAR, China; Japan; Korea; Singapore; level, international and regional assessments and Taiwan, China, these exams were an offer countries the opportunity to gauge how integral part of education strategies in the their education systems are performing rela- post–World War II period and served to allocate tive to others. Although the emphasis is often limited learning opportunities (Wong 2017). on the competitive nature of such assessments, They also played a central role in the push for the principal value of these assessments is the quality, by gauging teacher effectiveness and opportunity to benchmark. Both international influencing how teachers were trained. and national assessments provide data that can High-stakes exams helped to align edu- be used to examine aspects of education systems cation systems with economic development as well as the factors related to learning, which needs. As education resources were devel- can ultimately drive and inform policy responses oping in the postwar period, the focus was and support system-level accountability. At the on cognitive development in basic education national level, data on student learning can and technical and vocational education and serve as a gauge of how policies are work- training (TVET). In Japan, for example, edu- ing and how results vary across geographic cation policy was oriented to meet economic areas and demographic groups. Early Grade development needs. In the 1950s and 1960s, Reading Assessments (EGRAs) and Early Grade about 40 percent of senior high school stu- Mathematics Assessments (EGMAs), for exam- dents were in TVET. Exams for entry into ple, can provide detailed results on education high school and university were vital (Wong system performance and directly inform teach- 2017). ing methodologies. Systemic assessments also Exams are important as education systems inform central decisions regarding teacher and are developed, but countries need to be cogni- curricular policy (for example, identifying weak zant of risks, address them, and view exams areas and resources needed). At the school level, within the full system of assessment. King and data on student learning can help to distinguish Rogers (2014) examine the changes needed in high- and low-performing schools; identify areas Korea to promote a creative and innovative that require intervention; and, if disseminated, economy. They affirm that rigorous exams serve as a powerful tool for accountability and supported the development of cognitive skills parental choice by increasing the transparency and may have provided incentives to develop of schools’ results. At the individual level, feed- some noncognitive aspects, such as grit. But back from assessments can directly inform stu- they also highlight the heavy costs of the dents’ learning plans and efforts. exam-driven system, including low levels of Top Performing Systems regularly collect happiness, high stress, and private costs that and use information on student learning out- burden families (box 7.2). They call for a bal- comes at multiple levels. Their long history ance to manage these trade-offs to promote of exams in the civil service has translated creativity and innovation. 192 GROWING SMARTER BOX 7.2 Balancing the advantages and disadvantages of high-stakes exams Experience among East Asia and Pacific’s high per- participation rate in shadow education at the formers shows the value of meritocratic, standard- elementary level in the highest income bracket ized selection exams when coupled with good teach- and less than 49 percent in the lowest (Bray and ers, strategic vision, and labor markets that value Lykins 2012). In Thailand, the rise of shadow productive workers. At the same time, exams have education has affected the meaningfulness of potential negative effects: assessment results and adversely affected equity (OECD 2016a). • Student well-being. Highly competitive environ- ments driven by exams can lead to negative men- Countries are implementing reforms to manage tal health outcomes. In the Republic of Korea and these negative effects. Policy changes across countries other countries in the region, the focus on exams indicate a trend toward reducing the unintended effects has led to high stress (Wong 2017). of a singular-focus exam. In response to concerns, in • Crowding out. High-stakes exams narrow the 2013, Korea introduced test-free semesters as a pilot scope of education by incentivizing practices such program for the lower-secondary level. Since then, it as teaching to the test. High-stakes assessment has expanded the program across the system (OECD can also limit innovation in education, as teachers 2012). In 2014, Taiwan, China introduced exam-free are reluctant to try new methods (Looney 2009). pathways to secondary school (Wong 2017). Shanghai • Impact on equit y. Some studies show that has sought to reduce the pressure of the zhongkao delayed tracking can improve academic out- lower-secondary exam at the end of the ninth year. comes (Jakubowski and others 2016) and that Strategies recommended to reduce the dominance of early tracking (via exams) may negatively affect the test include testing more frequently, broadening outcomes for the poorest students, who may the domains tested, and reducing the stress and fear of be disproportionately tracked into vocational failure by having alternate routes to different levels and streams. A focus on exams also leads to “shadow types of education (Liang, Kidwai, and Zhang 2016). education” (private, supplementary learning Although these efforts are at early stages and their full offered outside of the mainstream education sys- impacts are not yet known, they indicate that policy tem), which many disadvantaged students can- makers need to consider balancing the weight of exami- not afford. Data from Korea show a 96 percent nations in a country’s assessment system. Most East Asia and Pacific economies mix of school-based exams and a central- still use exams for entrance decisions at the ized exam that includes more critical thinking secondary cycle, but many Top Performing skills. Efforts have aimed to reduce pressures Systems have removed or adapted high-stakes and some of the negative effects of exams, but exams at lower levels of education. Korea even exams at higher levels have been found removed middle school entrance exams in to create pressures early on. the 1960s and removed high school entrance exams as part of the High School Equalization Classroom assessments Policy of the 1970s. Hong Kong SAR, China removed the public assessment after primary Classroom assessments offer teachers an school in 2001. Singapore still maintains the opportunity to adjust and fine-tune instruc- primary school-leaving exam at the end of tion based on real-time data. Assessment grade 6. procedures that are utilized by teachers in the Shifts are also occurring in other countries. learning process in order to modify teaching In 2013 in Malaysia, the National Education and learning activities are considered forma- Blueprint replaced the high-stakes exam at tive assessments. They are widely used in the end of lower-secondary school with a the region (box 7.3). Classroom assessments USING STUDENT ASSESSMENTS TO IMPROVE INSTRUCTION AND LEARNING 193 BOX 7.3 What does the literature say about the use of formative assessments and student performance? Black and Wiliam (1998) find that formative assess- (2011) use data from New York City charter schools ment is linked to learning gains. Wiliam and others to look at data-driven instruction, which includes (2004) note that students benefited from “authentic how frequently assessments are conducted and how classroom tasks,” outperforming students whose the data are used. They find that high-achieving teachers did not use such tasks. Dobbie and Fryer schools make more use of assessment and data. offer immediate feedback on student per- inform policy. Policy makers view PISA per- formance to guide teaching strategies and formance as an indicator of a school system’s practices. They are particularly important effectiveness. Some of the impacts on policy given the emphasis on the execution of good and practice include actions related to estab- teaching practices and “micro-teaching” lishing or improving national assessment (a teacher training technique involving the systems, revising curriculum standards, pro- steps “plan, teach, observe, re-plan, re-teach, moting equity, setting strategies for improv- and re-observe”) as a means of raising learn- ing student engagement, setting system targets ing outcomes for all. and monitoring, and making changes to stu- A regional study conducted by the World dent tracking (Breakspear 2012). Bank in 2012 using a version of the bench- PISA can spur improvements in teaching marking tool Systems Approach for Better practices. Taiwan, China’s Happy Reading Education Results (SABER) finds that, program, launched in 2008, was a response although almost all countries surveyed had to low performance on PISA 2006. Changes a system-level framework for large-scale, included increasing the time allocated for system-level assessment and exam activities, reading instruction, increasing resources, only half had such a framework for class- and making changes to teacher develop- room assessment activities.2 “In almost all ment (National Center on Education and the of the economies … surveyed, respondents Economy 2014). In Vietnam, legal documen- reported that pre-service teacher training pro- tation on assessments was changed to drive grams provide teachers with courses or work- innovation, spurred by participation in PISA, shops on classroom assessment techniques” and education assessment became a key part (Patrinos 2012). Although SABER provides of curriculum reform beginning in 2015. insight into policy in the region, data on Barriers to impact include the low perceived implementation of classroom assessments are quality of the assessment, lack of or poorly not as readily available. targeted dissemination, and lack of policy- relevant analysis (Tobin and others 2015). Assessments other than PISA and Large-scale international assessments Trends in International Mathematics and Globally, the share of countries participat- Science Study (TIMSS) include the Pacific ing in PISA rose from one-fifth to one-third Islands Literacy and Numeracy Assessment of countries between 2000 and 2015; in East (PILNA) and the PISA for Development (to Asia and Pacific, 20 percent of countries par- be launched in 2018). These assessments ticipated in PISA in 2000, and 35 percent par- provide lower-income countries an oppor- ticipated in 2015 (Lockheed 2015). tunity to participate. PISA for Development The PISA assessment has increasingly increases the relevance of instruments been used to benchmark performance and for low-income countries by including 194 GROWING SMARTER out-of-school children and helps to build these tools can be implemented at a smaller the capacity of participating countries to scale, which can potentially lower costs, conduct large-scale student assessments. allow for more flexibility, and take less time (Wagner 2011). Their use has sometimes spurred systemic changes in teaching meth- National assessments odologies and curricula (table 7.2). Analysis The use of national assessments varies within of EGRA results also allows evaluation of the region (table 7.1). Top Performing Systems which interventions and teaching methods such as Japan and Korea reintroduced such have a positive impact on literacy (Graham assessments on a regular basis about a decade and Kelly 2017). ago. China formally rolled out a national Tonga and Vanuatu illustrate how EGRAs assessment for fourth- and eighth-grade stu- can spur literacy interventions. EGRA anal- dents in 2015. ysis in 2009 revealed low levels of reading Assessments are used more sporadically in and comprehension (with only 3 in 10 chil- other type of systems. Vietnam intermittently dren in Tonga and 2 in 10 in Vanuatu read- has assessed fifth graders, Mongolia has con- ing with comprehension). This evidence ducted the National Assessment of Primary prompted policy makers to adopt the Pacific Education since 2004, and the Lao People’s Early Age Readiness and Learning (PEARL) Democratic Republic uses the National program, which was designed to address Assessment of Learning Outcomes. school readiness and early grade literacy through community-based groups, public awareness, teacher training, and an early- Early Grade Reading Assessments years roadmap. EGRAs, which use oral assessment, have More fundamentally, EGRAs have high- increasingly been used in low- and middle- lighted a literacy crisis in Below-Average income countries to gauge levels of learning Performing and Emerging Systems. Graham in the system (see table 2.4 in chapter 2).3 and Kelly (2017) recommend that these Although more intensive to implement than countries implement a periodic and nation- traditional large-scale assessments (EGRAs ally representative system for early grade must be administered to individual children), literacy testing. TABLE 7.1 The use of national assessments varies within the region Country Assessment Year introduced Target Sample Subject areas China China National Piloted in 2008, Fourth- and eighth- Sample based Chinese, math, Assessment formally rolled out grade students science, physical in 2015; annual education, art, moral education Japan National Assessment of 2007 Sixth-year primary Alternates Japanese, math, Academic Ability and third-year middle between census science school students and sample Korea, Rep. National Diagnostic 2002 Year 3 students Census Reading, writing, Assessment of Basic math Competency National Assessment 2000, extended Years 6, 9, and 11 Census Korean language of Educational to all students in (assessment in sixth arts, math, social Achievement 2008 grade abolished in studies, science, 2013) English Sources: Liang, Kidwai, and Zhang 2016 for China; Kuramoto and Koizumi 2015 for Japan; OECD 2014 for the Republic of Korea. USING STUDENT ASSESSMENTS TO IMPROVE INSTRUCTION AND LEARNING 195 TABLE 7.2 Select countries have changed policy based on data from EGRAs Effect of EGRA data Countries Served as a wake-up call for education reform Cambodia, Indonesia, Philippines, Timor-Leste, Tonga, and Vanuatu Helped government to target regions for increased resources and Indonesia assistance, including by type of instruction in some districts Informed the development of new curricula and teacher training Timor-Leste and Vanuatu based on substantive feedback on specific issues faced by students such as phonetic awareness, decoding, knowledge of the alphabet Helped policy makers to develop clear and appropriate benchmarks Philippines for language, age, and region, enabling progress in tracking and accountability Source: Extracted from Graham and Kelly 2017. Note: EGRA = Early Grade Reading Assessment. Top Performing Systems in East PISA-participating East Asia and Pacific econ- Asia have linked assessments omies use teacher-developed tests; they use with teaching standardized tests less regularly. Of the economies in the region partici- Establishing the right policy pating in TIMSS 2015 (Hong Kong SAR, environment China; Indonesia; Japan; Korea; Malaysia; An assessment system encompasses the “group Singapore; Taiwan, China; and Thailand), of policies, structures, practices, and tools for nearly all use written classroom tests in math generating and using information on student at least sometimes (no more than 4 percent learning and achievement” (Clarke 2011, 4). of respondents in each country answered The 2012 SABER analysis finds that assess- “never”). Some 45–70 percent use written ment systems in the region were fairly well tests in “some lessons,” and 20–40 percent established and that many economies had use them in “about half” of lessons. No more policies for classroom assessments, exams, than 15 percent use tests in “every or almost and large-scale assessments; some also had every lesson,” with the notable exception of strong enabling contexts with clear frame- Thailand, where 54 percent do so. works and dedicated budgets and assessments The use of assessments by school leader- that were aligned with learning standards or ship is a key feature in four countries accord- the curriculum.4 Gaps included monitoring ing to the 2013 Teaching and Learning and quality assurance of classroom assessment International Survey (TALIS). The share of priorities, especially in lower-income coun- lower-secondary education principals who tries such as Cambodia, Lao PDR, and the reported that they use student performance Philippines. and student evaluation results (including In all four performance systems, the dis- national and international assessments) to semination and use of assessment results need develop the school’s educational goals and to be boosted. Box 7.4 provides a deeper look programs was 99.5 percent in Malaysia, at two systems, partly through the SABER 99.3 percent in Singapore, 95.3 percent in lens. Korea, and 93.0 percent in Japan; the survey average was 88.8 percent (OECD 2014). Teacher development systems must sup- Systematically linking assessments port teachers in conducting assessments to instruction and using their data. One of the most cru- Tests are frequently used to guide stu- cial aspects of ensuring the systematic use dent learning (figure 7.1). Almost all of assessments is providing teachers with 196 GROWING SMARTER BOX 7.4 Using SABER to benchmark assessment systems in Vietnam and Shanghai, China Shanghai’s education system is well developed changes. Vietnam moved from “emerging” to “estab- across types of assessments: It is rated “advanced or lished” in both classroom assessment and exams established” in all areas (Liang, Kidwai, and Zhang between 2009 and 2014. Efforts included training 2016) according to the Systems Approach for Better for stakeholders in assessment, widespread dialogue Education Results (SABER) system of ratings.a The on related policy and reform efforts, new policies and Shanghai education system is particularly advanced guidelines, a new competency-based assessment frame- in exams, which incorporate the zhongkao lower- work, participation in Programme for International secondary exam at the end of the ninth year (for Student Assessment (PISA) 2012, leadership from entry to high school) and the gaokao senior-second- the Ministry of Education and Training, and align- ary exam at high school graduation. The system is ment with curricular reforms (Gardner and Clarke rated as “established” in classroom assessment and 2015). Vietnam also conducted Early Development national and international assessments. Contribu- Instrument surveys in 2012 and 2014 to assess the tory aspects include formal policy documentation, school readiness of preschool children (Gardner and dedicated resources, alignment with curriculum stan- Clarke 2015; Ruby and Kent 2015). dards, systematic support of teachers, and quality assurance for the assessment. a. The SABER program collects comparable data on the policies and institutions of education systems around the world and benchmarks them Vietnam demonstrates how countries can use a against good practice. SABER rates policy areas on a four-point scale: from benchmarking approach such as SABER to spur policy “latent” to “emerging” to “established” and to “advanced.” FIGURE 7.1 Almost all PISA-participating East Asia and Pacific economies use teacher-developed tests and use standardized tests less regularly 100 80 to guide student learning % of schools using tests 60 40 20 0 Hong Kong Japan Korea, Rep. Macao SAR, Singapore Taiwan, B-S-J-G Vietnam Indonesia Thailand SAR, China China China (China) Top Performing Systems Above-Average Below-Average Performing Systems Performing Systems Teacher-developed tests Standardized tests Source: PISA 2015 (OECD 2016b). Note: No data available for use of standardized tests in Japan. B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China); PISA = Programme for International Student Assessment. USING STUDENT ASSESSMENTS TO IMPROVE INSTRUCTION AND LEARNING 197 adequate tools, guidance, knowledge, and teachers (OECD 2016a). Portfolio-based skills to implement and use them. Top assessment was introduced in 2012, but Performing Systems have succeeded in implementation was slow, because of teach- doing so by including assessment in teacher ers’ inadequate knowledge and skills in the training programs and providing clear guid- new methodologies required. ance and monitoring their use (box 7.5). Student assessments can also inform In Singapore, educational reform included teacher training and development. Ontario’s efforts to support assessment in the class- Teacher-Learning Critical Pathway uses stu- room, including study of teachers’ practices dent achievement data for each school to and design of a two-year professional devel- identify areas most in need of support. A “big opment program to support effective assess- idea” is identified, and faculty engage on the ment (Sui-chu Ho 2012). subject for six weeks; this process informs col- Innovative practices are emerging in lective efforts by the faculty in teaching and Vietnam. Portfolio-based formative assess- classroom assessment. Assessment of students ment was introduced under the Vietnam is ongoing throughout the period by teachers Escuela Nueva initiative and was institu- working together, which allows feedback and tionalized in 2014 for primary schools. The adjustments (and ultimately promotes teacher new guidance removed continuous test- learning) (OECD 2012). ing and indicated that teachers should use observation and communicate with parents Fostering accountability through (World Bank 2016). Teachers were trained assessments in these methods, which were built into training for school principals and educa- How to empower parents and other local tion administrators at the district and pro- stakeholders to improve a school’s effec- vincial levels. tiveness has been studied extensively in Failure to integrate assessments with the education and education economics lit- teacher development can hinder their erature (see, for example, Barrera-Osorio effectiveness. In Thailand, an obstacle to and others 2009). The main mechanisms implementing assessment-related reforms are school-based management (in which is inadequate capacity of and support for parents are involved in school decision BOX 7.5 Using formative assessments to inform teacher development and policy making in Shanghai, China Schools in Shanghai, China, use quizzes, oral tests, The system is well aligned. It includes guidelines presentations, and homework assignments for for- for conducting formative assessments that focus on mative assessments, which are integrated into other student growth rather than student comparison. areas, such as teacher development and policy Teachers are instructed not to use paper-based tests making. Two factors enable the use of such assess- for the youngest students (first and second graders). ments: (a) guidance and support for teachers and Formative assessments are also a focus of teacher (b) monitoring of the assessments’ quality, which development. The quality of teacher assessments is is used to inform teaching practices and future monitored and is a key area of focus during teacher assessments. Research projects by municipal and observation and performance evaluation. district-level teaching research offices are used to inform policy. Source: Liang, Kidwai, and Zhang 2016. 198 GROWING SMARTER making) and school choice (in which par- mandatory standardized assessments (the ents select a school for their children). Both exception is Japan, where the question pathways rely on the ability of parents to was not included on the PISA question- make informed choices. naire and no data are given ) (figure 7.2). Disseminating student achievement results Except in Macao SAR, China, at least half can have a strong impact on learning out- of schools use the assessments to compare comes, but doing so does not always have the their schools with others. desired results, for several reasons. Parents Few administrators share the achievement involved in school management may require results with the public, particularly at schools additional training to monitor school and attended by the poorest, suggesting inequity teaching practices (Duflo, Dupas, and Kremer in transparency (figure 7.3). 2011), or they may lack influence even if Accountability to parents may be hindered they are ostensibly included in school man- in countries where schools do not publicize agement (Parker 2005). Learning outcome achievement data. The 2012 PISA study col- data may not be fully understood or social- lected data on whether schools compete for ized among parents, especially in disadvan- other students (OECD 2013).5 In most coun- taged contexts (Bruns, Filmer, and Patrinos tries, the share of students attending schools 2011). In studies of large programs, research that compete for students is high (figure 7.4), methods differ, and the causal impact of dis- but many of these schools (especially in Japan seminating achievement results may be con- and Shanghai, China) do not share achieve- founded with that of other reforms (Bradley ment data publicly, leaving many parents and others 2000). disadvantaged when choosing schools, and In all East Asia and Pacific economies weakening the role of school choice as an that participated in PISA 2015, a major- accountability mechanism. ity of students attended schools with FIGURE 7.2 The use of standardized assessments varies within East Asia and Pacific 100 80 % of students 60 40 20 0 Hong Kong Korea, Rep. Macao SAR, Singapore Taiwan, B-S-J-G Vietnam Indonesia Thailand SAR, China China China (China) Top Performing Systems Above-Average Below-Average Performing Systems Performing Systems Attend schools that conduct mandatory standardized assessments Attend schools that use standardized assessments to compare school with other schools Source: PISA 2015 (OECD 2016b). Note: B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China). USING STUDENT ASSESSMENTS TO IMPROVE INSTRUCTION AND LEARNING 199 Parents sit on the governing boards of FIGURE 7.3 Poorer children in East Asia and Pacific attend schools many schools, according to PISA data. But that are less likely to publicize achievement results because many schools do not share achieve- ment data publicly, the lack of transpar- 100 ency hinders them from holding schools 90 % of students at schools that post accountable. 80 achievement data publicly The dissemination and use of assessment 70 data are considered key drivers in improving 60 quality in Chile (Ramírez 2012). Chile uses 50 a national exam (the Sistema de Medición de la Calidad de la Educación) to target 40 specific information to different audiences 30 for each of the assessment’s three purposes 20 (accountability, pedagogical support, and 10 informing policy), using school reports, 0 national reports, newspaper supplements, par- 1 2 3 4 5 ent reports, and online item banks. Many East (bottom) (top) Wealth quintile Asia and Pacific economies need to strengthen Vietnam Taiwan, China Japan accountability, in particular, by sharing learn- Thailand Indonesia Macao SAR, China ing outcome data. They could learn from Korea, Rep. Singapore Chile’s example. Hong Kong SAR, China B-S-J-G (China) Accountability can be supported through a systemic approach that need not rely on Source: PISA 2015 (OECD 2016b). Note: The top and bottom quintiles refer to children who come from families with the top 20 test data. Japan—at the lower end of the percent and the bottom 20 percent of household income, respectively. B-S-J-G (China) = Beijing, spectrum in the use of standardized testing Shanghai, Jiangsu, and Guangdong (China). FIGURE 7.4 Even when competing for students, many schools do not publicize their achievement data 100 90 80 70 % of students 60 50 40 30 20 10 0 a n a sia ia a d re m p. na in in in pa an ys Re po na hi ne Ch Ch Ch Ja ala ail ,C et ga a, do Th ai, R, R, Vi re M an Sin In SA SA gh Ko iw an ng ao Ta Sh ac Ko M ng Ho Attend schools that compete with other schools for students Attend schools that compete with other schools for students and publicize achievement results Source: PISA 2012 (OECD 2013). 200 GROWING SMARTER and publication of data results—has a strong communication and commitment, and that culture of accountability (OECD 2012). Its these skills are associated with higher wages system works because teachers collaborate, and employment probabilities (Acosta and teacher quality is known to peers, and all others 2017). students have a homeroom teacher through- out their education, giving a link to par- Few assessment systems measure ents and to students’ lives outside of school noncognitive skills (OECD 2012). Noncognitive skills are difficult to measure through traditional assessments, which are Keeping pace with shifting labor designed predominantly to assess content market priorities and evolving knowledge or cognitive skills. Many areas, beyond cognitive skills such as collaborative problem solving and The new economy requires skills for the creativity, encompass multiple interrelated modern era. Education curricula are adapt- components. Links between outcomes ing to these new demands (see, for example, and schooling may reflect factors other than Kautz and others 2014; Valerio and others schooling (Ercikan and Oliveri 2016). 2016). In East Asia and Pacific, this shift is Most policy frameworks and guidelines seen in movements toward competence-based in Hong Kong SAR, China; India; Korea; learning, student-centered learning, cross- Malaysia; Mongolia; the Philippines; Thailand, curricular studies, and personalized learning and Vietnam include transversal skills to vary- (Looney 2009). Singapore’s 1997 Thinking ing degrees (Care and Luo 2016). Few of these Schools, Learning Nation reform and Korea’s systems assess them, however. More clarity is 2015 curriculum reform responded to new needed in defining and assessing these skills. labor market demands (Ministry of Education of Korea, n.d). Examples of systems to assess The definition and terminology of these noncognitive skills skills is still evolving. Terms include non- cognitive, behavioral, higher-order, and Some systems have begun adapting assess- 21st century skills. Examples of these skills ment frameworks to measure noncogni- include creativity, collaborative problem tive skills. Singapore has shifted toward a solving, communication skills, flexibility “whole-student approach,” which uses proj- and adaptability, digital literacy, and conflict ect work as one method of assessment. The resolution. assessment is interdisciplinary, collaborative, The United Nations Educational, and group work–based; it includes written Scientific, and Cultural Organization’s and oral components (Conley and Darling- framework for transversal skills (skills Hammond 2013). relevant to a broad range of subjects and Shanghai’s Green Indicator System, disciplines) covers critical thinking, intra- launched in 2011, covers “student academic personal skills, interpersonal skills, global achievement, student learning motivation, citizenship, media and information liter- student academic workload, teacher-student acy, and other skills (UNESCO 2016). The relations, teachers’ instructional styles, prin- World Bank Skills Towards Employability cipals’ curriculum leadership, the impact and Productivity (STEP) surveys highlight of students’ socioeconomic background on the importance of such skills in the labor their academic achievement, students’ moral market. A recent STEP study finds that behavior, and annual progress on these indica- employers in the Philippines, especially tors” (Liang, Kidwai, and Zhang 2016, 96). employers that are modern and innova- PISA has recently included questions on tive, value socioemotional skills, such as collaborative problem solving. USING STUDENT ASSESSMENTS TO IMPROVE INSTRUCTION AND LEARNING 201 Conclusions and exams. They need to maintain a focus on recommendations equity and ensure that dissemination poli- cies fully support accountability. Education systems are shifting focus, moving • Above-Average Performing Systems have away from the narrow use of selection exams similar priorities. They need to continue that test for content knowledge and cognitive to use international assessments, ensure skills toward an emphasis on formative assess- alignment of student assessments with ments that measure noncognitive skills. The reforms to teacher policies and curricula, shift toward noncognitive outcomes does not and support the use of student assessments preclude valid and regular measurement of and their data in the classroom. basic cognitive outcomes, which is integral to • Below-Average Performing Systems need measuring and driving quality improvement. to define, support, and monitor the use of Experience in the Top Performing Systems assessments in the classroom; benchmark shows that it is important for countries to get assessment systems; ensure quality con- these fundamentals right. trol; and continue to use regular national Priorities for the four groups of countries and international assessments to drive differ (table 7.3): national commitment to improved educa- • For Top Performing Systems, a primary tional quality. challenge will be to ensure alignment • Emerging Systems should continue to use of assessments with changing curricula. tools such as EGRAs and EGMAs, as well as These systems need to continue to develop international and regional assessments, and and support means of measuring non- ensure that assessments are integrated with cognitive skills beyond multiple-choice teacher and curricular reforms. TABLE 7.3 Steps toward systematically strengthening assessments Action Below-Average Above-Average Policy Emerging Systems Performing Systems Performing Systems Top Performing Systems Assessments Periodically use early foundational skills Implement national assessments for Participate in regular international assessments to spur policy change and diagnostic purposes. benchmarking (PISA, TIMSS). target reforms every five years. Strengthen systems of assessment Engage in a national dialogue and secure Integrate assessment policies into (enabling environment, quality control, commitment to quality. reforms of curricula and teacher alignment). development systems. Disseminate assessment data to Integrate the use of assessment data into stakeholders for accountability. teacher development, alongside adequate support and quality monitoring. Reflect changing curricular priorities (such as noncognitive skills) in Link assessment use to equity, examining assessment design and methodology current practices for tracking and and provide support for change or accountability. research, guidance, training, quality, monitoring, and so forth. Use internationally benchmarked assessments to check system performance. Note: PISA = Programme for International Student Assessment; TIMSS = Trends in International Mathematics and Science Study. 202 GROWING SMARTER Notes on School-Based Management. Washington, DC: World Bank. 1. Paragraph 33 of the Incheon Declaration Black, P., and D. Wiliam. 1998. “Assessment states, “Moreover, there is a need for shared and Classroom Learning.” Assessment in understanding and viable strategies to Education: Principles, Policy, and Practice measure learning in ways that ensure that 5 (1): 7–74. all children and youth, regardless of their Bradley, S., R. Crouchley, J. Millington, and J. Taylor. circumstances, receive a quality and relevant 2000. “Testing for Quasi-Market Forces in education, including in human rights, arts, Secondary Education.” Oxford Bulletin of and citizenship. Such understanding can best Economics and Statistics 62 (3): 357–90. be cultivated through improved availability Bray, M., and C. Lykins. 2012. Shadow of systematic, reliable, and updated data Education: Private Supplementary Tutoring and information obtained through formative and Its Implications for Policy Makers in Asia. and/or continuous (classroom-based) Mandaluyong City, the Philippines: Asian assessments and summative assessments at Development Bank, Comparative Education different levels. Finally, quality also requires Research Centre (CERC). systems for managing teachers, governance, Breakspear, S. 2012. “The Policy Impact of PISA: accountability mechanisms, and strong public An Exploration of the Normative Effects of financial management.” International Benchmarking in School System 2. SABER is a set of tools that enables countries Performance.” Education Working Paper 71, to evaluate and benchmark education OECD, Paris. policies across a range of 13 topics, such as Bruns, B., D. Filmer, and H. A. Patrinos. 2011. teachers, early childhood development, school Making Schools Work . Washington, DC: autonomy and accountability, and student World Bank. assessments. See http://saber.worldbank.org/. Care, E., and R. Luo. 2016. Assessment of 3. EGMAs are also widely used. Transversal Competencies: Policy and Practice 4. The analysis evaluates the assessment systems in the Asia-Pacific Region. Paris: UNESCO. in three domains (enabling context, system Clarke, M. 2011. “Framework for Building an alignment with the rest of the education Effective Student Assessment System.” World system, and quality of assessments) in Bank, Washington, DC. http://files.eric.ed.gov Cambodia, China, Indonesia, Japan, /fulltext/ED553178.pdf. Korea, Lao PDR, Malaysia, Mongolia, Conley, D. T., and L. Darling-Hammond. 2013. the Philippines, Singapore, Thailand, and “Creating Systems of Assessment for Deeper Vietnam as well as in Hong Kong SAR, China Learning.” Stanford Center for Opportunity and Shanghai, China. Policy in Education, Stanford, CA. 5. PISA 2012 is used because PISA 2015 did Dobbie, W., and R. Fryer. 2011. “Getting Beneath not ask schools whether other schools in the Veil of Effective Schools: Evidence from the area compete for students (OECD 2013, New York City.” NBER Working Paper 17632, 2016b). National Bureau of Economic Research, Cambridge, MA. Duflo, E., P. Dupas, and M. Kremer. 2011. “Peer Referencesa Effects, Pupil-Teacher Ratios, and Teacher Acosta, P., T. Igarashi, R. Olfindo, and J. Rutkowski. Incentives: Evidence from a Randomized 2017. Developing Socioemotional Skills for Evaluation in Kenya.” American Economic the Philippines’ Labor Market. Washington, Review 101 (5): 1739–74. DC: World Bank. http://documents.worldbank Ercikan, K., and M. E. Oliveri. 2016. “In .org / curated/en/333521498727263689 Search of Validity Evidence in Support of / Developing-socioemotional-skills-for-the the Interpretation and Use of Assessments of -Philippines-labor-market. Complex Constructs: Discussion of Research Barrera-Osorio, F., T. Fasih, H. A. Patrinos, and on Assessing 21st Century Skills.” Applied L. Santibáñez. 2009. Decentralized Decision- Measurement in Education 29 (4): 310–18. Making in Schools: The Theory and Evidence Gardner, E. E., and M. M. Clarke. 2015. “READ Trust Fund Report 2008–2015.” a. Titles of publications that include “Korea” and “Taiwan” Working Paper 101527, World Bank, refer to the Republic of Korea and Taiwan, China, respectively. Washington, DC. http://documents.worldbank USING STUDENT ASSESSMENTS TO IMPROVE INSTRUCTION AND LEARNING 203 .org/curated/en/671151468187737988/READ http://documents.worldbank.org/curated -trust-fund-report-2008-2015. /en/692561468000587471/pdf/WPS7447.pdf. Graham, J., and S. Kelly. 2017. “Early Grade Looney, J. 2009. “Assessment and Innovation in Reading in East Asia and the Pacific.” Education,” OECD Education Working Paper Background paper for the World Bank 24, OECD Publishing, Paris. http://dx.doi flagship report, Growing Smarter: Learning .org/10.1787/222814543073. and Equitable Development in East Asia and Ministry of Education of Korea. n.d. “Education Pacific, World Bank, Washington, DC. Policies and Programs.” Ministry of Education, Hanushek, E., and L. Woessman. 2007. The Role Seoul, Republic of Korea. http://english.moe of Education Quality for Economic Growth. .go.kr/sub/info.do?m=040101&s=english. Washington, DC: World Bank. https://doi National Center on Education and the Economy. .org/10.1596/1813-9450-4122. 2014. “Global Perspectives: Explaining Jakubowski, M., H. A. Patrinos, E. E. Porta, and Taiwan’s Dramatic Improvement in PISA J. Wis′ niewski. 2016. “The Effects of Delaying Reading.” National Center on Education and Tracking in Secondary School: Evidence from the Economy, Washington, DC. http://ncee the 1999 Education Reform in Poland.” .org/2014/10/global-perspectives-explaining Education Economics 24 (6): 557–72. -taiwans-dramatic-improvement-in-pisa Kautz, T., J. Heckman, R. Diris, B. ter Weel, -reading/. and L. Borghans. 2014. “Fostering and OECD (Organisation for Economic Co-operation Measuring Skills: Improving Cognitive and and Development). 2012. Strong Performers Non-cognitive Skills to Promote Lifetime and Successful Reformers in Education: Lessons Success.” OECD Publishing, Paris. https:// from PISA for Japan. Paris: OECD Publishing. w w w. o e c d . o r g / e d u / c e r i / F o s t e r i n g - a n d http://dx.doi.org/10.1787/9789264118539-en. -Measuring-Skills-Improving-Cognitive-and ———. 2013. PISA 2012 Results. Paris: OECD -Non-Cognitive-Skills-to-Promote-Lifetime Publishing. -Success.pdf. ———. 2 0 1 4 . TA L I S 2 0 1 3 R e s u l t s : A n Kellaghan, T. 2004. “Public Examinations, International Perspective on Teaching and National and International Assessments, and Learning . Paris: OECD Publishing. http:// Educational Policy.” Educational Research dx.doi.org/10.1787/9789264196261-en. Center, St. Patrick’s College, Dublin. ———. 2016a. Education in Thailand: An OECD- King, E., and H. Rogers. 2014. “Intelligence, UNESCO Perspective. Paris. http://dx.doi Personality, and Creativity: Unleashing the .org/10.1787/9789264259119-en. Power of Intelligence and Personality Traits ———. 2016b. PISA 2015 Results. Paris: OECD. to Build a Creative and Innovative Economy.” Parker, C. E. 2005. “Teacher Incentives and Student Background paper for symposium “Achieving Achievement in Nicaraguan Autonomous HOPE: Happiness of People through Schools.” In Incentives to Improve Teaching: Education: Innovation in Korean Education Lessons from Latin America. Washington, DC: for a Creative Economy,” World Bank, Seoul, World Bank. Republic of Korea, November. Patrinos, H. A. 2012. “Strengthening Education Kuramoto, N., and R. Koizumi. 2015. “Current Quality in East Asia: System Assessment Issues in Large-Scale Educational Assessment and Benchmarking for Education Results in Japan: Focus on National Assessment of (SABER).” SABER Country Report Working Academic Ability and University Entrance Paper 66526, World Bank, Washington, DC. Examinations.” Assessment in Education: http://documents.worldbank.org/curated Principles, Policy, and Practice (August 31): / en/778841468245418246/Strengthening 1–19. -education-quality-in-East-Asia-System Liang, X., H. Kidwai, and M. Zhang. 2016. -Assessment-and-Benchmarking-for How Shanghai Does It: Insights and Lessons -Education-Results-SABER. from the Highest-Ranking Education System Ramírez, M.-J. 2012. “Disseminating and Using in the World. Washington, DC: World Bank. Student Assessment Information in Chile.” doi:10.1596/978-1-4648-0790-9. World Bank, Washington, DC. Lockheed, M. 2015. “Why Do Countries Ruby, A., and D. Kent. 2015. “Investing in Participate in International Large-Scale Learning Outcomes: Findings from an Assessments?” Policy Research Working Independent Evaluation of the Russia Paper 7447, World Bank, Washington, DC. Education Aid for Development (READ) Trust 204 GROWING SMARTER Fund Program.” World Bank, Washington, Data.” Policy Research Working Paper DC. http://documents.worldbank.org/curated 7879, World Bank, Washington, DC. /en/216471467995103312/pdf/103237-WP https://openknowledge .worldbank.org -P123662-Box394858B-PUBLIC-READ-Trust /handle/10986/25678. -Fund-Independent-Evaluation-Report-Final Wagner, D. A. 2011. Smaller, Quicker, Cheaper: .pdf. Improving Learning Assessments for Sui-chu Ho, E. 2012. “Student Learning Developing Countries. Paris: UNESCO Assessment.” UNESCO, Paris. http://www International Institute for Educational .unes cobkk.org/fileadmin/user_upload/epr Planning. http://www.literacy.org/sites/literacy /Quality/SLA_8August-final.pdf. .org/files/publications/213663e.pdf. Tobin, M., P. Lietz, D. Nugroho, R. Vivekanandan, Wiliam, D., C. Lee, C. Harrison, and P. Black. and T. Nyamkhuu. 2015. “Using Large-Scale 2004. “Teachers Developing Assessment for Assessments of Students’ Learning to Inform Learning: Impact on Student Achievement.” Education Policy: Insights from the Asia-Pacific Assessment in Education: Principles, Policy, Region.” ACER, Melbourne; UNESCO, and Practice 11 (1): 49–65. Bangkok. Wong, A. 2017. “Insights from East Asia’s High- UNESCO (United Nations Educational, Scientific, Performing Education Systems: Leadership, and Cultural Organization). 2015. “Incheon Pragmatism, and Continuous Improvement.” Declaration.” World Education Forum 2015, Background paper for World Bank flagship Incheon, Republic of Korea, May 19–22. report, Growing Smarter: Learning and ———. 2016. School and Teaching Practices for Equitable Development in East Asia and Twenty- First Century Challenges: Lessons Pacific, World Bank, Washington, DC. from the Asia-Pacific Region. Regional World Bank. 2016. “Implementation Completion Synthesis Report. Paris: UNESCO. and Results Report (Tf-13048) on a Grant in Valerio, A., M. L. Sanchez Puerta, N. Tognatta, the Amount of US$84.6 Million to the Socialist and S. Monroy-Taborda. 2016. “Are There Republic of Vietnam for a Global Partnership Skills Payoffs in Low- and Middle-Income for Education Vietnam Escuela Nueva Project.” Countries? Empirical Evidence Using STEP World Bank, Washington, DC. Charting the Course Ahead 8 The impressive record of success in education The Top Performing Systems in East Asia in some low- and middle-income countries in and Pacific took a particular path toward East Asia and Pacific is proof of concept that establishing these fundamentals and achiev- schooling in resource-constrained contexts ing learning and growth. They offer lessons can lead to learning for all. Other low- and that can help policy makers elsewhere in the middle-income countries—in East Asia and region to surmount the learning crisis affect- Pacific and elsewhere—can use the policies ing the millions of children who are out of and practices that underlie this success to school or in school but not learning. ensure that their students learn. Education remains a long-term process of acquiring background knowledge, skills, What has worked? habits, and behaviors. Current labor market What policies and practices will promote conditions extend the amount and refocus learning in schools? What should a coun- the types of knowledge and skills needed, try do if it wants to achieve high and equi- but they do not fundamentally alter them table learning outcomes? A focus on five or the process for acquiring them. Reading key policy domains and core elements ability is still the foundation for acquiring within them allowed the Top Performing all other types of knowledge—the cogni- Systems to promote learning. Coordinated tive equivalent of the opposable thumb. The efforts in these five areas (institutions, pub- bar for quantitative skills continues to rise lic spending, teachers, readiness to learn, dramatically, but mastery of the fundamen- and assessment) can improve the teaching tals of math, logic, and data analysis is still and learning experience in the classroom critical. Communication skills rest on mas- (figure 8.1). tery of grammar and vocabulary—and years of practice in oral and written expression. Behavioral skills and the ability to work in Aligning policies in five core teams improve through structured practice areas is key to success and feedback. Resilience and grit remain the For countries with Emerging and Below- glue that keeps other skills from falling to Average Performing Systems, the top prior- useless pieces. ity is to focus on fundamentals. The Pacific 205 206 GROWING SMARTER FIGURE 8.1 Five policy domains promote learning UT INSTIT IONS PUB ENT LIC SM S as da vel at ts. at ins uc oho sca pati ing om ith bl ss en cla cti on pr ass in ro w u e m i rn Co PE ive tru at r le on S ss on al s ogr ess fo form ery ose arge rtic lea Sp tra sic e urce icts hin nc on l re d d ling SE en ba so istr be en te du s t In ev agn nal l h p ark en m Ch se ta . ND d pub ca o s ss fro Di atio oug chm an eff li tio ch e t. a AS n ec c s n. oo - rn th n e t m tiv pe Be ING th ely nd a at r c n . i ar ef d al in e g te in ls d. rm LEARNING an ss a d te a ded o g Fo nit nd ildh pm act ices d es rly eve cro se th ach g a d te un r by pr by er. wh in lea . ch cla ea cla s. ar ack o t g d th te g ed g co s a c rn cu ve m oo en ors . A y t in an in ed in f in te in ce nd sie nd ers ch of ea on res e xt cle b g lit in rs rs er vid ng rv s o de pr d t se to er ad en s a ea s a ch ea ess bi tra ee che utt ro idi bse REA d e an n c ve ove edu rv de i nt le peri oal bs tice tea s a t ven nd e a er s p a cl p ov o ur om rs. m e hi lop th ca ice liv . ce a ex g ’ j ac w me ecti e c ro e oo Co ld m e tio s. an ina nee th ss rch ssr re e q um ti he p t n co el i o ul rac d ac om or be he s n’s nt ua n rd h lo ss rv ric p ph fro lity t ise DIN ys m of e Ra ica bi l rth RS r p e t ssr up o ak cla S s o r ES . HE e TO C S ep M A er Ke LEA TE s. En Ce ur ol o s et RN ch ha s t th a ll eb in a si c la ce cond in p iti o n s f o r le a r n i n g a r e Island countries and other Emerging • Concentrate effective, equity-minded pub- Systems must continue to focus on early lic spending on basic education. reading. Building basic cognitive skills for • Select and support teachers throughout primary and secondary students in econo- their careers to allow them to focus on the mies caught in the middle-income trap is classroom. paramount for moving on to imparting • Ensure that children are ready to learn in new skills. school. School systems need to redouble their • Assess students to diagnose issues and efforts in the following areas: inform instruction. • Align institutions to ensure basic condi- Top Performing Systems distinguish them- tions for learning. selves by achieving alignment across elements CHARTING THE COURSE AHEAD 207 and sustaining it. Above-Average Performing some plans are never implemented; in others, Systems attempt alignment and succeed in the plans themselves never materialize. some dimensions, but alignment is inconsis- No single element will produce the kinds tent in others. Below-Average Performing of success that Top Performing Systems Systems strive to emulate the success of the have achieved. Progress occurs as the result rest of the region. They have devised plans for of aligning all of these elements to promote alignment of elements—at least on paper—but learning (table 8.1). TABLE 8.1 Policy actions in moving from Emerging Systems to Top Performing Systems Action Below-Average Above-Average Policy Emerging Systems Performing Systems Performing Systems Top Performing Systems Public Prioritize public spending on basic Continue investing in basic education Continue investing in basic and upper- spending education and ensure high completion to ensure high completion while secondary education while diversifying of primary and lower-secondary school. expanding access to upper-secondary funding for vocational, technical, and school and above. tertiary education. Increase starting teacher salaries to attract qualified teachers. Strengthen teacher compensation Strengthen teacher career development policies to encourage good policies and nonfinancial benefits for Build simple but functional school performance and retain qualified teachers to continue improving their facilities to increase access to basic teachers. quality. education. Build simple but functional school Provide remedial measures such as Provide funding to support facilities to ensure access to basic extracurricular activities to enhance disadvantaged students in accessing education and increase access to learning of disadvantaged students. basic education. preprimary and upper-secondary education. Help disadvantaged students to access basic education; consider mobilizing the private sector to increase access to preprimary and upper-secondary education. Teachers Start the positive foundational cycle Take incremental steps to increase Establish rigorous criteria and multiple of attracting and ensuring reasonable- qualifications and quality, establishing filtering mechanisms, while ensuring quality teachers, while being realistic greater quality assurance and filtering that teaching is attractive in terms of about the system’s capacity to measures. salary, professionalism, and prestige. implement quality assurance measures balanced with supply of candidates. Deepen teacher networks in which Leverage high-performing teachers for more-advanced teachers provide professional development and support. Establish teacher networks (with a focus greater support through mentoring and on supporting new, weaker teachers coaching. Provide more autonomy for high- with mentoring and other support), performing teachers. lesson study, and an “open-door” Provide more advanced professional culture as key elements of professional development that develops deeper Create a highly developed professional development. critical thinking techniques. development framework with tailored, individualized approaches. Establish a basic framework of Create tracks of progression and professional development and how promotion for expert teachers that Deepen critical thinking, elaboration, teachers can reasonably allow them to remain in the classroom. and cognitive activation techniques. be reached. table continues next page 208 GROWING SMARTER TABLE 8.1 Policy actions in moving from Emerging Systems to Top Performing Systems (continued) Action Below-Average Above-Average Policy Emerging Systems Performing Systems Performing Systems Top Performing Systems Teachers Establish a mentality of teaching to Evolve system- and teacher-related (continued) learn with clear learning goals with policies in ways that reform and move curriculum and textbooks. all elements forward with alignment. Begin creating a cohesive, aligned Decrease classroom hours and increase system that supports teachers and supplementary activities. promotes focused, streamlined learning. Readiness to Develop policies and programs to ensure Ensure that clear quality standards exist, Engage multiple actors in the learn that both cognitive development and are disseminated, and are addressed in development and delivery of integrated physical health are the focus of early an evaluative process. early childhood services. childhood school readiness strategies. Assessments Periodically use early foundational skills Implement national assessments for Participate in regular international assessments to spur policy change and diagnostic purposes. benchmarking (PISA, TIMSS). target reforms every five years. Strengthen systems of assessment Engage in a national dialogue and secure Integrate assessment policies into (enabling environment, quality control, commitment to quality. reforms of curricula and teacher alignment). development systems. Disseminate assessment data to Integrate the use of assessment data stakeholders for accountability. into teacher development, alongside adequate support and quality Reflect changing curricular priorities monitoring. (such as noncognitive skills) in assessment design and methodology Link assessment use to equity, examining and provide support for change or current practices for tracking and research, guidance, training, quality, accountability. monitoring, and so forth. Use internationally benchmarked assessments to check system performance. Note: PISA = Programme for International Student Assessment; TIMSS = Trends in International Mathematics and Science Study. Increasing the role of the private Systems, where access to preprimary educa- sector tion is very limited, promoting public-private partnerships with or without public subsidies The private sector will continue to play a is likely to increase enrollment. Private pro- critical role in the transformation of edu- vision and financing is also likely to increase cation. A significant portion of preprimary enrollment in upper-secondary education. education is available through private pro- viders, and regional trends point to increas- ing enrollment in private institutions at the Keeping up with rapid changes in primary and secondary levels (see spotlight technology and labor markets 3). Governments can capitalize on the poten- Learning is critical to East Asia and Pacific’s tial impacts of public and private sector col- strategy for productivity-driven growth. laboration. In particular, private provision Policy makers therefore have to keep their and financing can increase access where pub- eyes focused on the next stage of education lic resources are limited. Especially in Below- development. Even for high performers, Average Performing Systems and Emerging a commitment to continuous improvement CHARTING THE COURSE AHEAD 209 BOX 8.1 The future requires both strong fundamentals and a wider view of skills All of the Top Performing Systems in East Asia and finds that the service content of manufacturing has Pacific invested heavily in human capital, technologi- grown and will continue to grow, eroding the dis- cal capability, and knowledge institutions that drove tinction between manufacturing and service jobs. growth. The virtuous circle through which govern- Furthermore, services are acquiring some of the pro- ments promised better jobs to reward ever-higher development characteristics that manufacturing is levels of investment in human capital depended fun- shedding. A main conclusion is that technological damentally on building successful national manufac- change fosters job creation as well as job displace- turing industries. ment, so fears of net job loss in the economy may Structural transformation continues today, but be ungrounded. Changes will bring a faster pace of productivity has increased (across and within sec- change in manufacturing and services, increasing the tors) only where countries sustained transformative, skill requirements in all jobs. high growth. Changing trends in technology and A broader set of skills will be needed for the next industry will continue to challenge education policy stage of development. Cognitive, behavioral, and in the region in the coming years. Two changes in creative skills complement technology. The value of particular stand out: the nature of work and the creativity and innovation rises as machines take over skills that jobs require. routine tasks. Education systems that promote cre- Artificial intelligence, machine learning, and ative thinking gain in value. Higher-level cognitive robotics may change the nature of work. Frey and skills are less amenable to automation or computer Osborne (2013) find that 47 percent of U.S. jobs replacement because of their nonroutine or non- are at high risk of being automated within one or algorithmic nature, as seminal work by Autor, Levy, two decades. Machine learning has the potential to and Murnane (2003) shows. Subsequent work shows extend the range of occupations and skill levels to that idiosyncratic and unstructured problems— relatively near-term computerization or automa- rather than “higher cognitive functions”—are most tion. The risk of automation varies inversely with resistant to imitation by machines and therefore most wages or salaries and average education in general, resistant to substitution (Murnane and Levy 2013). although very few tasks or employment categories Drawing on a full set of cognitive and behav- are exempt from the threat of automation. ioral skills and being able to integrate background Widespread net job loss is unlikely—and tech- information in novel ways are the challenges that lie nological change may create jobs—but one thing ahead for education systems. Building strong funda- is certain: all jobs will change. A World Bank mentals of cognitive, literacy, and numeracy skills report (Hallward-Driemeier and Nayyar 2017) early on remains more critical than ever before. will drive the vision of education systems. in countries that still have not achieved Policy makers should consider the impacts of fundamentals. rapid changes in technology and labor mar- Below-Average Performing and Emerging kets on education systems and skills develop- Systems need to start now—by taking action ment (box 8.1). to end or reduce functional illiteracy, focus- ing investment on engaging the private sector, recruiting and supporting better teachers, and A way forward maintaining a commitment to quality bolstered Even as technology and economic landscapes by adequate and appropriate assessment and shift, what made the high performers success- data on learning outcomes. A critical priority ful still matters for ensuring that education for Below-Average Performing Systems and systems provide good-quality learning for Emerging Systems is focusing on what made all. If the technological revolution will dis- high-performing systems successful and mov- rupt high performers, it will be catastrophic ing quickly to implement some of these lessons. 210 GROWING SMARTER No one size fits all reform agendas, but all profession and the accountability of teach- systems share some priorities: ers for good classroom performance is key. • Above-Average Performing Systems • Emerging Systems face the greatest chal- should not rest on their accomplishments. lenges. Resources are scarcest, few mea- Deepening the quality of the teaching sures of learning exist, and just getting all force and continuing to monitor equity students in school has been difficult. Efforts are in order. But these systems should also are best concentrated on ensuring that basic endeavor to tie learning to new and emerg- conditions are in place for learning in all ing needs, which includes introducing teach- schools and reviewing spending to ensure ing and measurement of noncognitive and that basic education is appropriately priori- 21st century skills. Redoubling the commit- tized. Emerging Systems should also commit ment to internationally comparable assess- to diagnosing cohort progress, especially of ment can keep public support for excel- early learners using early grade assessments, lence in education high, along with public and to using test results to inform and awareness of achievements. Deepening the improve basic reading and math instruc- availability of high-quality early childhood tion. They should continue to explore the education and development programs, espe- use of regional and international assess- cially for families that cannot afford them, ments for benchmarking. Second-order is another step. Also essential is to ensure challenges include channeling resources for that good options exist in the labor mar- equity and considering how to approach ket for capitalizing on knowledge and skills the range of teacher support policies that acquired in basic and postbasic education. can improve capacities for the long term. Amid all of these activities, further building As these policies are developed, Emerging institutional capacity cannot be neglected. Systems are well served to review the extent • Top Performing Systems provide compel- to which teacher training and professional ling examples of how the work of producing development focus on improving the quality high learning outcomes in schools is never of instruction. completed. Looking at how these systems • Below-Average Performing Systems are evolving reveals that promoting creativ- should review teacher development poli- ity and new forms of assessment, ensuring cies to ensure that training is prioritiz- that teachers remain motivated and grow ing improved instructional quality while in competency, and learning from other top building the institutional capacity for performers worldwide are core tasks. Their deeper and more comprehensive reforms. continued efforts demonstrate that conserv- Introducing selectivity, observation, col- ing and extending existing capacity and laboration, and feedback while creating expertise is a precondition for staying on top. incentives and career paths for teachers that reward teaching ability are likely to pay off substantially. Ensuring readiness A map of countries’ current to learn and broadening early childhood systems education and development services are The ability of East Asia and Pacific’s Top also critical. Developing and implementing Performing Systems to improve both school- systemwide national assessments of cohort ing and learning provides valuable lessons for progress should complement programs all countries struggling to address the learning for classroom assessment; assessment sys- crisis. Although these systems followed dif- tems should also include the regular use ferent paths to promote learning, all of them of internationally comparable assessments aligned and prioritized common elements for benchmarking and system account- across five policy domains. ability. Consideration should be given to Table 8.2 maps the current state of the making teachers’ jobs easier, through cur- region’s economies for these domains and ele- riculum and other reforms. Simultaneously ments. It provides a starting point for countries raising the attractiveness of teaching as a to take stock of the current situation in each CHARTING THE COURSE AHEAD 211 TABLE 8.2 The status of policies and practices that promote learning varies within the region Above- Average Below-Average Performing Performing Top Performing Systems Systems Systems Emerging Systems Pacific Island countries Hong Kong SAR, China Papua New Guinea Macao SAR, China B-S-J-G (China) Taiwan, China Timor-Leste Korea, Rep. Philippines Singapore Cambodia Indonesia Mongolia Myanmar Thailand Malaysia Vietnam Lao PDR Japan Policy Practice Align institutions Ensure that the basic conditions to ensure basic for learning are in place in all conditions for schools. learning. Concentrate Spend effectively. effective, equity- Concentrate public spending on minded public basic education. spending on basic education. Channel resources to schools and districts that are falling behind. Select and Raise the selectiveness of who support teachers becomes a teacher. throughout their Support new teachers by careers to allow observing classroom practices them to focus on and providing feedback. the classroom. Make teachers’ jobs easier by providing clear learning goals and uncluttered texts. Keep experienced teachers in the classroom and leading as peers and researchers. Center teacher training on classroom practice and the ability to teach the curriculum. Ensure that Focus on children’s physical and children are cognitive development from birth. ready to learn in Assess and improve the quality school. of early childhood education and development services. Coordinate across actors to deliver needed services. Assess students Benchmark learning through to diagnose issues participation in international and inform large-scale assessments. instruction. Diagnose cohort progress at every educational sublevel. Inform instruction with data from formative classroom assessment. Alignment is successful Alignment is attempted, Alignment plans exist in writing, but there is little or no No data. and sustained. but success is inconsistent. evidence of implementation, or no such plan exists. Note: B-S-J-G (China) = Beijing, Shanghai, Jiangsu, and Guangdong (China). 212 GROWING SMARTER domain and to envision a path forward. Top Change: An Empirical Exploration.” Performing Systems distinguish themselves Quarterly Journal of Economics 118 not only by achieving alignment across ele- (4, November): 1279–334. ments but also by sustaining it. Above-Average Frey, C. B., and M. A. Osborne. 2013. “The Future of Employment: How Susceptible Are Jobs Performing Systems attempt alignment, but it to Computerization?” University of Oxford, is not consistent in all domains. Below-Average Oxford, U.K. Performing Systems strive to emulate the suc- Hallward-Driemeier, M., and G. Nayyar. cess of the rest of the region. They have devised 2017. Trouble in the Making? The Future plans for alignment, but implementation is of Manufacturing-Led Development. lacking or plans have not yet materialized. Washington, DC: World Bank. Murnane, R., and F. Levy. 2013. The New Division of Labor: How Computers Are Creating the References Next Job Market. Princeton, NJ: Princeton Autor, D., F. Levy, and R. Murnane. 2003. University Press. “The Skill Content of Recent Technological Education Systems in Appendix East Asia and Pacific at a Glance This appendix provides summary informa- included in the radar charts). Data are for tion on 17 economies in East Asia and Pacific most recent year available unless otherwise (refer to box A.1 for definitions of data noted. BOX A.1 Radar charts in this appendix • Data in all radar charts refer to 2015, with a few latrine), ventilated improved pit latrine, pit latrine exceptions. Primary school enrollment refers with slab, and composting toilet. to 2014, except for Malaysia (2012), Papua • Improved water includes piped water on prem- New Guinea (2012), the Philippines (2013), and ises (piped household water connection located Vietnam (2013). Secondary school enrollment inside the user’s dwelling, plot, or yard) and refers to 2014, except for Cambodia (2008), the other improved sources of drinking water (public Philippines (2013), and Thailand (2012). Top taps or standpipes, tube wells or boreholes, pro- Performing Systems do not have radar charts as tected dug wells, protected springs, and rainwater their coverage of indicators is close to universal. collection). • Enrollments are net enrollments, except in China, • Access to Internet is defined as the share of where they are gross enrollments. people who used the Internet from any loca- • Improved sanitation facilities include flush / tion and from any kind of device in the previous pour flush (to piped sewer system, septic tank, pit 12 months. 213 214 GROWING SMARTER Cambodia FIGURE A.1 Average years of educational attainment in Cambodia, 1950–2010 TABLE A.1 Selected indicators for Cambodia Indicator Statistic 4.5 Average years of educational attainment Demographic and macroeconomic indicators 4.0 among population 25 and older Population (millions) 15.6 3.5 GDP (US$, billions) 18 GDP per capita (US$) 1,159 3.0 Adult literacy rate (%) 73.9 2.5 Unemployment rate (%) 0.4 Student enrollment 2.0 Preprimary 1.5 Number enrolled 187,450 % female 48 1.0 Primary 0.5 Number enrolled 2,178,916 0 % female 50 19 5 19 0 65 19 0 19 5 80 19 5 19 0 95 20 0 20 5 10 19 0 5 6 7 7 8 9 0 0 5 Secondary 19 19 20 19 Number enrolled 930,200 Total Females % female — Tertiary Source: Barro and Lee 2013. Number enrolled 217,364 % female — Number of teachers Preprimary 6,084 FIGURE A.2 Selected development indicators Primary 47,866 in Cambodia Secondary 30,258 Percent Tertiary — Education financing % of GDP 1.9 Primary school enrollment % of government spending 9.1 100 Source: EdStats data (World Bank, various years). 80 Note: GDP = gross domestic product; — = not available. 60 40 Secondary Access to Internet 20 school Overview enrollment 0 Cambodia’s National Strategic Development Plan, Rectangular Strategy, and Education Strategic Plan have driven the expansion of access to education over the past 20 years. Access to Net primary enrollment rates increased sig- improved Access to sanitation nificantly, and girls now have equal access to water educational opportunities. Cambodia edu- cates about 3.5 million students from prepri- Source: National statistical authorities and World Development Indicators data (World Bank, various years). mary through tertiary education and employs about 85,000 teachers in preprimary through secondary schools. 24 provincial or municipal education offices, and 193 district education offices. Schools Administration are grouped into local clusters to provide The Ministry of Education, Youth, and one another with technical and material Sport comprises a central headquarters, assistance. E D U C AT I O N S YS T E M S I N E A S T A S I A A N D PAC I F I C AT A G L A N C E 215 Compulsory education and testing TABLE A.2 Selected indicators for China (continued) Cambodia has nine years of compulsory edu- Indicator Statistic cation. Examinations that certify comple- Student enrollment (continued) tion and are necessary for progression are Secondary administered at the end of lower-secondary Number enrolled 86,127,200 school (grade 9) and upper-secondary school % female 46 (grade 12). Tertiary Number enrolled 43,367,392 % female — Curriculum and textbooks Number of teachers The Curriculum Development Department Preprimary 2,032,327 sets learning standards, implements and Primary 5,889,014 monitors textbook policy, and devel- Secondary 6,234,124 ops technical and vocational education Tertiary — curriculum. Education financing % of GDP 3.9 % of government spending 15.0 Teachers Average 2015 PISA scoresa Mathematics 531 Teacher education institutions comprise Reading 494 provincial and regional training centers for Science 518 primary school teachers and postgradu- Sources: EdStats data (World Bank, various years); National Bureau of ate programs at the National Institute of Statistics of China 2016; OECD 2016b. Education for upper-secondary school teach- Note: GDP = gross domestic product; — = not available; PISA = Programme for International Student Assessment. ers. In 2013 more than half of primary school a. Data are for Beijing, Shanghai, Jiangsu, and Guangdong provinces, teachers held an upper-secondary degree or which are home to 16 percent of China’s primary school students and 11 percent of its secondary school students. higher, up from a quarter in 2007, and more than 80 percent of secondary teachers had completed at least grade 12 (Tandon and Fukao 2015). FIGURE A.3 Average years of educational attainment in China, 1950–2010 China 8 Average years of educational attainment TABLE A.2 Selected indicators for China 7 among population 25 and older Indicator Statistic 6 Demographic and macroeconomic indicators 5 Population (millions) 1,371.2 GDP (US$, trillions) 10.8 4 GDP per capita (US$) 7,924 3 Adult literacy rate (%) 95.1 Unemployment rate (%) 4.7 2 Student enrollment Preprimary 1 Number enrolled 40,507,144 0 % female 46 19 5 19 0 19 5 19 0 19 5 19 0 85 19 0 95 20 0 20 5 10 19 0 5 6 6 7 7 8 9 0 0 5 19 20 19 Primary Number enrolled 95,958,040 Total Females % female 47 table continues Source: Barro and Lee 2013. 216 GROWING SMARTER FIGURE A.4 Selected development indicators Compulsory education and testing in China Percent China has nine years of compulsory educa- tion. Recent policies are aimed at universaliz- ing preprimary education, which is currently Primary school enrollment available largely through private institutions. 100 Upon finishing their ninth year of schooling, 80 students sit for the zhongkao, which deter- 60 mines admission to upper-secondary school. 40 Secondary Access to school Students who complete upper-secondary Internet 20 enrollment school are eligible to sit for the gaokao, which 0 determines college admissions. Curriculum and textbooks Access to Access to Curriculum and textbook design are relatively improved decentralized, with the Ministry of Education sanitation water setting standards for textbook develop- Sources: National statistical authorities and World Development ment. Once a textbook has been written, it Indicators data (World Bank, various years). is reviewed by either the national or provin- cial Textbook Review Commission, an expert group separate from the textbook publisher. Curricular standards, including lesson hours Overview and other guidelines, are set at the national China’s education system, once highly cen- level. Provincial authorities develop imple- tralized, is moving gradually toward a more mentation plans based on national-level guid- decentralized model, in which local authori- ance; these plans are subject to review before ties assume greater responsibility for cur- they can be implemented. riculum design, policy implementation, and funding. The change has been driven partly Teachers out of necessity and partly out of gains in capacity, as China seeks to educate nearly Teachers are required to be certified before 266 million students from preprimary applying for a teaching job. China is cur- through tertiary and to ensure that its 15 mil- rently moving to a new system in which the lion teachers in more than 500,000 schools qualification certificate is based on the results maintain high standards necessary to prepare of a national exam. Teachers must receive the country for future economic development. 350 hours of training over a five-year period to retain their teaching qualifications. Administration Hong Kong SAR, China The Ministry of Education, a governmental organ under the State Council, sets policy TABLE A.3 Selected indicators for Hong Kong SAR, China direction, allocates the budget, and drafts laws and regulations for the country’s educa- Indicator Statistic tion system. County-level education offices Demographic and macroeconomic indicators retain responsibility for delivering education, Population (millions) 7.3 including managing most of the financing, GDP (US$, billions) 309 setting locale-specific policy based on ministry GDP per capita (US$) 42,423 guidance, arranging standardized testing, and Adult literacy rate (%) — handling other administrative responsibilities. table continues next page E D U C AT I O N S YS T E M S I N E A S T A S I A A N D PAC I F I C AT A G L A N C E 217 TABLE A.3 Selected indicators for Hong Kong SAR, FIGURE A.5 Average years of educational China (continued) attainment in Hong Kong SAR, China, 1950–2010 Indicator Statistic 12 Demographic and macroeconomic indicators (continued) Average years of educational attainment Unemployment rate (%) 3.2 10 among population 25 and older Student enrollment Preprimary Number enrolled 178,119 8 % female 48 Primary 6 Number enrolled 332,531 % female 49 4 Secondary Number enrolled 393,952 2 % female 45 Tertiary 0 Number enrolled 298,643 19 5 19 0 19 5 19 0 19 5 19 0 85 19 0 95 20 0 20 5 10 19 0 5 6 6 7 7 8 9 0 0 5 % female — 19 20 Number of teachers 19 Total Females Preprimary — Primary 24,339 Source: Barro and Lee 2013. Secondary 30,405 Tertiary — Education financing school, students take the Hong Kong Diploma % of GDP 3.3 of Secondary Education Examination. There % of government spending 18.6 are far fewer spots for tertiary education than Average 2015 PISA scores there are qualified students. As a result, many Mathematics 548 students attend university abroad. Reading 527 Science 523 Sources: EdStats data (World Bank, various years); OECD 2016b. Curriculum and textbooks Note: GDP = gross domestic product; — = not available; PISA = Programme for International Student Assessment. The Curriculum Development Council, an independent advisory board of subject matter experts, develops curricula. The Curriculum Overview Development Council comprises committees Education is based largely on the British sys- on each subject studied in schools, as well as tem, imported under British rule. The system specialty areas, such as a committee on early educates about 1.2 million students from childhood education. The Education Bureau preprimary through tertiary education and publishes guidelines for the publishing and employs about 55,000 teachers in primary review of textbooks. and secondary schools. Teachers Administration All teachers are required to hold a postsecond- The Education Bureau manages education. ary degree. The Committee on Professional Development of Teachers and Principals (for- merly the Advisory Committee on Teacher Compulsory education and testing Education and Qualifications) advises the Compulsory education lasts from primary government on continuing teacher educa- through secondary school. Primary school lasts tion, including the development of a Teacher six years. Upon completion of upper-secondary Competencies Framework. 218 GROWING SMARTER Indonesia FIGURE A.6 Average years of educational attainment in Indonesia, 1950–2010 TABLE A.4 Selected indicators for Indonesia Indicator Statistic 8 Average years of educational attainment Demographic and macroeconomic indicators 7 among population 25 and older Population (millions) 254.5 GDP (US$, billions) 890 6 GDP per capita (US$) 3,499 5 Adult literacy rate (%) 95.1 Unemployment rate (%) 6.2 4 Student enrollment Preprimary 3 Number enrolled 5,349,040 2 % female 49 Primary 1 Number enrolled 29,838,440 0 % female 48 19 5 19 0 65 19 0 19 5 80 19 5 19 0 95 20 0 20 5 10 19 0 5 6 7 7 8 9 0 0 5 Secondary 19 19 20 19 Number enrolled 22,586,956 Total Females % female 51 Tertiary Source: Barro and Lee 2013. Number enrolled 6,463,297 % female — Number of teachers FIGURE A.7 Selected development indicators Preprimary 427,585 in Indonesia Primary 1,801,909 Percent Secondary 1,459,756 Tertiary 209,830 Primary school enrollment Education financing 100 % of GDP 3.3 80 % of government spending 17.7 Average 2015 PISA scores 60 Mathematics 386 Access to 40 Secondary Internet 20 school Reading 397 enrollment Science 403 0 Sources: EdStats data (World Bank, various years); OECD 2016b. Note: GDP = gross domestic product; — = not available; PISA = Programme for International Student Assessment. Access to Access to improved sanitation water Overview Sources: National statistical authorities and World Development Indonesia, the second-largest country in East Indicators data (World Bank, various years). Asia and Pacific, has a largely decentralized education system. The nature of the system reflects the challenges inherent in trying to Administration educate more than 64 million students across thousands of islands. The education system The Ministry of Education and Culture man- oversees more than 340,000 schools and ages 84 percent of schools; the Ministry institutions and employs about 3.9 million of Religious Affairs manages the other teachers in preprimary through tertiary 16 percent (OECD 2016b). Provincial and institutions. District Offices of Education manage the E D U C AT I O N S YS T E M S I N E A S T A S I A A N D PAC I F I C AT A G L A N C E 219 implementation of ministerial policy at TABLE A.5 Selected indicators for Japan (continued) their respective levels. Preprimary educa- Indicator Statistic tion is organized largely through civil soci- ety organizations, with some oversight by Student enrollment Preprimary local government education offices. Districts Number enrolled 2,880,706 are responsible for most spending on basic % female — education (OECD 2016b). The National Primary Education Standards Board prepares stan- Number enrolled 6,714,539 dards for schools, assessment, teachers, stu- % female 49 dent competency, and curriculum throughout Secondary the country. Number enrolled 7,227,485 % female 49 Tertiary Compulsory education and testing Number enrolled 3,862,460 Indonesia has nine years of compulsory edu- % female — Number of teachers cation. Students sit for tests at the end of pri- Preprimary 113,043 mary and secondary school. The results of Primary 408,247 these tests determine eligibility for subsequent Secondary 624,215 phases of schooling. Tertiary 168,064 Education financing Curriculum and textbooks % of GDP — % of government spending — The National Education Standards Board Average 2015 PISA scores sets curriculum standards at the national Mathematics 532 level, which local governments then adapt. Reading 516 Both the government and private publishers Science 538 develop textbooks. Nongovernment publica- Sources: EdStats data (World Bank, various years) and Japanese Ministry of tions must undergo review before they can be Education, Culture, Sports, Science, and Technology data; OECD 2016b. Note: GDP = gross domestic product; — = not available; PISA = Programme distributed to schools. for International Student Assessment. Teachers Overview In 2005, Indonesia passed a Teacher Japan has a relatively centralized education Law, designed to improve the quality of system. The Ministry of Education sets high teachers by introducing incentives, raising standards for teachers, schools, curriculum, standards for qualification, and increasing and many other facets of education. High stan- teacher pay. dards, combined with targeted financing, have produced one of the world’s highest-achieving education systems: Japan’s 21 million students Japan regularly test near the top of international assessments. Japan employs about 1.3 mil- TABLE A.5 Selected indicators for Japan lion teachers in preprimary through tertiary Indicator Statistic institutions. Demographic and macroeconomic indicators Administration Population (millions) 126.9 GDP (US$, trillions) 4.1 The Ministry of Education, Culture, Sports, GDP per capita (US$) 32,477 Science, and Technology (MEXT) is the Adult literacy rate (%) — central government organ responsible for Unemployment rate (%) 3.7 education. Every prefecture in Japan has a table continues Prefectural Board of Education. 220 GROWING SMARTER Compulsory education and testing only to see that textbooks are not written in a biased manner. Compulsory education lasts nine years, although nearly all students continue through Teachers upper-secondary education. Students who com- plete the lower-secondary program of study are Teachers in Japan are highly trained. MEXT eligible to sit for the high school entrance exam, requires all candidate teachers to attend a the results of which determine what school the certified university program. Prefectural edu- student will attend. Upper-secondary is not cation offices invest significant resources in compulsory but is free to students who pass training both new and experienced teachers. their entrance exams and attend. Experienced teachers are assigned to mentor new teachers; these master teachers do not teach while assisting apprentice teachers dur- Curriculum and textbooks ing their first year of instruction. MEXT develops Japan’s curriculum at the national level, with input from the Central Republic of Korea Council for Education as well as education experts from universities around the coun- TABLE A.6 Selected indicators for the Republic of try. The curriculum is revised every 10 years Korea to ensure that its content matches the needs Indicator Statistic of Japan’s students and workforce. Japan’s Demographic and macroeconomic indicators curriculum is highly focused, with most time Population (millions) 50.6 spent on a narrow range of core subjects. GDP (US$, trillions) 1.4 Until recently, all textbooks were subject to GDP per capita (US$) 27,222 MEXT approval before being introduced into Adult literacy rate (%) 98.0 the classroom. Given the highly structured Unemployment rate (%) 3.5 nature of the curriculum, which allows pub- Student enrollment Preprimary lishers to produce textbooks with less over- Number enrolled 1,278,960 sight, MEXT has recently taken a step back % female 48 from the textbook approval process, checking Primary Number enrolled 2,721,766 FIGURE A.8 Average years of educational % female 48 attainment in Japan, 1950–2010 Secondary Number enrolled 3,396,766 % female 45 14 Tertiary Average years of educational attainment Number enrolled 3,268,099 12 % female — among population 25 and older Number of teachers 10 Preprimary 93,024 8 Primary 165,786 Secondary 239,996 6 Tertiary 47,491 Education financing 4 % of GDP 5.1 % of government spending — 2 Average 2015 PISA scores Mathematics 524 0 Reading 517 19 5 19 0 19 5 19 0 19 5 19 0 85 19 0 95 20 0 20 5 10 19 0 Science 516 5 6 6 7 7 8 9 0 0 5 19 20 19 Total Females Sources: EdStats data (World Bank, various years) and Korean Ministry of Education data; OECD 2016b. Note: GDP = gross domestic product; — = not available; PISA = Programme Source: Barro and Lee 2013. for International Student Assessment. E D U C AT I O N S YS T E M S I N E A S T A S I A A N D PAC I F I C AT A G L A N C E 221 Overview who finish middle school can enter high school, although some high schools require an entrance Education is comparatively decentralized in exam, which is designed and administered by the Republic of Korea, with significant inde- the provincial or municipal Board of Education. pendence given to individual schools to set Students sit for the College Scholastic Ability curriculum and manage hiring. Korea invests Tests at the end of high school in order to gain significant resources in its education sys- admission to university. tem, and society views education as highly important. Korea educates about 11 million students from preprimary through tertiary Curriculum and textbooks education and employs about 550,000 teach- Within the Ministry of Education, the ers in preprimary through tertiary institutions. National Curriculum Division is responsible for providing the framework for curriculum Administration development. Individual schools are respon- sible for designing their own curricula within The Ministry of Education (formerly the national framework. the Ministry of Education, Science, and Technology) is the principal government Teachers body responsible for education. Offices of Education oversee policy implementation and Teachers in primary education are generally school management at the provincial, munici- trained at teachers colleges or departments of pal, and county levels. Boards of Education education within universities. Teachers who at the subnational level oversee administrative complete their course of study are awarded a and policy-related issues. teacher’s certificate, for which the Ministry of Education sets the standards. To begin work- ing at a public school, teachers must pass an Compulsory education and testing exam offered by a provincial office of educa- Korea has nine years of compulsory education, tion. Teachers have access to significant in- although most students continue through upper- service training opportunities. secondary and university education. Students Lao People’s Democratic Republic FIGURE A.9 Average years of educational attainment in the Republic of Korea, 1950–2010 TABLE A.7 Selected indicators for the Lao People’s Democratic Republic 14 Indicator Statistic Average years of educational attainment 12 Demographic and macroeconomic indicators among population 25 and older Population (millions) 6.8 10 GDP (US$, billions) 12 GDP per capita (US$) 1,812 8 Adult literacy rate (%) 58.3 Unemployment rate (%) 1.4 6 Student enrollment Preprimary 4 Number enrolled 175,492 % female 50 2 Primary 0 Number enrolled 850,466 % female 47 19 5 19 0 19 5 19 0 19 5 19 0 85 19 0 95 20 0 20 5 10 19 0 5 6 6 7 7 8 9 0 0 5 19 20 19 Secondary Total Females Number enrolled 640,231 % female 50 Source: Barro and Lee 2013. table continues next page 222 GROWING SMARTER TABLE A.7 Selected indicators for the Lao People’s FIGURE A.10 Average years of educational Democratic Republic (continued) attainment in the Lao People’s Democratic Republic, 1950–2010 Indicator Statistic Student enrollment (continued) 5.0 Tertiary Average years of educational attainment Number enrolled 130,191 4.5 among population 25 and older % female — 4.0 Number of teachers 3.5 Preprimary 9,791 Primary 35,206 3.0 Secondary 34,200 2.5 Tertiary 6,902 2.0 Education financing % of GDP 3.3 1.5 % of government spending 12.2 1.0 Source: EdStats data (World Bank, various years). 0.5 Note: GDP = gross domestic product; — = not available. 0 19 5 19 0 19 5 19 0 19 5 19 0 85 19 0 95 20 0 20 5 10 19 0 5 6 6 7 7 8 9 0 0 5 19 20 19 Overview Total Females The Lao People’s Democratic Republic edu- Source: Barro and Lee 2013. cates about 1.8 million students from prepri- mary through tertiary education and employs FIGURE A.11 Selected development indicators in about 85,000 teachers in preprimary through the Lao People’s Democratic Republic tertiary institutions. Percent Administration Primary school enrollment 100 The Ministry of Education administers the 80 system at the central level, but Provincial 60 Education Services are responsible for local ser- Access to 40 Secondary vice delivery, including staffing, inspection, and Internet 20 school school support. District Education Bureaus enrollment 0 support schools by providing pedagogical advisers. Village Education Development Committees promote the involvement of the community in school affairs. Public and pri- vate institutions can provide education, as Access to Access to improved long as they adhere to the national curriculum water sanitation approved by the Ministry of Education. Sources: National statistical authorities and World Development Indicators data (World Bank, various years). Compulsory education and testing Primary education, which lasts five years, was Curriculum and textbooks made free and compulsory by decree in 1996. The Research Institute for Educational Examinations that certify completion (and are Sciences, under the Ministry of Education, necessary for progression) are administered at oversees curriculum and textbook devel- the end of primary (grade 5), lower-secondary opment. In the context of the Education (grade 9), and upper-secondary (grade 12) Sector Development Framework 2009–15, school. the national curriculum was reformed. E D U C AT I O N S YS T E M S I N E A S T A S I A A N D PAC I F I C AT A G L A N C E 223 The changes affected early childhood educa- Overview tion and school readiness, nonformal educa- Macao SAR, China has one of the most tion, local curricula, and vocational skills. decentralized systems in the region. Most schools are privately run by a variety of social Teachers and religious groups. Despite the prevalence Lower- and upper-secondary teachers are of private education, most primary and sec- required to have completed 11 years of formal ondary schools are either free or subsidized. schooling, followed by three years of preservice Education is offered in Chinese, English, and teacher training for lower-secondary teach- Portuguese. Macao SAR, China educates ers and four to five years for upper-secondary about 100,000 students from preprimary teachers. In-service teacher upgrading centers through tertiary education and employs are located in various provinces. about 5,000 teachers in preprimary through secondary schools. Macao SAR, China Administration TABLE A.8 Selected indicators for Macao SAR, China The Education and Youth Affairs Bureau Indicator Statistic manages the education system. Demographic and macroeconomic indicators Population 587,606 GDP (US$, billions) 46 Compulsory education and testing GDP per capita (US$) 78,586 Macao SAR, China offers 15 years of free, Adult literacy rate (%) 96.5 compulsory education, from preprimary Unemployment rate (%) 1.5 through upper-secondary. It participates Student enrollment Preprimary in international assessments, including Number enrolled 14,552 the Programme for International Student % female 48 Assessment (PISA), where it has scored Primary among the top in recent rounds. Number enrolled 24,252 % female 51 FIGURE A.12 Average years of educational Secondary attainment in Macao SAR, China, 1950–2010 Number enrolled 30,088 % female 45 Tertiary 9 Number enrolled 30,771 Average years of educational attainment 8 % female — among population 25 and older Number of teachers 7 Preprimary 916 6 Primary 1,722 5 Secondary 2,629 Tertiary — 4 Education financing 3 % of GDP 2.0 % of government spending 13.4 2 Average 2015 PISA scores 1 Mathematics 544 Reading 509 0 19 5 19 0 19 5 19 0 19 5 19 0 85 19 0 95 20 0 20 5 10 19 0 Science 529 5 6 6 7 7 8 9 0 0 5 19 20 19 Sources: EdStats (World Bank, various years) and Macao Education and Total Females Youth Affairs Bureau data; OECD 2016b. Note: GDP = gross domestic product; — = not available; PISA = Programme for International Student Assessment. Source: Barro and Lee 2013. 224 GROWING SMARTER Curriculum and textbooks Overview Private schools are encouraged to develop A multicultural society with a history of colo- their own curricula. Public schools follow a nial rule, Malaysia has faced challenges in pro- government-developed curriculum. viding high-quality education to its 8 million students over the past several decades. It has Teachers invested significant resources in addressing these issues and focused top-level attention Teachers in private and public schools are with its National Education Blueprint 2013– required to have a postsecondary degree 2025. It employs about 580,000 teachers in from a teachers college or equivalent pro- preprimary through secondary schools. gram. The Education and Youth Affairs Bureau is responsible for checking teacher qualifications. Administration The Ministry of Education manages educa- Malaysia tion. It is organized into federal, state, district, and state levels. At the primary level, stu- TABLE A.9 Selected indicators for Malaysia dents are divided between schools instructing Indicator Statistic in Bahasa Melayu and schools instructing in Demographic and macroeconomic indicators Chinese or Tamil. Students who receive edu- Population (millions) 30.3 cation in Chinese or Tamil generally receive GDP (US$, billions) 296 an additional year of instruction after com- GDP per capita (US$) 9,766 pleting primary school to aid their transition Adult literacy rate (%) 93.1 to Bahasa Melayu instruction. Unemployment rate (%) 2.0 Student enrollment Preprimary Compulsory education and testing Number enrolled 934,418 % female 49 Malaysia requires just six years of compul- Primary sory education, although enrollment through Number enrolled 3,107,870 lower-secondary is nearly 100 percent. % female 48 It gives several nationwide exams during Secondary primary school to gauge student develop- Number enrolled 2,991,123 ment and progress. Upon completion of % female 50 lower-secondary school, students sit for a Tertiary test to determine their eligibility for upper- Number enrolled 817,587 secondary school. Depending on their score, % female — students are sent to academic or vocational Number of teachers schools. Malaysia also participates in vari- Preprimary 60,704 Primary 269,757 ous international assessments, including PISA Secondary 248,920 and Trends in International Mathematics and Tertiary — Science Study (TIMSS). Education financing % of GDP 5.0 % of government spending 19.7 Curriculum and textbooks Average 2015 PISA scores The Curriculum Development Division of Mathematics 446 the Ministry of Education is responsible for Reading 431 Science 443 formulating and disseminating the curricu- lum. Students in the upper-secondary levels Sources: EdStats data (World Bank, various years); OECD 2016b. Note: GDP = gross domestic product; — = not available; PISA = Programme choose their own curriculum outside of the for International Student Assessment. core classes. E D U C AT I O N S YS T E M S I N E A S T A S I A A N D PAC I F I C AT A G L A N C E 225 FIGURE A.13 Average years of educational Mongolia attainment in Malaysia, 1950–2010 TABLE A.10 Selected indicators for Mongolia 12 Indicator Statistic Average years of educational attainment Demographic and macroeconomic indicators 10 among population 25 and older Population (millions) 3.0 GDP (US$, billions) 11.8 8 GDP per capita (US$) 3,973 Adult literacy rate (%) 98.3 6 Unemployment rate (%) 5 Student enrollment 4 Preprimary — Primary 2 Number enrolled 251,204 % female 51 0 Secondary Number enrolled 281,961 19 5 19 0 19 5 19 0 19 5 19 0 85 19 0 95 20 0 20 5 10 19 0 5 6 6 7 7 8 9 0 0 5 19 20 19 % female 50 Total Females Tertiary Number enrolled 179,540 Source: Barro and Lee 2013. % female — Number of teachers Preprimary 4,907 FIGURE A.14 Selected development indicators Primary 8,901 in Malaysia Secondary 20,908 Percent Tertiary — Education financing % of GDP 4.6 Primary school enrollment % of government spending 12.2 100 Source: EdStats data (World Bank, various years). 80 Note: GDP = gross domestic product; — = not available. 60 Access to 40 Secondary Internet 20 school Overview enrollment 0 Mongolia educates about 700,000 students from primary through tertiary education and employs about 35,000 teachers in preprimary to secondary schools. Access to Access to improved sanitation water Administration Sources: National statistical authorities and World Development The Ministry of Education, Culture, and Indicators data (World Bank, various years). Science is responsible for all education- related supervision, from the preprimary through the tertiary level, as well as techni- cal and vocational education and training Teachers and nonformal education. Education and The Teacher Education Division within the Culture Departments in each province over- Ministry of Education assists teachers colleges see financing, administration, quality assur- and universities in developing curriculum and ance, and school administration staffing. The training programs. State Inspectorate of Education, Culture, and 226 GROWING SMARTER FIGURE A.15 Average years of educational participated in the 2007 round of TIMSS attainment in Mongolia, 1950–2010 but did not appear in any international com- parisons due to inability to meet sampling 12 criteria. It did not participate in subsequent rounds, and it has not participated in PISA. Average years of educational attainment 10 among population 25 and older Curriculum and textbooks 8 The Ministry of Education, Culture, and 6 Science is responsible for developing, approv- ing, and publishing textbooks and the 4 national curriculum. 2 Teachers 0 Teachers must have a four-year degree from a teachers college. 19 5 19 0 19 5 19 0 19 5 19 0 85 19 0 95 20 0 20 5 10 19 0 5 6 6 7 7 8 9 0 0 5 19 20 19 Total Females Source: Barro and Lee 2013. Papua New Guinea FIGURE A.16 Selected development indicators TABLE A.11 Selected indicators for Papua in Mongolia New Guinea Percent Indicator Statistic Demographic and macroeconomic indicators Primary school enrollment Population (millions) 7.6 100 GDP (US$, billions) 17.9 80 GDP per capita (US$) 2,268 60 Adult literacy rate (%) — Access to 40 Secondary Unemployment rate (%) 3 Internet 20 school enrollment Student enrollment 0 Preprimary Number enrolled 177,660 % female — Primary Number enrolled 1,426,884 Access to % female 46 Access to improved sanitation Secondary water Number enrolled 378,365 % female — Sources: National statistical authorities and World Development Indicators data (World Bank, various years). Tertiary Number enrolled — Science monitors the implementation of edu- % female — cation policy and regulation. At the school Number of teachers level, school boards (comprising teachers, Preprimary — Primary 31,586 parents, and representatives) are responsible Secondary 13,804 for managing school affairs. Tertiary — Education financing Compulsory education and testing % of GDP — % of government spending — Primary and lower-secondary school Source: EdStats data (World Bank, various years). are compulsory in Mongolia. Mongolia Note: GDP = gross domestic product; — = not available. E D U C AT I O N S YS T E M S I N E A S T A S I A A N D PAC I F I C AT A G L A N C E 227 Overview Administration Papua New Guinea educates about 1.9 mil- The Organic Law on Provincial and Local- lion students from the preprimary through Level Government of 1995 regulates decen- the secondary levels. It employs about tralization and focuses on district and local 45,000 teachers in primary and secondary communities. The National Department of schools. Education is responsible for preparing educa- tion plans, including curriculum, standards, and teacher education and training. Provincial governments are responsible for providing FIGURE A.17 Average years of educational services (operational costs). Communities attainment in Papua New Guinea, 1950–2010 are responsible for providing land for school infrastructure. 4.5 Average years of educational attainment 4.0 Compulsory education and testing among population 25 and older 3.5 Universal basic education is a policy goal, 3.0 but there are no legal provisions regard- 2.5 ing free or compulsory education. Up to grade 7, assessment is school based; the 2.0 Primary Education Certificate Examination 1.5 at the end of grade 6 certifies completion 1.0 of basic education. In grades 8–12, school- based and standardized assessments are 0.5 administered. 0 19 5 19 0 19 5 19 0 19 5 19 0 85 19 0 95 20 0 20 5 10 19 0 5 6 6 7 7 8 9 0 0 5 19 20 19 Total Females Curriculum and textbooks The National Department of Education Source: Barro and Lee 2013. prepares policies on curricula; local-level governments and district administrations are responsible for developing and produc- ing curriculum materials. At the primary FIGURE A.18 Selected development indicators level, teachers are involved in developing in Papua New Guinea components of the curriculum. More than Percent 800 languages are spoken in Papua New Guinea. Local languages are used in pri- Primary school enrollment mary education; English is the main lan- 100 guage of instruction at higher levels. An 80 outcome-based curriculum was introduced 60 in 2005. 40 20 Access to Access to Internet 0 sanitation Teachers Primary teachers train at community-based teachers colleges; secondary education teach- ers are trained at national teachers colleges Access to improved water or universities, which typically provide three- year training programs. Local communities Sources: National statistical authorities and World Development select teachers. Provincial inspectors rate Indicators data (World Bank, various years). them. 228 GROWING SMARTER The Philippines FIGURE A.19 Average years of educational attainment in the Philippines, 1950–2010 TABLE A.12 Selected indicators for the Philippines 9 Indicator Statistic Average years of educational attainment Demographic and macroeconomic indicators 8 among population 25 and older Population (millions) 100.7 7 GDP (US$, billions) 292 6 GDP per capita (US$) 2,899 Adult literacy rate (%) 96.4 5 Unemployment rate (%) 7.1 4 Student enrollment Preprimary 3 Number enrolled 1,165,771 2 % female — Primary 1 Number enrolled 13,686,643 0 % female 48 19 5 19 0 65 19 0 19 5 80 19 5 19 0 95 20 0 20 5 10 19 0 5 6 7 7 8 9 0 0 5 19 19 20 19 Secondary Number enrolled 6,766,952 Total Females % female 51 Tertiary Source: Barro and Lee 2013. Number enrolled 2,625,385 % female — Number of teachers FIGURE A.20 Selected development indicators Preprimary — in the Philippines Primary 435,385 Percent Secondary 194,373 Tertiary — Education financing Primary school enrollment % of GDP 2.7 100 % of government spending 13.2 80 Source: EdStats data (World Bank, various years). 60 Note: GDP = gross domestic product; — = not available. 40 Secondary Access to Internet 20 school enrollment Overview 0 The Philippines is in the middle of implement- ing an ambitious slate of reforms, including the extension of secondary education and the Access to shift to 13 years of compulsory education. improved Access to sanitation It educates about 24 million students from water preprimary through tertiary education and employs about 630,000 teachers in primary Sources: National statistical authorities and World Development Indicators data (World Bank, various years). and secondary schools. and resources. Under the central depart- Administration ment, regional and division offices manage The Department of Education is the main the education system at the local level, per- administrative body responsible for formu- forming school quality audits, providing lating and implementing education policy as resources, and managing administrative well as managing education infrastructure services. E D U C AT I O N S YS T E M S I N E A S T A S I A A N D PAC I F I C AT A G L A N C E 229 Compulsory education and testing TABLE A.13 Selected indicators for Singapore (continued) In 2012–13 the Philippines extended com- pulsory education to 13 years, including one Indicator Statistic year of kindergarten, and added two years Number of teachers (continued) of schooling to upper-secondary, delaying Secondary 15,560 entrance into tertiary education. It uses sev- Tertiary 6,051 eral nationwide exams to gauge student learn- Education financing ing, including the National Achievement Test, % of GDP 2.9 % of government spending 20.0 given at the primary and secondary levels. It Average 2015 PISA scores also participates in TIMSS, scoring near the Mathematics 564 bottom. Reading 535 Science 556 Curriculum and textbooks Sources: EdStats data (World Bank, various years); OECD 2016b. Note: Gender disaggregated data for enrollments not available in EdStats. The Curriculum Development Division of the GDP = gross domestic product; — = not available; PISA = Programme for International Student Assessment. Department of Education is responsible for managing and developing national curricu- FIGURE A.21 Average years of educational lum standards. The Instructional Materials attainment in Singapore, 1950–2010 Council, attached to the Department of Education, is responsible for textbook-related policies. 12 Average years of educational attainment 10 among population 25 and older Teachers All teachers in the Philippines must have 8 completed a four-year degree in education. 6 Secondary school teachers must also have tertiary-level instruction in the subject they 4 teach. Hiring takes place at the school level. 2 Singapore 0 19 5 19 0 19 5 19 0 19 5 19 0 85 19 0 95 20 0 20 5 10 19 0 TABLE A.13 Selected indicators for Singapore 5 6 6 7 7 8 9 0 0 5 19 20 19 Total Females Indicator Statistic Demographic and macroeconomic indicators Source: Barro and Lee 2013. Population (millions) 5.6 GDP (US$, billions) 293 GDP per capita (US$) 52,889 Adult literacy rate (%) 96.8 Unemployment rate (%) 3.0 Overview Student enrollment (number enrolled) One of Asia’s highest-performing educa- Preprimary — Primary 294,602 tion systems, Singapore is also one of the Secondary 232,003 smallest systems in the region, with only Tertiary 198,634 about 725,000 students enrolled in primary Number of teachers through tertiary and about 40,000 teachers. Preprimary — Singapore’s highly functioning system has Primary 16,893 served as an example for others in the region table continues and around the world. 230 GROWING SMARTER Administration TABLE A.14 Selected indicators for Taiwan, China (continued) The Ministry of Education governs educa- tion. In addition to controlling government- Indicator Statistic run schools, it supervises Singapore’s private Student enrollment (continued) schools. Singapore also has an extensive pri- Upper-secondary 694,025 vate school system, overseen by the Council Tertiary 1,309,441 for Private Education. % female 51 Number of teachers Preprimary 47,184 Compulsory education and testing Primary and junior-secondary 143,156 Upper-secondary 54,575 Singapore has nine years of compulsory Tertiary 48,096 schooling. At the end of sixth grade, stu- Education financing dents sit for the Primary School–Leaving % of GDP 3.8 Examination, which determines what course % of government spending 20.5 they take in secondary school. In addition to Average 2015 PISA scores national exams, Singapore participates in a Mathematics 542 wide range of international exams, including Reading 497 PISA and TIMSS, scoring first among all par- Science 532 ticipants in both examinations in 2015. Sources: OECD 2016b; Taiwan Ministry of Education, Taiwan Bureau of Statistics data. Note: Gender disaggregated data for enrollments not available from public Curriculum and textbooks ministry data for preprimary and upper-secondary. GDP = gross domestic product; PISA = Programme for International Student Assessment. The Curriculum Planning and Development Division of the Ministry of Education is Overview responsible for the curriculum framework. Taiwan, China is in the midst of reforming its Recent changes to the national curriculum education system, moving from decades of a have emphasized noncognitive competen- highly centralized system that focused on rote cies in addition to traditional academic memorization to a more student-centered sys- domains. tem that prioritizes the needs of a changing economic landscape. It educates about 4.4 mil- Teachers lion students from preprimary through tertiary Singapore has one teacher training institute, education and employs about 300,000 teach- the National Institute of Education. ers in preprimary through tertiary institutions. Taiwan, China Administration TABLE A.14 Selected indicators for Taiwan, China The Ministry of Education oversees the administration and regulation of the educa- Indicator Statistic tion system at the national, provincial, and Demographic and macroeconomic indicators municipal levels. It is also responsible for Population (millions) 23.5 private educational institutions, setting stan- GDP (US$, billions) 530 dards and requirements for their establish- GDP per capita (US$) 22,540 Adult literacy rate (%) 98.7 ment and administration. Unemployment rate (%) 3.8 Student enrollment Compulsory education and testing Preprimary 492,781 Primary and junior-secondary 1,861,086 Taiwan, China recently moved from 9 to % female 48 12 years of compulsory education. The table continues move was part of a broader reform aimed at E D U C AT I O N S YS T E M S I N E A S T A S I A A N D PAC I F I C AT A G L A N C E 231 modernizing the education system, includ- Qualification Exam to begin working at a ing updating and decentralizing curriculum public school. Taiwan, China recently estab- design, and reducing the testing burden in lished the Nationwide Teacher In-Service advancement to upper-secondary. Taiwan, Education Information Web, which provides China participates in international educa- information on training opportunities for tional assessments, scoring high in the most teachers to continue their skills develop- recent round of PISA. ment. Along with the curricular reforms, Taiwan, China is working to improve its training programs to move teachers from Curriculum and textbooks lecture-style classes to more student-centered Recent reforms have focused on decentral- lessons. izing and streamlining the curriculum. The curriculum was revised at the central level with the move to 12 years of compulsory Thailand education. It provides a cohesive frame- work and guidelines for schools rather than TABLE A.15 Selected indicators for Thailand mandates. The new curriculum moves away from rote memorization and teacher lec- Indicator Statistic tures and toward a more student-centered Demographic and macroeconomic indicators approach. Population (millions) 68.0 GDP (US$, billions) 395 GDP per capita (US$) 5,816 Teachers Adult literacy rate (%) 93 Unemployment rate (%) 0.9 All teachers are trained at either a teach- Student enrollment ers college or the education department Preprimary of a university. Upon completion of their Number enrolled 1,636,244 studies, candidates must take the Teacher % female 46 Primary Number enrolled 5,081,079 FIGURE A.22 Average years of educational % female 48 attainment in Taiwan, China, 1950–2010 Secondary Number enrolled 6,757,174 12 % female 46 Tertiary Average years of educational attainment 10 Number enrolled 2,235,450 among population 25 and older % female — Number of teachers 8 Preprimary — Primary 300,968 6 Secondary 240,005 Tertiary — 4 Education financing % of GDP 4.1 2 % of government spending 18.9 Average 2015 PISA scores 0 Mathematics 415 Reading 409 19 5 19 0 19 5 19 0 19 5 19 0 85 19 0 95 20 0 20 5 10 19 0 5 6 6 7 7 8 9 0 0 5 19 20 19 Science 421 Total Females Sources: OECD 2016b; UIS data. Note: GDP = gross domestic product; — = not available; PISA = Programme Source: Barro and Lee 2013. for International Student Assessment. 232 GROWING SMARTER Overview Administration Thailand has made the transition from a largely The Ministry of Education regulates and agrarian, low-income society to an upper- supervises both public and private schools. middle-income country. It educates about The Office of Higher Education Commission, 16 million students from preprimary through a department of the Ministry of Education, tertiary education and employs about 550,000 administers and controls public and private teachers in primary and secondary schools. universities. FIGURE A.23 Average years of educational Compulsory education and testing attainment in Thailand, 1950–2010 The constitution guarantees 12 years of free basic education; nine years of school atten- 8 dance is compulsory. Admission to upper- Average years of educational attainment 7 secondary school is regulated through an among population 25 and older entrance exam. Upon graduating from high 6 school, students need to pass the CUAS 5 (Central University Admission System). Half of the CUAS is the Ordinary National 4 Educational Test and the Advanced National 3 Educational Test results; the other half is the fourth-level grade-point average. 2 1 Curriculum and textbooks 0 With reforms in 2001 and 2008, Thailand shifted its content-based curriculum to a 19 5 19 0 19 5 19 0 19 5 19 0 85 19 0 95 20 0 20 5 10 19 0 5 6 6 7 7 8 9 0 0 5 19 20 19 Total Females standards-based approach (OECD 2016a). Source: Barro and Lee 2013. Teachers FIGURE A.24 Selected development indicators Teacher training is offered at universities and in Thailand teachers colleges. The mainstay of teacher out- Percent put is the government-run Rajabhat Universities (formerly Rajabhat Institutes), the traditional teacher training colleges in most provinces. Primary school enrollment 100 80 60 Timor-Lestea Access to 40 Secondary school TABLE A.16 Selected indicators for Timor-Leste Internet 20 enrollment 0 Indicator Statistic Demographic and macroeconomic indicators Population (millions) 1.2 GDP (US$, billion) 1.4 GDP per capita (US$) 1,131 Access to Access to Adult literacy rate (%) 58.3 improved sanitation Unemployment rate (%) 4.7 water table continues next page Sources: National statistical authorities and World Development Indicators data (World Bank, various years). a. An educational attainment line graph is not provided for Timor-Leste; those data are not available. E D U C AT I O N S YS T E M S I N E A S T A S I A A N D PAC I F I C AT A G L A N C E 233 TABLE A.16 Selected indicators for Timor-Leste a young government, and issues related to the (continued) language of instruction present challenges, Indicator Statistic but Timor-Leste is pushing forward with reform plans. Student enrollment Preprimary Number enrolled 16,206 Administration % female 50 Primary The Ministry of Education is responsible for Number enrolled 245,847 the design and implementation of education % female 48 policies. Centrally located schools offer all Secondary nine years of compulsory education; smaller Number enrolled 118,935 local schools generally offer only primary % female 55 education (first through fourth grade). Tertiary Number enrolled 18,553 % female — Compulsory education and testing Number of teachers Preprimary 615 Timor-Leste has nine years of compul- Primary 7,714 sory education, separated into three cycles Secondary 4,441 (grades 1–4, 5–6, and 7–9). Students sit for Tertiary — a national exam upon completion of ninth Education financing grade. Access rates for education are high % of GDP 6.6 through ninth grade but fall off for upper- % of government spending 7.7 secondary school. Timor-Leste has not par- Source: EdStats data (World Bank, various years). ticipated in any international education Note: GDP = gross domestic product; — = not available. assessments. It conducted three rounds of FIGURE A.25 Selected development indicators Early Grade Reading Assessments (EGRAs), in Timor-Leste in 2009, 2011, and 2017. Percent Curriculum and textbooks Primary school enrollment 100 Timor-Leste has struggled with curriculum 80 because of the lack of ubiquity of its two offi- 60 cial languages, Portuguese and Tetum. The lack 40 Secondary of a shared language has made it difficult to Access to Internet 20 school develop a comprehensive curriculum for all stu- enrollment 0 dents. Additionally, low levels of technical and pedagogical competence among teachers make delivery of the existing curriculum difficult. Access to Teachers Access to improved water sanitation Teachers in Timor-Leste are not highly quali- fied: less than 40 percent of teachers in basic Sources: National statistical authorities and World Development education have university education. Policy Indicators data (World Bank, various years). makers are seeking to address these issues with a variety of in-service training programs Overview as well as a teacher management system that Timor-Leste is still dealing with the legacy of tracks teachers’ academic and professional its independence movement and the manage- experience in order to deliver targeted train- ment of internal and external conflicts. Large ing. The lack of experience at the system level youth populations, significant refugee issues, has impeded progress. 234 GROWING SMARTER Vietnam FIGURE A.26 Average years of educational attainment in Vietnam, 1950–2010 TABLE A.17 Selected indicators for Vietnam Indicator Statistic 8 Average years of educational attainment Demographic and macroeconomic indicators 7 among population 25 and older Population (millions) 91.7 GDP (US$, billions) 193 6 GDP per capita (US$) 2,111 5 Adult literacy rate (%) 93.5 Unemployment rate (%) 2.3 4 Student enrollment 3 Preprimary Number enrolled 3,754,975 2 % female 47 Primary 1 Number enrolled 7,543,632 0 % female 49 19 5 19 0 65 19 0 19 5 80 19 5 19 0 95 20 0 20 5 10 19 0 5 6 7 7 8 9 0 0 5 19 19 20 19 Secondary Number enrolled 7,463,517 Total Females % female — Tertiary Source: Barro and Lee 2013. Number enrolled 2,466,643 % female — Number of teachers FIGURE A.27 Selected development indicators Preprimary 215,518 in Vietnam Primary 392,136 Percent Secondary 150,133 Tertiary 34,461 Primary school enrollment Education financing 100 % of GDP 5.7 80 % of government spending 18.5 60 Average 2015 PISA scores 40 Mathematics 495 20 Reading 487 Access to Access to 0 Science 525 Internet sanitation Sources: EdStats data (World Bank, various years); OECD 2016b. Note: GDP = gross domestic product; — = not available; PISA = Programme for International Student Assessment. Access to improved water Overview Sources: National statistical authorities and World Development Indicators data (World Bank, various years). Vietnam has one of the highest gross domes- tic product (GDP) growth rates in the region. Its education system has received much atten- Administration tion in recent years, thanks to its outstand- ing results on the last two PISA assessments. The Ministry of Education and Training Vietnam educates about 21 million students oversees the education sector. It is respon- from preprimary through tertiary education, sible for setting policy directions for all lev- employing about 800,000 teachers (pre- els of education. Provincial Departments of primary through tertiary). It has more than Education and Training and District Bureaus 30,000 primary and secondary schools. of Education and Training oversee school E D U C AT I O N S YS T E M S I N E A S T A S I A A N D PAC I F I C AT A G L A N C E 235 management, staffing, and distribution of References financial resources. Barro, R., and J.-W. Lee. 2013. “A New Data Set of Educational Attainment in the World, 1950– Compulsory education and testing 2010.” Journal of Development Economics 104 (C): 184–98. Vietnam has nine years of compulsory edu- National Bureau of Statistics of China. 2016. cation. Policies introduced in 2014 aim at China Statistical Yearbook 2016. Beijing: universalizing access to preprimary. About National Bureau of Statistics. http://www.stats 92 percent of children age three to five were .gov.cn/tjsj/ndsj/2016/indexeh.htm. enrolled in preprimary school in 2017. Access OECD (Organisation for Economic Co-operation to senior-secondary school (grades 10–12) and Development). 2016a. Education in is rationed through exams developed and Thailand: An OECD-UNESCO Perspective, delivered at the provincial level. Students Reviews of National Policies for Education. Paris: OECD Publishing. who complete upper-secondary school are ———. 2016b. PISA 2015 Results . Vol. I: eligible to sit for the Ký thi Trung ho . c phô ´c gia, a standardized test that certi- Excellence and Equity in Education. Paris: thông quô OECD Publishing. fies completion of secondary education and Tandon, P., and T. Fukao. 2015. Educating the determines college admissions based on score Next Generation: Improving Teacher Quality points. in Cambodia. Washington, DC: World Bank. Curriculum and textbooks Statistical websites The Ministry of Education and Training Japanese Ministry of Education, Culture, regulates curriculum and textbook design. Sports, Science, and Technology. “Education Vietnam’s major education reform, passed Statistics.” http://www.mext.go.jp/en by the National Assembly in 2014, mandates /publication/statistics/index.htm. a renovation of the national curriculum fol- Korean Ministry of Education. “Korean Education Statistics Service.” http://kess.kedi.re.kr/eng lowing a competency-based approach to /index. allow students to develop 21st century skills Macao Education and Youth Affairs Bureau. and the ability to apply knowledge. Together “Education and Youth Affairs.” http://portal with the roll-out of the new curriculum antic- .dsej.gov.mo/webdsejspace/internet/Inter ipated for 2019–22, Vietnam plans to move _main_page.jsp?langsel=E. from a single set of textbooks, currently pro- Taiwan Ministry of Education, Taiwan Bureau of vided by the Educational Publishing House, Statistics. “Education Statistics.” http://english an agency of the Ministry of Education and .moe.gov.tw/lp.asp?ctNode=11432&CtUnit=1 Training, to multiple sets of textbooks. 348&BaseDSD=16. UIS (UNESCO Institute for Statistics). “Thailand.” Institute for Statistics, Paris. http://uis.unesco Teachers .org/country/TH. World Bank. Various years. Education Statistics Ninety-five institutions offer preservice train- (EdStats) [database]. Washington, DC: ing for teachers. All teachers must meet mini- World Bank. http://datatopics.worldbank.org mum academic requirements, and secondary /education/. school teachers must have attended a teachers ———. Various years. World Development college or university to be certified. Teachers Indicators [database]. Washington, DC: World are required to participate in at least 30 days Bank. https://data.worldbank.org/data-catalog of in-service training a year. /world-development-indicators. ECO-AUDIT Environmental Benefits Statement The World Bank Group is committed to reducing its environmental footprint. In support of this commitment, we leverage electronic publishing options and print- on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, greenhouse gas emissions, and waste. We follow the recommended standards for paper use set by the Green Press Initiative. The majority of our books are printed on Forest Stewardship Council (FSC)–certified paper, with nearly all containing 50–100 percent recycled con- tent. The recycled fiber in our book paper is either unbleached or bleached using totally chlorine-free (TCF), processed chlorine–free (PCF), or enhanced elemental chlorine–free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://www.worldbank.org/corporateresponsibility.