Financing Higher Education

Public Disclosure Authorized Public Disclosure Authorized

Public Disclosure Authorized Putting Higher Education to Work

Skills and Research for Growth in East Asia

WORLD BANK EAST ASIA AND PACIFIC

Public Disclosure Authorized REGIONAL REPORT

Putting Higher Education to Work Skills and Research for Growth in East Asia WORLD BANK EAST ASIA AND PACIFIC REGIONAL REPORT Well known for their economic success and dynamism, countries in the East Asia and Pacifi c region must tackle an increasingly complex set of challenges to continue on a path of sustainable development. 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 report 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. WORLD BANK EAST ASIA AND PACIFIC REGIONAL REPORT iii World Bank East Asia and Pacifi c Regional Report

Putting Higher Education to Work

Skills and Research for Growth in East Asia © 2012 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org All rights reserved

1 2 3 4 14 13 12 11

This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The ﬁ ndings, interpretations, and conclusions expressed in this volume do not necessarily reﬂ ect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judge- ment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries.

Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruc- tion and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Ofﬁ ce of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: [email protected]. ISBN: 978-0-8213-8490-9 eISBN: 978-0-8213-8911-9 DOI: 10.1596/978-0-8213-8490-9 Library of Congress Cataloging-in-Publication data has been requested. Cover photo: Graduates attend the 2009 bachelor’s degree commencement ceremony at the Huazhong University of Science and Technology, Wuhan, Hubei province, China. © Feature China / European Pressphoto Agency. Cover design: Naylor Design, Inc. Contents

Foreword ...... xiii

Acknowledgments ...... xv

Abbreviations ...... xvii

Summary ...... 1 Despite impressive gains, higher education could contribute even more to East Asia’s development agenda ...... 1 Five disconnects ...... 2 Public policy and its three pillars ...... 2 Country priorities, policies, and reform ...... 4 Notes ...... 4 1 Higher Education for Growth through Skills and Research ...... 5 East Asia’s economic landscape ...... 6 Role and impact of higher education ...... 12 From higher education to growth: Skills and research ...... 17 Conclusion ...... 33 Notes ...... 34 2 Is Higher Education Meeting Its Promises? ...... 39 Higher education and skills for growth: The main issues ...... 39 Quantity of higher education graduates ...... 42 Quality of higher education graduates ...... 50 Research and innovation ...... 57 Conclusion ...... 63 Notes ...... 63 3 Disconnects in Higher Education ...... 67 The ﬁ rst disconnect: Between higher education and employers (skill users) ...... 67 The second disconnect: Between higher education and companies (research users) ...... 77

v vi CONTENTS

The third disconnect: Between higher education and research institutions (research providers) ...... 81 The fourth disconnect: Among higher education institutions themselves and between these institutions and training providers (horizontal disconnect across skill providers) ...... 82 The ﬁ fth disconnect: Between higher education and earlier education (schools) (vertical disconnect across skill providers) ...... 84 Conclusion and moving forward ...... 93 Notes ...... 97 4 Financing Higher Education ...... 101 Financing needs ...... 101 How to fund priority activities ...... 111 Summary of policy priorities ...... 124 Notes ...... 125 5 Managing Public Higher Education ...... 127 Global moves to autonomy ...... 128 Autonomy for low- and middle-income East Asia ...... 129 Two fundamental issues ...... 132 Moving forward ...... 143 Notes ...... 145 6 Providing Stewardship for Higher Education ...... 147 Providing effective coordination among government bodies ...... 148 Steering private delivery ...... 149 Encouraging effective university-industry links ...... 162 Stewardship of the internationalization of higher education ...... 166 Notes ...... 170

Appendixes ...... 173 A Number and Type of Higher Education Institutions in East Asia ...... 174 B Economic Indicators ...... 175 C Trends in Returns to Skill and Share of Skilled Workers, by Sector ...... 179 D Changes in Wage and Industry Education Premiums, by Subsector ...... 182 E Openness, Technology, and Demand for Tertiary Graduates, Regression Tables ...... 186 F Demand for Job-Speciﬁ c and Generic Skills in East Asia ...... 189 G Determinants of Skill Gap Indicators ...... 194 H Reasons for Skill Shortages in East Asia ...... 200 I Comparative Demand- and Supply-Side Indicators ...... 202 J Skill Gaps in East Asia ...... 204 K Skill Gaps across Professionals and Skilled Workers in Indonesia and the Philippines . . . .208 L Doctoral Degrees Earned in Science and Engineering ...... 210 M Simulations for Financing Higher Education ...... 211 References ...... 213

Boxes 1.1 Private and public beneﬁ ts of higher education ...... 13 1.2 Deﬁ ning skills ...... 15 CONTENTS vii

1.3 A snapshot of skills for innovation in Indonesia, the Philippines, and Thailand . . . . . 28 1.4 Skills for the future ...... 30 2.1 Preventing wasted talent in East Asia ...... 52 3.1 Poor diversifi cation in Cambodian higher education ...... 68 3.2 The rationale for public intervention in higher education ...... 95 4.1 Determinants of higher education access ...... 106 4.2 Evaluating the Student Loan Fund in Thailand ...... 111 4.3 Overseas scholarships for outstanding students from Thailand ...... 117 4.4 Competitive funds as an innovative fi nancing tool ...... 119 4.5 Matching funds in Hong Kong SAR, China, and Singapore ...... 122 4.6 Advantages of income-contingent loans ...... 122 4.7 The higher education contribution scheme in Australia ...... 123 5.1 Institutional autonomy defi ned ...... 128 5.2 Translating autonomy into more socially effi cient outcomes ...... 130 5.3 Addressing local labor market needs in the United States and Mexico ...... 130 5.4 East Asian autonomy ...... 132 5.5 Higher education autonomy outside East Asia ...... 134 5.6 Autonomy in Japan’s higher education before and after the National University Corporation Act of 2004 ...... 136 5.7 Governing boards worldwide ...... 138 5.8 Evaluating faculty performance in China ...... 140 5.9 Curriculum reform for East Asia ...... 142 5.10 National qualifi cations frameworks ...... 144 6.1 Expanding private higher education in the Republic of Korea ...... 155 6.2 Resource diversifi cation in China and Malaysia ...... 159 6.3 Technology licensing offi ces in three economies ...... 164 6.4 Creating university-industry links in Ireland ...... 165 6.5 Intermediary organizations ...... 166 6.6 E-learning and virtual universities as instruments of internationalization ...... 168

Figures 1.1 Conceptual framework ...... 7 1.2 A schematic of income groups and technology clusters ...... 8 1.3 Sectoral value added as a share of GDP, 1997 and 2007 ...... 10 1.4 Tertiary enrollment and per capita GDP, 2008 ...... 14 1.5 Tertiary enrollment and labor productivity, 2005 ...... 14 1.6 Educational attainment and skills profi ciency ...... 16 1.7 Share of tertiary-educated workers in technologically and nontechnologically innovative fi rms ...... 17 1.8 Correlation between STEM supply and patents, 2004–09 ...... 17 1.9 Trends in higher education R&D, 1996–2007 ...... 18 1.10 Trends in patents, 1996–2007 ...... 18 1.11 Trends in journal articles, 1995–2007 ...... 19 1.12 Trends in technology licensing, 1975–2007 ...... 19 1.13 Tertiary GERs in East Asia and some comparator economies, 1970–2010 ...... 20 1.14 Tertiary GERs in East Asia and OECD average, latest available year ...... 21 1.15 Trends in wage education premiums and educated workforce in selected East Asian economies ...... 23 1.16 Share of tertiary-educated workers by foreign ownership and export status of fi rms . . . 25 viii CONTENTS

1.17 Importance of generic skills by sector and export orientation, Indonesia, 2008 . . . . . 27 1.18 Labor market outcome indicators of TVET and university graduates, Mongolia, Indonesia, and Thailand ...... 29 1.19 R&D expenditure, East Asia and OECD ...... 31 1.20 Correlation between higher education R&D and scientifi c and technical journals . . . 34 1.21 Correlation between higher education R&D and patents ...... 34 2.1 Business climate obstacles and skill bottlenecks ...... 41 2.2 Time to fi ll professional vacancies ...... 41 2.3 Skill bottlenecks, technology, and openness ...... 42 2.4 Proportion of adult population with university qualifi cations ...... 43 2.5 Benchmarking quantity gaps in tertiary education, Cambodia, China, and Vietnam...... 45 2.6 Benchmarking quantity gaps in tertiary education, Mongolia and Thailand ...... 46 2.7 Benchmarking quantity gaps in tertiary education, Indonesia and the Philippines . . . 47 2.8 Tertiary gross completion rates and per capita income ...... 48 2.9 Tertiary gross enrollment ratio and number of journal articles, latest available year ...... 49 2.10 Enrollment shares in science and engineering, latest available year ...... 50 2.11 Predicted ratios of enrollment and completion in tertiary education, by key characteristics ...... 51 2.12 Tertiary unemployment rates and time to fi ll professional vacancies ...... 53 2.13 Key job-specifi c skill gaps in Vietnam ...... 55 2.14 Key job-specifi c skill gaps in Malaysia ...... 56 2.15 Average monthly salaries of Thai employees reporting a particular skill as a top-three defi ciency compared with employees not reporting skills as a top-three defi ciency ...... 57 2.16 Average monthly salaries of Malaysian employees (in manufacturing) with very good versus very poor skills, according to employer ...... 57 2.17 Share of workers needing no training, by education level in the Philippines, 2008 . . . 58 2.18 Royalty and license fee payments, 1995–2009 ...... 59 2.19 Royalty and license fee receipts, 1995–2009 ...... 60 2.20 Scientifi c and technical articles per million inhabitants, East Asia and the rest of the world, 2007 ...... 61 2.21 Leading ways of acquiring technological innovation in fi rms, Malaysia, Mongolia, and Thailand ...... 62 2.22 Sources of product innovation at fi rm level in Vietnam ...... 63 3.1 Five disconnects in higher education ...... 68 3.2 Proportion of tertiary student enrollments by fi eld of study, 2008 ...... 69 3.3 Science and engineering enrollment shares and wage education premiums in manufacturing, Cambodia and Mongolia ...... 70 3.4 Share of upper-secondary and tertiary students enrolled in TVET ...... 71 3.5 Employers’ perceptions on general and TVET tertiary education, Indonesia and the Philippines, 2008 ...... 72 3.6 Proportion of tertiary students enrolled in ISCED 6 programs ...... 73 3.7 Wage education premiums, Indonesia, the Philippines, and Thailand ...... 74 3.8 Trends in student-to-faculty ratios, 2001–07 ...... 75 3.9 Ratios of faculty with master’s degrees and PhDs, various years ...... 76 3.10 Intensity of university-industry links ...... 77 3.11 External collaboration for R&D activities, Thailand, 1999–2003 ...... 78 3.12 Number of scientifi c and technical journal articles and number of researchers in R&D ...... 80 CONTENTS ix

3.13 Secondary and tertiary gross enrollment ratios, 2007–08 ...... 84 3.14 Trends in shares of workers with secondary and tertiary education, four East Asian economies, various years ...... 85 3.15 Secondary education completion ...... 86 3.16 PISA 2009 scores for East Asia, United States, and OECD ...... 88 3.17 Relation between TIMSS scores and STEM enrollment shares ...... 89 3.18 Relation between TIMSS scores and journal articles ...... 90 3.19 Estimated survival rates for children ages 13–19, Cambodia, Indonesia, and Thailand ...... 91 3.20 Secondary and tertiary education completion rate by gender, Vietnam, Mongolia, and the Philippines ...... 92 3.21 Secondary and tertiary education completion rate, by ethnic group, Vietnam, 2008 ...... 92 3.22 Tertiary enrollment in proportion of secondary completion by ethnic group, Vietnam, 2008, and Thailand, 2006 ...... 92 3.23 Secondary and tertiary education completion, by urban or rural area, Vietnam, Mongolia, and the Philippines ...... 93 3.24 Tertiary enrollment in proportion to secondary completion by urban or rural area, selected economies ...... 93 3.25 Relationships between disconnects and policies ...... 96 4.1 Research and development spending in tertiary education as a share of GDP and journals, latest available year ...... 105 4.2 Research and development spending in tertiary education as a share of GDP, latest available year ...... 105 4.3 Change in ratio of tertiary enrollments and completions for the poorest and richest quintiles, selected East Asian economies ...... 107 4.4 Indonesia: Tertiary education expenditure as a percentage of annual household income ...... 109 4.5 Thailand: Private expenditure in higher education, by income quintile ...... 110 4.6 Public tertiary expenditure as a share of GDP, latest available year ...... 113 4.7 Public expenditure per tertiary pupil as a share of GDP per capita, 2006–07 ...... 114 4.8 Public expenditure per tertiary pupil as a share of GDP per capita, 1998–2009 ...... 115 4.9 Public tertiary education expenditure as a share of total public education expenditure, 2006–07 ...... 116 4.10 Overall public education expenditure as a share of GDP, latest year ...... 116 4.11 Research and development spending and total public spending in tertiary education, latest available year ...... 117 4.12 Tertiary spending per student as a share of GDP per capita and STEM enrollment share, latest available year ...... 118 4.13 Tertiary spending per student and journals per million people, latest available year ...... 118 5.1 Accountability relationships in an institutional autonomy setting ...... 129 6.1 Private higher education enrollment share, by region or country ...... 150 6.2 Private higher education shares ...... 151 6.3 Rating of graduates from public and private tertiary institutions, Indonesia and the Philippines, 2008 ...... 152 6.4 Higher education GER and enrollments, Republic of Korea, 1971–2008 ...... 154 B6.1 Enrollment by education level, Republic of Korea, 1971–2008 ...... 155 6.5 PHE enrollment share, Malaysia, Thailand, and Vietnam, 1994–2009 ...... 155 6.6 Tertiary students studying abroad, 2007 ...... 167 x CONTENTS

6.7 Asian students studying abroad in tertiary schools in the top-fi ve receiving countries, 2007 ...... 167 C.1 Return to skill by sector over time (tertiary and above), Indonesia ...... 179 C.2 Proportion of skilled workers by sector over time (tertiary and above), Indonesia . . . 179 C.3 Return to education by sector over time (tertiary and above), Philippines ...... 179 C.4 Proportion of educated workers by sector over time (tertiary and above), Philippines ...... 179 C.5 Return to skill by sector over time (tertiary and above), Thailand ...... 180 C.6 Proportion of skilled workers by sector over time (tertiary and above), Thailand . . . 180 C.7 Return to education by sector over time (tertiary and above), Vietnam ...... 180 C.8 Proportion of educated workers by sector over time (tertiary and above), Vietnam . . 180 C.9 Return to skill by sector over time (tertiary and above), Cambodia ...... 180 C.10 Proportion of skilled workers by sector over time (tertiary and above), Cambodia ...... 180 C.11 Return to skill by sector over time (tertiary and above), China ...... 181 C.12 Proportion of skilled labor by sector over time (tertiary and above), China ...... 181 C.13 Return to skill by sector over time (tertiary and above), Mongolia ...... 181 C.14 Proportion of skilled labor by sector over time (tertiary and above), Mongolia . . . . . 181 D.1 Change in unstandardized weighted industry and skill premiums by sector, Indonesia, 1996–2007 ...... 182 D.2 Change in unstandardized weighted industry and skill premiums by sector, Philippines, 1988–2006 ...... 183 D.3 Change in unstandardized weighted industry and skill premiums by sector, Indonesia, 1990–2004 ...... 183 D.4 Change in unstandardized weighted industry and skill premiums by sector, Vietnam, 1992–2006 ...... 184 D.5 Change in unstandardized weighted industry and skill premiums by sector, Cambodia, 1997–2007 ...... 184 D.6 Change in unstandardized weighted industry and skill premiums by sector, China, 1995–2005 ...... 185 D.7 Change in unstandardized weighted industry and skill premiums by sector, Mongolia, 1998–2007 ...... 185 F.1 Key job-specifi c and generic skills in a sample of East Asian economies (according to employers) ...... 189 F.2 Key skills in a sample of East Asian economies (according to employees) ...... 192 H.1 Reasons for skill shortages, by sector, Philippines ...... 200 H.2 Reasons for skill shortages, by sector, Indonesia ...... 200 H.3 Three main causes of vacancies in manufacturing in Thailand identifi ed by fi rms ...... 201 H.4 Three main causes of vacancies in manufacturing in Malaysia identifi ed by fi rms ...... 201 H.5 Main obstacles for recruiting, Vietnam ...... 201 H.6 Most important reasons for vacancies in manufacturing in Mongolia ...... 201 J.1 Skill gaps identifi ed by employers ...... 204 J.2 Skill gaps according to employees ...... 207 K.1 Key generic skill gaps (according to employers), Philippines ...... 208 K.2 Key job-specifi c skill gaps (according to employers), Philippines ...... 208 K.3 Share of respondents identifying a gap in generic skills, Indonesia ...... 209 K.4 Share of respondents identifying a gap in behavioral skills, Indonesia ...... 209 M.1 Architecture of the simulation model ...... 211 CONTENTS xi

Tables 1.1 Per capita GDP averages, upper-income economies ...... 9 1.2 Manufacturing value added, upper-income economies ...... 9 1.3 Per capita GDP averages, middle-income economies ...... 10 1.4 Manufacturing value added, middle-income economies ...... 10 1.5 Per capita GDP averages, lower-income economies ...... 11 1.6 Manufacturing value added, lower-income economies ...... 11 1.7 Regression coeffi cients of technological and openness variables in a sample of East Asian economies ...... 25 1.8 Importance of technical, thinking, and behavioral skills for professionals ...... 26 1.9 R&D expenditure, 1996–2007 ...... 31 1.10 Composition of R&D expenditure ...... 32 2.1 Regression coeffi cients of technological and openness variables in a sample of East Asian economies ...... 43 2.2 Tertiary education quantity analysis by country groups ...... 44 2.3 Share of fi rst university degrees in science and engineering ...... 49 B2.1 2009 PISA reading achievement for poorest quintile, Shanghai, China ...... 52 2.4 Comparative skill gaps among professionals ...... 54 2.5 Number of patents granted by the USPTO, selected years ...... 58 2.6 World Intellectual Property Organization patent fi lings by origin and offi ce, 2007 ...... 59 2.7 Ranking of universities, East Asia, 2007, 2008, and 2009 ...... 60 2.8 Rankings of universities over time, 2004–09 ...... 61 2.9 Yearly average number of publications, 1980–2005 ...... 62 3.1 Earned science and engineering doctoral degrees, selected region or economy and selected fi eld, 2006 or most recent year ...... 73 3.2 Student-to-faculty ratios in tertiary education, 2007 ...... 75 3.3 Academic qualifi cations of faculty in a sample of East Asian economies ...... 76 3.4 Eighth-grade TIMSS scores for mathematics, selected East Asian economies, 1999, 2003, and 2007 ...... 87 3.5 Eighth-grade TIMSS scores for science, selected East Asian economies, 1999, 2003, and 2007 ...... 87 3.6 Intensity of disconnects by income and technology cluster group ...... 94 4.1 Gap between projected per-student expenditure needed and current levels, Indonesia ...... 103 4.2 Gap between projected per-student expenditure needed and current levels, Vietnam ...... 103 4.3 Gap between projected per-student expenditure needed and current levels, the Philippines ...... 103 4.4 Gap between projected per-student expenditure needed and current levels, Mongolia ...... 104 4.5 Snapshot of public university revenue breakdown by proportion of funding source, selected economies ...... 105 4.6 Vietnam: Tertiary education costs per month, 2009 ...... 106 4.7 Mongolia: State Training Fund recipients, by program area ...... 108 4.8 China: Net payment in regular universities, 2008 ...... 108 4.9 Block grants in East Asia ...... 119 4.10 Average tertiary tuition fees ...... 121 5.1 Substantive and procedural autonomy ...... 133 xii CONTENTS

5.2 Autonomy among higher education institutions, East Asia ...... 135 B5.6 Institutional autonomy of higher education institutions in Japan, 2003 and 2007 ...... 136 5.3 University governance, East Asia ...... 139 5.4 External quality assurance, East Asia ...... 141 6.1 Private higher education shares and income groups ...... 151 6.2 Government regulations on private higher education, East Asia ...... 157 A.1 Number and type of higher education institutions in East Asia ...... 174 B.1 GDP growth of upper-income East Asian economies ...... 175 B.2 Savings and investment of upper-income East Asian economies ...... 175 B.3 Exports of upper-income East Asian economies ...... 175 B.4 Leading export sectors, upper-income East Asian economies ...... 176 B.5 GDP growth of middle-income East Asian economies ...... 176 B.6 Exports of middle-income East Asian economies ...... 176 B.7 Savings and investment of middle-income East Asian economies ...... 177 B.8 Leading export sectors, middle-income East Asian economies ...... 177 B.9 GDP growth of lower-income East Asian economies ...... 177 B.10 Exports of lower-income East Asian economies ...... 178 B.11 Savings and investment of lower-income East Asian economies ...... 178 B.12 Leading export sectors, lower-income East Asian economies ...... 178 E.1 Openness, technological innovation, and the demand for skills ...... 186 E.2 Openness, technological innovation, and the demand for skills: Robustness to technological variables ...... 186 E.3 Openness, technological innovation, and the demand for skills: Robustness to alternative samples ...... 187 E.4 Openness, technological innovation, and the demand for skills: Robustness to alternative samples ...... 188 G.1 Determinants of time to fi ll skilled vacancies ...... 194 G.2 Determinants of managerial perceptions of skills and education of East Asian workforce ...... 194 G.3 Determinants of managerial perceptions of skills of workforce: Robustness to additional variables ...... 195 G.4 Determinants of time to fi ll skilled vacancies: Robustness to additional variables . . . 196 G.5 Determinants of managerial perceptions of skills of workforce: Robustness to different geographical areas ...... 197 G.6 Determinants of managerial perceptions of skills of workforce: Robustness to different geographic areas, by income levels ...... 198 G.7 Determinants of time to fi ll skilled vacancies: Robustness to different geographical areas ...... 198 G.8 Determinants of time to fi ll skilled vacancies: Robustness to different geographical areas, by income levels ...... 199 I.1 Comparative demand- and supply-side indicators, by country ...... 202 L.1 Science and engineering doctoral degrees earned in selected regions and locations, by fi eld (2000 or most recent year) ...... 210 Foreword

xpanding employment and increas- improved to become more responsive and rel- ing productivity are at the top of the evant to the needs of the labor market and Eagenda for policy makers across the the economy as a whole. Using innovative globe. For countries seeking to spur growth, ﬁ rm surveys and the latest available evidence creating jobs and raising productivity are from the region, the authors shed light on the primary concerns. For young people, too, functional skills that workers must possess to these issues are a priority. Students and grad- be employable and to support ﬁ rms’ competi- uates everywhere are asking for more and tiveness and productivity. They also examine better opportunities—to study and work, how higher education systems can produce and to learn and create new knowledge and the commercially applicable research that enterprises. will help countries apply, assimilate, adapt, With skilled labor and technological capa- and develop the new technologies that will bility increasingly becoming the touchstones drive growth. of competitiveness in an open and integrated Though this volume focuses specifically world environment, the role of higher educa- on the developing countries of East Asia, its tion in economic growth is thus taking on methodologies, messages, and analysis will be a greater significance. Not only do higher important resources for students, researchers, education institutions help impart the behav- and policy makers who study and shape the ioral, cognitive, and technical skills that delivery of higher education and training in make workers effective in the labor market, other regions around the world. The authors they are increasingly valued as the engines of offer valuable and succinct guidance on some research that can drive innovation, entrepre- of the most effective policy measures being neurship, and productivity. deployed by national and regional govern- Realizing the potential of higher educa- ments, by firms, and by universities them- tion to spur growth is a priority for East selves to enhance the contribution that higher Asia. Putting Higher Education to Work: education systems can make to economic Skills and Research for Growth in East Asia change. is a comprehensive look at higher education Higher education will continue to be a in East Asia—how it has changed, how it core issue for the World Bank and its client will continue to evolve, and how it can be countries, and it will also constitute a central

xiii xiv FOREWORD

pillar of the labor and educational agenda help East Asia reach this objective in an for years to come. I am confi dent that this increasingly competitive global environment. volume, the fi rst in the East Asia and Pacifi c Regional Report Series, will help the region’s James W. Adams economies embrace the challenge of achiev- Vice President ing rapid growth led by gains in productivity. East Asia and Pacifi c Region The measures proposed in this volume should World Bank Acknowledgments

he preparation of this study was led Philippines), Vicharn Panich (former Chair- by Emanuela di Gropello under the man of the Thai Higher Education Commis- Tguidance of Emmanuel Jimenez and sion), Annie Koh (Dean, Ofﬁ ce of Executive Eduardo Velez Bustillo. The volume was and Professional Education, and Academic written by Emanuela di Gropello (lead Director, Singapore Management University), author), Prateek Tandon, and Shahid Yusuf, and Phonephet Boupha (Director General of with significant contributions from many Higher Education, Lao People’s Democratic others. Sonali Ballal and Eleanor Wang Republic). The study was edited by Bruce provided outstanding research inputs. The Ross-Larson. volume beneﬁ ted from many commissioned Finally, this volume incorporates the valu- background papers written by interna- able comments received by policy makers, tional and regional experts outside the core academics, and international donors during team. The core team is grateful for analyti- several regional and country consultation cal work done by David Chapman, Dandan events that took place from June 2010 to July Chen, Richard Doner, Hal Hill, HRInc., 2011. A Global Development Learning Net- Patarapong Intarakumnerd, Daniel Levy, Vu work virtual consultation took place in June Hoang Linh, Giang Thanh Long, Wanhua 2010 to discuss the initial findings of the Ma, Bagyo Moeliodihardjo, Ka Ho Mok, study with 180 participants from the region. Hena Mukherjee, Kaoru Nabeshima, Ree- Other regional consultations occurred in hana Raza, Omporn Regel, Bryan Ritchie, Seoul, Korea, at the Global Human Resource Chris Sakellariou, Edita Tan, Le Viet Thuy, Forum (October 2010); in Phuket, Thailand, Kin Bing Wu, Weiping Wu, and Yesim Yil- at the Conference on Governance and Financ- maz. These background papers are available ing of Higher Education in East Asia (October at http://www.worldbank.org/eap/highered. 2010); in Jakarta, Indonesia, at the Regional The World Bank internal peer reviewers Skill Conference (March 2011); and in Bali, were Andreas Blom, Alberto Rodriguez, Jamil Indonesia, at the East Asia Summit Educa- Salmi, and Jee-Peng Tan. Excellent comments tion Ministers’ Meeting (July 2011). The task were also received from Luc Weber (Rector team is grateful for the reactions and com- Emeritus, University of Geneva), Mike Luz ments of the delegations at these and other (former Undersecretary of Education for the country events.

Abbreviations

ABD Asian Development Bank ALL Adult Literacy and Life Skills Survey DGHE Directorate General of Higher Education EU European Union GDP gross domestic product GER gross enrollment ratio GPA grade point average HECS Higher Education Contribution Scheme HEI higher education institution ICS investment climate survey ICT information communication technology ILO International Labour Organization IMF International Monetary Fund ISCED International Standard Classiﬁ cation of Education IT information technology KIC Knowledge-Integrating Community MIT Massachusetts Institute of Technology MNC multinational corporations NQF national qualiﬁ cations framework NSB National Science Board NUS National University of Singapore OECD Organisation for Economic Co-operation and Development PHE private higher education PHEI private higher education institution PISA Programme for International Student Assessment POSTECH Pohang University of Science and Technology PPP purchasing power parity PROPHE Program for Research on Private Higher Education S&E science and engineering SAR Special Administrative Regions

xvii xviii ABBREVIATIONS

SJTU Shanghai Jiao Tong University SLF Student Loan Fund SME small and medium enterprise SMK sekolah menengah kejuruan (vocational secondary schools), Indonesia SMU sekolah menengah umum (general secondary schools), Indonesia SOE state-owned enterprises STEM science, technology, engineering, and math STF State Training Fund TAMA Technology Advanced Metropolitan Area THES Times Higher Education Supplement TIMSS Trends in International Mathematics and Science Study TLO technology licensing ofﬁ ce TTI Technology Transfer Initiative TVET Technical and vocational education and training UIL university-industry link UIS UNESCO Institute for Statistics USPTO U.S. Patent and Trademark Ofﬁ ce VHLSS Vietnam Household Living Standards Survey WB World Bank WDR World Development Report WIPO World Intellectual Property Organization

Country abbreviations used in fi gures ARG Argentina AUS Australia AUT Austria BEL Belgium BRA Brazil CAN Canada CHE Switzerland CHL Chile CHN China DEU Germany DNK Denmark ESP Spain FIN Finland FRA France GBR United Kingdom GIB Gibraltar HKG Hong Kong SAR, China IDN Indonesia IND India ISR Israel ITA Italy JPN Japan KHM Cambodia KOR Korea, Republic of LAO Lao People’s Democratic Republic MEX Mexico ABBREVIATIONS xix

MYS Malaysia NLD Netherlands NOR Norway NZL New Zealand PHL Philippines RUS Russian Federation SGP Singapore SWE Sweden THA Thailand TUR Turkey VNM Vietnam ZAF South Africa

Note: These three-letter country codes are part of the International Organization for Stan- dardization (ISO) 3166 standard to represent countries.

Currency equivalents (rates effective April 20, 2011)

Thai Baht Cambodian Riels US$ 1 = 29.960 THB US$ 1 = KHR 3998 THB 1 = US$ 0.334 KHR 3998 = US$ 0.0002

Chinese Yuan Singapore Dollars US$ 1 = 6.525 CNY US$ 1 = SGD 1.24 CNY 1 = US$ 0.153 SGD 1 = US$ 0.8

Vietnamese Dong Mongolian Tugriks US$ 1 = VND 2092 US$ 1 = MNT 1226.5 VND = US$ 0.0048 MNT 1 = US$ 0.0008

Indonesian Rupiah Lao Kip US$ 1 = IDR 8654.5 US$ 1 = LAK 8045 IDR 1 = US$ 0.0001 LAK 1 = US$ 0.0001

Philippine Pesos Special Drawing Rights US$ 1 = PHP 43.279 1XDR = US$ 1.602 PHP 1 = US$ 0.023

Fiscal year January 1–December 31

Summary

Despite impressive gains, higher Access has increased dramatically in low- education could contribute even and middle-income East Asia, but higher more to East Asia’s development education is not yet fulfilling its potential. agenda Low- and middle-income East Asia has been expanding access to higher education over the East Asia is a model region that has grown past 20 to 30 years, going in many cases from rapidly, but its low- and middle-income coun- very low enrollment rates to enrollment rates tries face the challenges of maintaining growth of 20 percent or more. Fluctuating between and climbing the income ladder, both requir- 10 and 50 percent, tertiary gross enrollment ing improvements in productivity. Higher rates of low- and middle-income East Asian education is critical in this effort because it countries are on par with those of countries provides the high-level skills and research to of similar income levels but are still below apply current technologies and to assimilate, those of upper-income countries. A key vul- adapt, and develop new technologies, two nerability to sustained growth of low- and 1 drivers of productivity. It can thus be a key middle-income East Asia is in developing and driver of growth. Individuals with more years deploying enough of the right types of skills of higher education score higher on measures and research for a more competitive global of skill competencies than do individuals with economy. Higher education can reduce this few or no years of higher education. And aca- vulnerability by sufficiently providing the demic, technical, thinking, and behavioral skills and research to increase productivity skills and productivity are shown to be posi- and innovation. tively related. Several indicators of innovation But higher education today does not suf- also support the need for higher education. An fi ciently provide its graduates with the skills innovative fi rm is associated with an increase that fi rms need to increase their productivity. of about 25 percentage points in its share of The quantity of higher education graduates is workers with more than 12 years of school- still too low for the labor market in countries ing. And countries that have more science like Cambodia, China, and Vietnam. More and engineering graduates and that engage in important than quantity, however, is quality. more higher education research tend to have Across low- and middle-income East Asia, better innovation outcomes. employers expect workers—particularly those

1 2 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

with higher education—to possess the techni- earlier education institutions, and other skill cal, behavioral, and thinking skills to increase providers. Higher education outcomes are their productivity and growth. They need the product of the interactions between all science, technology, engineering, and math these actors, and failing to consider the links (STEM) skills. They also need the problem- between higher education institutions and solving and creative skills to support a higher- the wider world around them leads to poor value-added manufacturing sector and the performance and poor outcomes. business, thinking, and behavioral skills for a At least five disconnects are evident in higher-productivity service sector. Employer East Asia’s higher education systems, roughly perceptions and wage skill premiums point ranked for their impact on higher education to gaps in all these groups of skills in newly outcomes: hired professionals across the region. • A gap between higher education institu- Compounding the quality issues is higher tions and the skill needs of employers education’s exclusion of capable and talented • A weak research and technology nexus students because of their socioeconomic sta- between higher education institutions and tus, ethnicity, and rural residence. Across companies the region, ethnic minorities, in particular, • A separation between teaching and appear to exhibit shortfalls in both tertiary research institutions (or more generally enrollment and completion. teaching and research functions) Higher education also fails to provide the • A disconnect among higher education type of research needed to boost techno- institutions themselves and between these logical upgrading in fi rms. Governments are institutions and training providers urging universities to go beyond simply pro- • A separation between higher education viding skills to support innovation through institutions and earlier education institu- research and technology. Research enables tions (schools) universities to produce ideas for the business community, thereby contributing to knowl- These disconnects are pervasive throughout edge and technological innovation through low- and middle-income East Asia, but their basic and applied research and technology intensity varies across income and technology transfer. But international rankings and cluster groups, tending to be most severe in research output indicate that low- and low-income–lower-technology cluster coun- middle-income East Asian higher education tries. The intensity of the various disconnects systems are not providing research of ade- within each country will also vary, giving rise quate quality. Even mere university involve- to diversifi ed diagnostics by country. ment in technology adaptation and upgrading is limited in lower- and middle-income East Asia, with the possible exception of China. Public policy and its three pillars In Malaysia, Mongolia, and Thailand, for Most disconnects are due to information, instance, universities are mentioned as lead- capacity, and incentive constraints that have ing in acquiring technological innovations (in been poorly addressed, suggesting market a broad sense) by only 1–2 percent of fi rms. and policy failures. Public policy has the potential to address them (although not all Five disconnects the policies are strictly related to higher education, and not all actors and interactions at A major reason higher education fails to do the core of higher education systems will be its job is that its institutions have been man- equally amenable to policies). Reforms in the aged as disconnected individual institutions. following areas should have the highest pri- Higher education needs to be seen as a “sys- ority in low- and middle-income East Asia: tem” including both institutions and the stakeholders that interact with them. Such • Financing adequately the aspects of higher a system includes fi rms, research institutes, education that correct for externalities SUMMARY 3

and market failures, such as research, for teaching and research are allocated science, technology, engineering, math, across institutions and targeting scholarships and scholarships and loans for the poor and loans better. Being innovative in rais- and disadvantaged—all within a coherent ing resources means attracting more private fi nancing framework funds. Variable fee policies, combined with • Managing public higher institutions effective loan schemes, are one effective way by supporting more autonomous and to mobilize private resources while protect- accountable institutions ing access for the poor and disadvantaged. • Providing better stewardship for the higher Public-private matching grant schemes are education system, especially by putting in other options successfully applied in some place the right incentives for the private upper-income East Asian countries.2 sector to thrive, ensuring links between industry and providers, and handling the Better management of public interaction between domestic and interna- institutions tional higher education Public tertiary institutions are critical in More effi cient spending and fi nancing East Asia because 70 percent of all students of higher education are enrolled in them. Yet decision-making autonomy in public sector institutions In most countries, public spending goes to remains underdeveloped. Academic and institutions regardless of whether they are procedural autonomy are particularly mis- addressing public goods (such as research), aligned, with greater autonomy in academic externalities (as in STEM), or equity con- than procedural issues. Accountability cerns. Too often it is not allocated in a way structures fall particularly short in devel- that is performance based. This lack of pri- oping relationships with nongovernment oritization and these ineffi ciencies contribute stakeholders (employers, faculty members, to many of the disconnects by underfunding students). Having insufficient autonomy of research, STEM skills, scholarships, and to select the staff members they want and other equity-related measures. Improved to decide on their academic programs public spending can address many of these makes it diffi cult for tertiary institutions issues. For instance, it can help tackle to deliver what fi rms need—perpetuating the disconnect with earlier education by the disconnect between skill providers and supporting student transitions from second- users. And the lack of accountability of ary to tertiary education through scholar- university management to representative ships and loans. It can also help tackle the university boards makes it less compelling disconnects between universities and fi rms for institutions to fulfi ll the needs of skill in research and technology by supporting or research users. higher funding for research in universities By setting the right incentives for pub- combined with performance-based funding. lic institutions through appropriate auton- The precise challenges vary across the indi- omy and accountability, governments vidual countries, but all countries face some can address many disconnects. Although common imperatives. First, countries need autonomy is most urgent for middle-income to be selective in deciding their priorities for countries, the beneficial effects of auton- funding. Second, they need to identify priori- omy are valid at all income levels. It is also ties for public spending (research, inclusive- essential for all countries to complete their ness, and so forth). Because public resources accountability frameworks by delegating are scarce, countries should be as efficient greater power and responsibilities to insti- as possible in allocating these resources and tution boards and by providing students the be more innovative in raising funds. Greater information to choose and the opportunity efficiency means being more selective and to move across institutions, while ensuring performance based in the way public funds continuous accountability to governments. 4 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

Exercising stewardship of the higher and thus to different immediate priorities. education system Addressing skill gaps through higher-quality graduates and greater inclusiveness should Beyond managing the public sector, higher be the fi rst immediate goal of lower-income education departments need to coordinate countries, followed by gradually increasing actors and connections not under their full the quantity of graduates and starting to build control but critical to the performance of research capacity in some university depart- the sector. Exercising stewardship requires ments relevant to the economy’s needs. For the capacity to coordinate higher education middle-income countries, it will be critical departments with other departments and to have both a solid skill base and a stronger ministries, steer private higher education capacity for innovation through skills and institutions, support links between higher research. The priorities will differ across education institutions and fi rms in skills and countries. Indonesia and the Philippines, for research, and handle the interaction between instance, should focus on improving gradu- the domestic and international higher edu- ate quality and inclusiveness, while building cation markets. The lack of interaction research capacity in a few universities. China between higher education institutions and should continue developing its skill base (both fi rms to some extent refl ects a lack of infor- quality and quantity) and further scale up its mation on what works and a lack of legal research effort and impact. and fi nancial incentives to connect. Discon- The underlying disconnects, constraints, nects between skill providers and users are and priorities will dictate the policy pillars, also related to a private sector not fulfi lling levers, and specifi c policy measures for each its potential because of poor policy, regula- country and country group. The report illus- tory, and information incentives. Both issues trates the policy priorities overall and by point to stewardship failures. income group and technology cluster. With Governments can improve their steward- this suggested set of priorities as a start, indi- ship by ensuring that private and public pro- vidual country diagnostics of disconnects viders complement each other, especially in (including better assessments of the intensity meeting the skill needs of employers. They of the various disconnects in each country) can ensure favorable policies, clear and effi - and their causes are then needed to inform cient regulation and information, and better more refi ned policy measures. Moving from access of both public and private providers policy to reform will require careful con- to student loans (and competitive funding sideration of the political economy and the for research). Governments can also connect broader economic policies affecting higher fi rms and providers of skills and research by education to decide on the appropriate con- sharing best practices—from collaborating in tent, sequencing, and pace of reform. curriculum development to setting up university incubators—and by offering the incentives to make these university-industry links Notes work (bringing in intermediaries and providing matching funds). 1. Higher education is defi ned broadly to include all public and private formal institutions of learning that take place beyond upper- Country priorities, policies, secondary education. and reform 2. Under a coherent fi nancing framework, private funding would not only complement public The fi nal goals of higher education are much funding in fi nancing some of these activities the same across countries. But the chal- but also support system expansion and diver- lenges, disconnects, and constraints—which sifi cation (targeting some of the other country are related not just to higher education— priorities, such as increasing enrollment or vary, leading to different intermediate goals service-related disciplines). Higher Education for Growth through Skills and Research 1

fundamental question facing East Asia’s drive toward greater productivity, Asia, especially its low- and middle- growth, and technological development. A income economies, is how to sus- This book introduces a conceptual frame- tain or even accelerate the growth of recent work for the analysis of higher education in decades. From 1950 to 2005, for example, lower- and middle-income countries in East the region’s real income per head rose seven- Asia (the target country group).2 The book fold. With aging populations, these econo- takes a broad definition of higher educa- mies will need to derive an increasing share tion to include all public and private formal of growth from productivity improvements institutions of learning beyond the upper- rather than from physical factor accumula- secondary level. These institutions award tion to drive growth.1 formal academic degrees, diplomas, or pro- The recent global economic and ﬁ nancial fessional certiﬁ cation and include, but are not crisis has served only to lend urgency to East limited to, universities, two- and four-year Asia’s search for avenues to higher produc- colleges, institutes of technology, religious- tivity and competitiveness in an increasingly based educational institutions, online and global market, ultimately leading to growth. distance learning, foreign branch campuses, Investment in human capital, physical invest- and other collegiate-level institutions (such as ment in research and development (R&D), vocational, trade, or career) (see appendix A technological progress, and the increase in for a detailed list of higher education institu- total factor productivity arising from scale tions in East Asia). economies and agglomeration effects are all At the very core of the conceptual frame- elements of the search. These investments work is the idea that higher education in low- will help East Asia reap the rewards of glo- and middle-income East Asia has the poten- balization and rapid technological develop- tial to lift productivity and competitiveness ment, promote within-sector productivity by providing the high-level skills demanded growth, and provide the necessary incentives by the labor market and by launching or for further labor reallocation toward high- expanding robust research needed for inno- productivity sectors. vation and growth. As important is the need Investing in education—particularly to consider higher education as a system com- higher education—is a crucial part of East posed of the higher education institutions

5 6 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

themselves; the other skill and research showing the vast differences of productivity users and providers that interact with them; and growth across countries in the region. To the underlying policies that support higher show how low- and middle-income countries education institutions; and the interactions can advance, this chapter then introduces the among higher education institutions, users, role of higher education in equipping individ- and providers. While it is important to note uals with skills and producing research that that higher education provides several non- can lead to greater productivity and growth. economic benefits such as nation building Chapter 2 continues with a diagnostic of and socialization, this book focuses on the higher education in skills and research in economic benefi ts of higher education as they low-, middle-, and upper-income economies relate to skills and research. in East Asia. Chapter 3 shows how failures The book argues that higher education to deliver on skills and research in lower- is failing to deliver skills for growth and and middle-income countries are related to research for innovation because of wide- critical disconnects between (a) higher edu- spread disconnects between higher education cation and (b) users and providers of skills institutions and other skill and research users and research. It demonstrates how these dis- and providers. These disconnects undermine connects are related to problems with poor the very functioning of the higher education information, low capacity, and weak incen- system. The main assumption of the report is tives. Given these challenges, chapters 4–6 that to deliver labor market skills to higher provide policy recommendations to address education graduates, these institutions (a) these problems and mitigate the disconnects must have characteristics that are aligned through better financing of higher educa- with what employers and employees need tion (chapter 4), better management of public and (b) must be well connected among them- higher education institutions (chapter 5), and selves and other skills providers. Similarly, to better stewardship for the higher education deliver research that can enhance innovation system (chapter 6). and productivity, higher education institutions need to have a strong role in research East Asia’s economic landscape provision and have strong links with fi rms and other research providers. The economies of East Asia can be divided Getting the system to work well requires into three income groups, which beyond a adequate information, capacity, and incen- certain income per capita tend to share some tives that are closely related to financial common characteristics in terms of eco- resources, public higher education manage- nomic structure, human development, and ment, and stewardship for higher education business climate. The fi rst income group is systems. Government and households have made up of Hong Kong SAR, China; Japan; a critical role to play at the policy stage, the Republic of Korea; Singapore; and including holding institutions accountable Taiwan, China. These economies also have for results and providing public and private a sophisticated economic structure and resources. The disconnects are ultimately advanced human development indicators. illustrative of weaknesses and failures in China, Indonesia, Malaysia, Mongolia, the the way fi nancial resources and institutions Philippines, and Thailand3 represent the sec- are managed. Prompt public intervention is ond middle-income group of East Asia. More required because no country in East Asia has precisely, according to the income classifi - reached high-income status without a strong cation adopted by the World Development higher education system. Report,4 fi ve of these countries are lower- Figure 1.1 illustrates this conceptual middle-income economies, and Malaysia framework as presented throughout the is an upper-middle-income economy. This chapters of this report. This first chapter group is fairly heterogeneous, but coun- presents the economic landscape in East Asia, tries in it generally share more developed HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 7

FIGURE 1.1 Conceptual framework

Chapters Goals 1 Productivity Growth

Higher education outcomes 1/2 Skills Research

System disconnects 3 Higher education institutions (HEIs) Causes of disconnects Disconnected actors

employers

companies as research users

research institutions low capacity weak incentives

poor information other HEIs and training providers

earlier education

Finance 4

Management of public sector 5

Stewardship 6

Source: Authors’ elaboration. economic structures, human development into three technology clusters on the basis indicators, and business climate than the of the skill and high-tech intensity of their lower-income economies have. Finally, the products and exports.6 Together, these lower-income group comprises Cambodia, metrics serve as a proxy for measuring an the Lao People’s Democratic Republic, and economy’s level of innovation in supporting Vietnam.5 These countries are late starters, productivity and its development of science as is apparent from their per capita gross and technology. As a result, this proxy pro- domestic products (GDPs) and develop- vides an assessment of an economy’s produc- ment levels. tivity and ability to move up the value chain In addition to the three income groups, within the service, manufacturing, non- the East Asian economies can be divided manufacturing industry, and agricultural 8 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

sectors. Although the relationship between with high levels of innovation and productiv- income and technology is similar, there is ity and thus form part of the top technology not a one-to-one match between the groups, cluster. While less manufacturing oriented, with economies within income groups per- Hong Kong SAR, China’s level of innova- forming at slightly different levels of techno- tion in the service sector also positions it in logical capacity (fi gure 1.2).7 that cluster, but at the bottom. Throughout The fi rst group’s economies grew rapidly the past two decades, exports as a share of from the mid-1960s to the mid-1990s and GDP have grown signifi cantly (table B.3 in have achieved high income levels (table 1.1). appendix B). They coincide therefore now with the group The share of services is also signifi cant and of high-income East Asian economies. Each has been increasing, reaching at least 60 per- of them, with the exception of Hong Kong cent of GDP in 2007 (fi gure 1.3). SAR, China, focused on manufacturing, China, Indonesia, Malaysia, Mongo- which accounts for around a quarter of GDP lia, the Philippines, and Thailand compose (table 1.2). Growth was sourced mainly from the middle-income country group. They capital accumulation, supplemented by gains are all in the under-US$7,000 per capita in factor productivity, with investment largely range (table 1.3) in constant nominal GDP. fi nanced by domestic savings. Indonesia, the Philippines, and especially Overall, these economies are East Asia’s Mongolia trail in manufacturing value technological leaders in a wide range of added (table 1.4) and investment. Electron- medium- and high-tech manufacturing ics and electrical products are the leading industries, including electronics and electri- export subsectors in China, Malaysia, the cal products, automobiles and parts, ship- Philippines, and Thailand, while Indone- building, and machinery (table B.4 in appen- sia and Mongolia still export mainly pri- dix B), but they also have service sectors mary and agro-based products (table B.8

FIGURE 1.2 A schematic of income groups and technology clusters

Korea, Rep. Japan

top Taiwan, China Singapore Hong Kong SAR, China

upper China

Malaysia middle Thailand middle Philippines lower

technology clusters Indonesia

Vietnam

low Cambodia Mongolia Lao PDR

low-income economies middle-income economies high-income economies

Source: Authors’ elaboration. Note: Income level is given by GDP per capita in 2009. Economies’ position in the fi gure refl ects their ranking by income and technology. HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 9

TABLE 1.1 Per capita GDP averages, upper-income economies

Nominal PPP (constant 2005 (constant 2000 US$) international $) Economy 1990–99 2000–09 1990–99 2000–09 Change (%) Japan 35,310 38,368 27,462 29,841 9 Singapore 18,269 26,500 29,201 42,358 45 Hong Kong SAR, China 23,115 29,188a 27,133 34,262a 26 Taiwan, China 11,102.2 16,044a 17,476 25,255a 45 Korea, Rep. 8,974 13,593 14,813 22,439 51

Source: World Development Indicators (WDI) database. Note: PPP = purchasing power parity. a. Hong Kong SAR, China, and Taiwan, China, average for 2000–08. in appendix B). Unlike those economies in TABLE 1.2 Manufacturing value added, upper-income economies the upper-income group, when they produce 1990–99 2000–09 or export electronics, this group’s coun- Economy (average % of GDP) (average % of GDP) Change (%) tries are mainly assemblers and processors Korea, Rep. 27.1 27.2 0 of electronic products, operating at a far Singapore 26.0 24.4 −6 lower level of technological competency and Taiwan, China 27.1 23.3a −14 at lower points in the value chain. Services Japan 24.0 21.1a −12 account for about 35–55 percent of GDP, Hong Kong SAR, and their share has been increasing in most China 9.4 3.9b −59 of these countries. This sector (as well as Source: WDI database. agriculture) is also working at lower lev- a. Japan and Taiwan, China, average for 2000–08. els of productivity. From that perspective, b. Hong Kong SAR, China, average for 2000–07. and because they cannot yet be classiﬁ ed as innovative, although China,8 Malaysia, and to a lesser degree Thailand began encourag- Cambodia, Lao PDR, and Vietnam are ing investment in R&D in the 1990s, they the lower-income East Asian countries, are part of another technology cluster. but they have moved onto a rapid-growth Within this lower cluster there are, how- trajectory (table 1.5). Although Lao PDR’s ever, substantial differences between coun- and Cambodia’s investment rates remain tries and countries’ subgroups. China, below 20 percent (table B.11 in appendix Indonesia, Malaysia, the Philippines, and B), Vietnam’s are similar to those of its Thailand started climbing the technologi- neighbors at earlier stages of development. cal ladder in the 1970s and have facilitated Vietnam and Cambodia are building man- technology assimilation by becoming increas- ufacturing capacity (table 1.6) but have yet ingly open and following the example of the to rise beyond the low-tech, labor-intensive East Asian frontrunners by promoting indus- stage of manufacturing, as clearly reﬂ ected trialization and infrastructure development in their exports (table B.12 in appendix through high levels of investment (table B.7 in B). Lao PDR is further behind in terms of appendix B). In contrast, Mongolia is much manufacturing capacity. Services represent more of a late starter and, as such, part of about 40 percent of GDP in all countries, an even lower cluster together with lower-in- and they are still working at low productiv- come countries. As will be further illustrated ity levels. These three countries, together below, even within the five countries with with Mongolia, are part of the lower tech- some developing innovative capacity, there nology cluster, with Vietnam sitting in are significant differences between China, front, but with generally smaller differ- clearly ahead, and countries such as Indone- ences in technological capability across sia and the Philippines, trailing behind. countries. 10 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 1.3 Sectoral value added as a share of GDP, 1997 and 2007

a. 1997 b. 2007

Korea, Rep. Korea, Rep. Hong Kong SAR, Hong Kong SAR, China China Singapore Singapore economies high-income economies Japan high-income Japan Mongolia Mongolia Philippines Philippines Indonesia Indonesia China China economies economies middle-income middle-income Thailand Thailand Malaysia Malaysia Cambodia Cambodia Lao PDR Lao PDR

economies Vietnam Vietnam low-income economies low-income 0 20 40 60 80 100 0 20 40 60 80 100 sectoral value added (% of GDP) sectoral value added (% of GDP) services manufacturing nonmanufacturing industry agriculture

Sources: WDI database.

TABLE 1.3 Per capita GDP averages, middle-income economies

Nominal PPP (constant 2005 (constant 2000 US$) international $) Economy 1990–99 2000–09 1990–99 2000–09 Change (%) Malaysia 3,382 4,539 8,619 11,567 34 Thailand 1,792 2,320 5,070 6,563 29 China 628 1,482 1,766 4,164 136 Philippines 901 1,094 2,385 2,897 21 Indonesia 767 945 2,602 3,204 23 Mongolia 432 580 1,930 2,593 34

Source: WDI database. Note: PPP = purchasing power parity.

This picture conveys the main challenges facing East Asian low- and middle-income TABLE 1.4 Manufacturing value added, middle-income economies economies. The members of the East Asian 1990–99 2000–09 technology clusters are competing in export Economy (average % of GDP) (average % of GDP) Change (%) markets—often ﬁ ercely with each other—to Thailand 29.5 34.4 17 enlarge the gains from industrialization and China 32.9 32.6 −1 trade, seeking to move up the value chain in Malaysia 27.0 28.8 7 manufacturing while increasing the produc- Indonesia 23.7 28.1 19 tivity of their growing service sectors. Philippines 23.3 22.5 −3 Japan leads the top technology cluster (and Mongolia 18.5 4.6 −75 upper-income group), followed by Hong Kong Source: WDI database. SAR, China; Korea; Singapore; and Taiwan, HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 11

TABLE 1.5 Per capita GDP averages, lower-income economies

Nominal (constant 2000 US$) PPP (constant 2005 international $) Economy 1990–99 2000–09 1990–99 2000–09 Change (%) Vietnam 299 530 1,187 2,109 78 Cambodia 239a 407 823a 1,400 70 Lao PDR 263 399 1,088 1,648 51

Source: WDI database. Note: PPP = purchasing power parity. a. Cambodia average for 1993–99.

China. These last four have some way to go TABLE 1.6 Manufacturing value added, lower-income economies before they achieve per capita income parity with Japan. Korea most closely approxi- 1990–99 2000–09 Economy (average % of GDP) (average % of GDP) Change (%) mates Japan in manufacturing breadth, while Hong Kong SAR, China, and Singapore are Vietnam 15.2 20.3 34 a the furthest removed. All have one thing in Cambodia 11.1 18.0 62 Lao PDR 14.2 14.6b 3 common: they are at the technological frontiers in their respective industries. Moreover, Source: WDI database. a. Cambodia average for 1993–99. Japan, Korea, and Taiwan, China, are among b. Lao PDR average for 2000–08. the most innovative industrial economies, pushing the technological frontiers in electronics, petrochemical, metallurgical, auto- Malaysia, Thailand, the Philippines, motive, and other fi elds, such as services and and Indonesia, in that order, have also built agriculture. Singapore is attempting to join up competitive manufacturing industries, the other members of the cluster through with electronics, textiles, automotive, and research in biotechnology and electronics, resource-based industries the most important. with some success. Its innovations are more Electronics and electrical engineering are conspicuous in services (such as managing the most important for the fi rst three coun- hotels and industrial parks, and urban plan- tries, whereas light manufacturing and pro- ning), water purifi cation, logistics, and use cessing activities are of greater signifi cance of information and communication technol- to Indonesia. All these countries have relied ogy. All fi ve economies are exemplars for the extensively on foreign direct investment to rest of East Asia. build manufacturing capacity and to master Middle-income countries of the middle production technology. Multinational corpo- technology cluster range in size from rations and their joint ventures account for China with more than 1.3 billion people to a sizable share of production for export (the Malaysia with 28 million. With China com- most advanced and competitive segment of fortably in the forefront of the technology industry),9 though production skills are now cluster, all these nations have acquired sig- widely diffused, with domestic manufacturers nifi cant manufacturing capabilities. China often able to compete with multinationals. is the world’s second-ranked industrial Nevertheless, members of the middle tech- nation and the largest exporter. In less than nology cluster are still some years and, in the two decades it has emerged as the leading case of Indonesia and the Philippines, decades producer of products ranging from steel, away from acquiring the technological capa- glass, and cement to electronics, photovol- bilities of the leading industrial nations. taic cells, and household appliances. China’s Except China (which is more diversifi ed and export competitiveness speaks to its rapidly has greater industrial depth), the other coun- maturing manufacturing capabilities and tries are largely focused on assembly, process- strengthening grasp of production technolo- ing, testing, and relatively low-value-adding gies across several sectors. operations, and their indigenous technological 12 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

capabilities are limited to simpler down- for these economies is to increase productivity stream activities. Their service sectors are in all sectors and break into manufacturing. also still low productivity. And although they Climbing the technology ladder is another are achieving high levels of effi ciency in pro- immediate objective. duction, their indigenous technological capa- Given this economic context, how can bilities are limited to simpler downstream low- and middle-income East Asia develop activities. Even China, whose exports overlap higher levels of productivity in the short with those of countries in the Organisation run? And how can economies develop the for Economic Co-operation and Development technological capacity they need to increase (OECD), still operates in the lower-quality productivity in the medium run, which has and value-adding ends of manufacturing.10 shown to be so important in determining the While China is closing the technology gap in higher productivity levels of upper-income virtually all fi elds of manufacturing—and in East Asia? Skills, which enhance capacity to some areas is approaching the point where it apply, adapt, and create new technology, and can become an innovator—its domestic capa- research, which enhances capacity to develop bility still lags behind that of leading indus- new technology, will be two key drivers. And trial nations,11 leaving substantial room for higher education can supply both. catch-up in some areas. Some countries in Southeast Asia (four Role and impact of higher are in the middle cluster) began taking a seri- education ous interest in innovation as a new growth driver after the crisis of 1997–98, because of Skills are positively related to innovation and lost growth momentum and declining private productivity (and so growth), as discussed investment. Innovation captured the imagi- below. This is one reason why in low- and nation of policy makers in Malaysia, and to middle-income East Asia—whether play- a lesser degree in Thailand and even China ing technological catch-up or moving from (which did not experience a comparable catch-up to creation (both parts of a broad growth slowdown). Countries clearly show defi nition of innovation in this report)—the differences, though, because while firms importance of higher education as a source in China are beginning to experiment with of scientifi c, technical, and analytical skills real innovation, fi rms in Indonesia and the is increasing. A well-trained and highly edu- Philippines see it as a more distant prospect. cated workforce underpins growth: skilled Most middle-income countries’ biggest labor can deploy fl exibly, achieve high lev- gains at this stage are from applying, assim- els of productivity, apply existing technolo- ilating, and adapting new technology. Busi- gies, and engage in innovation as a means nesses have rationally shown little interest to increase a nation’s competitiveness and in developing new technologies when there growth. At the same time, East Asian mar- is so much low-hanging fruit to be picked. kets are absorbing a larger share of exports For most companies, the demand for inno- from the region itself; thus, adapting technol- vation is the same as learning technologies ogies while customizing products, processes, already developed (in country or outside) and design will be of greater signifi cance and and adding to production and technological will call for the kind of support higher edu- capabilities. cation can provide.12 The lower-income and low technology As economies move up the technology lad- cluster includes the latecomers, countries that der and the gap between some of them and began industrializing in the 1990s and that the leading industrial economies narrows, have not progressed much beyond light man- their need for education and skills at all lev- ufacturing and simple assembly operations. els grows, particularly at the tertiary level.13 They are likely to remain in the learning Thus a range of tertiary institutions take on mode for some time. The primary challenge larger responsibilities—and they can help HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 13

accelerate industrial change—because they Whether tertiary education is a more sig- are the source of an increasing share of entre- nifi cant predictor of growth than primary preneurs, managers, and skilled workers. education in developing countries continues Additionally, a few research universities begin to be debated.14 There is, however, some contributing to innovation through basic consensus that tertiary education is posi- research that generates ideas or upstream tively associated with economic growth and applied research and technology transfer (or GDP regardless of a country’s development both) that initiates the process of transmut- level. Very crudely, this relationship is appar- ing knowledge and ideas into applications ent from fi gure 1.4, which shows a positive with potential commercial relevance. correlation between GDP and enrollments Higher education provides several noneco- in higher education. While causality cannot nomic benefi ts, such as nation building and be established (given the many other factors socialization (box 1.1). But this report focuses that matter, including the composition and on the contribution of higher education to the quality of higher education itself), no coun- productivity that allows economies to remain try or region has achieved, in the long term, competitive in a global market. high-income status without first crossing

BOX 1.1 Private and public benefi ts of higher education

As a proponent of liberal education, the Oxford Higher education also generates economic ben- scholar Cardinal Newman advocated for higher efi ts to society. Countries with a large labor force of education as a place for cultivating universal knowl- individuals with higher education have higher pro- edge rather than developing vocational training and ductivity and higher tax payments. This also lowers research. He believed students should study the clas- dependence on public welfare programs. The United sics and philosophy because these courses had the States spends US$800–2,700 less per year on social ability to “strengthen, refi ne, and enrich the intellec- programs for graduates of higher education than for tual powers.” With this “habit of mind,” individuals high school graduates. All these factors contribute to would be equipped to think clearly and articulate a country’s economic growth. their thoughts in any profession.a While Cardinal In addition to economic benefi ts, higher educa- Newman’s view is an important mission of higher tion provides several social benefi ts. Higher educa- education, the benefi ts of higher education extend tion institutions are good settings for individuals to from individuals to society and from economic to socialize with peers. Individuals with higher educa- social benefi ts. tion tend to have higher standards of living and bet- There is much evidence of the economic benefi ts ter well-being. They also tend to be in better health, of investing in higher education. Individuals who are less likely to smoke, and are less likely to engage attend higher education have higher average earn- in criminal activities. The benefi ts of higher education ings, are more likely to be employed, and are less also extend across generations: children of parents likely to experience poverty than individuals with- with higher education are more exposed to read- out higher education. Data on lifetime earnings of ing, have higher cognitive skills, and are better able U.S. workers by education level show that relative to to concentrate. Finally, higher education promotes high school graduates, individuals with some college nation building, because citizens with higher educa- education have on average 17 percent higher earn- tion are more likely to vote, to donate blood, and ings, and those with a professional degree have more to participate in community service. These benefi ts than three times the earnings. Thus, individuals who show how higher education can enhance the quality invest in higher education can expect their future of life for individuals and countries. economic returns to exceed the costs of tuition, fees, Source: Baum and Payea 2005. and forgone earnings from not working. a. Newman 1976. 14 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 1.4 Tertiary enrollment and per capita GDP, 2008

100 Korea, Rep. Finland

United States 80 New Zealand Australia Denmark Iceland Spain OECD Sweden Italy Belgium 60 Netherlands Japan United Kingdom Ireland Europe and Central Asia Hong Kong SAR, China Austria France Mongolia Switzerland Thailand Upper-middle-income countries 40

tertiary (% gross) school enrollment Philippines Middle East and North Africa Middle-income countries China 20 Indonesia Lao PDR Cambodia Sub-Saharan Africa Low-income countries 0 0 10203040506070 GDP per capita (current US$, thousands)

Source: EdStats database.

FIGURE 1.5 Tertiary enrollment and labor productivity, 2005 a “respectable” higher education threshold. And ﬁ gure 1.5 reports a positive rela- 100 48 tion between labor productivity and higher education. The relation between higher education 80 36 and skills is well documented. Results of seven participating countries of the Adult 60 Literacy and Life Skills Survey undertaken 24 by the OECD show a strong positive cor- 40 relation between cognitive skills (generally

1990 US$, thousands) considered a combination of academic and 12 20 output per worker (constant thinking skills—see box 1.2) and educa-

tertiary school enrollment (% gross) tional attainment, with individuals having more years of higher education consistently 0 0 scoring at higher skill levels than indi- China Japan viduals having fewer or no years of higher Indonesia Thailand Malaysia CambodiaPhilippines Korea, Rep. education (figure 1.6). This finding is tertiary gross enrollment ratio (left axis) consistent with those of past studies that output per worker (right axis) educational attainment is a leading contributor of cognitive skills such as literacy and 15 Sources: International Labour Organization data; WDI database. numeracy. HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 15

BOX 1.2 Defi ning skills

At least three main categories of skills can be and typically include a combination of academic identifi ed: academic skills, generic (or life) skills, and thinking skills. and technical skills. Academic skills are directly These three skill categories can also be divided measured and are generally associated with sub- internally: there are both threshold-level academic, ject areas (math, literacy, English). They are typi- generic, and technical skills, and higher-order aca- cally taught in schools and measured through demic, generic, and technical skills. The level and standardized tests. Generic (or life) skills refer to combination of required skills vary by fi rm and job a broader set of skills transferable across jobs and function. The nature of the skill acquisition process from education to employment. They generally implies that a comprehensive approach to skill devel- include thinking (critical and creative thinking, opment is needed. Formal education is one criti- and problem solving), behavioral (communica- cal actor. High-quality and relevant school-based tion, organization, teamwork, time management, formal education has a strong role in providing all ability to negotiate confl ict and manage risks, and three types of skills discussed. Primary education leadership), and computing skills. Technical skills systems can provide basic academic and generic are those associated with one’s profession. They skills. Secondary education systems can provide are generally considered a mix of specifi c knowl- more advanced academic and generic skills, as well edge and skills to perform specifi c jobs. Cognitive as some technical skills. And tertiary education sys- skills are often used as a further categorization tems can provide all three types, of a higher order.

In turn, skills are positively related to Beyond supporting the application of cur- productivity, labor market outcomes, and rent technology, higher education can foster growth. Earlier research on the contribu- innovation (in this report defi ned broadly as tion of human capital to economic growth the ability of economies to both create and using regression analysis yielded equivocal adapt new knowledge) through the skills results on the relation between education that it provides. Those individuals with more and growth. More recent analysis that fac- education of better quality have a higher tors in the quality of education and moves probability of starting a technology-intensive beyond the criticized measures of educa- business, hiring skilled workers, and engag- tional achievement (which largely relied on ing in innovation.20 This is also illustrated years of schooling) has been, however, more by the positive relation at the firm level successful in establishing a relationship run- between technological innovation, defi ned ning from the level of education to the level in a broad sense,21 and ratio of workers with of output and from the quality of education higher education found in a sample of fi rm to growth of GDP.16 Along these lines, recent surveys of the region (fi gure 1.7).22 The cor- evidence confi rms that there is a direct rela- relation would be stronger if one were able tion between cognitive skills, on the one to control for the quality of higher education hand, and productivity and growth, on the graduates, considering their skill sets. Unfor- other.17 An emerging literature has also dem- tunately, that cannot be done. Nevertheless, onstrated the link between cognitive ability it is known that there is a positive relation and labor market outcomes.18 Beyond cog- between professionals with management nitive skills, noncognitive skills (intended skills, innovation, and productivity,23 or that mostly as behavioral skills) are increasingly a higher share of graduates with science, shown to have a positive impact on labor technology, engineering, and math (STEM) market outcomes.19 skills is generally correlated with higher 16 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 1.6 Educational attainment and skills profi ciency this group has attained high innovation and income status without making considerable a. Mean numeracy scores levels strategic investments in higher education. 500 Korea, for example, has invested aggres- 450 5 sively in government research institutes and 400 university-affiliated science parks, as well 350 4 as pursued policies leading to high enroll- 300 3 ment in higher education (particularly in 250 2 science and engineering) since the 1970s.25

scale scores 200 Similarly, Singapore has devoted much capi- 150 1 tal to developing world-class scientifi c and 100 technological capabilities through higher 50 education.26 Hong Kong SAR, China, has 0 12341 2341 2341 2341 2341 234 pursued an agenda to develop itself as a hub Switzerland Norway Canada United Bermuda Italy for higher education, pouring resources into States improving quality, developing staff, fostering b. Mean problem-solving scores levels 500 greater links between academia and indus- 450 4 try, and incentivizing research. And Taiwan, 400 China, has attributed investment in science, 350 3 engineering, and higher education as a key 300 27 2 driver of its economic growth. Long-term 250 trends in R&D undertaken by higher edu- 200 scale scores cation compared with long-term trends in 150 1 some innovation outcome indicators, while 100 not perfectly aligned, point to a positive cor- 50 relation between higher education and inno- 0 12341234123412341234 vation outcomes in upper-income East Asia NorwaySwitzerland Canada Bermuda Italy (and other advanced economies) (figures 5th percentile 1.9–1.12). 25th percentile The relation between higher education and mean and 0.95 confidence interval for mean 75th percentile innovation is also widely illustrated in several 5th percentile studies that show the positive relationship between economic growth and higher levels Source: OECD 2005. of education as measured by variables such as Note: The numbers on the x axis denote: 1 = Less than upper secondary; 2 = Upper secondary; 3 = Post-secondary, non-tertiary; 4 = Tertiary type B or higher. tertiary gross enrollment ratio (GER), science test scores, R&D, and number of scientists and engineers per capita.28 innovation outcomes (fi gure 1.8). Firm inno- This evidence suggests that low- and vation surveys undertaken for this report middle-income economies will need to start in Indonesia, the Philippines, and Thailand prioritizing higher education if they want to showed that the active innovators are those grow. As a matter of example, while cover- with higher levels of R&D expenditures, age of higher education has been rising fairly more highly qualifi ed staff, and located in steadily since the 1970s in most economies more R&D-intensive industries.24 (fi gure 1.13), the GER still remains below par The top technology cluster in East Asia can in developing East Asia (fi gure 1.14). Korea also readily illustrate the relation between possesses one of the highest tertiary GERs higher education, innovation, and growth in the world: almost 100 percent of the adult through both skills and research. High- school-age population is enrolled. Japan’s income economies in East Asia are among tertiary GER has reached 60 percent. The the world’s most innovative. No member of tertiary GERs of Hong Kong SAR, China; HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 17

Japan; and Korea are on par with (or even FIGURE 1.7 Share of tertiary-educated workers in technologically higher than) those of several high-income and nontechnologically innovative fi rms countries in Europe and North America. But the relatively poor performance of most low- 60 and middle-income East Asian economies Indonesia has kept the developing-Asia region’s average 50 tertiary GER signiﬁ cantly below the OECD Korea, Rep. average, at a paltry 20 percent. 40 Quantity, however, is not necessarily a Philippines goal in itself. Indeed, as seen above, similar 30 GDP and innovation outcomes are related Thailand to very different quantity levels. And as China 20 shown in ﬁ gure 1.6, the variation in skills Malaysia share of tertiaryshare in workers and attainment indicates that other fac- Vietnam

technologically innovative firms (%) innovative technologically 10 Cambodia tors may inﬂ uence skill proﬁ ciency (besides attainment).29 What matters is the capacity 0 of higher education to provide labor market 0 102030405060 skills and research for productivity and inno- share of tertiary workers in vation. On both counts, lower- and middle- nontechnologically innovative firms (%) income countries have a long way to go, as will be seen. Source: Almeida 2009b, on the basis of World Bank Investment Climate Surveys (ICSs), various years.

From higher education to growth: Skills and research FIGURE 1.8 Correlation between STEM supply and patents, 2004–09 As a provider of high-quality skills relevant to current and future labor market needs, 45 effective higher education systems improve Hong Kong SAR, China 40 human capital formation and allow entrepreneurs, managers, and skilled workers to 35 CHE AUT 30 perform well, thus supporting technological FIN mastery, productivity, and competitiveness. 25 DNK SWE These systems also help develop countries’ 20 Korea, Rep. AUS DEU technological capability by undertaking 15 GBR (per million people) BEL NZL FRA research, supporting technology transfer, 10 ITA NLD patents granted by USPTO, 2009 and providing workers with (and upgrading) 5 NOR Philippines BRA ARG Indonesia ESP the skills for innovation. 0 RUS Skills and research support longer-run 0 5 10 15 20CHL 25MEX 30 35 40 science and engineering graduates, latest year, 2004–09 productivity and competitiveness by reorient- ing, upgrading, and diversifying national eco- Sources: U.S. Patent and Trademark Offi ce data; WDI database. nomic structures. Cross-country regressions have shown that a college degree is associated with higher individual earnings, higher productivity, and higher wages.30 One com- In pursuing these twin tracks, countries mentator has written, “Tertiary education need to consider opportunities of access helps countries build globally competitive to and inclusiveness of higher education. economies by developing a skilled, produc- Whatever higher education coverage targets tive, and flexible labor force and by creat- they choose, they must be able to draw from ing, applying and spreading new ideas and the widest talent pool possible, ensuring technologies.”31 that the most able students are not excluded FIGURE 1.9 Trends in higher education R&D, 1996–2007

10 higher education R&D expenditure (% of GDP)

0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Finland Singapore United Kingdom Japan Korea, Rep.

Hong Kong SAR, China United States Italy Brazil China

Thailand Malaysia India Mongolia Philippines

Sources: UNESCO Institute for Statistics (UIS) Data Centre; WDI database.

FIGURE 1.10 Trends in patents, 1996–2007

10 patents granted by USPTO (per million people)

0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

United States Hong Kong SAR, China Finland Japan Korea, Rep.

Singapore United Kingdom Italy Malaysia China

Brazil Thailand India Philippines Indonesia Vietnam

Source: U.S. Patent and Trademark Offi ce data. 18 FIGURE 1.11 Trends in journal articles, 1995–2007

1,000

800

600

400

200 scientiﬁc and technical journal articles (per million people)

0 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 Finland Singapore United Kingdom United States Italy Japan Korea, Rep. Hong Kong SAR, China Brazil China Malaysia Thailand India Mongolia Vietnam Philippines Lao PDR Cambodia Indonesia

Source: WDI database.

FIGURE 1.12 Trends in technology licensing, 1975–2007

350

300

250

200

150

100 royalty and license fees, receipts

per million people (BoP, current US$ millions) 50

0 1975 1979 1983 1987 1991 1995 1999 2003 2007 Finland United States Singapore United Kingdom Japan Korea, Rep. Italy Malaysia Hong Kong SAR, China Brazil Thailand China India Indonesia Mongolia Philippines Cambodia

Source: WDI database. 19 FIGURE 1.13 Tertiary GERs in East Asia and some comparator economies, 1970–2010

a. Low-income economies 60

20 gross enrollment ratio (%)

1970 1975 1980 1985 1990 1995 2000 2005 2010 Lao PDR Vietnam Ghana Bangladesh Cambodia Nepal

b. Middle-income economies 60

20 gross enrollment ratio (%)

1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 Lao PDR Vietnam Ghana Bangladesh Cambodia Nepal Chile Mongolia Thailand Malaysia Brazil Philippines Mexico China Indonesia India Pakistan

c. High-income economies 100

40 gross enrollment ratio (%)

1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 Lao PDR Vietnam Ghana Bangladesh Cambodia Nepal Korea, Rep. Finland United States Japan United Kingdom Hong Kong SAR, China

Source: EdStats database.

20 HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 21

from higher education because of inequities FIGURE 1.14 Tertiary GERs in East Asia and OECD average, latest of access or long-term disadvantage. This is available year not only an ethical choice but also an economic one. OECD Countries will not fi nd a one-size-fi ts-all East Asia and Pacific (developing only) target for access. For example, to support a Japan more focused innovation agenda, they may Hong Kong SAR, China assume a strong focus on developing centers Korea, Rep. of excellence and very highly skilled gradu- Malaysia ates, a strategy that may contrast with the Thailand idea of more widely available higher educa- China tion. Others looking to improve their higher Indonesia education system’s labor market relevance Philippines may make investments and assign priori- Mongolia ties in ways that are more compatible with Vietnam broader access. Still, the critical economic Lao PDR Cambodia criteria guiding the system’s expansion are 0 20 40 60 80 100 the current and future labor market absorp- gross enrollment ratio (%) tion capacity, the need for a critical mass to support innovation, and budget constraints. Sources: EdStats database; UIS Data Centre. Note: The East Asia and the Pacifi c average also includes the Democratic People’s Republic of The next two sections delve deeper into the Korea, Papua New Guinea, and Timor-Leste. roles of skills and research.

Higher education as a provider of skills they can deteriorate if the possibilities for lifelong learning are not well developed. As suggested by the deﬁ nitions of skill cate- Actors outside the formal education sys- gories (box 1.2), skills are produced in many tem also have a large role. Quality nonformal different ways, dynamically, and through education and training can provide academic, multiple actors. Preemployment education generic, and technical skills to out-of-school and training, on-the-job training, work and populations and can complement formal edu- life experience, and peer learning all con- cation with additional generic or technical tribute to skill development that could be skill instruction. It can also provide opportu- useful on the job.32 Whereas academic skills nities to update academic and technical skills are normally acquired through formal and over time, particularly for the non-school-age nonformal educational institutions, generic population. Firm and on-the-job training can or life skills are acquired in various ways. complement both formal and nonformal edu- Early-childhood parental education, spe- cation and training by providing additional ciﬁ cally targeted curricular and pedagogical job-relevant technical and generic skills. This approaches, on-the-job training, and work type of training can also provide opportuni- experience all develop and enrich these ties to maintain the existing generic and tech- types of skills. Technical skills are generally nical skills of workers. provided through targeted upper-secondary Within this broad skill framework, higher and tertiary training programs, on-the-job education plays a crucial role—perhaps no training, and learning-by-doing. more clearly than in skill provision. Higher Skill acquisition is thus a cumulative and education institutions provide the basis for dynamic process that occurs throughout the the range of skills needed for both mature life cycle. It starts at birth with parental edu- and developing economies. Tertiary gradu- cation and continues through the course of ates enter the workforce with cognitive, school education, training, and experience. technical, social, and behavioral skills And just as these skills can grow over time, honed at university that allow them to 22 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

bring advanced knowledge to bear on com- systems that shape them. Policy makers in plex problems, use that knowledge to work East Asia are reexamining how higher educa- toward their solution, perform research, tion systems should prepare graduates to take and develop ideas of more productive ways their places in the labor force. They are also of performing. It is during higher education asking how graduates should be equipped to that more mature students have the capacity, deal with changing labor force structures and ability, and time to learn sophisticated client demands from employers in ways that can orientation, communication, problem solv- meet both the current and future needs of the ing, and creativity skills, not only through economy. close links with particular careers (for exam- A better perspective of higher education’s ple, business and communication), but also potential for delivering skills requires know- across careers through the use of well-crafted ing the skills—particularly tertiary—that teaching-learning methodologies. While low- and middle-income East Asian labor many practical skills will be acquired on markets need, as seen in demand both for the job, higher education also offers a criti- tertiary graduates and for specifi c functional cal opportunity to its students to apply aca- skills. demic skills to more concrete and practical cases through case studies and other meth- Trends in demand for tertiary graduates odologies, with wider breadth than more In the long term, demand for tertiary gradu- specifi c on-the-job training would provide. ates has been generally on the rise in the Other research indicates that higher educa- region, as seen in the steep increase in wage tion instructors can teach students relevant premiums for completing a tertiary edu- technical and behavioral skills that they will cation in Cambodia, China, Mongolia, need to know and use as industrial actors, and Vietnam, as well as gradually increas- without actually doing industrial research ing ratios of tertiary education workers in themselves.33 Cambodia, China,36 and Vietnam (figure These points refl ect in some part a chang- 1.15). In Mongolia, the decreasing trend of a ing concept of the role of higher education. tertiary-educated workforce clearly indicates As technological structures and the nature quantity gaps. Alongside sharper increases in of industry evolve, academic qualifi cations the number of workers with a tertiary educa- are increasingly taken as indicators of a par- tion, the slightly increasing or fl at premiums in ticular level of academic competence and the Philippines, Indonesia, and Thailand indi- of the skills to deal with the demands of a cate sustained demand for such graduates.37 fast-changing work environment. Employ- Demand is particularly dynamic in the ser- ers expect tertiary graduates to possess the vice sector. Trends in tertiary education pre- academic, generic, and technical skills to miums and the tertiary-educated workforce increase their productivity and growth.34 by sector reported in appendixes C and D38 Increasingly, employers also expect a smaller show that demand for tertiary graduates has group of workers to possess the ability to been generally stronger in services, though think, to be creative, and to have the capac- often sluggish in manufacturing (but with ity to spur innovation. This is consistent differences across countries). The evolution of with emerging research on academic knowl- tertiary education premiums has been rather edge transfer, which has found that skilled sector specifi c in Indonesia, the Philippines, graduates bring to industry attitudes and and Thailand, with generally decreasing abilities for acquiring knowledge and using returns in agriculture, mixed performance in it in novel ways.35 manufacturing (decreasing in Indonesia and Keen to bolster their productivity, East the Philippines, fl at in Thailand), and increas- Asian economies are giving new consideration ing in services (at least in Indonesia39 and the to the knowledge and skills of their workers, Philippines). The evolution of education pre- and consequently the education and training miums has been broad-based in Cambodia, FIGURE 1.15 Trends in wage education premiums and educated workforce in selected East Asian economies

a. Cambodia b. Vietnam 100 60 50 80 40

60 30 20 40 10 0 skilled workforce (%) skilled workforce 20 (%) skilled workforce –10

skill and proportion premium of 0 skill and proportion premium of –20 1997 2003–04 2007 1992 1998 2004 2006 year year

c. Mongolia a 90 d. China 70 80 70 60 60 50 50 40 40 30 30

skilled workforce (%) skilled workforce (%) skilled workforce 20 20

skill and proportion premium of 10 skill and proportion premium of 10 0 0 1998 2002 2007 1999 2005 year year

e. Philippines f. Indonesia 80 100 70 80 60

50 60 40 30 40 20 skilled workforce (%) skilled workforce (%) skilled workforce 20 10 skill and proportion premium of skill and proportion premium of 0 0 1988 1991 1994 1997 2001 2004 2006 1994 1999 2001 2003 2005 2007 year year

g. Thailand 140

120

100 skilled 1 skill premium 1 80 skilled 2 skill premium 2 60 skilled 3 skill premium 3 skilled workers 40

skill and proportion premium of 20

0 1990 1994 2000 2004 year

Source: di Gropello and Sakellariou 2010, on the basis of labor and household surveys (various years). Note: For Cambodia, skilled 1 = workforce with at least lower-secondary education; skilled 2 = workforce with at least upper-secondary education; skilled 3 = workforce with tertiary education and above; skill premium 1 = wage premium for workforce with at least lower-secondary education compared to workforce with less education; skill premium 2 = wage premium for workforce with at least upper-secondary education compared to workforce with less education; skill premium 3 = wage premium for workforce with at least tertiary education compared with workforce with less education. For other countries, skilled 1 = workforce with at least upper-secondary education (secondary education for the Philippines); skilled 2 = workforce with at least tertiary education; and skill premium 1 = wage premium for workforce with at least upper-secondary education (secondary education for the Philippines) compared to workforce with less education; skill premium 2 = wage premium for workforce with at least tertiary education compared to workforce with less education. a. One needs to interpret the trends carefully in the Chinese case, given the much shorter time covered by the data and the less updated information.

23 24 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

China, Mongolia, and Vietnam (with ser- their overseas subsidiaries various manage- vices showing an edge, particularly if com- rial, organizational, and technical innova- pared with the upward trends in educated tions that would not otherwise be diffused to workers in this sector). Apart from Mongo- the host country, and (b) that highly skilled lia, the share of tertiary-educated workers labor is needed to adapt and further diffuse has increased in all sectors across countries, these innovations. This second point is partic- with a generally faster increase in services. ularly the case in Vietnam (table 1.7), where Tertiary education premiums have generally foreign firms employ more highly skilled increased the most in subsectors such as busi- graduates and play a larger role in technology ness, finance, and information technology development. (IT) services, transport and telecommunica- The results also underscore the broad tion, and trade (appendix D). and well-documented positive relationship Given services’ signifi cant employment and between technological development, also GDP share, these trends indicate that tertiary including technological assimilation and graduates need to possess the skills demanded adaptation, and tertiary education (this by the sector. To the extent that services’ relation goes both ways, from the tertiary- GDP and employment shares grow—aligning educated workforce to technological inno- more with the East Asian high-income (see vation, and vice versa), which is clearer in fi gure 1.2) and OECD economic structure— middle-income countries (likely explained by and the trend continues toward more skill-in- their somewhat more developed innovation tensive subsectors,40 services will increasingly systems). By contrast, the net effect of exports drive demand for graduates with a tertiary on the employment of tertiary graduates heav- education. It will be important to ensure the ily depends on the relative importance of the delivery of service-related careers in business, pressures of international competition and fi nance, transport and telecommunications, the drive to specialize and within specializa- or even only some generic social science tion on the importance of low-skill- intensity careers (at the university and college levels) products relative to high-skill-intensity ones. that support workers’ mobility and fl exibility In China and Vietnam, the net effect is nega- among service-related jobs. tive (fewer tertiary graduates employed in Tertiary graduates also need to meet the export sectors), but in Cambodia, Indonesia, needs of technologically intensive and open and Thailand, the effect is clearly positive manufacturing: within the sector, foreign- (more tertiary graduates employed in export owned enterprises, more technologically sectors). The pressure exerted by Chinese intensive firms, and to some extent, more exports largely accounts for the negative export-oriented fi rms employ a greater num- overall outcomes in low- and middle-income ber of tertiary graduates. Beyond sector com- countries in table 1.7. position effects, fi rm surveys41 offer evidence Manufacturing demand for tertiary grad- of a robust positive correlation among for- uates depends on trends in openness indi- eign direct investment, measures of techno- cators, among other factors. To the extent logical adaptation and innovation, and share that globalization continues and stimu- of tertiary-educated workers at the fi rm level, lates higher foreign direct investment45 and confi rming the presence of skill-biased tech- import penetration—in turn related to nical change; the role of exports in driving higher access to technology and new work- demand for tertiary-educated workers is less ing practices—demand for tertiary gradu- clear-cut (fi gure 1.16).42 ates in manufacturing may increase (despite More sophisticated cross-section regres- the sector’s slower value-added upgrading sion analysis (appendix E) confirms these in most of the region).46 results.43 Although this type of analysis can The association among foreign direct imply only association,44 it suggests two ini- investment, technology, and tertiary edu- tial points: (a) that foreign investors bring to cation in East Asian manufacturing is also HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 25

important when looking forward, particu- FIGURE 1.16 Share of tertiary-educated workers by foreign larly because it has the potential to shape ownership and export status of fi rms future growth and development patterns.47 In a context in which low-income countries a. Foreign/nonforeign ﬁrms 60 need to break into manufacturing and engen- Indonesia der a technological capability, and middle- 50 income countries need to move further up the manufacturing value chain (while continuing 40 to develop their technological capability), Thailand governments must achieve an alignment of 30 foreign direct investment, technology, and China Philippines higher-level skills. 20 Vietnam foreign-owned ﬁrms (%)

Doctoral graduates may therefore be share of tertiary workers in Malaysia important, because they are key contributors 10 in conducting research, fostering innovation, Cambodia and sharing knowledge. Furthermore, the 0 0 102030405060 experience of the leading East Asian econo- share of tertiary workers in nonforeign-owned firms (%) mies suggests that developing indigenous b. Exporting/nonexporting firms technological capabilities requires a steady 60 increase in the stock of scientists and engineers who help in assimilating and adapting Indonesia 50 foreign technology. The experience of Japan, Korea, Rep. Korea, and Taiwan, China, suggests that if an economy is to rapidly assimilate technology, 40 one-third or more of its university graduates 30 must have studied science and engineering. Thailand The positive relation between STEM skills Philippines and innovation has been illustrated above. exporting firms (%) 20 Malaysia

share of tertiary workers in Cambodia China Trends in demand for functional skills 10 Vietnam Recent employer and employee surveys provide a benchmark of the generic and tech- 0 0 102030405060 nical skills the region needs (box 1.2). To share of tertiary workers in nonexporting ﬁrms (%) the extent that these apply to professionals and managers (expected to be educated at Source: Almeida 2009b, based on World Bank ICSs, various years.

TABLE 1.7 Regression coeffi cients of technological and openness variables in a sample of East Asian economies

Korea, Low Middle Middle Variable Cambodia Vietnam Philippines Indonesia China Thailand Malaysia Rep. income incomea incomeb Foreign ownership 0.628 0.400*** −0.231 −0.118 0.191*** 0.0121 0.131 0.419*** 0.459*** 0.0721 0.0785 Technology innovation −0.161 0.0549 0.497*** 0.247* 0.342*** 0.134** 0.00724 0.0509 0.015 0.297*** 0.280*** Exports 1.531*** −0.0503 −0.0276 0.388** −0.406*** 0.142** 0.336 0.015 0.002 −0.115** −0.0958**

Source: Appendix E. Note: Dependent variable is the share of workers with more than 12 years of schooling. a. All East Asian economies except low-income countries, Malaysia, and Republic of Korea. b. All East Asian economies except low-income countries and Republic of Korea. Signifi cance level: * = 10 percent; ** = 5 percent; *** = 1 percent. 26 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

the tertiary level), these surveys reveal the countries but are becoming a priority in coun- skills expected from tertiary graduates.48 A tries such as Cambodia, Mongolia, Thailand, share of skilled production and nonproduc- and Vietnam. While there are no obvious tion workers is also educated at the tertiary differences across countries, creativity and education level (particularly college educa- IT skills tend to have a higher relative impor- tion and technical and vocational education tance within upper-middle-income countries. and training), making these other categories A further analysis of data for Indonesia and of workers also relevant. (Charts by coun- the Philippines shows that skill demands try derived from employer and employee are lower overall for skilled production and surveys—Investment Climate Surveys, and nonproduction workers—and that most of fi rm skill surveys—are included in appen- the relative priorities in terms of skills are dix F.) A short summary is provided in maintained for the Philippines, while chang- table 1.8, which ranks (from 0 to 7) the ing more substantially in Indonesia. Overall, relative importance of each skill within each basic academic skills, practical knowledge of country, largely for professionals. Differ- the job, IT, teamwork, and ability to work ences between professionals and other rel- independently become particularly relevant, evant workers’ categories are alluded to in a whereas English, leadership, and creativity couple of countries. decrease in importance. According to the surveys and as illustrated The relative importance of generic skills in table 1.8, employers and employees in East varies by sector, trade orientation, and for- Asia are giving particular emphasis to some eign ownership. Behavioral skills appear to job-specifi c skills and several thinking and be more important in services. (Further dis- behavioral skills, refl ecting a change in skill aggregation of skill demand49 in other studies demand. Job-specifi c technical skills—both shows that behavioral skills, such as ability theoretical and practical—are important in to work independently, initiative and leader- most countries, with an edge for experience ship skills, communication skills, and team- and practical job-related knowledge. Think- work skills, tend to have higher importance ing and behavioral skills, and to a lesser extent in services than in manufacturing in Indone- IT skills, are also important. Problem solving sia, Malaysia, the Philippines, and Vietnam, and creativity receive signifi cant emphasis in particularly communication skills.) Several most countries, as do communication and generic skills are, however, also critical in leadership skills. English skills are subject to manufacturing. Problem-solving skills and signifi cant fl uctuations in importance across creativity (at least for a few workers in the

TABLE 1.8 Importance of technical, thinking, and behavioral skills for professionals

Skill Vietnama Cambodia Indonesia Malaysiab Philippines Thailandb Mongolia Average Technical 7 4 5 7 7 5 5 5.7 Communication 6 5 7 5 5 4 4 5.1 English 5 5 3 4 3 7 7 4.9 Problem solving — 7 5 4 6 4 3 4.8 Leadership — 6 4 4 6 4 4 4.7 Information technology — 3 4 6 3 6 6 4.7 Creativity — — 6 5 4 4 4 4.6 Work attitude 7 5 6 4 4 3 3 4.6

Source: Appendix F. Note: Ranking from 0 to 7 of the relative importance of each skill for employers. a. Relates to college graduates. b. Relates to professionals and other skilled workers. — = not available. HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 27

case of creativity) appear to be much more FIGURE 1.17 Importance of generic skills by sector and export important in the exporting than nonexport- orientation, Indonesia, 2008 ing sector in Indonesia (fi gure 1.17) and the a. Skills by sector Philippines, as well as negotiation and leader- 80 ship skills for professionals and managers in 70 both countries, highlighting the pressure of 60 international competition. Foreign-language 50 and communication skills are also particularly important in Vietnamese foreign-owned 40 enterprises. 30 Leading fi rms that attach importance to 20 skills “very important” (%) innovation look for several attributes in their 10 new hires. They value industry experience share of firms rating workers' behavioral 0 and a “big picture mindset,” and, in a small manager/ skilled workermanager/ skilled worker professional professional but important group of workers, creativity and the ability to “think outside the box.”50 manufacturers non-education services Similar opinions were reflected in the firm behavioral skills (general) ability to work independently risk taking and initiative surveys undertaken in Indonesia, the Philip- communication negotiation team orientation b. Skills by export orientation pines, and Thailand (box 1.3). 100 Across sectors, practical knowledge also 90 appears crucial, particularly so for low- and 80 lower-middle-income countries. This comes 70 out clearly from fi rm surveys in Indonesia, 60 the Philippines, and Vietnam (appendix F), 50 which emphasized the practical knowledge of 40

the job (even more than theory) and general “very important” (%) 30 experience in the fi eld. 20 As a further illustration of the importance share of firms rating workers' skills 10 of practical orientation, labor force surveys 0 suggest that fi rms often prefer workers with nonexporter exporter nonexporter exporter technical and vocational education and train- manager/professional skilled worker ing (TVET) skills. High shares of such gradu- basic skills (math and literacy) thinking skills ates are, for instance, employed as profes- behavioral skills computer skills English skills sionals in Indonesia and Mongolia (the share is even higher than for university graduates Source: di Gropello, Kruse, and Tandon 2011. in Indonesia). Managers and professionals with TVET command, relative to unskilled workers, similar or even higher salary premi- the need for certain skills that only university ums (especially in Mongolia, because of its graduates can offer. lower technology demands) than managers The data in fi gure 1.18 also show that a and professionals with a university educa- signifi cant share of tertiary education gradu- tion (fi gure 1.18). In Mongolia skilled work- ates, particularly those educated at TVET ers with TVET are also paid much more than and college level, fi nd employment as skilled skilled workers with university degrees. The production and nonproduction workers— situation is different in Thailand where the and that therefore demand for skills at these much higher share of university graduates levels is also one of the benchmarks. in professional occupations may indicate a In sum, employers expect profession- combination of better university education, als and other skilled workers to have poor primary and secondary education, and some key technical and generic skills, and 28 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 1.3 A snapshot of skills for innovation in Indonesia, the Philippines, and Thailand

To understand how much Indonesia’s higher educa- job experience was mentioned and was presumably tion system is contributing to innovation at the fi rm relevant in most cases. level, the World Bank interviewed 12 Jakarta-based As to the skill levels required, all fi rms (except fi rms in October 2009. The fi rms were drawn from one nonrespondent) emphasized the importance of services and manufacturing. Service fi rms included relevant industry-specifi c technical knowledge and a providers of mining services, education, fi nancial broad understanding of the company’s general opera- services, research, and logistics; manufacturing tions. Several referred to the importance of under- fi rms included pharmaceutical, wireless technology, standing international best practice. Some respon- and palm oil–processing companies. Respondents dents stressed the importance of generic skills such were asked questions on the education levels of the as “curiosity,” “proactivity,” and “creativity”; sig- top manager and employees, R&D expenditure, nifi cantly, these remarks originated from fi rms with staff training, relationships with universities, and a general commitment to R&D and education. innovation activities (which included details of the Similar surveys were carried out in the Philip- innovation, the person who introduced it, the req- pines and Thailand. Philippine respondents empha- uisite education and skill levels, and the constraints sized the importance of a strong base of core skills encountered). Respondents were also asked to pro- in all new employees, and fi rms in IT and manufac- vide any general recommendations on innovation turing underscored the importance of job-specifi c policy as it relates to Indonesia’s higher education technical skills as well. In Thailand, surveyed man- system. agers, particularly for exporting fi rms, complained Responses varied about the importance of educa- of weak language skills (English), lack of creativity tion levels in the innovation process. Two respon- (workers seemed to do only what they were told by dents regarded PhDs as required: unsurprisingly, the their managers), and lack of teamwork skills. As a wireless technology manufacturer and one of the result, most new employees of surveyed fi rms under- pharmaceutical fi rms. The majority of the remaining went some form of in-house training. respondents thought that a master’s degree was suffi cient. Two of the fi rms reported that the requisite Sources: Doner, Intarakumnerd, and Ritchie 2010; Hill and Tandon 2010; Tan skills were obtained through outsourcing. On-the- 2010.

requirements for some of these skills are orientation, communication, and initiative even higher in services and open sectors. skills, as well as knowing a foreign language. This clearly has implications for higher edu- At the same time, technology and engineer- cation, which, beyond providing the training ing careers, as well as problem solving, some needed for some careers, will also need to command of foreign language, and IT skills, inculcate the functional skills for workplace among other things, will help manufacturing requirements.51 ﬁ rms respond to the requirements of foreign The dynamism of demand for tertiary direct investment and competitive export graduates in services emphasizes the impor- markets and therefore need to be possessed tance of tailoring curriculum design and ped- by tertiary educated workers. And there is agogical approaches of tertiary education to evidence that critical and creative thinking labor-market needs. Service-related careers and management skills are already requested in business, ﬁ nance, transport, and telecom- by employers from a group of employees to munications will remain important, as well move their technology frontier forward. as some broader social science tracks, gained Finally, possessing practical knowledge is a at both university and college levels. Skills plus in all countries, particularly in low- and more typically applicable to services include lower-middle-income countries that need to higher-order behavioral skills such as client develop higher technological capability. HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 29

FIGURE 1.18 Labor market outcome indicators of TVET and university graduates, Mongolia, Indonesia, and Thailand

a. Occupation shares by type of higher education b. Occupation premium by type of higher educationa

managers managers

professionals professionals clerical/service/sales clerical

skilled workers Mongolia Mongolia service/sales unskilled

unemployment rate skilled workers

managers managers professionals professionals clerical/service/sales clerical skilled workers Indonesia Indonesia service/sales unskilled

unemployment rate skilled workers

managers managers professionals professionals clerical/services/sales clerical

Thailand skilled workers Thailand

unskilled service/sales

unemployment rate skilled workers

01020 30 40 50 60 70 0 20406080100 percent percent higher vocational/diploma university higher vocational/diploma university

Sources: Labor force surveys, latest years available. a. Excluded category is unskilled workers.

The demand for skills will continue tackling this problem (box 1.4) in its own changing in East Asia. As countries move way. As in industrial countries, greater cli- up the value-added chain, the types of skills ent orientation and teamwork, new ways needed will continue evolving to a greater of working, and greater computer use are focus on more sophisticated technical skills likely to accelerate demand for new skills in and on the high-level generic skills that East Asia.52 Moving forward, higher educa- are increasingly driving labor productivity tion sectors will need to provide the skills (such as analysis, problem solving, and com- relevant to the labor markets of today with munication). The experience in the United a vision of tomorrow—they cannot stand States shows that as countries’ demand for still. Skills for the service sector, high- interactive and analytical skills increases, order generic skills, and skills for innovation demand for manual and routine cognitive not only are needed now but also will be in skills falls. The European Union (EU) is increasing demand moving forward. 30 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 1.4 Skills for the future

The European Union has formally recognized that further learning, and skill development. These are lifelong learning—and the skills to enable it—will communication in the mother tongue, communica- be a major determinant of the bloc’s future innovation in foreign languages, competence in math and tion, productivity, and competitiveness. In the con- basic competence in science and technology, digital text of increasing internationalization and regional competence, learning to learn, social and civic com- integration, and continuous technological upgrad- petence, a sense of initiative and entrepreneurship, ing, it is taking steps to ensure that its education and and cultural awareness and expression. training systems develop a workforce that will not The framework emphasizes competence in the only be able to keep its job-specifi c skills up to date basic skills of language, literacy, numeracy, and and relevant but also possess certain generic skills information and communication technology as the that will allow workers to better adapt to change. foundation of learning, while stressing the impor- The EU in 2006 adopted the European Framework tance of critical thinking, creativity, initiative, prob- for Key Competences for Lifelong Learning, the aim lem solving, risk assessment, and decision making. of which is to inform curricular development at all With the framework, the EU aims to develop a work- levels of education. It identifi es the key competences force better able to adapt to changing circumstances that European citizens will require for “personal and technologies, innovate, create jobs, and make its fulfi llment, social inclusion, active citizenship, and education and training systems more relevant to pro- employability in [the] knowledge-based economy.” ductivity and growth. The framework defi nes eight competences that young people should possess by the end of their formal education to equip them for their working life, Source: The Expert Group on New Skills for New Jobs 2010.

Higher education as a producer of research in 2007. Singapore had been increasing its spending on R&D, at 2.6 percent that year. The most critical growth-oriented objective By comparison, R&D in Southeast Asia of higher education is producing enough amounted to less than 1 percent of GDP. proﬁ cient and innovative graduates for the The fastest rate of growth was in China, ris- labor market, but the importance of higher ing from 0.6 percent in 1996 to 1.5 percent education in directly supporting techno- in 2007 (and 1.7 percent in 2009). Given its logical development and innovation is also rapid economic growth during this period, growing. The presence of a few research the increase in the volume of resources com- universities also becomes a key priority for mitted was huge. growth. The low R&D spending rate by low- and There is a close positive relation between middle-income countries (lower and middle R&D and innovation. Expenditure on R&D technology cluster) is also much lower than (usually as a share of GDP) is a common the OECD average (ﬁ gure 1.19), which only metric that provides a reading on a country’s Japan, Korea, and Singapore exceed. acquisition of technological capacity, and it is Mirroring the overall pattern in OECD much higher in the East Asia upper-income countries, firms account for two-thirds or innovative economies. By this yardstick, more of R&D spending in a majority of East Japan was the largest spender in East Asia in Asian economies (table 1.10). 2006 (3.4 percent of GDP), followed closely Private and public funding have a comple- by Korea with 3.2 percent (table 1.9). The mentary role in R&D. As the technological position of these two countries switched gap narrows, increasing attention to process HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 31

TABLE 1.9 R&D expenditure, 1996–2007 percent of GDP

Economy 1996 2000 2002 2004 2005 2006 2007 Vietnam — 0.1 0.2 — 0.3 — — Philippines — — 0.2 — 0.1 — — Indonesia — 0.1 — 0.1 0.1 — — China 0.6 0.9 1.1 1.2 1.3 1.4 1.5 Thailand 0.1 0.3 0.2 0.3 — 0.2 — Malaysia 0.2 0.5 0.7 0.6 — 0.6 — Korea, Rep. 2.4 2.4 2.5 2.9 3 3.2 3.5 Singapore 1.4 1.9 2.2 2.2 2.4 2.3 2.6 Japan 2.8 3 3.2 3.2 3.3 3.4 3.4

Source: WDI database. — = not available.

innovation by fi rms prepares the ground for FIGURE 1.19 R&D expenditure, East Asia and OECD competition strategies alive to the need for continuous innovation, because improved OECD countries processes can be integrated more readily East Asia and Pacific into the operations of a fi rm and because the Japan returns accrue quickly. Once process innova- Singapore tion, which is generally incremental, gath- ers momentum and its utility is widely per- Hong Kong SAR, China ceived, R&D gains stronger adherence (most Korea, Rep. importantly from management) both inside Malaysia the fi rm and outside. It also becomes better Thailand integrated into the operations of an entire China industry. Firms can and should then start Indonesia to play a larger role and support university Philippines research. Thus, encouraging fi rms to improve Mongolia existing processes and products through sys- Vietnam tematic research must be a major strand of Lao PDR government policy aimed at stoking interest in innovation. Such encouragement can Cambodia complement efforts to augment research in 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 universities and public research institutions % of GDP with the help of public fi nancing. In particular, in lower- and middle-income countries Source: WDI database (latest year, 2002–07). the public sector can have an important role in fi nancing basic early-stage applied research and technology transfer through universities upgrading achieved through technological and research institutions, when private initia- change embodied in equipment and through tive is still limited, but positive externalities interaction with overseas buyers and suppli- are clear.53 R&D spending during the begin- ers. During the rapid-growth phase of Japan, ning stage of development can be thought more than 30 percent of R&D was devoted of as part of the effort to assimilate and to learning.54 The role of public financing internalize foreign technology as well as to continues even later with the continuous con- build the foundations of a national innova- tribution of universities and research centers tion system. It supplements the technological to basic research and their potential role in 32 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

high-tech R&D, often too costly to be under- of ideas and engage in applied research (see, taken by firms. Indeed, government and among other indicators, the higher-ranked higher education research maintain a signifi - universities in table 2.7, chapter 2). But uni- cant share of overall R&D in the economies versities cannot substitute for fi rms, and sev- of the top technology cluster (table 1.10), even eral challenges exist in setting up effective more so in Hong Kong SAR, China, and Sin- university-industry links (reviewed in the gapore. And in Singapore, the role of higher rest of this report). Research by universities education was even larger 15 years ago, illus- and even the R&D by fi rms themselves will trating its importance in leading R&D effort yield meager returns if businesses have lim- and economic development.55 ited faith in the usefulness of innovating and Firms and universities have a comple- are not persuaded of the value of making mentary role. Research universities not only innovation into a routine. The type of uni- add to the fund of knowledge, but universi- versity-industry link will also change across ties and other tertiary institutions can also country groups. help countries of the lower technology clus- Economies of scope and scale provide fur- ter raise their technological capabilities and ther justifi cation for research universities. An countries of the middle technology cluster important assumption of this study is that go beyond technology assimilation to inno- there are economies of scope between teach- vation by assisting firms in assimilating ing and research, and economies of scale in and upgrading technology through provid- research,56 making support to research in at ing consulting services, hosting incubation least a few higher education teaching institu- facilities, and customizing foreign technolo- tions an effective option (and more effective gies for local requirements, among other than using only research centers). Although approaches. Because most small and medium the link between teaching and research has enterprises in the lower and middle technol- been hotly debated, results from three meta- ogy cluster carry out very little R&D (and analyses fi nd a positive correlation suggesting even large fi rms only do a modest amount that teaching and research are not mutually of applied research), universities can step in exclusive goals in education.57 A compari- to help narrow the technology gaps between son of the world’s highest-ranked universities technology clusters by enhancing technolog- reveals several similar characteristics, such as ical capabilities. conducting research, teaching innovative cur- Among the region’s high-income coun- ricula, and integrating research into under- tries, several research universities are sources graduate teaching.58

TABLE 1.10 Composition of R&D expenditure percent

Economy Business enterprises Government Higher education Private nonprofi t Vietnam 14.5 66.4 17.9 1.1 Philippines 68.0 19.1 11.1 1.8 Indonesia 14.3 81.1 4.6 0.0 China 71.1 19.7 9.2 0.0 Thailand 43.9 22.5 31.0 2.6 Malaysia 71.5 10.4 18.1 0.0 Korea, Rep. 77.3 11.6 10.0 1.2 Hong Kong SAR, China 48.3 2.2 49.5 0.0 Singapore 65.7 10.4 23.9 0.0 Japan 77.2 8.3 12.7 1.9

Source: UIS Data Centre. Note: Hong Kong SAR, China, and Malaysia (2004); the Philippines and Thailand (2003); Vietnam (2002); Indonesia (2001); other countries (2005–08). HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 33

There are certainly several arguments that in the United States and Japan provide fur- research facilitates teaching, with benefits ther evidence of product-specifi c economies to both students and professors. Students of scale. A study of more than 300 com- who are actively engaged in research acquire prehensive universities in the United States knowledge and experience in their fi elds. They fi nds product-specifi c economies of scale for gain exposure to research methodology, data conducting research in undergraduate and analysis, critical analyses, and presentation of graduate public and private universities. Sim- the fi ndings. All these are greatly needed skills ilarly, the study also fi nds product-specifi c (as seen above). Research also provides stu- economies of scope in these institutions.64 dents with credibility and concrete evidence Another study of nearly 1,900 U.S. higher of what their professors teach. In developing education institutions further illustrates that countries, professors can often relay more cur- institutions that combine undergraduate rent research to students than textbooks.59 teaching, graduate teaching, and research Indeed, other fi ndings suggest that teach- are more effi cient than single-output insti- ing receives larger beneficial effects from tutions. In this study, product-specific research than research does from teaching economies of scale for research are most and that these benefits are greater at the apparent in public graduate institutions.65 graduate than the undergraduate level.60 Finally, a study of 94 private universities There is, however, also evidence of the ben- in Japan fi nds product-specifi c economies efi ts that research receives from teaching, of scale for conducting research in large and studies suggest three factors are impor- universities.66 tant: (a) teaching provides young researchers Innovation is clearly not fully captured with opportunities to present their ideas; (b) by patents, journals, or even technology teaching and students can stimulate ideas for licensing since mere technological upgrad- new research; and (c) classrooms can act as a ing may be an even more important objec- forum for academics to clarify and close gaps tive. But a simple relation between science in their research.61 In addition, a U.S. study and technology journals and R&D under- found that teaching has a positive effect on taken at the university level on the one side, research up to eight hours of teaching per and patents and R&D undertaken at the week, indicating that previous inconclusive university on the other, indicates that there results could be because teaching and research is a positive correlation between university have a curvilinear relationship, rather than research and innovation (fi gures 1.20 and the previously accepted linear relationship.62 1.21), pointing to (at least potential) benefi - Institutions that provide teaching and cial effects of R&D carried out at the uni- research gain several benefits. Research versity level. increases an institution’s image and reputation, and thus attracts high-quality stu- Conclusion dents and faculty. Teaching and research share similar skills such as creativity, Higher education has the potential to critical thinking, and diligence. Students deliver skills and research for productiv- exposed to research are more likely to ity and innovation. While the final goals conduct research, including in areas that of higher education are the same across have few researchers. Students and faculty countries, the different conditions faced engaged in research can foster collaboration by income and technology cluster groups and learning of research. Finally, present- dictate slightly different immediate broad ing research results not only reinforces the priorities. Low-income and low-technology- research but also generates ideas for further cluster countries striving toward middle- investigation.63 income status must focus their immediate Results of several empirical studies of the attention on higher education as skill pro- costs of conducting research in universities ducers and develop their human capital 34 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 1.20 Correlation between higher education R&D and depend on their position within the tech- scientifi c and technical journals nology cluster. For instance, developing the technological and engineering capacity of 1,400 workers and building some limited research 1,200 capacity for technology upgrading may be suffi cient for countries such as Indonesia 1,000 Singapore and the Philippines at their stage of techno- 800 logical development but would not be suf- 600 fi cient for China. The bottom line is that all Japan countries need to start moving up within 400 per million people, 2007 Korea, Rep their income and cluster group and beyond, Mongolia China 200Philippines Malaysia Vietnam and higher education can help them to do scientific and technical journal articles Thailand LAO PDR 0 so. But does it? 0.0 0.2 0.4 0.6 0.8 1.0 higher education research and development expenditure, latest year 2004–07 (% of GDP) Notes Sources: UIS Data Centre; WDI database. 1. The earlier performance was driven primarily by rapid accumulation of physical capital, gains in labor and total factor productiv- FIGURE 1.21 Correlation between higher education R&D and ity, and generally favorable institutional and patents policy environments (IMF 2006). 2. The terms higher education and tertiary edu- 45 cation are used interchangeably in this book. HKG 3. Thailand very recently transited to upper- 40 middle-income status but was still a middle- 35 middle-income economy according to the AUT CHE 30 gross national income per capita Atlas method FIN applied to 2009 data. Hence, it is considered 25 CAN a middle-middle-income economy in this 20 JPN KOR book. This classifi cation is also justifi able AUS 15 GBR from the fact that all available higher educa- (per million people) FRA ISR 10 BEL tion indicators are for 2009 or earlier. patents granted by USPTO 2009 ITA 4. The World Development Report uses the 5 MYS MEX ESP ZAF BRA gross national income per capita Atlas method IDN IND THARUS ARG CHL 0 applied to 2009 data (latest available update). 0.0PHL 0.1 0.2 0.3 0.4 0.5 0 0.7 0.8 higher education research and development expenditure, latest year 5. Vietnam very recently transited to lower- 2004–07 (% of GDP) middle-income status but was still a lower- income economy according to the gross Sources: UIS Data Centre; U.S. Patent and Trademark Offi ce data; WDI database. national income per capita Atlas method applied to 2009 data. Hence, it is considered a lower-income economy in this book. This classifi cation is also justifi able from the fact base. To further enhance innovation, as a that all available higher education indicators second goal they must start building some are for 2009 or earlier. research capacity in higher education. Mid- 6. See Castellacci and Archibugi (2008) for an adaptation of the concept from the conver- dle-income countries aspiring toward high- gence club literature and an attempt at group- income status must focus urgently on both ing countries. further developing the skills of their labor 7. While the income group is the main classifi ca- force and fostering research through higher tion followed in the book, reference is often education as a research provider. How made to technology clusters when technologi- much they should aim to do in relation to cal capacity is a critical variable for the under- technology and innovation will very much standing of some issue or when technological HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 35

clusters are more closely aligned to economy nontechnologically innovative fi rms within performance. countries that are relevant. By the same 8. China’s technological prowess in bio- token, the fi rm sample of Indonesia distorts technology and nanotechnology might enable somewhat the results by including more it to cross this threshold in the next decade. fi rms close to the technology frontier. 9. Because of this technical advancement, export- 23. Yusuf 2003. ing fi rms and multinational corporations tend 24. These surveys, undertaken in 2009–10, sought to hire more educated workers than nonex- to understand the extent to which higher edu- porting or domestic fi rms (Yilmaz 2009). cation systems in these countries contribute to 10. Schott 2006. innovation at the fi rm level. The fi rms, which 11. China still lacks core technologies in most were in services and manufacturing, included advanced areas of electronics and transport providers of mining services, fi nancial services, equipment, for example. research and logistics, pharmaceuticals, and 12. Kapur and Crowley 2008. wireless technology as well as rubber proces- 13. Vandenbussche, Aghion, and Meghir (2006) sors and palm oil processors. Respondents were draw attention to the greater returns from asked questions on the education levels of the investment in skills and research as a country top manager and employees, R&D expendi- approaches the technological frontier. ture and staff training, relationships with uni- 14. Gittleman and Wolff 1993; Sianesi and Van versities, and innovation activities. The last Reenen 2003. included details of the innovation, the person 15. Boudard 2001; Kirsh and others 1993. who introduced it, requisite education and skill 16. Dowrick 2003; Permani 2009. levels, and the constraints encountered when 17. Hanushek and Wößmann 2007. introducing the innovation. Respondents were 18. In Canada, for instance, higher reading scores also asked for general recommendations on at age 15 lead to higher future wages (OECD innovation policy as it related to their national 2010b). Improving the quality of educa- higher education system. tion improves test scores in the short term 25. Mok 2010. (Jakubowski and others 2010) and labor 26. Mok 2010. market success in the medium term (Bertschy, 27. T.-C. Lin 2004. Cattaneo, and Wolter 2009). 28. Barro and Sala-i-Martin 1995; Gittleman and 19. Borghans and others (2008) and Heckman, Wolff 1993; Lederman and Maloney 2003. Stixrud, and Urzua (2006) highlight the 29. OECD 2005. importance of noncognitive and cognitive 30. Bloom, Hartley, and Rosovsky 2004; abilities on social and economic success. Macerinskiene and Vaiksnoraite 2006. 20. Entrepreneurial performance is associated 31. Salmi 2009. with the quality of formal schooling (Berry 32. Indeed, school is considered only the fourth and Glaeser 2005; Glaeser 2007; van der source of skills for college-level workers in Sluis, van Praag, and Vijverberg 2008). Vietnam, after on-the-job experience, training, 21. The technology variable refers to fi rms having and previous experience (World Bank 2008); “introduced a new technology that substan- whereas in Indonesia, school is still consid- tially changed the way the main product was ered the most important source of skills for produced in the three years prior to the sur- managers and professionals, but it comes after vey” (defi nition adopted in the World Bank’s on- the-job exposure and previous experience Investment Climate Surveys). This defi nition, for nonproduction and skilled production work- which is related to process innovation, leaves ers (di Gropello, Kruse, and Tandon 2010). room for both adaptation of an existing tech- 33. Nelson 1987. nology (developed domestically or imported 34. Computerization has increased the demand from abroad) and innovation within the fi rm. for college preparation to undertake non- 22. Almeida (2009b) gives details on method- routine cognitive tasks as machines take over ology and sample. The different samples routine tasks (Autor, Levy, and Murnane are not directly comparable, however. The 2001). The scarcity of skilled and technical Indonesia sample, for instance, covers a workers and of R&D facilities is a major rea- greater number of large and foreign-owned son for the slow pace of industrial upgrading fi rms than some other samples. It is the in Southeast Asia (Tan 2010). comparisons between technologically and 35. Senker 1995. 36 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

36. One needs to interpret the trends carefully in 41. The main data set used here is a large fi rm- the Chinese case, however, given the much level database collected by the World Bank, shorter time covered by the data and the less Investment Climate Surveys, covering eight updated status of the information. developing countries in East Asia. The sur- 37. They also indicate some emerging constraints veys were conducted in 2002–05, and the in the overall absorption capacity for new samples were designed to be representative tertiary-educated graduates. of the population of fi rms according to their 38. The regression framework for estimating edu- industry and location within each country. cation (and industry) wage premiums follows The fi nal sample has 9,776 fi rms distrib- the approach used by Goldberg and Pavcnik uted across a wide range of sectors (manu- (2005) and other researchers. Specifi cally, for facturing 77 percent, construction 1.37 per-

each year the log of worker i’s wage (ln(wijt)) cent, services 20 percent, and agro-industry ′ is regressed on worker i’s characteristics (H ijt) 0.8 percent). Manufacturing covers a range such as gender and age; on whether, based on of industries, such as auto and auto com- his or her education, the worker is skilled ponents, beverages, chemicals, electronics,

or unskilled (Sijt); and on a set of industry j food, garments, leather, metals and machin- indicators (Iijt) refl ecting worker i’s industry ery, nonmetallic and plastic materials, paper, affi liation: textiles, and wood and furniture. 42. The relationship with technological innova- ln(w ) = H′ b + S · I sp + I wp + e ijt ijt H ijt ijt jt ijt jt ijt tion was shown in fi gure 1.4.

where spjt represents the sectoral return to 43. The model, fully detailed in Almeida (2009b), education (or education premium) of sector j is related to a vast literature linking foreign

at time t, and wpjt represents the industry investment, trade, and technology with skills. premium. See, among others, Berman, Bound, and 39. In Indonesia about 65 percent of service fi rms Machin (1998); Fajnzylber and Fernandes confi rm to having seen skill requirements (2004); Feenstra and Hanson (1997); and increase over these past 10 years (compared Tybout (2000). with about 50 percent of manufacturing fi rms), 44. This limitation is despite the effort to control and 95 percent of fi rms (in both manufacturing for as many observable and unobservable and services) think skill requirements will con- variables as possible. tinue to rise over the next 10 years (di Gropello, 45. Such investment would continue the upward Kruse, and Tandon 2010). trend experienced in some countries. In 40. Increases in value added within services, Thailand, for example, foreign direct invest- though only gradual, are visible in Cambodia, ment as a share of GDP increased from only Indonesia, the Philippines, and Vietnam and 1 percent in 1990 to 5 percent in 2008. where the public administration and other Concurrently in Indonesia, foreign ownership services subsector has been generally decreas- in manufacturing rose from 22 percent to 37 ing, with corresponding increases in the more percent (WDI database various years). skill-intensive fi nance and business, trans- 46. The evidence suggests that the generally low- port and communication, and trade and value-added food, textile, timber, and fur- tourism subsectors. In these cases the public niture subsectors employ most workers in administration and other services subsector Indonesia and Vietnam, while other subsec- includes a mix of services with both higher- tors that add higher value are developing only educated workers (health, education) and slowly (mainly chemicals in Indonesia and lower-educated workers (private household machinery and equipment in Vietnam). The services, lower public administration levels, Philippines focused more on other, higher- nonformal). Further, despite the recent global value-added, manufacturing subsectors, but economic downturn, fi nance and banking the direction of change is not clear, with an subsectors in Singapore; trade and tour- increase in the employment share of other ism subsectors in Cambodia, Lao PDR, and manufacturing but a decrease of the machin- Vietnam; and the communication subsector ery and transport subsector. in the Philippines are seen as emerging or 47. As mentioned previously, one can look at quickly recovering, implying that these sec- the relation both ways: from technology and tors will need increasingly skilled workers foreign direct investment to demand for ter- (Asia Business Council 2009). tiary education, or from tertiary education to HIGHER EDUCATION FOR GROWTH THROUGH SKILLS AND RESEARCH 37

technological innovation and attraction of 52. And they are likely to create demand for new foreign capital. skills related to higher education workers, as 48. This book uses Investment Climate Surveys, illustrated by the results of alternative speci- related skill modules, and two employer and fi cations checking robustness of the tertiary employee skill surveys carried out in Indonesia education determinants to different tech- and the Philippines. This book analyzes how nological variables (see appendix E), which employers (and employees where possible) show a positive association among R&D, rate the importance of generic (thinking, use of computers, use of the Internet, and behavioral, IT), technical (job-specifi c), and in share of workers with more than 12 years of a few cases subject-based (academic) skills for schooling. doing their jobs. The analysis is carried out for 53. Although without the existence of at least Cambodia, Indonesia, Malaysia, Mongolia, some proactively innovating fi rms (or fi rms the Philippines, Thailand, and Vietnam. Some demanding innovation), increasing the supply results are not strictly comparable across of ideas alone will rarely bear any fruit. countries because of different samples’ com- 54. Mok 2010. positions and sizes, occasionally different skill 55. Mok 2010. defi nitions (which have been kept as compa- 56. Cohn, Rhine, and Santos 1989; Hashimoto rable as possible), and in some cases slightly and Cohn 1997; Koshal and Koshal 1999. different units of measure. Even so, it is pos- 57. Allen 1996; Feldman 1987; Hattie and Marsh sible to derive some main fi ndings. 1996. 49. See di Gropello, Kruse, and Tandon 2011; di 58. Salmi 2009. Gropello, Tan, and Tandon 2010; Malaysia 59. Jenkins and others 1998. Investment Climate Surveys 2002, 2007; 60. Smeby 1998. World Bank 2008. 61. Zaman 2004. 50. Wadhwa and others 2007. 62. Mitchell and Rebne 1995. 51. Learning such skills will be coordinated with 63. Zaman 2004. other education and training levels and types, 64. Koshal and Koshal 1999. since, as illustrated previously, skill acquisi- 65. Cohn, Rhine, and Santos 1989. tion is cumulative and diverse. 66. Hashimoto and Cohn 1997.

Is Higher Education Meeting Its Promises? 2

he previous chapter noted that higher of gaps in other types of skills—thinking and education has the potential to lift behavioral—which tertiary education is not Tproductivity and competitiveness. providing well enough. And the importance This chapter analyzes how much higher edu- of most of these generic skills will keep on cation is fulfi lling that potential in low- and growing as the share of the service sector middle-income East Asia, delivering to its grows and international competition intensi- students sufficient skills for growth while fi es. As the share of the service sector contin- fostering innovation through research. Dis- ues to grow and the importance of practical entangling education’s impact from other knowledge persists, countries will also most factors, however, is sometimes hard, as will likely have to maintain a balanced tertiary be seen from survey and other data used in education sector among technical and voca- this chapter. tional education and training (TVET) institu- The main broad conclusion of this chapter tions, social sciences, and STEM fi elds. is that higher education in low- and middle- Moreover, beyond weaknesses in deliv- income East Asia is not delivering the skills ering high-level skills, universities are also and producing the research that is required to shown to be weak in terms of quality rank- address labor market and innovation needs. ings and research and technology develop- With skills, the challenges are multiple. ment outputs in all low- and middle-income Some countries urgently need to grow their countries, thus increasing the magnitude of higher education systems in terms of enroll- the higher education challenge. ment. In most countries there is scope to enhance equitable access to widen the talent Higher education and skills for pool, and the share of graduates in science, growth: The main issues technology, engineering, and mathematics (STEM) remains too low to support much Investment Climate Surveys (ICSs)—which technological capability. At the same time, provide large national samples of fi rm-level beyond pervasive gaps in technical skills in data across the industrial, manufacturing, most countries (including the capacity to apply and to a lesser extent, service sectors of the knowledge), both employers and employees region—point to macroeconomic instability, are increasingly recognizing the signifi cance regulatory economic and policy uncertainty,

39 40 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

and corruption, as well as the skills and technology adoption, a fact already appar- education of workers, as the main obsta- ent from simple correlations (fi gure 2.3) and cles to operating in a country (fi gure 2.1). from more sophisticated multivariate regres- East Asia as a region is second only to North sions (appendix G).2 The regression analysis Africa—and at the level of Latin America confi rms that both (a) technology innovation and Africa—for the importance of skills as or adoption and (b) openness (measured by an obstacle.1 export orientation) are positively correlated Time to fi ll professional and skilled vacan- with measures of skill gaps, even with the cies is a standard indicator to measure dif- inclusion of such control variables as fi rm size fi culties in matching skill demand and supply. and fi rm ownership and alternative specifi - It takes about six weeks to fi ll professional cations including location, manager educa- vacancies in both Malaysia and Thailand, and tion, access to fi nance, and several indicators more than four weeks in China and Mongolia of technological intensity and firm perfor- (fi gure 2.2a), which is high relative to a high- mance.3 Notwithstanding the small size of the income country such as the Republic of Korea country samples, which requires particular and a low-income country such as India (but care in interpretation, the effects of technol- still lower than Brazil). These figures are ogy on gaps are particularly evident in Indo- also higher than for the United States, where nesia, Malaysia, the Philippines, Thailand, recent data from the 2010 Job Opening and and Vietnam (table 2.1). Overall, the relation Labor Turnover Survey indicated that 3.5 between skill gaps and technology is apparent weeks are needed to fill a skilled vacancy. in all income groups but somewhat stronger An update of the vacancy data for the coun- for middle-income countries.4 tries for which updated ICSs or other surveys The export sector faces skill gaps, par- (fi gure 2.2b) are available indicates a slight ticularly in China, Indonesia, the Philippines, improvement in Malaysia but the contrary and Thailand.5 These gaps may already be in Indonesia, the Philippines, and Thailand. refl ected in some of these countries’ export For Indonesia and the Philippines, which use performances, particularly the declining more extensive employer skill surveys, this share of relatively high-value-added prod- result reveals possibly higher skill inadequa- ucts (though direct connections cannot be cies in services. made).6 These results also illustrate that Skill gaps have also increased in lower- skill gaps in the export sector are more of an income countries—as illustrated by the fact issue for middle-income countries than low- that even in the Lao People’s Democratic income ones. An upper-income country like Republic and Cambodia, where skills were Korea experiences both lower skill gaps and initially a lesser obstacle (fi gure 2.1), the lat- a weaker relation among skill gaps, technol- est ICS data confi rm that in both countries ogy, and openness. about 40 percent of fi rms considered skills Skill constraints may have become an issue to be at least a moderate obstacle for their in many parts of East Asia when measured as business (with almost 20 percent considering obstacles to business and time to fi ll skilled them at least a major obstacle). These results vacancies, but to what extent is it the fault of show that lower- and middle-income coun- education in general and higher education in tries have been unable to raise skill supply particular? and quality, though demand has strengthened Of course, higher education has a key and wage premiums are overall attractive. role to play in skill supply. As economies In a worrisome trend, finding the right and demand for skills grow, the quantity skills becomes harder the greater the export and the quality of tertiary graduates gener- orientation and technological intensity of ally adjust to the rising needs of the region’s manufacturing firms, where many tertiary labor markets and innovation systems— graduates work. In manufacturing, skill gaps coverage of higher education is higher in grow with the fi rms’ degree of openness and upper-middle-income and upper-income FIGURE 2.1 Business climate obstacles and skill bottlenecks FIGURE 2.2 Time to fi ll professional vacancies

a. Perceptions of business climate obstacles 2.0 a. First round surveys, various years 9 8 1.5 7 6 5

1.0 weeks 4 index 3 2 1 0.5 0

China India Brazil Vietnam Mongolia Malaysia Thailand IndonesiaPhilippinesKorea, Rep. Cambodia 0 East Asia other China Lao PDR Malaysia Mongolia Thailand Vietnam Cambodia Indonesia PhilippinesKorea, Rep. corruption macroeconomic instability b. First and second round surveys, various years economic and regulatory policy skills and education of available uncertainty workers 8 7 b. Firms reporting skills of workers as an obstacle 60 6

50 5

40 4 weeks

30 3 percent 2 20 1 10 0 0

China Mongolia Indonesia Thailand Malaysia Lao PDR Malaysia Vietnam Mongolia Thailand Philippines Korea, Rep.Cambodia Philippines Indonesia major or very severe obstacle previous year most recent year moderate, major, or very severe obstacle 41 Source: Almeida 2009a, based on World Bank Investment Climate Surveys (ICSs), Sources: World Bank IC Surveys (various years), di Gropello, Kruse, and Tandon (2011), and di Gropello, various years. Tan, and Tandon (2010). 42 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 2.3 Skill bottlenecks, technology, and openness Quantity of higher education graduates a. Share of ﬁrms reporting skills of workforce as an obstacle, by technological intensity This ﬁ rst section looks at the issue of quan- 35 tity of higher education graduates. The next Malaysia Thailand 30 China section then turns to quality issues. Mongolia 25 Indonesia Vietnam 20 15 Philippines Higher education for the labor market Lao PDR

innovation (%) innovation 10 Korea, Rep. Even though tertiary gross enrollment ratios Cambodia 5 (GERs) are below par in several countries

firms engaging in technological 0 (as shown in chapter 1), three points must 0 5 10 15 20 25 30 35 firms not engaging in technological innovation (%) be made. First, despite overall poor performance in East Asia’s low- and middle- b. Share of firms reporting skills of workforce as an obstacle, by income countries, they are nonetheless Internet use aligned with other low- and middle-income 70 Mongolia countries in other regions. These results may Lao PDR 60 help explain why employers in most coun- Indonesia Thailand 50 tries in the region fi nd the current quantity Malaysia Vietnam of university graduates, or quantity of edu- 40 Philippines cation in general, much less of an issue than 30 Cambodia quality in skill gaps (appendix H). 20 Korea, Rep. Second, the GER and current proportion of tertiary graduates may not be fully aligned

firms using the Internet (%) firms using the Internet 10 when the GER has been subject to recent 0 0 102030405060changes (such as recent growth). The propor- firms not using the Internet (%) tion of adults in low- and middle-income East Asia with a university degree tends to be pro- c. Share of firms reporting skills of worksforce as an obstacle, by portional to the country’s GER. For example, exporting intensity as a result of its longer-term increases in the 70 Mongolia Thailand GER, the Philippines also has a higher ratio 60 Indonesia China of workers with university qualifications 50 Vietnam Philippines (figure 2.4). In contrast, the gaps between 40 Lao PDR Malaysia GERs and the ratio of university graduates in Cambodia 30 Korea, Rep. Mongolia and Thailand suggest that enroll- 20 ment growth has been more recent. exporting firms (%) 10 Third, “below par” does not mean that 0 010203040506070 the lack of supply of tertiary graduates is an nonexporting firms (%) immediate cause of skill gaps in all countries. While demand for higher education is ris- Source: Almeida 2009a, based on World Bank ICSs (various years). ing in all countries in East Asia, the trends in education premiums in Indonesia and the Philippines, for example, point to a very East Asian economies. But to what extent different labor market than in Vietnam. In can the skill gap in low- and lower-middle- the longer run, higher education coverage income East Asia be explained by too few will likely continue growing (all the more if tertiary graduates? And looking beyond head services continue to grow), but the growth counts to quality, is higher education deliver- imperative is not the same for all countries— ing to its graduates the skills that allow them an important factor when one assesses the to be productive and innovative? key country-level issues of higher education. IS HIGHER EDUCATION MEETING ITS PROMISES? 43

TABLE 2.1 Regression coeffi cients of technological and openness variables in a sample of East Asian economies

Low- Middle- Middle- Lao Korea, income income income Variable PDR Cambodia Vietnam Mongolia Philippines Indonesia China Thailand Malaysia Rep. economies economiesa economiesb

Technological innovationc 0.013 −0.001 0.046** −0.008 0.062** 0.066** 0.003 0.017 0.063* 0.025 0.032** 0.031** 0.035** Technological innovationd 0.050 0.457** −0.006 0.081* 0.018 −0.040 0.050 0.131*** 0.104*** Opennessc −0.01 −0.017 0.026 −0.008 0.089** 0.026 0.042 0.048* −0.005 0.045 0.012 0.044*** 0.037*** Opennessd 0.091 0.284** 0.121** 0.065 −0.082 −0.096 0.091 0.144** 0.105***

Source: Appendix G. a. All East Asian economies except low-income economies, Malaysia, and the Republic of Korea. b. All East Asian economies except low-income economies and the Republic of Korea. c. Dependent variable is perceptions on skill bottlenecks. d. Dependent variable is time to fi ll skilled vacancies. Signifi cance level: * = 10 percent; ** = 5 percent; *** = 1 percent.

These points call for an analysis of demand- FIGURE 2.4 Proportion of adult population with university side and supply-side factors in each country qualifi cations when assessing how much the number of tertiary graduates may be a constraint, which 20 follows here. Several critical indicators are reviewed for each country. The main demand-side (labor 15 market) indicators include the level and trend of the tertiary education premium (for the 10

overall adult population and specifi c genera- percent tions), the time required to fi ll professional vacancies, the unemployment rate of tertiary 5 graduates, and the proportion of tertiary graduates working in professional occupations (the “professionalization” rate). Edu- 0 cation premiums, employment, and vacancy Vietnam Mongolia Thailand indicators can be used to measure the capac- Cambodia Philippines Indonesia ity of absorption of the surrounding economy and, related to that, the scarcity or abun- Source: Sakellariou 2010b. dance of tertiary graduates. For instance, Note: Cambodia and Indonesia data are for 2009; Vietnam data are from 2008; Mongolia data are high unemployment rates, low professional- from 2007; Philippines and Thailand data are from 2006. Adult population is 22–65 years of age. ization rates, and short required times to fi ll skilled vacancies, combined with decreasing or low education premiums, may suggest the trends by country group experiencing similar presence of constraints on the labor market trends are reported in table 2.2.) Combined, side (limited demand for tertiary graduates), these indicators allow assessment of both the and vice versa. situation of tertiary graduates in the labor The main supply-side indicators include market and the magnitude of their supply to the level and trends in the ratio of the identify possible pressure points. This analy- tertiary-educated workforce and the tertiary sis is undertaken with the understanding that GER. (The values of all the demand-side and it is a simplifi ed exercise and that the trends supply-side indicators mentioned in the text in some labor market indicators may also are reported in appendix I, and their main indicate the presence of other issues beyond 44 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

TABLE 2.2 Tertiary education quantity analysis by country groups

Country group Indicators Severity of quantity gaps Suggested response Cambodia, China, Higha or quite sharply increasing education premium Overall quantity gaps Supply push to increase Vietnam across sectorsb (including new graduates) number of graduates High tertiary professionalization ratec and low confi rmed by demand and unemployment rated supply indicators Signifi cant time required to fi ll professional vacanciese Quantity gaps particularly strong in Cambodian Education premium for youth stable or increasingf manufacturing Low ratios of tertiary educated workers Low tertiary GER Mongolia, Thailand High (quite sharply increasing in Mongolia) education Overall quantity gap in Focus on tertiary-educated premium current tertiary-educated workers (rather than supply Low unemployment rate and high professionalization workers (but suffi cient new push) rate graduates for labor market) confi rmed by demand and Signifi cant time required to fi ll professional vacancies supply indicators Education premium for youth stable (Mongolia) or decreasing (Thailand) Low (in relation to GER) or decreasing ratio of tertiary-educated workers (particularly in Mongolian manufacturing) Tertiary GER high regionally Indonesia, Decreasing or fl at education premium, but higher and Quantity gaps (including new Reallocation across sectors Philippines increasing education premiums on tertiary graduates in graduates) in specifi c sectors Policies to enhance demand Indonesian manufacturing and Philippine services (Indonesian manufacturing; for tertiary workers Signifi cant unemployment and low professionalization Philippine services) confi rmed rates by demand and supply data, but no overall quantity gap Education premium for Philippine youth decreasing In Indonesia lack of overall Short time required to fi ll professional vacancies quantity gap illustrated by Relatively high ratio of tertiary-educated workers in the demand data coexisting with Philippines (though remaining fl at in services); a lower supply data on the low side ratio in Indonesia (even decreasing in manufacturing)g GER relatively high regionally in the Philippines but low internationally; low GER in Indonesia Malaysiah Fairly high but decreasing time to fi ll professional No clear quantity gaps vacancies according to demand University graduates marginal and decreasing cause of indicators (at least for new vacancies graduates), although current GER might be higher GER relatively high regionally but low internationally

Sources: Appendixes C and I. Note: Magnitude is assessed in relation to regional and international levels. a. The education premiums are lower in China and Vietnam because of their much lower starting point (because of later market liberalization). b. The increase is even more marked in Cambodian manufacturing. c. This relies on the proportion of tertiary graduates working as professionals (admittedly an imprecise indicator of local absorption potential because it is normal to have a share of tertiary graduates working as skilled production or nonproduction workers). d. The professionalization rate is somewhat lower in Cambodia because of a more basic economic structure and because of insuffi cient diversifi cation of tertiary graduates. e. Apart from Vietnam possibly, but the time required there is likely to be higher for managerial positions and certain professional categories (according to more recent evidence). f. This indicates that rewards to higher education are not decreasing as supply increases. g. A high and increasing premium, combined with a decreasing ratio of tertiary workers in manufacturing in Indonesia, is evidence of quantity gaps in the sector. The same reasoning is valid, though the evidence is a bit less clear-cut, for Philippine services. h. Data for Malaysia are too scarce to perform a complete analysis. IS HIGHER EDUCATION MEETING ITS PROMISES? 45

quantity-related ones (such as labor market FIGURE 2.5 Benchmarking quantity gaps in tertiary education, distortions and poor quality). Cambodia, China, and Vietnam Figures 2.5–2.7 provide a visualization for each country using several key indica- a. Cambodia tors from the table 2.2. From these data, one tertiary education premium can broadly distinguish four main groups 90 of countries. In the fi rst group (Cambodia, China, and Vietnam), most of the demand- tertiary gross 100 trends in education side and supply-side indicators suggest that enrollment ratio the overall quantity of tertiary graduates is premium still insufficient to respond to the needs of 8 3 the economy and a supply push is thus war- 20 ranted to increase the number of graduates. tertiary-educated days to fill While not all indicators are consistent (some workforce professional vacancies of the caveats are explained in table 2.2), 52 these countries are generally characterized professionalization rate by levels and trends in the demand-side and supply-side indicators that indicate a situation b. China of quantity gaps (supply-side constraints). As a result, for all three countries, the right side tertiary education premium of the graphs tends to be more developed 60 visually than the left. Quantity gaps are particularly strong in tertiary gross trends in education Cambodia’s manufacturing sector, judg- enrollment ratio premium 20 36 ing from the particularly steep increase in the tertiary education premium in that sector combined with a very low and stagnant 24 25 tertiary-educated workforce (appendix C). tertiary-educated days to fill professional vacancies This fi nding suggests the need for Cambodia workforce to widen its very low curriculum diversifi cation (discussed in chapter 3). Otherwise the professionalization rate system faces a strong risk of persisting in producing unemployed or poorly employed c. Vietnam graduates and in failing to meet manufacturing demand for graduates. tertiary education premium One needs to be careful when interpreting trends in rates of return in China, given the tertiary gross 55 trends in education short period under discussion and the lack of enrollment ratio premium more recent evidence. Anecdotal evidence (for 12 40 instance, in newspapers) suggests that over the past few years tertiary graduates may 13 13 have seen stagnating returns and increasing tertiary-educated days to fill professional vacancies rates. The recent postcrisis recovery should, workforce however, have put things back on track. In the case of Vietnam, the evidence con- 84 fi rms the perception of employers about the professionalization rate lack of suffi cient tertiary graduates as a key Source: Appendix I. reason for skill gaps (appendix H). 46 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 2.6 Benchmarking quantity gaps in tertiary education, tertiary-educated workers is stronger in Thai- Mongolia and Thailand land, but the decrease in tertiary-educated workers is more notable in Mongolia (par- a. Mongolia ticularly in manufacturing), likely because of tertiary education premium emigration.7 70 These fi ndings highlight a need to focus on the number of tertiary-educated workers in tertiary gross trends in education these two countries rather than on a supply enrollment ratio premium 50 41 push for new graduates (beyond sustaining the current GER). In Thailand the problem appears to be how to provide tertiary educa- 33 23 tion to the current adult population to cover tertiary-educated days to fill the transition period; in Mongolia much of workforce professional vacancies the problem is how to retain current tertiary workers in the country. The charts for both 69 countries are relatively more symmetric than professionalization rate the charts for the other countries. The third group of countries—Indonesia b. Thailand and the Philippines—experience some tertiary education quantity gaps limited to criti- tertiary education premium cal sectors such as services (the Philippines) 100 and manufacturing (Indonesia), but no immediate overall quantity gap. tertiary gross trends in education The Philippines seems to experience an enrollment ratio premium insuffi cient supply of tertiary graduates in ser- 40 17 vices. This is because of higher and increasing education premiums on such graduates in 21 30 services (appendix C), combined with quan- tertiary-educated days to fill tity being a bit more of an issue in this sector workforce professional vacancies (according to employers), as well as a fl at ratio 68 of tertiary-educated workers in services. professionalization rate More evident is the supply gap of tertiary graduates in Indonesian manufacturing, Source: Appendix I. according to high and increasing rates of return in that sector (appendix C), combined with the higher importance of quantity con- In the second group, Mongolia and Thai- straints (according to employers) and a low land have gaps in the current quantity of and even decreasing ratio of tertiary-educated tertiary-educated workers but not in the ter- workers in that sector. tiary GER. The relevant fi gures are demand- But in both countries, at least for the time side indicators suggesting the need for more being, reallocation across sectors would be tertiary graduates in the labor market, com- suffi cient. For overall labor market indica- bined with relatively low or decreasing ratios tors (the demand side) for tertiary-educated of tertiary-educated workers on the one side, workers, neither country apparently has an but quite high tertiary GER on the other overall immediate tertiary education quantity (and decreasing returns for tertiary-educated gap. To illustrate this point, the charts for the youth in Thailand). In both cases, suffi cient two countries have a somewhat more devel- GERs are also confi rmed by the employers’ oped left-hand side. Beyond this bump, they perception that universities are producing differ. While the quantity issue is confi rmed enough graduates (appendix H). The gap in by supply-side indicators for the Philippines IS HIGHER EDUCATION MEETING ITS PROMISES? 47

(relatively high ratio of tertiary-educated FIGURE 2.7 Benchmarking quantity gaps in tertiary education, workers and tertiary enrollment), a different Indonesia and the Philippines picture emerges for the same indicators in Indonesia. The relatively low tertiary supply- a. Indonesia side indicators suggest that Indonesia would tertiary education premium need more of an overall supply push, but it is 84 currently constrained by a sluggish demand tertiary gross trends in education for skills, as also illustrated by a longer-term enrollment ratio 8 premium wage compression on tertiary workers. 16 –5 Broader policies to enhance the demand for tertiary workers may also be needed (assum- 21 10 ing that demand does not rise drastically in tertiary-educated days to fill the future). Visually speaking, as a result of workforce professional vacancies 45 both supply and demand constraints, the fi g- ure for Indonesia is the narrowest of all. professionalization rate Data for Malaysia, on its own in the fourth group, are too scarce to perform a complete b. Philippines analysis. The still-signifi cant but decreasing tertiary education premium length of time to fi ll professional vacancies, 70 combined with very low (and decreasing) signifi cance of the shortage of new university tertiary gross trends in education graduates as a reason for vacancies, but still enrollment ratio 30 premium relatively low tertiary GERs, points to pos- 5 sible demand constraints in relation to new graduates. However, without data on the 27 15 days to fill tertiary-educated current stock of tertiary-educated workers professional vacancies (such as returns, ratio, and employment per- workforce formance) one cannot conclude if the “likely” 59 relatively low ratio of tertiary-educated work- professionalization rate ers is demand constrained (as in Indonesia) or supply constrained (as in Thailand). Source: Appendix I. Even in East Asian countries that seem to need an immediate boost to the number of tertiary graduates, it seems clear that their 1.084 for tertiary over primary is the differ- absorption capacity is lower than that of ence in coeffi cients for tertiary and primary developed countries or Latin America when in the estimated earnings equation. Dividing one compares return levels with the tertiary 1.084 by the difference in years of education education stock. In Indonesia and Thailand, between tertiary and primary education (10 tertiary graduates earn about 100–120 per- years) implies an annualized return of about cent more than primary education gradu- 10.8 percent. Similarly, to derive an estimate ates, which is a substantial gain.9 Figures are of the return to tertiary over senior second- broadly similar in other East Asian econo- ary (instead of primary), one would calculate mies, though lower for China and Vietnam the difference between 1.084 and 0.508, that because of their lower starting points (caused is, 0.576; annualized, this implies an average by late liberalization). return of 14.4 percent for each year of tertiary A simple equivalence obtained between education over senior secondary. education premiums as calculated in this book These are significant annual rates of and the more traditional annualized rates of return comparable with high-income Asian return allows comparison of magnitudes with and most OECD economies,10 but they are other countries. For Indonesia the estimate of lower than Latin America’s returns.11 When 48 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

one relates these numbers to the much higher (the largest share of the overall GER) and coverage of tertiary education in OECD completion rates suggests weaknesses in how countries and high-income East Asia, and the lower- and middle-income countries in the equivalent coverage levels in Latin America, region are retaining and supporting students. these findings also indicate increasing, but This hypothesis is confi rmed by a compari- still relatively limited, capacity of absorption son of gross completion rates and income of tertiary graduates in low- and middle- levels: most countries are underperforming income East Asia.12 in relation to their income per capita (see Finally, beyond enrollment, completion fi gure 2.8). matters. Regional tertiary completion rates The preceding analysis suggests that for the International Standard Classifi cation the lack of tertiary enrollment opportuni- of Education (ISCED) 5A13 programs sug- ties for young people or adults in Cambo- gest room to improve the retention of tertiary dia, China, Thailand, and Vietnam may be students. China and Indonesia have ISCED insufficient for the skills required by the 5A completion rates of less than 20 percent, labor market. In Indonesia and the Philip- lower than middle-income countries outside pines constraints are more sector-specifi c. the region; Malaysia and the Philippines per- Does this breakdown also hold good for form a bit better, and Mongolia performs innovation indicators? much better (fi gure 2.8). Rather than increasing enrollment rates (and another reason not Higher education for innovation to do it), this evidence confi rms that Indonesia and the Philippines should focus on enhanc- While more innovative economies—such ing student retention (to allow efficiency as Japan and Korea—also have higher edu- improvements). China seems to have scope to cation coverage, in the longer run the rela- improve both enrollment and completion. tion between higher education coverage and These relative trends have persisted over innovation outputs is very loose. Japan, for time, with regional high-income countries instance, has better innovation outputs than consistently exhibiting much higher comple- Korea (see chapter 1) despite lower coverage. tion rates than others. Although one must And lower development levels also show no compare different cohorts with care, the such relationship (fi gure 2.9).14 Indeed, above difference between ISCED 5A enrollment a certain threshold level, too high a GER is associated with lower innovation outputs: Mongolia and the Philippines, for example, seem to have prioritized education quantity FIGURE 2.8 Tertiary gross completion rates and per capita income over quality. Two further points can therefore be made. First, this evidence suggests 60 that while there may be a minimum coverage 50 threshold to spur innovation, the volume of tertiary education graduates matters much 40 Mongolia less than the composition of the graduates and their skills. Second, this evidence also 30 raises the possibility of a quantity-quality Philippines trade-off in higher education. 20 China Malaysia Because the potential capacity of fi rms to absorb and develop technology depends upon

gross tertiarygross (%) completion 10 Cambodia Indonesia STEM skills of employees, degrees in science 0 Lao PDR and engineering ﬁ elds are a better indicator of 5.5 6.5 7.5 8.5 9.5 10.5 11.5 building capabilities than the overall volume GDP per capita (natural log) of tertiary education graduates. STEM shares Source: Yilmaz 2009. are too low in several countries. As reported IS HIGHER EDUCATION MEETING ITS PROMISES? 49

in chapter 1, the experience of leading East FIGURE 2.9 Tertiary gross enrollment ratio and number of journal Asian economies suggests that indigenous articles, latest available year technological capabilities require an abun- dance of scientists and engineers in the earlier 60 stages of industrialization. Indeed, more than half of students earn science and engineer- 50 ing degrees in China, Japan, Singapore, and Thailand. Between half and one-third of stu- 40 dents earn science and engineering degrees in 30 Hong Kong SAR, China; Korea; and Taiwan, China; and about one-fourth in Indonesia 20 and the Philippines (table 2.3).15 Another source of data (figure 2.10) 10 conﬁ rms that less than one-fourth of students earn their degrees in science and 0 engineering in most of East Asia’s low- and China middle-income countries. While the East Vietnam Lao PDR Malaysia Thailand Cambodia Indonesia Mongolia Asian average is slightly higher than the G7 Philippines average (25 percent compared with 22 per- gross enrollment ratio journal articles cent),16 these shares remain a challenge in several countries. Source: United Nations Educational, Scientifi c and Cultural Organization (UNESCO) Institute for Statistics (UIS) Data Centre. Note: Journal articles published per million people. Equity in higher education

In addition to issues of quantity of higher TABLE 2.3 Share of first university degrees in education graduates for the labor market science and engineering and innovation, many of East Asia’s low- percent and middle-income countries are constraining innovation simply by limiting the pool Economy Percent Year of talent accessing and completing tertiary Philippines 26 2004 education. GERs and completion rates mask Indonesia 27 2006 disparities in access by socioeconomic sta- China 56 2004 tus, race or ethnicity, and rural and urban Thailand 69 2001 location, suggesting inclusiveness issues. Korea, Rep. 46 2004 Disparities are particularly acute in Cam- Taiwan, China 41 2005 bodia, Indonesia, Vietnam, and, to a lesser Hong Kong SAR, China 38 2004 extent, the Philippines (figure 2.11). Inas- Singapore 59 2004 much as these disparities are correlated with Japan 63 2005 ability, the potential for innovation is not Sources: National Science Board 2008; data for the Philippines are from the World Bank Knowledge Assessment Methodology data set (http://www. necessarily affected, but if some capable and worldbank.org/kam). talented females or ethnic minority students Note: The National Science Board’s defi nition of science and engineers contains science, engineering, agriculture, and health (and possibly math) are excluded, for example, potential for and counts only the completion. The standard UNESCO Institute for innovation can be seriously affected. And Statistics defi nition picks up only science and engineering and is based on enrollment data. there is evidence that disparities in access and completion are indeed not related to ability but rather to economic, cultural, or the poorest quintile and ethnic minority stu- yet supply-side factors that have prevented dents in Asia and Latin America perform as otherwise talented youth from getting access well as other students after controlling for to and thriving in higher education. Box 2.1 student and school background variables. makes this case by showing that students of The exceptionally high number of resilient 50 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 2.10 Enrollment shares in science and engineering, latest nearly ﬁ vefold higher. In Thailand, majority available year Thais are twice as likely to enroll and three times more likely to complete the tertiary 50 cycle than ethnic minorities. Time series 45 data indicate that these differences in these 40 three countries have generally persisted, 35 though in recent years in Vietnam minorities 30 25 have increased their enrollment—though

percent 20 not completion—rates. Controlling for other 15 variables,18 the negative correlation between 10 rural location and tertiary access is still quite 5 strong in all countries. And while disparities 0 by gender are less acute, females are still penalized in Cambodia and China, whereas Lao PDR VietnamMongolia Malaysia Cambodia PhilippinesIndonesia Korea, Rep.Singapore males are lower completers in the Philippines and Thailand. Finally, the poorest quintile Hong Kong SAR, China still faces lower access to and completion of higher education. In Cambodia, for example, Source: UIS Data Centre. controlling for other socioeconomic factors, students from the top income quintile are more than 50 percent more likely to enroll children in East Asia’s high-income econo- in higher education than students from the mies also shows that even economically and lowest income quintile, thus excluding many socially disadvantaged children can perform talented poor students. well on international tests of math and read- To sum up this section, a few countries ing. These studies show that lower academic such as Cambodia, China, and Vietnam ability is not caused by lower innate abil- present a case for further expanding tertiary ity but rather attributable to school factors education to address immediate labor market and student background. While some stu- needs. Most low- and middle-income coun- dents in East Asia’s high-income economies tries, however, prompt an economic effi ciency are able to overcome their disadvantaged argument for a focus on decreasing dropouts backgrounds, lower- and middle-income during the tertiary cycle (that is, aiming for economies must increase their number of higher completion rates), supporting more resilient children by providing better edu- inclusive access to increase the talent pool cational opportunities to disadvantaged accessing higher education, and increasing students. In this way, economies will have enrollment in STEM fi elds. a better skilled labor force that will lead to So much for the quantity side of the analy- increased productivity for the individual sis. What about quality? and the country. Ethnic minorities appear to exhibit signifi cant disparities in both tertiary enroll- Quality of higher education ment and completion. This disparity is most graduates apparent in Cambodia where the Khmer The quality (and relevance) of education and majority dominates tertiary enrollment by training appears to be much more of a bind- a ratio of eight to one,17 and its completion ing constraint than the quantity of students in rates are several orders of magnitude higher all employers’ surveys, except for Vietnam.19 than among minorities who enroll. Similarly, For example, some 30–40 percent of fi rms in Vietnam enrollments by the majority Kinh report quality to be an important or very are almost fourfold higher than enrollments important issue in Indonesia and the Philip- by minority groups, and completions are pines. And 15–35 percent of fi rms recognize FIGURE 2.11 Predicted ratios of enrollment and completion in tertiary education, by key characteristics

120 a. Quintile 5/quintile 1 b. Male/female 140

100 120

80 100

80 60 60 40 40

20 20

0 0

China China Vietnam Vietnam Thailand Indonesia Mongolia Thailand Indonesia Mongolia Cambodia Philippines Cambodia Philippines low-income middle-income low-income middle-income countries countries countries countries

c. Majority/minority d. Urban/rural 20 20 18 18 16 16 14 14 12 12 10 10 8 8 6 6 4 4 2 2 0 0 Cambodia Vietnam Thailand

Vietnam Thailand Cambodia Indonesia Mongolia low-income middle-income countries countries enrollment completion

51 Source: Sakellariou 2010a, based on labor force and household surveys (latest year). 52 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 2.1 Preventing wasted talent in East Asia

Regression analyses provide a useful tool for TABLE B2.1 2009 PISA reading achievement for understanding how different factors contribute to poorest quintile, Shanghai, China an outcome. In particular, they provide evidence that supporting disadvantaged groups can prevent Regression model (1) (2) wasted talent in East Asia. The following table of Poorest quintile −30.52*** −5.60 Programme for International Student Assessment [5.31] [3.46] (PISA) 2009 reading achievement for students in Controlling for student Shanghai, China, shows that students’ background background characteristics No Yes characteristics—including mother’s education, Constant 562.43*** 504.69*** number of books at home, and other measures [2.24] [4.95] of wealth—are important factors in determining Observations 5,115 5,094 reading achievement. This suggests that students’ R-squared 0.02 0.23 background characteristics can explain much of Reading achievement among the variation in reading scores, whereas simply poorest quintile 531.91 499.09 being in the poorest quintile is not a signifi cant indi- Source: World Bank 2010a. cator of performance.a Similarly, results of regres- Note: Standard errors are noted in brackets. sion analyses looking at the determinants of grade Signifi cance level: *** = 1 percent. point average (GPA) for fi rst-year university students a disadvantaged socio-economic background and in Vietnam indicate that ethnicity and household perform much higher than would be predicted by income are not signifi cant predictors of GPA after their background.”e PISA 2009 fi nds that on aver- controlling for parental background, geographical age, 31 percent of disadvantaged students in the location, and other wealth-related measures.b And, Organisation for Economic Co-operation and Devel- always in Vietnam, ethnicity also loses signifi cance opment are resilient. East Asia has among the high- after controlling for various family wealth measures, est number of resilient students from disadvantaged parental background, and quality of education in backgrounds, with six of the top seven economies analyses looking at the determinants of primary and in number of resilient students coming from East secondary test scores.c Similarly, a World Bank study Asia. Shanghai, China, has the highest number of of PISA 2006 reading scores for six Latin American all economies, with about 75 percent of its disadvan- countries and three high-income countries, including taged students. It is followed by Hong Kong SAR, Korea, fi nds that school characteristics rather than China, at 72 percent; Korea at 56 percent; Macao wealth contribute to student achievement. Results SAR, China, at 50 percent; Singapore at 48 percent; show a small and generally insignifi cant relationship and Japan at 42 percent. Although these economies between wealth and achievement within schools— may have better education systems to support disad- but a large and significant relationship between vantaged groups, the lower levels of resilient children wealth and achievement among schools, suggesting in Thailand (26 percent) and Indonesia (24 percent) that school quality contributes to cognitive achieve- suggest these economies need to expand access to ment and that low-income students perform worse quality education systems.f on cognitive achievement because of the poor qual- d a. World Bank 2010a. ity of their schools. b. Linh, Thuy, and Long 2010. Results of the PISA 2009 also highlight the prev- c. World Bank 2010g. alence of resilient children among disadvantaged d. OECD 2010a. e. OECD 2010a. groups. “Resilient students are those who come from f. OECD 2010a.

the lack of required core or technical skills as unemployment rates, and fairly long times to one of the three main causes of vacancies in ﬁ ll professional positions suggests skill mis- Malaysia, Mongolia, and Thailand. matches between the labor market and the The coexistence of significant demand tertiary education system in some countries for professionals, relatively high tertiary (figure 2.12). These mismatches may show IS HIGHER EDUCATION MEETING ITS PROMISES? 53

FIGURE 2.12 Tertiary unemployment rates and time to fi ll professional vacancies

China Vietnam Mongolia Indonesia Thailand Malaysia Cambodia Philippines unemployment rate of tertiary graduates (percent) weeks to ﬁll professional vacancies

Sources: Sakellariou 2010b; World Bank ICSs (various years).

either that the matching process is not work- The most important and the weakest skills ing properly or—what is of greater interest were also compared by country relative to here—that higher education graduates simply current demand for that skill (see table 2.4). do not have the right skills. This latter reason For instance, information technology (IT) is likely the case in China, Mongolia, and the gaps in the Philippines, while important in Philippines, for example.20 themselves, become less of an issue when compared with current demand for IT skills, Gaps in thinking, technical, and but gaps in technical skills in the Philippines behavioral skills become more important. One limitation is the absence of standard- Gaps in generic and technical skills for ized comparable tertiary testing, which would newly hired professionals21 (who to a large have allowed objective skill comparisons to extent possess tertiary qualifications in be made among countries. East Asia) as reported by employers and Gaps in thinking skills are multiple. Most employees themselves point to clear skill countries cite thinking skills as a gap. Look- gaps among tertiary graduates. Such gaps ing at employers’ and employees’ opinions are assessed simply in terms of lack of skills from ICS data and the skill surveys, there to perform the job. Because all skills are (or is particular emphasis on creative-thinking will be) relevant in the near future, given gaps in Malaysia and Thailand,23 and on the comparison of skill importance across both creativity and problem-solving gaps in countries, both current and future needs Indonesia and the Philippines.24 Thailand has are considered relevant. (Detailed results of the largest deﬁ ciencies, with 20–30 percent skill gaps by country are shown in appen- of employers putting creative thinking and dix J.) The main gaps by country, where problem solving among the three most serious the intensity of each gap is assessed relative gaps. Basic analytical skills—which include to other gaps in the country, are shown in academic and problem-solving skills—are the table 2.4.22 There is no direct comparison most important gap in Cambodia. All these across countries. skills, already in high demand, will become 54 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

TABLE 2.4 Comparative skill gaps among professionals

Creativity Information English Leadership Communication Problem Work Technical Numeracy/ technology solving attitude skills literacy Cambodia — Decision Lack of making analytical skills Vietnam —— — —

Mongolia

Philippines

Indonesia

Thailand

Malaysia

Source: Appendix J (employer and employee surveys). Note: The darker the shade, the stronger the gap (within each country only). Dotted cells indicate gaps that became less serious, and hashed cells indicate gaps that became more serious, in relation to the current demand for that skill. — = not available

more so as the push to innovation intensifi es worse perception of their professional com- in the region. munication skills than employers. Overall, Gaps in behavioral skills are also per- gaps in behavioral skills are bound to be par- vasive. Gaps in leadership, initiative, and ticularly problematic in services, where they decision-making skills are given particu- are of importance and where most tertiary lar emphasis in Indonesia, the Philippines, graduates work. Thailand (where 30 percent of employers put Gaps in IT and computer skills come out leadership among the three weakest skills), clearly in most countries.26 These trends are and Cambodia. For Cambodia, Indonesia, in line with the fi nding that gaps are higher and the Philippines this fi nding may point to in the exporting and technologically inten- a positive relation between countries that are sive sector of the economy. While countries generally considered “more open” politically such as Cambodia and Mongolia do not con- and the importance of these skills (when one sider IT skills among the most critical, these compares the evidence on the most important skills will likely become more important as and weakest skills).25 international and technological integration Gaps in communication and negotia- continues. Employees’ opinions on skill gaps tion skills are also relevant, particularly are aligned with employers’ opinions (except in V ietnam, in the employer and employee in Mongolia where employees view these surveys. But some of the samples cover only gaps as less serious). Gaps in creativity and manufacturing, where these skills are less IT skills are clearer in middle-income and important, and so may underestimate these upper-middle-income countries than in low- skills’ importance and gaps. Indeed, com- income ones, refl ecting in part their higher munication and negotiation are particu- importance in this setting. This does not larly important and also present a relatively mean, however, that low-income countries important gap in Indonesia (all the more in should ignore these skills because sooner or relation to current needs) and the Philippines, later jobs will require them. Moreover, break- which have the most sector-balanced surveys. ing into manufacturing is already requiring In Malaysia and Mongolia employees have a higher levels of creativity. IS HIGHER EDUCATION MEETING ITS PROMISES? 55

Gaps in technical skills are also evident, FIGURE 2.13 Key job-specifi c skill gaps in Vietnam particularly in practical job knowledge. While gaps in generic skills are on the rise, gaps in job-specifi c technical skills (practical and theoretical knowledge of the job) remain job-related technical skills pervasive, all the more in relation to current needs. This is particularly the case in Vietnam, practical knowledge where only a small share of college graduates of the job are said to possess good technical knowl- theoretical knowledge edge of their job (fi gure 2.13), and Indonesia, of the job where technical gaps are signifi cant in relative terms (and more so for employers). But even good writing/verbal skills in Malaysia, though it may not be the largest gap, a signifi cant number of employees still foreign language/ feel that they do not have the needed techni- communication cal or professional skills (fi gure 2.14), and the problem grows in relation to the importance given to these skills. punctuality In the three countries for which the theory versus practice differentiation exists independence (Vietnam, Indonesia, and the Philippines, as illustrated in appendix J), practical knowledge is even more of a gap than theoretical experience knowledge of the job. This gap is stronger 0 5 10 15 20 25 30 for college graduates than for secondary percent TVET graduates in Vietnam (figure 2.13). technical school professional school college The employer skill surveys of Indonesia and the Philippines point to better specifi c skills Source: World Bank 2008, based on 2003 MOLISA-ADB survey on labor market. (including practical knowledge) acquired Note: Characteristics possessed by recruits of diff erent education level (percentage). in tertiary TVET than in tertiary academic education. Some surveys focusing on China, such and intellectual property.28 Similarly, a 2008 as one on foreign employers by McKinsey McKinsey survey of Chinese fi rms found that (China), find that poor technical skills are 44 percent of executives cited insuffi cient tal- a critical issue for engineering graduates.27 ent as their main challenge for reaching global It concluded that only 10 percent of science ambitions and that China has seen a widen- and engineering graduates had acceptable ing gap in the skills mismatch between fi rms technical skills. (The same report notes that and employees at all employment levels.29 the quality of university graduates is a serious issue in Vietnam, where fi rms report that Academic skills only 10 percent of the graduating class of engineers had the potential to become “effec- Academic skills appear to be an important tive employees.”) constraint, according to employers, more so A 2010 study by the American Chamber in low-income countries. Literacy is weak of Commerce in China found that 28 percent in Vietnam, and literacy and numeracy in of U.S. companies based in China reported Cambodia. But employers (and employees) recruiting talented management-level indi- in more advanced countries such as Malaysia viduals as a top-fi ve business challenge, sec- and Thailand also complain about numerical ond only to regulatory problems, but harder skill gaps: about 20 percent of Thai employ- than problems in transparency, bureaucracy, ers, for example, rank numerical skills among 56 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 2.14 Key job-specifi c skill gaps in Malaysia international integration continues, particularly in Vietnam because of its recent acces- sion to the World Trade Organization or in English language proﬁciency Indonesia because of its planned increase in exports. Gaps in English skills are felt par- professional communication skills ticularly strongly in Thailand.

social skills Wage skill premiums teamworking Results from employer and employee skill leadership skills modules in Thailand (fi gure 2.15) and Malay- time management skills sia (fi gure 2.16) further suggest that employers are willing to pay a premium for certain adaptability skills, thereby confi rming scarcity. In Thai- land skills such as English skills, numerical creativity/innovation skills skills, leadership, and even creativity skills are numerical skills indeed commanding higher wages. In Malay- sia professionals with very good professional problem solving communication, technical skills, and IT skills information technology skills command, as expected, the largest wage premiums in relation to professionals with very technical/professional skills poor levels of these skills.30 0 102030405060 70 80 percent Retraining Source: World Bank Malaysia Investment Climate Survey 2007 (employee module). Critical skill gaps associated with preem- Note: Skills most lacking among employees (proportion of respondents indicating particular skill “one of the three most lacked”), both professionals and skilled workers. ployment education and practical knowledge are also confi rmed by the need to retrain, especially younger workers. This need the three weakest skills—or worse than tech- appears significant in the three countries nical skills. for which this information was gathered in Malaysian and Thai employers are correct the employer surveys—Indonesia, the Phil- in highlighting these gaps, which are tradi- ippines, and Vietnam.31 In Indonesia, for tionally better measured through standard- instance, employers rate that 30 percent of ized international testing. While standardized their employees 30–45 years of age need data are available only up to the secondary retraining, and the share increases to almost level, the following chapters illustrate in 50 percent in the under-30 employees. (The detail that education levels are connected and situation is even worse in the Philippines.) that low preparation in earlier levels trans- In Indonesia a significant proportion of lates into defi ciencies later on. professionals (about 20–25 percent), largely English is the weakest academic skill in trained at the higher education level, need most lower- and middle-income countries retraining. Most university graduates, regard- according to employers and employees. Of less of position, need retraining in the Phil- easier assessment by employers—with the ippines (fi gure 2.17). Similarly, in Vietnam expected exception of the Philippines and about 30 percent of those surveyed reported Malaysia, both of Anglo-Saxon tradition, signifi cant need for training college graduates and less expected of Cambodia—English (retraining, new training, or upgrading). appears to be the number one weakness in all It is important to acknowledge, or remem- countries. According to employees it is also ber here, that these skill gaps are not the sole still a critical skill gap in Malaysia. English responsibility of higher education. Indeed, is bound to continue to gain importance as skills such as creativity, problem solving, IS HIGHER EDUCATION MEETING ITS PROMISES? 57

communication, and most academic skills FIGURE 2.15 Average monthly salaries of Thai employees typically start to be taught very early in the reporting a particular skill as a top-three defi ciency compared with education cycle, and others such as practi- employees not reporting skills as a top-three defi ciency cal knowledge are also taught on the job. It is thus a responsibility of the whole skill English language proficiency development system to support better skills. Higher education has, however, a critical information technology skills responsibility in both the failure and the technical/professional skills solution. Evidence from Indonesia and the numerical skills Philippines, for instance, shows that gaps problem solving between employers’ expectations and actual leadership skills skills are stronger for professionals than for creativity/innovation skills other worker categories in East Asia (as is the time to ﬁ ll vacancies at that level). This is professional communication skills particularly the case for skills such as creativ- teamworking ity, problem solving, leadership, negotiation, social skills common skillcommon deficiency most) (least to and English, which, though higher in profes- adaptability sionals, are lower in relation to the expec- 5 6 7 8 9 10 11 12 13 tations that employers and employees alike average monthly salary (baht, thousands) have for these workers.32 In fairness, tertiary top three skill deficiency not top three skill deficiency education also prepares workers for non- professional careers, and thus gaps may be somewhat overestimated. But skilled produc- Source: World Bank Thailand IC Survey 2004 (employee module). tion and nonproduction workers need some of the same skills as professionals, making gaps nonetheless relevant. And the fact that FIGURE 2.16 Average monthly salaries of Malaysian employees secondary TVET may prepare these work- (in manufacturing) with very good versus very poor skills, according ers better than college is alarming. Finally, as to employer chapter 1 showed, it is also clear that higher education can potentially greatly improve numerical skills most of these skills through the right peda- gogy33 and play a critical complementary role teamworking to other skill providers. Unfortunately, this is adaptability skills not happening. time management skills

Research and innovation technical/professional skills

Beyond providing high-level skills, tertiary English language proﬁciency institutions, particularly universities, can support technological capability and inno- communication skills vation through basic and applied research information technology skills and technology transfer in all three technology clusters. To assess how well they are problem-solving skills progressing, this chapter reviews indicators leadership skills of patenting, licensing, and university quality as ranked by various criteria. (Univer- creativity/innovation skills sities are only part of the innovation sys- 0500 1,000 1,500 2,000 2,500 3,000 tem, so their role cannot be fully isolated average monthly salary (ringgit) from that of other actors. They can, how- very good very poor ever, potentially support many innovation outcomes.) Source: World Bank Malaysia IC survey, 2007 (employee and employer modules). 58 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 2.17 Share of workers needing no training, by education TABLE 2.5 Number of patents granted by the level in the Philippines, 2008 USPTO, selected years

Economy 1992 2000 2008 postgraduate Japan 23,151 32,922 36,679 university Singapore 35 242 450 tertiary technical Taiwan, China 1,252 5,806 7,779 postsecondary vocational technical Korea, Rep. 586 3,472 8,731 secondary school diploma Malaysia 11 47 168 primary Thailand 2 30 40 incomplete primary China 41 163 1,874 Indonesia 9 14 19 0204060 80 100 percent Philippines 7 12 22 Vietnam000 Source: di Gropello, Tan, and Tandon 2010. Source: USPTO data.

Patenting roughly the same in 1992, moving ahead of the Southeast Asian countries. Vietnam— Patenting, while far from ideal as an indi- together with Mongolia, Cambodia, and Lao cator of the productivity of research and PDR, not listed in the table—did not receive development (R&D), is generally the pre- any patents during this period. ferred metric,34 and patents granted by the A somewhat different impression is U.S. Patent and Trademark Offi ce (USPTO) conveyed by table 2.6. As of 2007, Japan are preferred for several reasons.35 First, filed the largest number of patents with because the criteria for submission, exam- the World Intellectual Property Organi- ination of patents, and decision to award zation, followed by China, with Korea in patents differ across countries, the num- third place. If measured by country of ori- ber of patents granted by any one coun- gin, Japan still leads, but Korea and China try is not directly comparable with that of switch places. another, and the quality of the patents differs. Using data from a specifi c patent offi ce Licensing eliminates this incompatibility. Because the United States has been the major market for Royalty and license fee payments are taken East Asian economies, using data from the as a proxy for technology absorption and USPTO is appropriate. Second, applying to technology production. China and Singapore a foreign patenting offi ce is more expensive. stand out partly because of the presence of Therefore, only high-quality patents are multinational corporations. Korea also pur- submitted for approval by USPTO, reduc- chases technologies from abroad, though the ing the noise in the data. pace of increase has leveled off since 2005. The number of patents granted to East Other economies in East Asia are not active Asian economies in 1992, 2000, and 2008 in the market for technology because high- is shown in table 2.5. Japan led the fi eld all tech industries in these countries are domi- three years by a wide margin, with more nated by multinational corporations that than 36,000 patents awarded in 2008.36 In generally do not need to license technologies 2008 Korea was a distant second followed by (fi gure 2.18).37 Taiwan, China, though the number of pat- Japan is the major technology provider ents accelerated after 1992. to East Asia in the sense that it maintains a China ranked fourth in 2008. As a mark of surplus in this area, though it became a net its acceleration, in 2008 it received four times provider of technology only in the past few as many as those granted to Singapore, from years.38 Korea and Singapore are emerging IS HIGHER EDUCATION MEETING ITS PROMISES? 59

technology providers. Other countries’ TABLE 2.6 World Intellectual Property Organization patent fi lings by receipts are insignifi cant (fi gure 2.19). origin and offi ce, 2007 In sum, Korea could emerge as a net Patent fi lings provider of technology in the near future. Patent fi lings by offi ce by origin Singapore is actively engaged in the trade of Economy Total Resident Nonresident Total technologies in both directions through foreign affi liates. China is currently the leading Japan 396,291 333,498 62,793 501,270 Singapore 9,951 696 9,255 3,538 importer. Other East Asian economies are Korea, Rep. 172,469 128,701 43,768 174,896 not importing technologies from abroad or Malaysia 2,372 670 1,702 1,144 generating much by themselves. Thailand 1,388 877 511 1,049 China 245,161 153,060 92,101 160,523 Indonesia 4,606 282 4,324 308 Quality of universities Philippines 3,265 231 3,034 310 The quality and research capacity of univer- Vietnam 0 0 0 13 sities play a key role in determining techno- Source: WIPO 2009. logical capability and innovation, though Note: Data for Indonesia and the Philippines are for 2006. many other factors are at work. The top technology cluster and major innovator in East Asia has the higher-quality universities. FIGURE 2.18 Royalty and license fee payments, 1995–2009 China has been making positive progress in both broader and more closely university- 14,000 related innovation and technology indica- 12,000 tors, while all other low- and middle-income countries are making much slower progress 10,000 in all indicators. 8,000 One measure of whether a university is “world class” is how it ranks on interna- 6,000 tional indexes. Currently, two of the most dollars (000) U.S. 4,000 respected are the Shanghai Jiao Tong Univer- sity (SJTU) rankings and the Times Higher 2,000 Education (THE) (formerly the Times Higher 0 Education Supplement) rankings.39 The 1995 1997 1999 2001 2003 2005 2007 2009 Singapore China Korea, Rep. Thailand SJTU ranking methodology focuses almost Hong Kong SAR, China Indonesia Malaysia Philippines exclusively on research-related dimensions of university performance. Its criteria include Source: World Development Indicators (WDI) database. the numbers of Nobel laureates among a university’s faculty and alumni, the number of articles published in the journals Nature as a measure of the attractiveness of that and Science, performance on the academic university as a home to world-class instruc- citation index, and university size. Because tors and researchers. it contains broader elements in its ranking The relatively low quality of universities criteria, the THE ranking includes mea- in East Asia is reflected in the 2009 THE sures of the quality of teaching and skill ranking, which lists just nine East Asian provision in universities, such as surveys of universities in the top 50 (table 2.7). The employers to determine their perception of nine are in Japan; Singapore; Hong Kong the quality of universities’ graduates and SAR, China; Korea; and China, which is the graduate employment rates and salaries. only middle-income economy among them. The criteria also include an indicator on the Thailand’s leading university is ranked 138, number of international faculty on a uni- Malaysia’s 180, the Philippines’ 262, and versity’s payroll, which can be interpreted Indonesia’s 351. 60 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 2.19 Royalty and license fee receipts, 1995–2009 SJTU ranking changes in the top 500 institutions and THE ranking changes in 3,500 the top 2100 in the 2004–09 period confi rm steep improvements for Chinese universities 3,000 (table 2.8). The data show that China has 2,500 been improving systematically since 2004, the year of the fi rst SJTU ranking, doubling 2,000 its top 500 institutions from 16 to 30 in a 1,500 fi ve-year period.

U.S. dollars (000) U.S. Published scientific output provides 1,000 additional evidence that most East Asian 500 economies lag behind OECD countries. Unsurprisingly, members of the lead- 0 1995 1997 1999 2001 2003 2005 2007 2009 ing technology cluster score the highest Korea, Rep. Singapore China Hong Kong SAR, China in terms of articles. They benefit from a Malaysia Thailand Indonesia Philippines solid base of primary and secondary education and have high-quality tertiary edu- Source: WDI database. cation. They invest heavily in R&D (and,

TABLE 2.7 Ranking of universities, East Asia, 2007, 2008, and 2009

Regional rank International rank School Economy 2009 2007 2008 2009 University of Tokyo Japan 1 17 19 22 University of Hong Kong Hong Kong SAR, China 2 18 26 24 Kyoto University Japan 3 25 25 25 National University of Singapore Singapore 4 33 30 30 Hong Kong University of Science and Technology Hong Kong SAR, China 5 53 39 35 Osaka University Japan 6 46 44 43 Chinese University of Hong Kong Hong Kong SAR, China 7 38 42 46 Seoul National University Korea, Rep. 8 51 50 47 Tsinghua University China 9 40 56 49 Peking University China 10 36 50 52 Chulalongkorn University Thailand 19 223 166 138 Universiti Malaya Malaysia 29 246 230 180 University of Indonesia Indonesia 31 395 287 201 Mahidol University Thailand 34 284 251 220 Universitas Gadjah Mada Indonesia 37 360 316 250 University of the Philippines Philippines 39 398 276 262 Universiti Kebangsaan Malaysia Malaysia 40 309 250 291 Universiti Sains Malaysia Malaysia 43 307 313 314 Universiti Teknologi Malaysia Malaysia 44 401–500 356 320 Universiti Putra Malaysia Malaysia 46 364 320 345 Bandung Institute of Technology Indonesia 47 369 315 351

Source: THE World University Rankings. IS HIGHER EDUCATION MEETING ITS PROMISES? 61

TABLE 2.8 Rankings of universities over time, 2004–09

Ranking 2004 2005 2006 2007 2008 2009 SJTU top 500 Japan363432333131 Singapore222222 China 16 18 19 25 30 30 THE top 200 Japan 5 10 11 11 10 11 Singapore222222 China 9 10 10 10 10 11

Source: SJTU Academic Ranking of World Universities.

as noted, receive more patents than other FIGURE 2.20 Scientifi c and technical articles per million regional economies). Each is a source of a inhabitants, East Asia and the rest of the world, 2007 respectable number of scientiﬁ c and technical papers when data are scaled to the total a. East Asia populations. With the notable exception of Singapore Singapore, even the region’s leaders, such as Japan Japan and Korea, produced fewer articles Korea, Rep. countries per million inhabitants than the world’s top high-income Malaysia performers—Finland, the United Kingdom, Thailand Mongolia and the United States (ﬁ gure 2.20). Philippines income middle- China is trying hard to move up to the top countries Indonesia cluster by giving the highest priority to tech- China nology development through investment in Vietnam tertiary education and R&D, and over these Lao PDR

countries Cambodia

past 15 years it is closing the gap, as measured lower-income 0 200 400 600 800 1,000 by the sharp increase in patents granted and scientiﬁc and technical articles per the number of papers published in interna- million inhabitants tional science and engineering journals. These b. Rest of world rose from 9,000 in 1995 to 41,000 in 2004.40 Or, over a longer time frame, from 2,694 in Finland 1980–84 to 48,552 in 2000–05 (table 2.9). United Kingdom

countries United States But there is still a way to go to catch up with high-income Chile upper-income economies when these ﬁ gures Brazil are compared with population. South Africa Mexico income middle-

The other East Asian economies also countries assign priority to innovation in their policy India Pakistan statements but are struggling to demonstrate Ghana results. They are, in fact, much further behind, Nepal

trailing also several other middle-income countries Bangladesh countries outside the region. From 2000 to lower-income 0 200 400 600 800 1,000 2005, Korea produced an annual average of scientiﬁc and technical articles per million inhabitants more than 20,000 scholarly publications, over 10 times more than the number of publications produced in Thailand and Malaysia. Source: WDI database. 62 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

TABLE 2.9 Yearly average number of publications, 1980–2005

Economy 1980–84 1985–89 1990–94 1995–99 2000–05 Singapore 253 597 1,142 2,501 5,177 Taiwan, China 642 1,644 4,326 8,608 13,307 Korea, Rep. 341 1,043 2,756 9,813 21,471 Malaysia 259 298 421 745 1,221 Thailand 394 446 557 926 2,059 Chinaa 2,694 6,244 10,365 21,205 48,552 Indonesia 104 141 198 366 524 Philippines 237 207 246 329 474

Source: Yusuf and Nabeshima 2010. a. Includes Hong Kong SAR, China.

FIGURE 2.21 Leading ways of acquiring technological innovation in fi rms, Malaysia, Mongolia, and Thailand

Malaysia, 2007

Thailand, 2006

Mongolia, 2004

0204060 80 100 percent clients and suppliers machinery within firm or from parent company personnel and consultants international and domestic licensing universities and public institutions other

Source: World Bank ICSs (various years).

Most important, universities and other technology upgrading,42 indicative of a tertiary entities are not supporting tech- disconnect between firms and universi- nology assimilation and upgrading as they ties, which is further explored in the next should. In the top technology cluster, only chapter. a few research universities are sources of In Malaysia, Mongolia, and Thailand, ideas and engage in applied research with universities (and other public research commercial applications. In all other institutions) are mentioned as a leading countries, with the possible exception way of acquiring technological innova- of China,41 there is very limited engage- tion (defi ned in a broad sense) by only 1 ment of universities in applied research or 2 percent of fi rms (fi gure 2.21). In Viet- and technology development, or even only nam only a marginal percentage of fi rms IS HIGHER EDUCATION MEETING ITS PROMISES? 63

declared using research centers or uni- FIGURE 2.22 Sources of product innovation at fi rm level in versities as sources of product innovation Vietnam (fi gure 2.22). A related fi rm survey in the Philippines also noted that innovations in 3.0% new product lines for almost all the com- 2.8% 6.6% panies surveyed were based on their own R&D staff work, not university research. Strong engagement of universities in technology upgrading remains a challenge also in upper-income economies in East Asia and elsewhere, and it is evident that firms remain the main actors in R&D, but university contribution is higher. As an example, universities are cited as the source of technological innovation in the 87.6% United Kingdom by about 27 percent of fi rms (with 11 percent mentioning universities as a high and medium source)43 and in 44 Singapore by more than 10 percent. This implemented by enterprises themselves suggests that universities in lower- and cooperation with local enterprises or institutes/universities middle-income East Asia could do more to implemented mainly by other enterprises or institutes/universities support technological capability (compat- cooperation with foreign enterprises or institutes/universities ibly with the broader constraints of their technology cluster). Source: Vietnam Chamber of Commerce and Industry 2009.

Conclusion Looking ahead, it is clear that lower- and Notes middle-income East Asian countries will need to do a much better job at providing 1. These fi ndings are somewhat conditioned by the skills required by the labor market and the ICS samples, which mainly cover manufacturing fi rms and a generally signifi cant pro- to support innovation, while also starting portion of large and foreign-owned fi rms. The to build higher research potential of their bias, however, can cut both ways. For instance, universities. The precise challenges and pri- when including services in the Philippine fi rm orities vary by income group and country. sample—as done in the Employer Skill Survey Challenges are deeper and more widespread (di Gropello, Tan, and Tandon 2010)—skills in lower-income countries and decrease at become much more of an obstacle, revealing higher income levels. China’s better per- possibly higher skill gaps in this sector. In the formance from the perspective of R&D for Indonesia Skill Survey (di Gropello, Kruse, innovation is aligned with its leadership and Tandon 2011), however, which incorpo- position in the middle technology cluster, rates a greater number of small and domestic but skill-related gaps are persistent. Gaps fi rms, skills become less of an obstacle. 2. See Almeida (2009a) for details on the data relative to upper-income groups are evident sample and methodology. The surveys were in all dimensions. The immediate urgency conducted in 2002–05, and the samples were for lower- and middle-income East Asia is to designed to be representative of the popu- better understand why its higher education lation of fi rms according to their industry is not performing as it should, which is dis- and location within each country. The fi nal cussed in chapter 3 through the perspective sample consists of 10,215 fi rms in a wide of disconnects. range of sectors (manufacturing 78 percent, 64 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

services 20 percent, construction 1.3 per- 10. OECD 2007a. cent, agro-industry 0.85 percent, and other 11. In Latin America, rates of return for tertiary 0.11 percent). Manufacturing covers several education fl uctuate at 14–28 percent (di industries—auto and auto components, bev- Gropello 2006). erages, chemicals, electronics, food, garments, 12. One cannot rule out that relatively lower leather, metals and machinery, nonmetal- returns in relation to the stock of tertiary grad- lic and plastic materials, paper, textiles, and uates also indicate lower graduate quality. wood and furniture. 13. UNESCO’s ISCED 5A refers to the fi rst stage 3. Interestingly, while technological innovation of tertiary education not leading directly to or adoption loses signifi cance after research an advanced research qualifi cation (equiva- and development is included (likely because of lent to a bachelor’s degree). ISCED 6 pro- the high correlation between these two vari- grams are designed to prepare graduates for ables), technological innovation and Internet faculty and research posts and are the second and computer use reinforce themselves in stage of tertiary education (equivalent to a the explanation of skill gaps, indicating a doctoral degree). cumulative effect on skill gaps of indicators 14. These fi ndings hold when comparing tertiary- of technological innovation, adoption, and educated workers with journal articles. application. 15. The situation is much worse in the Philippines 4. This relation is on the increase in Lao PDR according to the number of students receiv- and Cambodia, according to the last ICS. ing advanced degrees in STEM fi elds. In 2004 5. A positive relation between time to fi ll only 315 received a master’s degree in engi- skilled vacancies and openness in China and neering, 203 in math and computer science, Indonesia, which does not translate into a and 153 in natural sciences. Extremely few positive relation with skills as an obstacle, received doctoral degrees: 13 in natural sci- initially looks like a puzzle. But, at least for ences, 6 in engineering, and 6 in math and China, where the export-oriented sector computer science (Tan 2010). remains quite low skilled, it may simply sug- 16. UNESCO Institute for Statistics Data Centre, gest the presence of some other not strictly 2009 data. skill-related labor market–related factor, 17. Sakellariou 2010a on the basis of labor force which has an impact on time to fi ll vacancies and household surveys. but is less likely to be picked up in the skill as 18. All results are predicted results from a regres- an obstacle variable. sion analysis with controls to isolate the 6. Although having many possible causes, some effects of each single variable. sectoral trends are troubling. In Indonesia 19. The response varies, according to how the metals and metal products, chemicals and surveys pose the question. chem ical products, textiles, garments, and 20. This conclusion does not rule out quality electrical machinery and equipment have issues also in other countries, where the pres- all lost signifi cant export shares since 2006 ence of quantity gaps may simply decrease the (International Monetary Fund 2010 Article IV incidence of unemployment even if quality is Consultation: Indonesia). In the Philippines low (likely to be the case in Cambodia and the heavy metals industry (including chemi- Vietnam), or limited potential for domestic cals and machinery) has lost ground, as mea- absorption may explain while unemployment sured by share of manufacturing value added, is more of an issue than vacancies (Indonesia) falling from 44 percent in 2003 to 36 percent without discarding quality issues. in 2007 (Tan 2010). In Thailand the electronic 21. In Malaysia, Mongolia, and Thailand, gaps and electrical subsector’s share of exports fell relate to both professionals and skilled pro- from 30.5 percent in 2000 to just more than duction and nonproduction workers. 25 percent in 2007 (Doner, Intarakumnerd, 22. For Vietnam, results are directly comparing and Ritchie 2010). skills possessed by graduates with different 7. Judging from the continuous, steeply decreas- education levels. ing ratio of tertiary-educated workers in that 23. See also Abelmann, Chang, and Ayudhaya sector (appendix C of this book). 2000. 8. di Gropello, Kruse, and Tandon 2011. 24. This is also shown for Indonesia in di Gropello, 9. di Gropello and Sakellariou 2009. Kruse, and Tandon (2011). IS HIGHER EDUCATION MEETING ITS PROMISES? 65

25. For Indonesia this is confi rmed by the disagg- 34. Scotchmer 2004. regation of the most important behavioral 35. Other indicators include enrollment in post- skills included in di Gropello, Kruse, and graduate programs, number of PhDs awarded, Tandon (2011). number of graduate schools and research 26. Cambodia is an exception for IT; such skills institutes, number of scientifi c publications, may still not fully be on the radar screen. and number of multinational corporation 27. Farrell and Grant 2005. R&D centers. 28. AmCham-China 2010. 36. In fact, among foreign countries, Japan 29. Lane and Pollner 2008. receives the most patents (60 percent), fol- 30. Beyond the gaps underlined by employers, as lowed by Germany. In any year, about half visible in fi gure 2.16, technical and profes- the USPTO-granted patents are to foreigners. sional skills are among the largest skill gaps 37. Even though multinational corporations reported by employees in Malaysia. have tended to locate their labor-intensive 31. See di Gropello, Kruse, and Tandon 2011; di downstream operations in Southeast Asia, Gropello, Tan, and Tandon 2010; and World they generally hire more educated and skilled Bank 2008 for further information. workers because, on average, the skill coeffi - 32. This is visible from comparing appendix F of cients of their activities are higher than those this book and appendix K of this book. of domestic industries. 33. A study of higher education and information 38. Mok 2010. and communication technology skills, for 39. Salmi and Saroyan (2007) discuss these rank- instance, found that employers in the United ings’ limitations. Kingdom have increasing confi dence in the 40. Chapman 2010. information and communication technology 41. In China university engagement in tech- skills that workers have developed through nology upgrading has been growing (Mok higher education: employers are pursuing the 2010), but specifi c data on sources of tech- greater use of nontechnical graduates in roles nological innovation at fi rm level are not such as technical support through the “sub- available. stitution” of university-educated employees 42. Mok 2010. into roles that would previously have been 43. Laursen and Salter 2004. fi lled by workplace trainees (Round 2003). 44. Mok 2010.

Disconnects in Higher Education 3

he analysis of skill demand and skill should because these different actors are not gaps in earlier chapters indicated clear adequately connected. In other words, higher Tshortcomings in the quality and rele- education in lower- and middle-income East vance of higher education in low- and middle- Asia is not being managed as a system but income East Asia. Although higher education instead as individual disconnected institu- has the potential to be a leading contributor tions. And because many education policy to regional competitiveness, it is not realiz- makers regard these disconnects as outside the ing its potential. It suffers from signiﬁ cant immediate purview of the tertiary education weaknesses in delivering relevant skills and system, they have not used them in formulat- contributing to technological capability ing policies to improve higher education, thus and innovation—and ultimately growth— keeping reforms inconsistent, piecemeal, and through higher productivity. As analyzed in incomplete—in a word, ineffective. this chapter, this failure reflects five main disconnects among skill and research provid- The fi rst disconnect: Between ers and users (see ﬁ gure 3.1): between higher higher education and employers education institutions (HEIs) and skill users; (skill users) between universities and firms (now seen as research users) in the technology arena; This first section reviews disconnects between teaching and research (or HEIs and between HEIs and their main characteris- research providers); among HEIs themselves tics in terms of pedagogy, curriculum, and and HEIs and tertiary-level nonformal or degrees—and employers, who constitute the enterprise-based skill provision; and between skill users. higher education and prior education levels. These two last categories represent discon- Curricula and pedagogy nects between skill providers. At the core of the higher education system Many low- and middle-income East Asian are the HEIs and other critical institutional countries’ curricular and pedagogical actors that supply and use skills and research, approaches do not meet the needs of services and the interrelation among all these actors. and manufacturing, nor do they allow stu- Higher education is not delivering as it dents to acquire the needed generic skills,

67 68 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 3.1 Five disconnects in higher education

employers

companies as research users

research institutions low capacitylow weak incentives weak poor information other HEIs and training providers

earlier education

Higher education institutions (HEIs) Causes of disconnects Disconnected actors

BOX 3.1 Poor diversifi cation in Cambodian higher education

Nearly 60 percent of all tertiary students study estimates of current or future demand for university business, social sciences, or law, but fewer than 25 graduates are lacking. In addition, there is no reliable percent are in agriculture, education, engineering, tracking of the employment outcomes of recent uni- health, or the hard sciences, even though many of versity graduates. these latter sectors provide—and will continue to The mismatch between graduate supply and provide—the most jobs. For example, fewer than demand has serious economic and social implica- 4 percent of students enroll in agriculture, even tions for Cambodia, such as high structural unem- though the sector accounts for 29 percent of the ployment, with a pool of university graduates seeking country’s gross domestic product (GDP) and sup- jobs but without the skills demanded by employers. ports the livelihood of 59 percent of the population. The mismatch can also constrain productivity and What accounts for this extreme polarization? economic growth because employers fill positions Recent evidence suggests that most students select with poorly qualifi ed (and less productive) workers or their subject based on the advice and wishes—and, to scale back their growth ambitions. It can also stifl e a a lesser extent, interests—of their family rather than country’s attempts to diversify its sources of growth. on their own future labor market prospects, opportu- Cambodia can take some immediate steps in fi ll- nities, and outcomes. A 2008 survey on youth employ- ing the information gaps. The government and uni- ment reported that only one in fi ve graduating second- versities could undertake tracer studies of university ary students based their decision on what to study on graduates and follow their employment experience. the job market; 7 of 10 followed their parents’ advice. The results would provide information on the demand In the same survey, only one of three university and for skills for all jobs and those requiring higher skills vocational training students chose their fi eld of study and then signal overall graduate unemployment levels because of market need. (Among university and voca- and supply-demand mismatches in particular fi elds. tional student respondents, 59 percent chose their Additionally, universities could review their course fi eld because of an interest in the subject.a) curricula and teaching practices to better equip stu- Lack of information seems to be one reason for dents with the types of skills employers demand. these decisions, and students know little about demand for higher-level skills. No labor market infor- Source: HRINC 2010. mation systems survey labor market demand, and a. BDLINK Cambodia Co. 2008.

often because teaching methods and prac- speak for the majority of classroom sessions, tices are outdated. In Cambodia one survey with little student interaction, teamwork, found that teachers usually adopt a “teacher- or time for problem solving.1 Similar fron- centered” approach to instruction. Teachers tal, discipline-based practices also abound DISCONNECTS IN HIGHER EDUCATION 69

in China, Indonesia, Japan, Malaysia, and appear to have a much more even distribu- Vietnam.2 This approach is usually ineffec- tion of students across disciplines. Engineer- tive in higher education, especially compared ing and manufacturing, on the one hand, with a “student-centered” approach that and social sciences, business, and law, on encourages student interaction. the other, are the two most popular groups Data on fi elds of study indicate that most of disciplines in these economies, but nei- tertiary systems in the region’s low- and ther accounts for more than 40 percent of middle-income countries have an uneven tertiary students. Moreover, the difference distribution of students across disciplines. in shares of students enrolled in these fi elds Cambodia (box 3.1), the Lao People’s Demo- and other disciplines is not nearly as large cratic Republic, and Indonesia notably have as in the low- and middle-income countries an extremely large share of tertiary students (fi gure 3.2). pursuing degrees in (a) social sciences, busi- This lack of diversification has implica- ness, and law or (b) humanities, in general— tions for the responsiveness of their educa- at around 70 percent, 60 percent, and 50 tion system to new labor market demands. percent, respectively (fi gure 3.2). Far fewer Low enrollment in science, technology, engi- students are in other fi elds. neering, and math (STEM) fi elds is already High-income economies such as Japan, a serious constraint for manufacturing in the Republic of Korea, and Singapore Cambodia and Mongolia, as illustrated by

FIGURE 3.2 Proportion of tertiary student enrollments by fi eld of study, 2008

30 percent

0 Cambodia Lao PDR Vietnam Mongolia Indonesia Thailand Malaysia Korea, Rep. Singapore Japan

low-income countries middle-income countries high-income countries

education humanities and arts social sciences, business, and law science engineering and manufacturing agriculture health services

Source: UIS (UNESCO Institute for Statistics) Data Centre. 70 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

the current quantity gaps. While wage skill sufficient skills. Even countries with an premiums continue to rise in these countries economic structure more focused on ser- in manufacturing, no apparent increase in vices show gaps between education supply STEM fi elds is visible (fi gure 3.3), illustrat- and labor market demand. Policy makers in ing systemic rigidities. And in these and other the Philippines, for example, know that the countries this disconnect underestimates the courses of its higher education system have true needs for STEM skills in the economy not kept track with changes in the fi nancial for innovation purposes. services industry.3 In a country with limited course offerings, such as Cambodia, many students can- Misalignment of institutional mix not obtain jobs because of a lack of opportunity to study certain fi elds and thus gain Higher education systems are fairly varied in East Asia, with a mix of colleges, universi- FIGURE 3.3 Science and engineering enrollment ties, and technical and vocational education shares and wage education premiums in and training (TVET) institutions provid- manufacturing, Cambodia and Mongolia ing professional certificates and diplomas through to bachelor’s and postgraduate a. Cambodia degrees. All countries have public and pri- 140 vate institutions, and several have nontraditional delivery, such as distance higher 120 education. (Appendix A illustrates the mix 100 in each country.) Evidence suggests that the institution- 80 degree mix is misaligned with labor market needs in several countries, with implica- percent 60 tions for the quantity and quality of tertiary 40 education graduates. TVET delivery is, for instance, still suboptimal or subpar in some 20 countries. And Cambodia,4 Mongolia,5 and 6 0 Vietnam, for instance, have put very limited 2001 2002 2003 2004 2005 2006 2007 2008 focus on college (as opposed to university) b. Mongolia degrees (though college degrees may provide 70 valuable intermediate skills relevant to manu- 7 60 facturing and services). Most countries also have still relatively low shares of postgradu- 50 ate students. 40 TVET/non-TVET percent 30 Often difficult employment prospects for 20 graduates and relatively low tertiary completion rates have made the TVET subsec- 10 tor an attractive alternative to university 0 education in many low- and middle-income 1998 2000 2002 2004 2006 2008 countries. Historically, East Asian parents science and engineering enrollment share and youth have tended to prefer academic tertiary education wage premium in manufacturing training, leading to unemployed university graduates alongside unfilled positions in linear (tertiary education wage premium in manufacturing) basic trades. Governments are shifting a greater propor- Sources: di Gropello and Sakellariou 2010; UIS Data Centre. tion of those seeking tertiary education into DISCONNECTS IN HIGHER EDUCATION 71

TVET, notably in middle-income countries. FIGURE 3.4 Share of upper-secondary and tertiary students In China 40 percent of upper-secondary stu- enrolled in TVET dents and nearly 50 percent of tertiary students are enrolled in TVET (figure 3.4). In Malaysia nearly 60 percent of tertiary students Korea, Rep. high- are in TVET institutions. Nearly a third of income countries Thailand’s upper-secondary students pursue Malaysia TVET qualiﬁ cations, and almost 20 percent of tertiary students do. Indonesia has recently Thailand considered having 70 percent of its upper- secondary students in TVET programs. China Tertiary TVET has several advantages. It shapes the skill set of the next generation countries Indonesia

of workers, provides students with read- middle-income ily employable skills, and can help make up Philippines student undersupply in some fi elds. A recent World Bank study on the Philippines, for Mongolia example, found that postsecondary TVET institutions have higher labor market rele- Vietnam vance and adaptability (fl exibility and responsiveness to changes) than universities. Similar countries Cambodia results hold in Indonesia (fi gure 3.5). low-income It is thus unclear why countries such as 0 10 20 30 40 50 60 Cambodia and Mongolia, which need more percent practical skills, have such low shares of upper secondary tertiary TVET tertiary enrollment, especially because Mongolia, at least, is happy to accept TVET Source: Yilmaz 2009. workers (as seen in the employment and Note: The Philippines includes postsecondary students. remuneration analysis shown in chapter 1). Still, skill-delivery gaps remain in TVET. Postsecondary TVET graduates in the used for students’ stipends, and less than Philippines, for example, lack relevant cer- 1 percent of that TVET budget was used tifi cations in some technologically advanced for developing infrastructure and acquir- fi elds, are of varying quality, and often need ing training equipment, both of which are retraining.8 Employers there, while acknowl- severely inadequate.10 These fi ndings suggest edging higher overall relevance to labor an urgent need to improve the quality and market needs, highlighted the limitations of relevance of current TVET education. postsecondary TVET education in several Given the strong social demand for higher fields of study and the quality of teaching education, coupled with skill shortages across (see fi gure 3.5). In Indonesia employers felt a spectrum of sectors, countries could also do the variety of fi elds of study to be less strong better at realigning their higher education than in universities (see figure 3.5). Also institutions in relation to their core mission in Indonesia those possessing a technical and the length of degrees provided with the diploma lacked some generic skills.9 needs of the skill users. In Mongolia, as may be the case elsewhere, ineffective use of resources and out- Colleges/universities dated equipment undermine the quality Second-tier universities or community col- of TVET. Public spending on TVET was leges have different but equally important roughly 6 percent of total public spending on functions, which are to serve commu- education. Moreover, nearly half of that was nity and local development and develop a 72 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 3.5 Employers’ perceptions on general and TVET tertiary education, Indonesia and the Philippines, 2008

a. Indonesia b. Philippines 100 14 90 12 80 70 10 60 8 50 40 6 by strength 30 by weakness 4 20 10 2

share of university institutions 0 share of university institutions 0

cost

cycle length teacher skills industry link fields of studyLM relevance facilities quality LM adaptability length of studiesteaching quality research capacity quality of teaching quality ofresearch facilities capability links with industry teacher qualification relevance to LM needs variety of fields of study

adaptability to changing LM needs

c. Indonesia d. Philippines 16 100 14 90 80 12 70 10 60 8 50 40 6 30 4 20

institutions by strength 2 10 institutions by weakness share of postsecondary TVET 0 share of postsecondary TVET 0

cost

adaptability to changing LM needs

Sources: (a), (c) di Gropello, Kruse, and Tandon 2011; (b), (d) di Gropello, Tan, and Tandon 2010.

solid basis of some core generic skills and Postgraduate degrees technical skills in some subjects. Mongo- At the other end of the educational spec- lia, however, has only 11 schools offering trum, the share of students enrolled in preliminary and intermediate training for postgraduate studies may also not yet be technical skills.11 This number is probably suffi cient in some East Asian countries. In insuffi cient to fulfi ll the needs of its labor 2000 Asia—principally East Asia—already market, given current gaps in English and had a large lead over other developing even information technology that could regions in the number of doctoral degrees be well provided in the context of shorter awarded overall and probably in science generic degrees.12 and engineering disciplines (see appendix DISCONNECTS IN HIGHER EDUCATION 73

TABLE 3.1 Earned science and engineering doctoral degrees, selected region or economy and selected fi eld, 2006 or most recent year

Share of science and engineering (%) Physical and All science and biological Mathematics and Region or country All fi elds engineering sciences computer sciences Engineering Asia 85,441 44,552 35 1 46 China 36,247 22,953 32 — 53 India (2005) 17,898 7,537 74 — 13 Japan 17,396 8,122 20 — 52 Philippines (2004) 1,748 56 0 23 13 Korea, Rep. 8,657 3,779 22 5 60 Taiwan, China 2,614 1,643 19 11 57 Middle East 5,759 2,902 45 8 20 Sub-Saharan Africa 2,064 679 37 0 21 Europe 97,840 53,119 44 8 26 United States 50,544 31,198 39 7 22

Source: National Science Board 2008. — = not available.

L). Since 2000 the margin has widened FIGURE 3.6 Proportion of tertiary students enrolled in considerably in East Asia’s favor mainly ISCED 6 programs because of a surge of degrees in China. The number of doctorates remains small, however, in relation to the stock in Europe Korea, Rep. 1.2 high- income (table 3.1 and appendix L), and almost all country countries in the region have fewer than 2 Malaysia 1.5 percent of their tertiary students studying Thailand 0.3 at the International Standard Classifi cation of Education (ISCED) 6 level13 (fi gure 3.6), China 0.8 in contrast to about 5 percent in western Indonesia 1.7 Europe and North America.14 countries Indeed, some countries’ higher educa- middle-income Philippines 0.4 tion systems may not be producing suffi cient Mongolia 1.4 postgraduate students. There is no ideal ratio of postgraduate students: Korea, for exam- Vietnam 2.6 low- income ple, has a fairly low ratio of students enrolled country 0 0.5 1.0 1.5 2.0 2.5 3.0 in advanced degrees and still manages to percent achieve enviable innovation outcomes. Many fi rms may not be interested in hiring work- Source: Yilmaz 2009. ers with the profile of researchers, in particular in middle- and lower-income countries. Still, trends in education premiums Reasons for the disconnect suggest higher education systems may not be producing enough postgraduate students A fairly widespread regional skill disconnect to meet the growing demand in some coun- is apparent between what the current insti- tries (fi gure 3.7). Beyond fi rms’ needs, this tutions deliver and what skill users need. growing demand is seen in the low share of Why? There seem to be three main groups highly qualified faculty in many countries of reasons, related to information, capacity, (discussed further below). and incentives. 74 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 3.7 Wage education premiums, Indonesia, the Philippines, themselves are regularly invited to visit cam- and Thailand pus, to provide input.15 Still, despite a greater emphasis on formalizing such links in recent years—not just for research but also for skill 60 development and training as in the Philip- 50 pines and Vietnam—they have remained mostly informal and piecemeal. 40 Overall, information on graduate employment, labor markets, and skills is still weak 30 in the region (as seen for Cambodia earlier),

percent contributing to lack of relevance of curricula 20 and pedagogies. Institutions do not systematically attempt to gather recent graduates’ feed- 10 back about the workplace relevance of their courses and training programs, which would 0 Thailand Philippines Indonesia allow those institutions to make changes in 1996 1998 2000 2006 2009 curricula and programs. In particular, only a few countries have effectively carried out and used graduate tracer studies.16 Source: Sakellariou 2010b. Note: Postgraduate in relation to undergraduate; all results signifi cant at 5 percent or 1 percent. Nor do labor force and enterprise surveys fulfi ll their potential. While labor force surveys are conducted periodically in most coun- Poor information tries and provide data on that area, including HEIs may be unable to provide the skills educational attainment, employment, and that respond better to labor markets’ needs unemployment, they provide very limited because they lack information on demand. information on the tertiary sector and its rel- Instruments to provide institutions (and stu- evance to the labor market, because most do dents) with labor market information are not include information on public or private virtually absent, and mechanisms to chan- education and type of training program.17 nel inputs of fi rms in curriculum design and Equally, enterprise (or census) surveys do not implementation are limited. collect information on skill levels; nor do they Low- and middle-income East Asian ask employers about the types of programs countries have weak links between industry and skills that are more relevant to their busi- and higher education in curriculum design. ness needs or employees about the skills they Many countries recognize the importance of most need or lack in their jobs. strong links, but few have set up the neces- Lack of university-industry links and sary mechanisms. High-income economies in other sources of information refl ects coun- East Asia and elsewhere tend to have stronger try-specifi c conditions and legal frameworks, university-industry links. (This is also visible but it also more widely relates to lack of from fi gure 3.10, which captures university- capacity and incentives, the principal cause industry links in research and technology.) for the broader disconnect between HEIs These links are often the fruits of collabo- and skill users. ration to ensure that pedagogy and curricula are aligned with what firms need. In New Low capacity Zealand, for example, the Tertiary Education Pervasive lack of human capacity in higher Commission requires HEIs to work closely education makes it hard to respond to with local businesses and industry training labor market demand. The lack of quali- organizations to identify current and future fied human resources has widespread skill needs. In Korea all polytechnics have implications for the relevance and quality to undertake regular visits to companies to of higher education, all the way from cur- keep their curricula updated, and companies riculum design to teaching and to research, DISCONNECTS IN HIGHER EDUCATION 75

also affecting the quality and quantity of TABLE 3.2 Student-to-faculty ratios in tertiary university-industry links. education, 2007 Academic faculty has a critical role in skill Economy Student-to-faculty ratio provision. First, they train future primary, secondary, and tertiary teachers who in turn Cambodia 23:1 shape the quality and relevance of the entire China 19:1 Indonesia 15:1 national education system. Second, they pro- Korea, Rep. 16:1 vide skills to future high-level research, tech- Lao PDR 25:1 nical, managerial, and administrative person- Malaysia 20:1 nel who will lead government, business, and Mongolia 29:1 industry. Third, they are key incubators of the Philippines 23:1 innovation and creativity that will enhance Singapore 13:1 national productivity and competitiveness.18 Thailand 37:1 Lower- and middle-income East Asia are Vietnam 30:1 suffering from two main faculty-related Sources: Chapman 2010; UIS Data Centre. constraints: higher and growing student-to- faculty ratios, and a low share of faculty with FIGURE 3.8 Trends in student-to-faculty ratios, 2001–07 graduate degrees. The importance of these two challenges varies across countries. As a result of the past decade’s rapid 40 expansion of East Asia’s tertiary enrollments, which grew far faster than faculty numbers, 35 student-to-faculty ratios are high in lower- 30 and middle-income East Asia. Most low- and middle-income countries in East Asia have 25 much higher student-to-faculty ratios than 20 the Organisation for Economic Co-operation ratio and Development (OECD) average of 15 to 1 15 19 (table 3.2), having climbed or stayed rela- 10 tively constant in most countries (except for the Philippines; fi gure 3.8). High student-to- 5 faculty ratios affect teaching by leading to 0 disproportionally high teaching loads and to 2001 2002 2003 2004 2005 2006 2007 less time for personal interaction with stu- Cambodia Lao PDR Vietnam Mongolia dents and for professional development. They also affect the level of curriculum and degree Philippines Indonesia China Thailand diversifi cation (particularly the possibility of Malaysia Korea, Rep. Japan offering college education and STEM fi elds, which require smaller classes). Indirectly they Sources: Chapman 2010; UIS Data Centre. also affect research by decreasing the time left for research. As student-to-faculty ratios have risen, very few had doctorates (table 3.3 and fi gure many universities and institutions have hired 3.9). Worryingly, the share of faculty with lower shares of faculty with graduate degrees. doctorates has decreased since. About 53 This second constraint affects (a) the capacity percent of the faculty in Indonesia lacks mas- of universities to offer postgraduate degrees; ter’s degrees, as does 60 percent in the Phil- (b) teaching, to the extent that a master’s ippines.21 This proportion has been decreas- degree is correlated with teaching;20 and (c) ing in the Philippines. This is in contrast research (more on this below). In Vietnam, with Korea; Mongolia; Taiwan, China; and for example, only 46 percent of academic Thailand, which have more than 70 percent staff in 2005 had graduate qualifi cations, and of faculty with at least a master’s degree. 76 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

TABLE 3.3 Academic qualifi cations of faculty in a sample of East Asian economies percent

Total Public HEIs Private HEIs Share with Share with Share with Share with master’s Share with master’s Share with master’s Economy PhD degree PhD degree PhD degree Low-income economies Cambodia (2008/09) 8 52 8 49 7 55 Vietnam (2005/06) 14 32 13 32 23 32 Vietnam (2008/09) 10 37 — — — — Middle-income economies Mongolia (2004/05) 20 65 — — — — Mongolia (2007/08) 20 66 23 65 15 68 Philippines (2009/10) 10 36 13 54 7 29 Indonesia (2007) 7 40 — — — — Thailand (2005) — — 26 59 16 66 High-income economies Korea, Rep. (1994) 59 24 — — — — Taiwan, China (1992) 46 26 — — — — Japan (2005) 41 17 — — — —

Sources: Cambodia: HRINC 2010; Indonesia: World Bank Indonesia Higher Education Sector Assessment 2009; Japan: Newby and others 2009; Mongolia: National Statistical Offi ce data; Philippines: Commission on Higher Education 2010 data; Taiwan, China, and Republic of Korea: China Higher Education Reform 1997/2009; Thailand: Commission on Higher Education 2008 data and World Bank 2009a; Vietnam: MOET 2005, 2010. — = not available.

FIGURE 3.9 Ratios of faculty with master’s degrees and PhDs, other, middle-income countries face a fac- various years ulty quantity-quality trade-off affecting their capacity to deliver high-quality teaching in a 100 diversiﬁ ed institutional setting. Lower- income countries face a strong double quantity-quality 80 challenge (particularly Vietnam).

60 Weak incentives Even when sufﬁ cient capacity exists, lack

percent 40 of incentives for public institutions to produce graduates with the skills needed by 20 ﬁ rms may ultimately hamper all attempts to improve education relevance. In general, 0 private institutions have more incentives Japan Vietnam to produce skills relevant to ﬁ rm needs to Mongolia Indonesia Thailand Cambodia Philippines Korea, Rep. Taiwan, China attract and keep students, whereas public faculty with PhD faculty with master’s degree institutions have fewer, stemming from lack of clear accountability for results either to the government or to other stakeholders. Source: Authors’ extrapolation from table 3.3. Public institutions do not therefore focus on the outcomes of the education process. They are often “protected” by supportive In Mongolia and Thailand, high propor- financing policies that shield them from tions of faculty with a master’s degree have, having to show results or compete against however, been achieved at the cost of a very other institutions. And within public insti- high student-to-faculty ratio. One way or the tutions, even highly qualified faculty may DISCONNECTS IN HIGHER EDUCATION 77

not deliver on skills if not held adequately FIGURE 3.10 Intensity of university-industry links accountable to parents and students. Lim- ited autonomy, too, in academic and admin- a. East Asia istrative areas, generates another disincentive to tailor programs to the needs of the Singapore local community. Korea, Rep. Japan economies high-income The second disconnect: Between Hong Kong SAR, China higher education and companies Thailand (research users) Philippines Mongolia Engagement is weak. The limited contribu-

economies Malaysia tion of universities to technology adaptation middle-income and upgrading in companies is suggestive of Indonesia a disconnect between companies and uni- China versities in research and technology mostly Vietnam low- applicable to lower- and middle-income income Cambodia economies countries but also to some upper-income 01234567 ones. Figure 3.10 provides a snapshot of index university-industry links in East Asia. The b. Comparators data include both skill-related and research- United States related links and, as such, cannot confer the full picture of university-industry links in United Kingdom economies research and technology. In particular, in high-income Finland the ﬁ gure, were one to consider only research South Africa and technology, the links would be lower in Pakistan Korea and Malaysia, where skill-related links Mexico are fairly developed. By contrast, were one to India consider only research and technology, they economies middle-income would be higher in Japan. Chile Even so, the picture that emerges is one Brazil where high-income economies in East Asia Nepal and elsewhere, China, and to some extent Ghana

Malaysia have stronger university-industry economies low-income Bangladesh links than lower- and middle-income coun- 01234567 tries. Links in several middle-income coun- index tries of the region are also weaker than in their peers elsewhere, such as Brazil, Source: KAM, 2009 data. Chile, and South Africa. The low intensity of links in Mongolia is aligned with its position in the lower technology cluster together universities and research institutes as a “not with Cambodia and Vietnam. unimportant” source of information.23 Firms In Thailand recent innovation surveys con- in petroleum, fabricated metal, telecommu- fi rm that fi rms have very limited cooperation nications, and computer industries view uni- with universities in research and development versities as important partners in adapting or (R&D) activities (fi gure 3.11). refining and troubleshooting technologies. However, some industries in Thailand This is understandable, given that many of show some interaction.22 Firms in food, these subsectors are technology intensive. textiles, printing, synthetic rubber and plas- Low engagement is also confirmed in tics, telecommunications, and R&D regard Indonesia and the Philippines.24 At most, 78 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 3.11 External collaboration for R&D activities, Thailand, 1999–2003

other ﬁrms other government agencies competitors technical service providers business service providers universities private nonproﬁt public research institute parent/associate companies foreign-owned suppliers locally owned suppliers customers buyers 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 index 1999 2001 2003

Source: Doner, Intarakumnerd, and Ritchie 2010. Note: In 1999 and 2001, research institutes and universities are in the same category. They are separate from 2003. Scores: 0 = unknown, 1 = not at all, 5 = very intensely.

faculty members in these two countries have programs) is more widespread, as it is in informal relationships with firms, mainly Japan, for instance.25 consulting and providing occasional techni- Finally, Korean universities, which have cal assistance and training. proven to be good trainers, are much less of a Collaboration is somewhat stronger in fertile source of ideas for established fi rms or Malaysia (where, according to the 2007 for spin-offs.26 Investment Climate Survey, about 13 percent Some economies are, though, work- of fi rms have declared to have looked for col- ing more closely together. The authori- laboration with universities on technological ties in Taiwan, China, for instance, have innovation) and China. launched a program for promoting the But even in China, demand from the excellence of universities and have made enterprise sector is still low. According to university-industry links one of the crite- a national survey of science and technology ria for evaluating universities (and their activities by the National Bureau of Statis- faculty and students).27 Small and medium tics, only about a sixth of large and medium enterprises in Taiwan, China, collaborate manufacturing enterprises had any form of more with public universities than private collaboration with Chinese universities dur- ones and are gradually shifting the focus of ing 2000–02, and less than 3 percent of pat- collaboration from process innovation to ents are jointly fi led by fi rms and universi- product innovation, suggesting an increase ties. Among private fi rms, only about half in technological capabilities of these enter- of surveyed firms collaborated with uni- prises and the economy’s technological versities formally in 2007, though informal maturity. University-industry links are also collaboration (sharing of equipment, labo- both stronger and on the rise in Japan and ratory space, and development of training Singapore.28 DISCONNECTS IN HIGHER EDUCATION 79

Similarly, since 2001, and in line with its Lack of capacity will hamper both collabo- position as leader of the middle technology ration itself and the results of this collabora- cluster, the Chinese government has encour- tion. Some of these capacity issues relate to aged closer university-industry collabora- the lack of skills of higher education gradu- tion, making the core missions of research ates who stay in university (becoming fac- universities teaching, research, and commer- ulty) or move to fi rms. In addition, faculty cialization of technology.29 Because of this who lack business acumen and entrepreneur- push, patenting by universities has increased ial talent may be reluctant to enter the busi- dramatically, and patents granted to univer- ness world, even if they have made promising sities now account for close to 30 percent of fi ndings and generated new ideas that can be all patent grants in China.30 Chinese uni- commercialized. And lack of creativity and versities are beginning to work with indus- entrepreneurial skills on the fi rm side will be try through contracts for technology services clearly unsupportive of innovative collabora- (particularly for technology transfer), patent tion. Capacity constraints are likely to apply licensing and sales, and university-affi liated even more strongly to basic research. enterprises. Revenues from these activities A 2009 survey of a small sample of Indo- account for a quarter to a third of universi- nesian fi rms identifi ed poor technical quality ties’ research budgets. State-owned enter- of domestic graduates as a key constraint, prises have been the most frequent users of along with lack of awareness of technologi- universities’ technology services, because cal development abroad (most likely because these enterprises and universities are public of poor foreign-language skills). Firms also entities, making such collaboration easier. pointed to the lack of original, commercially Still, universities’ technology links with pri- relevant research originating in universities vate firms have increased significantly in and the lack of a national or regional research recent years and now account for 40 percent hub within the country.34 of the total. (Foreign fi rms rarely collaborate Similarly, fi rms in Thailand’s sugar indus- with domestic universities because they gen- try were interested in cellulosic ethanol, erally rely on home- or advanced-country given the current concern with energy con- research.) servation, but no Thai university possessed Thailand has separated university staff the expertise to help fi rms with R&D. These from the civil services to increase the fl exibil- fi rms ended up collaborating with Japanese ity of salaries and improve the relationship universities through the intermediation of a with firms through the Education Reform Japanese trading company.35 Over time, a Act of 1999.31 In the Higher Education range of factors—weak academic capacity, Development Project, launched in 2001, the mismatched supply and demand between government identifi ed seven areas of focus to universities and multinational corporations, improve research capabilities,32 as more tech- and a political system marked by fragmented nology-intensive industries, such as engineer- bureaucracies—have diminished the cred- ing and the electronic and electrical sectors, ibility of universities as a viable innovation capture larger shares of exports. partner for fi rms in Thailand.36 Elsewhere, an electronics firm in the Reasons for the disconnect Philippines identifi ed the lack of qualifi ed personnel for research and of facilities and Again, problems related to information, tools for simulations and modeling in a par- capacity, and incentives are reasons for the ticular field as the major impediments to disconnect. Taking capacity fi rst, a critical innovation.37 issue relates to the capacity of universities to Clearly, capacity to undertake research undertake meaningful research and capac- and support technological development in ity of fi rms not only to identify but also to fi rms is constrained by low ratios of faculty use the knowledge available at universities.33 with doctorates. Table 3.3 and fi gure 3.9 80 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

illustrate that the low ratio of faculty with and on the kinds of industries and firms doctorates in lower- and middle-income that can more successfully interact with East Asia is in sharp contrast to the more universities. than 40 percent of faculty with PhDs in As an illustration, among other factors, of Japan, Korea, and Taiwan, China. Coun- information failures, according to a survey of tries with higher PhD ratios among their manufacturing fi rms in Beijing, key barriers faculty also have higher researcher ratios to collaboration some years ago included lack (take Japan and Korea), translating into of efficient communication channels with higher yearly scientifi c and technical pub- universities, uncertainty of market perspec- lications (fi gure 3.12). While not the best tive for research results, high cost to com- proxy for technological support to fi rms, mercialize results of research, and immature this indicator nonetheless captures some technology from academic research.38 Private potential for it. Other elements, however, enterprises in Shanghai have reported sev- play an equally or even more important role eral major problems (ranked by propor- in constraining university-industry links in tion of responses): university R&D lagging research and technology. behind market trends (22.5 percent), high Information failures are multiple and can costs associated with outsourcing to univer- be related to firms’ lack of knowledge of sity (16.6 percent), lack of communication existing technologies (inside or outside the channels (13.8 percent), diffi culty reaching country) generally and university offerings mutually agreeable profi t-sharing schemes particularly. By contrast, universities may (9 percent), and immature technology and not be fully aware of industry’s R&D and lack of marketability of academic research technology needs. More crucially, there is a (7.7 percent).39 Because Beijing and Shang- lack of information among both fi rms and hai have the most active university-industry universities as well as governments on what links, these results are likely representative types of university-industry links are most of fi rms across the country. effective and in which economic settings, on Finally, without adequate incentives, uni- the pros and cons of different approaches, versities and faculty may not be interested or even able to relate to enterprises.40 On the fi rm side, lack of incentives to innovate FIGURE 3.12 Number of scientifi c and technical journal articles will also lead to a lack of links. In fact, risk- and number of researchers in R&D averse firms with little exposure to new technologies and competition and a lack of technically skilled labor are often unwilling 1,400 to bear the risks of or put up the fi nancing 41 1,200 to bring new technologies to market. As an illustration of weak incentives 1,000 for university-industry links, research in SGP Vietnam has found that the weak nature 800 of the relationships between universities and fi rms could be attributed to four con- 600 straints: a credibility gap between industry KOR JPN

per million people, 2007 per million people, 400 and academia from both sides; bureaucratic regulations and attitudes not conducive to scientific and technical journal articlesscientific and technical 200 MYS innovative partnerships and links; insuf- VNM THA LAO CHN fi cient understanding of intellectual prop- - KHM PHL IDN 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 erty rights and related matters, possibly researchers in R&D per million people, latest year (2001–08) constraining partnership efforts; and inadequate incentive structures and financial Source: WDI (World Development Indicators) database. support programs.42 DISCONNECTS IN HIGHER EDUCATION 81

The third disconnect: Between to bolster research output and capacity. higher education and research For instance, the national congress created institutions (research providers) the Commission on Science, Technology, and Engineering in 2007 to study the state Teaching and research have close comple- of the innovation system and recommend mentarities. As chapter 1 illustrated, several strategies for improving science and tech- authors have stated that there needs to be a nology capabilities. The commission called link between tertiary education and research, for the creation of 20 councils, institutes, while some have argued that the two ele- and administrative units outside the uni- ments are intrinsically intertwined, because versity system to pursue research in priority active researchers can provide high-quality sectors. Yet the older trend continues. The teaching, and interactions between research- Philippine Atmospheric, Geophysical, and ers and students help improve research. Astronomical Services Administration, Whether students come into contact with the country’s climate forecasting agency; research, the role that research plays in the the National Academy of Science; and the academic program, or the extent to which Philippine Council for Industry and Energy the most recent developments in research Research and Development are all adminis- are reflected in the curriculum, suggests tered as governmental departments, rather economies of scope and, to some extent, than as university partners. And although scale between teaching and research at the awards and research grants to these insti- tertiary level. tutions increased from ₱247.2 million in Most lower- and middle-income East 2006 to ₱639.6 million in 2008,46 none of Asian economies, however, show a wide dis- the funds went to universities. connect between teaching and research.43 Universities in Malaysia and Thailand Universities in many countries remain conduct more research than those in Cam- the domain of teaching, whereas research bodia and the Philippines, but research insti- institutes, the government, and the private tutes still conduct most research. In Malaysia sector conduct practically all the research. only 17 of the 254 R&D agencies are within Links between these two communities are universities,47 spending US$94.8 million in weak, if they exist. In a given country, not 2005, compared with private sector entities’ all universities need to undertake research, US$439 million. of course, but teaching-research syner- In Thailand HEIs account for about 30 gies should be exploited in at least a few percent of total research expenditure, com- institutions. pared with 23 percent in government and 45 Country examples of the disconnect. Uni- percent in the private sector. Most research versities in Cambodia, for example, conduct conducted at HEIs is short-term applied very little research, and their connections research, as opposed to the research that with research institutes are weak. A recent is undertaken by the private sector and by study noted that research capacity is at a research institutes, which provide 89 percent very early stage of development because of of the experimental research undertaken in the inadequacy of research budgets, qualiﬁ ed the country.48 Universities engage in research researchers, and supporting infrastructure projects mainly as part of consulting services such as research facilities or laboratories.44 and technical or analytical services, because Research and technical institutes conduct public funding levels are too low to fund cap- most of the country’s research. ital investment. This keeps down the num- Teaching and research functions are often ber of academics committed to long-term kept separate in the Philippines: universities research: only 20 percent of Thai academics and colleges have traditionally focused on conduct research continuously.49 teaching, and other institutions pursued Even in China, where several universities research.45 Recent efforts have been made have emerged as centers of excellence for 82 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

both teaching and research in recent years, developed according to the Humboldtian research institutes and government bodies vision of the unity of research and teach- still conduct most research, leaving teaching, emphasize blending these two elements ing to HEIs. Such separation is rooted in the across all disciplines and require that teach- country’s recent history. Between 1949 and ing be informed by research.53 The OECD 1979, the science and technology system fol- has noted that even Eastern European coun- lowed a model under which research, includ- tries, in which there was a clear separation ing all innovation activities, was conducted between teaching and research characteristic by research institutes, manufactured by fac- of socialist regimes, have been restructured tories, and distributed by distributors. A mul- posttransition according to the Humbold- titude of central ministries coordinated these tian vision.54 Closer to Asia, New Zealand’s elements, creating a vertically rather than parliament has passed a law that only those horizontally integrated system dependent on actively involved in research can teach at centralized, top-down allocations for neces- degree level. sary inputs.50 In recent years, however, links between Reasons for the disconnect HEIs and those conducting research have strengthened in China. National programs The overall approach to managing teaching speciﬁ cally designed to elevate the importance and research is partly at fault. Most coun- of research in higher education have been tries have been adopting a model of separa- implemented since 2002. Project 985 aims to tion between teaching and research that has turn China’s top universities into world-class not been conducive to enhancing research research universities by introducing elements in universities. This approach has further of competition for resources. Competition for debilitated the already weak capacity of “985” designation is ﬁ erce because selected universities to undertake research, while institutions receive substantial central gov- there are few incentives for universities and ernment funding, with matching funds from research institutions to collaborate on, say, provincial governments. common projects, with the consequence that But Chinese universities have yet to useful synergies between research at HEIs become key drivers of innovation, particu- and research institutes are not exploited.55 larly relative to public research institutes, Beyond constraints in resources, the lack of indicating that separation between teaching incentives to pursue research in universities and research functions is still prominent. (alone or in collaboration with research insti- Universities have consistently spent less than tutes) is thus also to blame. And even more public research institutes, accounting for so is the lack of incentives to pursue mean- less than 10 percent of total R&D expendi- ingful and productive research (resources tures from 1997 to 2006.51 In addition, with need to be well used). expanding corporate R&D, higher education’s share in both national R&D expen- The fourth disconnect: Among diture and personnel has in fact trended higher education institutions downward in recent years, accounting for themselves and between 8.5 percent and 14.6 percent, respectively, in these institutions and training 2007. Moreover, interaction between public providers (horizontal disconnect research institutes and universities largely across skill providers) focuses on recruiting university graduates.52 Non-Asian trends. East Asia’s divide HEIs themselves (and those institutions with between teaching and research functions other skill providers at about the same level contrasts with trends elsewhere. The Ger- or postemployment level) are not well con- man and Swedish higher education systems, nected to each other in much of lower- and DISCONNECTS IN HIGHER EDUCATION 83

middle-income East Asia, and this has led commonly referred to as a national quali- to weak skill transmission to graduates. fi cations framework, can help provide sub- Indeed, because the links and complemen- stantial links among institutions. When tarities among institutions have generally well developed, it can provide more effec- not been well understood or built upon by tive overall governance to address fragmen- institutions and governments, the develop- tation and ensure that students and work- ment of higher-level skills and competencies ers can move horizontally and vertically has been inhibited. Although some encour- between education and training levels58 and aging steps have been taken, the connection the formal/nonformal education and train- between different HEIs and these institu- ing system, through a strengthened skills tions and, for example, fi rm training, on the certifi cation. one hand, and the development of different National qualifications frameworks are skill acquisition pathways, on the other, has at an incipient stage in lower- and middle- not been well or coherently managed across income East Asia59 (and not fully developed the region. And it is one of the reasons that also in some upper-income economies), and the skill supply system has been unable to the underlying concept has yet to take hold: ensure more flexible, efficient, and even instead of a more modern focus on outputs equitable skill acquisition. and adherence to a market-oriented policy The situation in Mongolia is illustra- agenda, the skill provision systems in East tive of fragmentation between TVET and Asia remain focused on inputs. Instead of non-TVET providers. Despite the demand stressing competencies acquired, East Asian from employers for vocational skills, public education systems emphasize inputs and the investment in TVET is limited, and links institutions that teach the skills. Job mobil- between universities and TVET skill pro- ity and thus labor market effi ciency have suf- viders are scarce. Facilities and machinery fered. Malaysia is one step ahead of other are often outdated and faulty, faculty rarely lower- and middle-income countries in this share their expertise with TVET institu- dimension. tions, the different skill providers’ curricula Fragmentation among skill providers has are not aligned, transfers between university also arisen from a lack of focus on concrete and TVET tracks are not easy, and private competencies. Indeed, most countries in the providers are not part of a regulatory and region have no competence-based train- monitoring framework.56 ing systems, which would shift the empha- Thailand is similar: students have little sis from what courses a trainee or student mobility across institutions because of a has taken and when to what the trainee or lack of formal links among different skill student can do. Such systems focus on the providers. Curricula between university and skills needed for performance in a job and vocational tracks are, for instance, rarely put pressure on instructors and institutions aligned, making it difficult for students to deliver these skills. They are usually mod- to pursue multiple paths to skill acquisi- ular and, in theory, facilitate fl exible entry tion. And the story repeats itself across the and exit and recognize different routes for region.57 skill acquisition. Lack of incentives to focus on the skills that matter, lack of links with Reasons for the disconnect industry in curriculum development, and lack of capacity to adopt a competency-based What has caused this fragmentation among approach may have prevented the develop- skill providers? Part of the underlying ment of such an approach in East Asia. In problem appears to be the lack of a well- turn, the absence of such an approach cre- developed and unifying framework for dif- ates fewer incentives for institutions to relate ferent skill providers. Such a framework, with industry. 84 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

The fi fth disconnect: Between Secondary and tertiary enrollments higher education and earlier Secondary and tertiary enrollments are education (schools) (vertical necessarily related (fi gure 3.13). While the disconnect across skill relationship is not necessarily linear, with providers) Malaysia and Thailand performing better The quantity and quality of earlier educa- than expected,60 countries with higher ter- tion, particularly secondary education, have tiary enrollment ratios tend to have higher a strong infl uence on tertiary education, and secondary enrollment ratios. Similarly, the transition points across education levels are share of workers with tertiary education is critical. But low- and middle-income East highly conditioned by the share of workers Asian countries show a disconnect in this with secondary education (fi gure 3.14). The area, too. This is illustrated by the fact that difference between the share of workers with the outcomes of earlier education levels and tertiary education and the share of workers the transition between levels are constrain- with at least secondary education indicates ing higher education outcomes one way or that the share of workers with secondary another in the different countries, suggest- education has had to grow even faster than ing a lack of comprehensive vision of the the share of workers with tertiary education. education sector. While secondary enrollment is not yet a constraint across low- and middle-income East Asia, it may become one as countries seek to extend coverage. In some countries, such FIGURE 3.13 Secondary and tertiary gross enrollment ratios, as Cambodia and Lao PDR, low secondary 2007–08 enrollment is already a constraint to getting a wider talent pool. While low- and middle-income countries Cambodia have either improved or maintained their secondary completion levels (figure 3.15), Lao PDR a comparison of tertiary enrollment rates with secondary completion rates61 indicates Malaysia that secondary completion may already be a binding constraint for expanding tertiary Thailand education in Cambodia, China, Indonesia, and Thailand. Thus, expansion may need to Indonesia be related to effi ciency improvements in secondary education and in the transition from China secondary to tertiary education.

Philippines Tertiary and earlier academic Hong Kong SAR, China performance Academic performance at the tertiary level Mongolia is closely related to earlier academic performance. Indeed, performance in secondary Japan education has been shown to be a strong 0 20 40 60 80 100 120 predictor of both access and outcomes in gross enrollment ratios tertiary education. Research has shown, secondary tertiary for instance, that students who have taken academically advanced courses in high Source: UIS Data Centre. school are more likely to perform better and DISCONNECTS IN HIGHER EDUCATION 85

FIGURE 3.14 Trends in shares of workers with secondary and tertiary education, four East Asian economies, various years

a. Cambodia b. Vietnam 45 45 40 40 35 35 30 30 25 25 20 20 percent percent 15 15 10 10 5 5 0 0 19972003–04 2007 1992 1998 2004 2006

c. Philippines d. Indonesia 70 60

60 50 50 40 40 30 percent percent 30 20 20

10 10

0 0 1991 1994 1997 2001 2004 2007 1994 1997 1999 2001 2003 2005 2007 skilled (secondary and above) skilled (tertiary and above)

Source: Labor force surveys, various years. complete college than those who have not.62 stronger predictor of college completion than This ﬁ nding is consistent with prior research a student’s socioeconomic status.65 Finally, that found that advanced mathematics and results of international assessments across algebra in the eighth grade are strong pre- 132 countries found a positive correlation dictors of whether students enrolled in a between math and reading scores of 15-year- four-year college and completed a bachelor’s olds and the quality of higher education in degree.63 their countries (based on university rankings, According to one U.S. study, gradua- number of researchers, and number of pat- tion rates for better prepared Latino stu- ents). Indeed, after including GDP per capita dents were on par with white students.64 in the regression model, the study found that And researchers have found that the quality math scores of 15-year-olds were a better of high school education (measured by the predictor of higher education quality than quality of coursework and test scores) was a GDP.66 86 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 3.15 Secondary education completion strong performer in math according to the TIMSS—has been gradually declining, and Thailand has not done much better. The Thailand Philippines is the weakest TIMSS performer among the tested Asian countries. In Indo- nesia the significant improvement in Pro- Philippines gramme for International Student Assess- ment (PISA) scores between 2003 and 2006 67 Mongolia in reading and math subjects was partly

countries reversed between 2006 and 2009, under-

middle-income scoring its signiﬁ cant gaps in relation to the Indonesia OECD average in 2009 (figure 3.16). The country also needs to reverse its stagnating TIMSS results. China Second, countries with a higher TIMSS score in math and science have higher STEM Vietnam enrollment shares in tertiary education later (ﬁ gure 3.17). The relation may be even clearer countries

low-income if one considers the quality of these STEM Cambodia skills (as seen in the poor quality of engineers 0 10 20 30 40 50 60 70 in several countries). A similar relation holds percent between TIMSS scores and journal publica- earliest year latest year tions (ﬁ gure 3.18): higher TIMSS scores are associated with more publications. Source: Sakellariou 2010b. While only illustrative, this evidence sug- Note: Data for China are for large cities only, so they are not directly comparable to the other gests that the low or relatively low (in rela- surveys. Years are 2005 for China (earlier year missing); 1998/2006 for Indonesia, Mongolia, the Philippines, Thailand, and Vietnam; and 1999/2007 for Cambodia. Figures for China and Thailand tion to high-income East Asia) TIMSS scores are underestimated. of Indonesia, Malaysia, the Philippines, and Thailand are constraining these countries from achieving better STEM and innovation A couple of indicators show the same outcomes in tertiary education. interrelation and how it can constrain in East Asia. First, in alignment with the low Tertiary inclusiveness quality of tertiary graduates, lower- and and prior education middle-income East Asian economies show weaknesses in literacy and numeracy skills Tertiary inclusiveness is clearly related to at the lower-secondary stage. The most inclusiveness at earlier education levels. The recent Trends in International Mathematics inequities in access to tertiary education, and Science Study (TIMSS) data (tables 3.4 documented in chapter 2, at least partially and 3.5) highlight gaps of middle-income originate in the lack of access to earlier edu- Asian economies in numeracy and science cation levels (ﬁ gure 3.19). Beyond showing relative to high-income Asian and non-Asian the cumulative effect of dropouts at different economies—and more broadly the inter- levels and grades on the completion of the national average—suggesting continuous secondary cycle, these plots also demonstrate gaps in numerical skills. While the interna- variation in primary and secondary education tional average and the performance of some access and survival rates by rural and urban high-income economies, such as the United populations, wealth, and gender: larger gaps States, Korea, and Taiwan, China, have been between the two lines in each graph indicate increasing, that of Malaysia—traditionally a greater levels of inequity, whereas perfectly DISCONNECTS IN HIGHER EDUCATION 87

overlapping lines indicate complete equity TABLE 3.4 Eighth-grade TIMSS scores for mathematics, selected between groups. East Asian economies, 1999, 2003, and 2007 Cambodia shows signifi cant inequities in Economy 1999 2003 2007 access to and survival in prior education levels across all dimensions. Indonesia seems to Taiwan, China 585 585 598 Korea, Rep. 587 589 597 have a relation between high inequity across Singapore 604 605 593 urban and rural areas in tertiary and sec- Hong Kong SAR, China 582 586 572 ondary education (while inequities of access Japan 579 570 570 to tertiary education by income level seem to United States 502 504 508 have an impact later). In Thailand, under- International average 487 466 500 scoring the relation between tertiary and Malaysia 519 508 474 prior education levels, the generally lower Thailand 467 — 441 levels of inequity in access to tertiary educa- Indonesia 403 411 397 tion also seem to refl ect lower levels of ineq- Philippines 345 378 — uity at prior education levels (beyond other Sources: Gonzales, Guzman, and others 2004; Gonzales, Williams, and others 2008; Mullis and reasons). The gender inequities in favor of others 2000. Note: Economies are ranked by their score in 2007. females in secondary education are repli- — = Not available. cated at the tertiary level, as clearly visible in Mongolia, the Philippines, and Vietnam (fi gure 3.20). TABLE 3.5 Eighth-grade TIMSS scores for science, selected East Asian economies, 1999, 2003, and 2007 Gender Gender inequities now cut both ways, Economy 1999 2003 2007 given steady improvements for females in Singapore 568 578 567 access to secondary and higher education Taiwan, China 569 571 561 in several countries, including Malaysia, Japan 550 552 554 Mongolia, the Philippines, Thailand, and Korea, Rep. 549 558 553 Vietnam. There is much evidence on the Hong Kong SAR, China 530 556 530 benefits of promoting girls’ education. United States 515 527 520 Results of PISA 2006 show the benefi ts of International average 488 473 500 Malaysia 492 510 471 investing in girls’ education: girls outper- Thailand 482 — 471 formed boys in math and science in some Indonesia 435 420 427 countries but consistently outperformed Philippines 345 377 — boys in reading in all OECD countries and Sources: Gonzales, Guzman, and others 2004; Gonzales, Williams, and others 2008; Mullis and others some developing countries. It is impera- 2000. tive for countries to ensure that their girls Note: Economies are ranked by their score in 2007. — = Not available. have access to education—including access to higher education.68 However, although in some countries, such as Cambodia and probably China, females remain discrimi- interpreting their results, but this possibility nated against, in others, such as Thai- is quite unlikely, particularly in Mongolia and land, it is now males who may be falling the Philippines, given the signifi cant relative behind at the entrance of lower- and upper- gap in completion. This gap suggests other secondary education. issues related to retaining males in the sec- The inequities in completion disfavor- ondary cycle. Indeed, opportunity costs are ing males in Mongolia, the Philippines, and potentially high among the poor and higher Vietnam may in theory be related to having for males than females in the secondary admitted a wider mixed-ability pool of males school age range, leading to dropout of other- in the fi rst place, requiring some caution in wise talented students.69 88 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 3.16 PISA 2009 scores for East Asia, United States, to cultural and awareness issues, but the and OECD absence of comprehensive policy measures for ethnic minority groups means they are not being tackled (primary and secondary equity 72 Shanghai, China interventions focus on fee exemptions). While they start early, inequities by ethnic group are most often further exacerbated Singapore in the transition between secondary and tertiary education, raising additional issues for Hong Kong SAR, China addressing this disadvantage.73 Even the few beneﬁ ts and policies in favor of these groups Korea, Rep. initiated at the secondary level are generally not pursued further, creating new barriers to Taiwan, China entry. This was still the case in Vietnam in 2008 and in Thailand in 2006 where ethnic groups had a much worse transition rate than Japan the majority (ﬁ gure 3.22).74

Thailand Urban or rural location Inequities in other areas are worrisome. Indonesia The significant losses of rural students at the entrance of lower and upper secondary United States in Indonesia and Thailand, going beyond poverty gaps, are particularly problematic. They suggest barriers to entry, such as lack OECD average of schools, which become an impediment 200 300 400 500 600 for possibly bright students. Other data sug- PISA scores gest that in Vietnam, Mongolia, and the math reading Philippines inequities are already evident in secondary education and are refl ected then in tertiary education (fi gure 3.23). Also in Source: OECD 2010b. these three countries, as well as in Indone- sia, inequities are further exacerbated in the Ethnicity transition between secondary and tertiary Inequities by ethnic group seem to be education (fi gure 3.24), suggesting insuffi - sharper than gender inequities. Although cient action to retain rural students through- data on access to primary and secondary out the education cycle.75 education by ethnic group are not always reliable in labor force or household surveys, Income evidence suggests discrimination in sev- Income remains the biggest constraint in eral countries.70 Vietnam’s minorities have Cambodia for both access to and survival attendance rates of 89 percent for primary through the secondary cycle. Poor students education, 85 percent for lower-secondary are likely to drop out or not enroll because of education, and 52 percent for upper-second- high direct and opportunity costs. They can ary education (as of 2008), well below the access only a few need-based scholarships (at national average. all levels) and generally go to lower-quality In Vietnam inequities by ethnic group are schools.76 visible in completion (fi gure 3.21). They are likely to refl ect both differences in access and in quality of schools.71 Several studies point DISCONNECTS IN HIGHER EDUCATION 89

Tertiary entrance examinations FIGURE 3.17 Relation between TIMSS scores and STEM enrollment Ethnic minorities, rural students, and other shares groups are disadvantaged in their access to high-quality higher education. Entrance a. Math 650 examinations used to determine places for KOR 600 SGP students at universities are inherently linked HKG to issues of access and equity because suc- 550 cessful performance in these examinations 500 MYS is the result of prior academic experience. 450 Evidence for this ﬁ nding is seen notably in TIMSS math 400 IDN PHL Vietnam and China. In Vietnam students 350 from disadvantaged groups score lower marks 300 on the National University Entrance Exami- 15 20 25 30 35 40 45 50 nation (without this being related to ability, Tertiary STEM enrollment share (%) as shown in chapter 2) and are less likely to b. Science 600 SGP enroll in high-quality tertiary education insti- KOR tutions.77 While the Ministry of Education 550 HKG and Training sets the cutoff marks needed to 500 enroll in tertiary institutions for mathemat- MYS ics, physics, and chemistry, higher-quality 450 IDN universities require better entrance examina-

TIMSS science 400 PHL tion scores. In China aspiring tertiary students must 350 take a national entrance examination in 300 addition to the senior secondary diploma 15 20 25 30 35 40 45 50 examination to enter tertiary education. Tertiary STEM enrollment share (%) Some recent research has noted signifi cant disparities in access across provinces despite Sources: EdStats database; TIMSS 2007 database. similar test scores: one study reported that applicants from Beijing needed a slightly lower score to enter tertiary education education cycle itself.82 Autonomous govern- than those in Hunan.78 Other research has ment offi cials have the right to determine the shown that students from areas with higher education plans, curricula, and language of concentrations of tertiary education oppor- instruction at the primary and secondary tunities need lower scores than students levels following government principles and from other provinces. These fi ndings may goals (such as the universalization of com- result from the quota system, which favors pulsory education). This has led to schools students from regions with stronger tertiary using minority languages for instruction in systems.79 all subjects up to the end of senior secondary education. (Chinese is taught only as a sec- Steps taken ond language.)83 Even when countries have started to address Despite a policy to improve minority inequities, outcomes have not always turned enrollment in mainstream institutions in out as expected because of weaknesses in place since 1951, most ethnic minority stu- policy design.80 China is well known for hav- dents are still concentrated in ethnic univer- ing instituted policies to ease the plight of sities, which are weak in science and engi- minority and disadvantaged groups, includ- neering and tend to direct minorities into ing interventions targeted beyond the educa- languages and teacher studies—fields that tion sector,81 and has been generally active traditionally do not allow much upward in supporting ethnic minorities through the economic mobility. Moreover, when ethnic 90 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 3.18 Relation between TIMSS scores and journal articles secondary education quality.85 Low- and middle-income East Asia may not be mak- a. Math 700 ing enough effort at that level. Curricula may not be sufficiently well 650 KOR SGP aligned between the secondary and tertiary 600 JPN cycles, the case in Mongolia and Thailand.86 550 This misalignment may be partly responsible 500 MYS for some of the dropouts between secondary 450 TIMSS math THA and tertiary education. Yet countries such as 400 IDN Korea and Singapore have been successful at PHL 350 creating a solid higher education system by 300 aligning their secondary and tertiary systems 0 200 400 600 800 1,000 well.87 Journal articles (per million people) Inclusiveness of tertiary education clearly b. Science 650 depends on inclusiveness early on, requiring 600 a bold and comprehensive approach to tackle KOR JPN disadvantage. There is, for instance, still too 550 HKG SGP little use of distance education at the second- 500 MYS ary level, and scholarships to cover transport 450 costs to urban centers, to address the urban- IDN 88 TIMSS science 400 rural gap, are limited. And comprehensive PHL packages, including afﬁ rmative action poli- 350 cies, to address the multiple disadvantages 300 0 200 400 600 800 1,000 faced by ethnic minority groups are extremely Journal articles (per million people) rare. Moreover, when they are implemented, they do not always have the desired effects.

Sources: EdStats database; TIMSS 2007 database. Beyond the speciﬁ c measures, implementing comprehensive approaches to tackle quality and inclusiveness is constrained by gaps in information, capacity, and incentives. minorities are enrolled in mainstream institu- Fortunately, information is now improving tions, they may have difﬁ culty in performing through the use of international standard- well.84 These results may be related to weak- ized tests in primary and secondary educa- nesses in quality of teaching of the Chinese tion. Equity statistics are, however, still weak language in the secondary education pro- in some countries.89 Policy makers need to vided to ethnic minorities, highlighting the be well informed with disaggregated data as need for standards and alignment between to why certain groups are more vulnerable cycles. than others and why these groups tend to be Reasons for the disconnect. The evi- poorer—and have a deeper understanding of dence suggests that policy makers in low- what has determined the success of groups and middle-income East Asia may not be who have been able to improve their circum- putting enough emphasis on the connec- stances despite long-term disadvantage. tions between education levels, or fully Equally, equipping students with informa- understanding these interrelationships— tion about the tertiary education application and so the need to address them early and process is critical. In many countries (and not consistently. A comprehensive vision of the just in East Asia), the lack of information for education sector is missing. Math results in students on college choices or procedures, secondary education, for example, clearly as well as adult guidance, is a significant influence outcomes later, requiring at nonmonetary barrier to access, particularly minimum a coordinated effort to improve for disadvantaged groups whose parents DISCONNECTS IN HIGHER EDUCATION 91

FIGURE 3.19 Estimated survival rates for children ages 13–19, Cambodia, Indonesia, and Thailand

a. Cambodia, 2004, by gender d. Thailand, 2002, by gender 100 100

75 75

50 50

25 25 percentage completed percentage percentage completed percentage 0 0 0123456789101112 01234567 8 9 10 11 12 grade grade male female male female

b. Cambodia, 2004, by urban/rural location e. Thailand, 2002, by urban/rural location 100 100

75 75

50 50

25 25 percentage completed percentage percentage completed percentage

0 0 01234 56789101112 01234567 8 9 10 11 12 grade grade urban rural urban rural

c. Cambodia, 2004, by income quintile f. Thailand, 2002, by income quintile 100 100

75 75

50 50

25 25 percentage completed percentage percentage completed percentage 0 0 01234 56789101112 0123456789101112 grade grade poorest quintile richest quintile poorest quintile richest quintile

g. Indonesia, 2003, by urban/rural location h. Indonesia, 2003, by income quintile 100 100

75 75

50 50

25 25 percentage completed percentage percentage completed percentage 0 0 01234567 8 9 10 11 12 0123456789101112 grade grade urban rural poorest quintile richest quintile

Sources: di Gropello 2006. 92 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 3.20 Secondary and tertiary education completion rate by FIGURE 3.22 Tertiary enrollment in proportion of gender, Vietnam, Mongolia, and the Philippines secondary completion by ethnic group, Vietnam, 2008, and Thailand, 2006

70 100

60 90 80

50 70 60 40 50 40

percent 30 30

20 10

ratio of tertiaryratio secondary to enrollment completion 0 10 Vietnam Thailand minorities majority 0 Vietnam Mongolia Philippines Sources: VHLSS 2008; Thailand 2006 Labor Force Survey. Note: Secondary completion rates are calculated in relation to the 17–21 secondary completion rate, females age range; tertiary secondary completion rate, males enrollment rates are calculated in relation to the 18–22 age range. tertiary completion rate, females tertiary completion rate, males are much less knowledgeable about higher education. Sources: Household and labor force surveys, most recent years. Note: Tertiary education is proxied by university education in Mongolia and the Philippines; Understanding the interrelationships secondary education goes up to only grade 10 in the Philippines. Secondary completion rates between educational levels and planning and are calculated in relation to the 17–21 age range; tertiary completion rates are calculated in relation to the 22–28 age range. implementing successful packages is also constrained by lack of institutional and human capacity, as well as the right ﬁ nancing and nonfinancing policies at all levels. Finally, FIGURE 3.21 Secondary and tertiary education completion rate, by governance overall, including accountability ethnic group, Vietnam, 2008 mechanisms and the management arrangements of the government in relation to the 60 education sector, also influences the way 50 interrelationships are understood and built upon. For instance, in the Philippines and 40 Thailand the separation of higher education 30 management from the ministries of educa- 20 tion (through the Ministry of University Affairs in Thailand and the Commission of completion rate (%) rate completion 10 Higher Education in the Philippines) has not 0 been conducive to an overall vision of the secondary, secondary, tertiary, tertiary, education sector. In Thailand this separa- minorities majority minorities majority tion has by now been addressed.90 But also in other countries where policy and opera- Source: VHLSS 2008. Note: Secondary completion rates are calculated in relation to the 17–21 age range; tertiary comple- tional responsibilities are confused, an inte- tion rates are calculated in relation to the 22–28 age range. grated approach has been difficult. These DISCONNECTS IN HIGHER EDUCATION 93

considerations lead to the next and last sec- FIGURE 3.23 Secondary and tertiary education completion, by tion of this chapter. urban or rural area, Vietnam, Mongolia, and the Philippines

Conclusion and moving forward 80 This chapter has argued that failure to 70 deliver on skills and research reflects five 60 main disconnects among skill and research providers and users at the core of higher 50 education systems. With limited systematic 40 quantitative or qualitative information, it is diffi cult to make generalizations on com- 30 parative intensity of disconnects across (%) rate completion countries, but the evidence presented in the 20 chapter points to these disconnects being 10 present all around lower- and middle-income 0 East Asia (table 3.6). Cambodia, Indonesia, Vietnam Mongolia Philippines Mongolia, the Philippines, and Vietnam secondary completion rate, rural secondary completion rate, urban have been repeatedly used to illustrate some tertiary completion rate, rural tertiary completion rate, urban of these disconnects. China, Malaysia, and to a lesser extent Thailand tend to perform Sources: Household and labor force surveys (most recent years). better along most dimensions than other Note: Tertiary education is proxied by university education in Mongolia and the Philippines; countries. In particular, China, in line with secondary education goes up to only grade 10 in the Philippines. Secondary completion rates are calculated in relation to the 17–21 age range; tertiary completion rates are calculated in relation to its position of leading the middle technology the 22–28 age range. cluster, is being more successful in building links with fi rms in the technology arena FIGURE 3.24 Tertiary enrollment in proportion to secondary than other lower- and middle-income coun- completion by urban or rural area, selected economies tries, while Malaysia has an edge in linking skill providers. Upper-income countries, 80 refl ecting their better skill and innovation outcomes, appear to have many less or less- 70 intense disconnects, though links between HEIs and fi rms in research and technology 60 remain more of a challenge for them as well. 50 This fact indicates that some disconnects are likely to be more diffi cult than others to 40 address. 30 This chapter has also argued that many of the disconnects in higher education in low- 20 and middle-income East Asia have their origin in information, capacity, and incentive 10 ratio of tertiaryratio secondary to enrollment completion gaps, which point to market and policy fail- 0 ures. Public policy can address these discon- Vietnam Mongolia Philippines Indonesia nects. Not all the policies are strictly related rural urban to higher education,91 and not all actors and interactions at the core of higher education Sources: Household and labor force surveys (most recent year). systems will respond the same way to higher Note: Tertiary education is proxied by university education in Mongolia and the Philippines; secondary education goes up to only grade 10 in the Philippines. Secondary completion rates are education policies, but they clearly have a calculated in relation to the 17–21 age range; tertiary enrollments are calculated in relation to the 18–22 age range. 94 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

TABLE 3.6 Intensity of disconnects by income and technology cluster group

Lower income Middle income Upper income Disconnect LTC LTC LMTC MTC UMTC TTC Disconnect 1: HEIs and employers Disconnect 2: HEIs and companies in More advanced in Japan; Taiwan, research and technology China; Singapore Disconnect 3: HEIs and research institutions Less advanced in Korea, Rep. Disconnect 4: HEIs among themselves and with training providers Disconnect 5: HEIs and earlier education Incomplete institutions (schools) evidence

Source: Authors’ elaboration. Note: The darker the area, the larger is the disconnect. LTC = lower technology cluster; LMTC = lower-middle technology cluster; MTC = middle technology cluster; UMTC = upper- middle technology cluster; TTC = top technology cluster.

tremendous role. Figure 3.25 shows a simpli- conduct research, hire high-quality faculty, fi ed picture of the potential impact of three develop technical and engineering courses, main pillars of higher education policy— and install laboratories. Financing decisions fi nancing, management, and stewardship— can also provide HEIs with incentives to on the fi ve disconnects through higher capac- improve performance. For example, firms ity, stronger incentives, and better or more and governments can improve skill outcomes information. To every policy pillar are asso- by allocating funds to universities according ciated critical policy levers that are the object to graduation and employment indicators. of the next chapters. The fi gures are simpli- Finally, funding for scholarships can help fi ed because only the most obvious relation- increase the pool of talented students getting ships between policy pillars and disconnects access to university. are shown—in fact, the policy pillars have So public fi nancing can address the dis- the potential to infl uence all disconnects with connects. Greater and more effi cient public the right policy levers. Figure 3.25 illustrates spending, combined with additional pri- how some of the policy pillars work. vate resource mobilization, can tackle the Many of the disconnects are related disconnect with earlier education by sup- to financing. For instance, disconnects in porting student transitions from secondary research and from secondary to tertiary to tertiary education through scholarships education suggest that too few financial and loans (addressing capacity constraints). resources may be available for research and Greater and more effi cient public spending scholarships. Low enrollment in STEM fi elds can also tackle the disconnects between uni- in several countries, a disconnect with skill versities and fi rms in skills and research and users, also suggests underspending in these between teaching and research provision by fi elds. Developing or attracting faculty with supporting higher funding for research and PhDs may require fi nancial decisions, such STEM skills in universities combined with as offering scholarships to brilliant students performance-based funding (addressing to study in prestigious universities abroad. capacity and incentive constraints). These It may also require offering internationally policy levers are explored in depth in the competitive faculty salaries as is done at the next chapter. top universities in Singapore; Hong Kong Several information, capacity, and incen- SAR, China; Korea; and Japan. Japan pays tive constraints and the related disconnects its top-level academics over three times more are related to management. For instance, than China does.92 Adequate fi nancing can limited information on skill and institutional thus improve higher education’s capacity to outcomes—or the research and technology DISCONNECTS IN HIGHER EDUCATION 95

needs of firms—suggests a market failure departments of higher education but whose not yet corrected. Even if institutions receive performance is critical to the sector. In suffi cient funds for highly qualifi ed faculty, other words, governments need to exercise insufficient autonomy to select staff and appropriate stewardship of the overall sys- decide on their academic programs makes it tem. The lack of interaction between HEIs diffi cult for them to deliver what fi rms need. and fi rms to some extent refl ects a lack of A lack of accountability of university man- information on what works and legal and agement to representative university boards fi nancial incentives to connect. Disconnects may also not be conducive to universities between skill providers and users are also fulfi lling the needs of skill or research users, related to a private sector not fulfi lling its as would be the lack of competition among potential because of overregulation. Stew- institutions. ardship will require expanding the capacity Management of the public sector can help to coordinate higher education depart- tackle the disconnect between HEIs and skill ments with other departments, actors, and and research users through the right mix of ministries; private HEIs; the links between institutional autonomy and accountability. HEIs and fi rms in skills and research; and The right accountability structures will also increasingly the links between the domestic help address other disconnects. Qualifica- and international higher education markets. tion frameworks have, for instance, a direct Adequate stewardship of the higher educa- impact on improving the incentives and infor- tion system can help solve the disconnects mation for coordination among HEIs. Chap- between skill users and providers through ter 5 reviews policies related to the manage- effective promotion and regulation of the ment of the public sector. private sector (addressing incentive, infor- Beyond managing the public sector, mation, and even capacity constraints). It governments also need to coordinate and can also address the disconnects between handle actors and connections not under fi rms and providers in skills and research the direct administrative authority of the by sharing best practices on what works

BOX 3.2 The rationale for public intervention in higher education

The overall policy direction, oversight, regulation, addressing scale economies needed to fi nance centers and quality control of the system. The government of excellence. The government may need to dedicate has the primary responsibility for establishing the an increasing share of its resources to create the frame- main strategic directions, providing the core regula- work for innovation and excellence that the market tions, and setting up the accountability and quality alone may fail to create because of benefi ts diffi cult to assurance framework needed to address opportunis- internalize or prohibitive initial investment. tic behavior and asymmetric information. Capacity Funding the access of key vulnerable groups that building can also be provided. are left out but deserve an opportunity to partici- Accurate and timely information about the sup- pate. There continues to be a strong scope for public ply, demand, and value of particular skills and other intervention on equity grounds. education outcomes and outputs. This function can Funding minimum teaching needs to provide entail supporting learning outcome assessments; basic coverage that the market may fail to provide for education management information systems; and positive spillovers, while giving institutions an incen- fi rm, labor force, and household surveys. tive to perform without necessarily stifl ing them with Funding “public goods,” such as R&D and sci- rules. This calls for funds to be provided according ence, technology, engineering, and math, and to transparent and performance-based criteria. 96 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 3.25 Relationships between disconnects and policies

a. Eﬀective ﬁnancing can address many disconnects of the disconnects by improving capacity and incentives higher education and research institutions

higher education and earlier education

higher capacity Financing higher education stronger and employers incentives

higher education and companies (as research users)

HEIs between themselves and with training providers

b. Eﬀective management of public HE disconnects institutions can address many of the disconnects by improving incentives higher education and and information research institutions

higher education and earlier education

better information higher education and Management employers stronger incentives higher education and companies (as research users)

HEIs between themselves and with training providers

(continued next page) DISCONNECTS IN HIGHER EDUCATION 97

FIGURE 3.25 continued

disconnects c. Eﬀective stewardship of the higher education system can address many of the disconnects by improving higher education and incentives, information, and capacity research institutions

C higher education and earlier education

C Stewardship higher higher education and capacity (PS; IM) employers Coordination (C) C Private-Sector (PS) better information PS higher education UILs UILs and companies International Market UILs (as research users) (IM) stronger incentives IM PS C HEIs between themselves and with training providers

Source: Authors’ elaboration. Note: Only stronger relations shown; the intensity of shade indicates intensity of impact.

and by putting in place the right legal and 6. World Bank 2008. financial incentives (addressing incentive 7. Smeaton and Hughes (2003) deﬁ ne interme- and information constraints). Stewardship- diate skills as those located above routine related policies are reviewed in chapter 6. skills and below professional skills that are In light of these considerations and the required to operate effectively in jobs with a relatively high level of skills and fair degree of above diagnostic, it is clear that public policy responsibility. Examples of intermediate skills can have a profound role in improving higher include effective communication (in local and education outcomes by tackling the discon- foreign languages); personal skills (interest, nects. Box 3.2 reviews the case for public motivation, interpersonal skills, leadership); intervention in higher education. computer skills; and technical knowledge of The next three chapters delve into the some disciplines (such as economics, law, and three most critical public policy areas: ﬁ nanc- business). ing, management of public institutions, and 8. di Gropello, Tan, and Tandon 2010. stewardship of the higher education system. 9. di Gropello, Kruse, and Tandon 2011. 10. World Bank 2010d. 11. World Bank 2010d. Notes 12. Appendix A shows that other countries in the 1. HRINC 2010. region, such as Cambodia and Vietnam, may 2. Kennedy and Lee 2008; World Bank 2008. also not have achieved a healthy balance among 3. Government of the Philippines 2007. universities, colleges, and non-degree-granting 4. HRINC 2010. skill providers, judging from the types of HEIs 5. World Bank 2010d. and workers’ gaps in information technology, English, or even work attitude skills, indicating 98 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

that this poor mix may be contributing to the univer sity input into creating and developing skill shortages prevalent in the region (though regional science parks, though only in recent China, with its large number of “tier III” insti- years (Mok 2010). tutions, appears better positioned to deliver 27. Mok 2010. these types of skills). 28. Mok 2010; Motohashi 2005. 13. The ISCED 6 programs are designed to pre- 29. Approximately 150 research universities have pare graduates for faculty and research posts been so designated. and are the second stage of tertiary education 30. Only about 10 percent of these patents are (equivalent to a doctoral degree). directly marketable, however (Wu 2010). 14. According to United Nations Educational, 31. Doner, Intarakumnerd, and Ritchie 2010. Scientifi c and Cultural Organization (UNESCO) 32. These were agricultural biotechnology; envi- Institute data. ronmental hazardous waste management; 15. Mok 2010. environmental science, technology, and man- 16. These studies are commonly used to elicit agement; energy and environment; chemistry; information from recent graduates—about postharvest technology; and petroleum and three to six months after convocation— petroleum technology (Doner, Intarakumnerd, regarding fi eld of study, employment, and job- and Ritchie 2010). search methods and, if working, information 33. Kodama and Suzuki 2007. about the employer, starting pay, and relevance 34. Hill and Tandon 2010. of training received to the job. These studies 35. Doner, Intarakumnerd, and Ritchie 2010. can provide useful information to institutions 36. See Brimble and Doner (2007) and Schiller and and policy makers on the relevance of differ- Brimble (2009) for more on these arguments. ent academic programs to the job market, as 37. Tan 2010. well as on skills most needed in the workplace, 38. Guan, Yam, and Mok 2005. which would allow them to make adjustments 39. Wu 2010. to curriculum and training programs to make 40. The importance of incentives is also evident graduates more employable. from Wright and others (2008), who fi nd 17. Only Thailand and Vietnam provide a dis- that faculty in second-tier European univer- aggregation of tertiary education by fi eld of sities have limited time and few incentives study in their household surveys. to conduct applied research and engage in 18. Chapman 2010. patenting, even if this activity is subsidized. 19. Underneath these aggregate data is much For purposes of recognition and promotion, variation across institutions, academic disci- publication in scientifi c journals takes prece- plines, and types of schools, as well as within dence. countries. 41. Carney and Zheng (2009) contend that 20. Chapman (2010) provides support for a Singapore fi rms’ risk aversion has generated positive correlation between teaching and a confl icting innovation incentives and ulti- graduate degree. mately undermined innovative activity. A 2009 21. In the Philippines, the relative lack of qualifi - study of the Malaysia Innovation Climate cations among tertiary faculty results from a Survey by the fi rm Alpha Catalyst found that combination of several factors, including the Malaysian fi rms, particularly in manufactur- fact that lower-level schools were upgraded ing, were unwilling to invest in new product to HEIs and had the faculty only to instruct processes without guaranteed success. at the secondary level; moreover, low com- 42. Ca 2006. pensation to faculty does not make teaching 43. Although an exceptional case, an upper- at the tertiary level attractive. income country like Korea, where most 22. Doner, Intarakumnerd, and Ritchie 2010. applied research is still carried out by public 23. The survey scale metric was as follows: 0 = or private research institutes or by chaebol, unknown, 1 = not important, and 5 = very which have historically received the state’s important. strong fi nancial support to set up their own 24. Hill and Tandon 2010; Tan 2010. in-house research units, also shows this type 25. Wu 2010. of disconnect (Mok 2010). 26. Mok 2010; Sohn and Kenney 2007. The 44. HRINC 2010. Korean government has also encouraged 45. Tan 2010. DISCONNECTS IN HIGHER EDUCATION 99

46. Tan 2010. parental education, and have higher levels of 47. Yilmaz 2009. income (Linh, Thuy, and Long 2010). 48. Yilmaz 2009. 72. World Bank 2010g. 49. Yilmaz 2009. 73. In Mongolia, for example, Kazakh students 50. Wu 2010. demonstrate lower transition rates to tertiary 51. Wu 2010. education, partly because the language of 52. Wu 2010. instruction (Mongolian) is introduced in 53. OECD 2008a. classrooms in Kazakh provinces only at the 54. OECD 2008a. beginning of the secondary level (World Bank 55. As a matter of fact, there is even often a sepa- 2010d). ration between teaching and research within 74. Strictly speaking, these are not transition universities themselves. rates, because the tertiary gross enrollment 56. World Bank 2010d. ratio is an average of the different grades, 57. World Bank 2009a. so using the gross enrollment ratio underes- 58. These include different types of institutions timates transition. The trends across groups and courses (such as short and long courses). are likely to be the same, however. Caution is 59. Indonesia is still at the pilot stage of devel- also needed because these indicators are com- oping a framework based on the Australian paring different student cohorts. model. The Philippines is relatively more 75. In Vietnam, for instance, the evidence on advanced in the design and implementation determinants of access for rural students is of a model more focused on addressing the clear. Among regions, young people in South needs of vocational and technical education. Central Coast are more likely to enroll in ter- Although only in its initial stages of imple- tiary education than those in Red River Delta. mentation, Thailand’s Commission on Higher By contrast, young people in North East are Education is aiming to institute a national less likely to go to colleges than those in Red qualifi cation framework over eight subject River Delta, and young people living in urban groups over the next fi ve years. areas are more likely to attend college than 60. This performance is owing to a large share those in rural areas. Among the determinants of TVET tertiary education in Malaysia and of tertiary education access, both head of distance tertiary education in Thailand. household and head of household’s spouse’s 61. Because these rates involve different cohorts, education levels have strong effects: if they caution is needed. have only primary education or no formal 62. Hoachlander, Sikora, and Horn 2003. schooling, children living in such households 63. Ingles and others 2002. are less likely to go to college. Also, those 64. Fry 2004. living in households whose heads have high 65. Fry 2004. school degrees or tertiary degrees are more 66. Michaelowa 2007. likely to go to college and university (Linh, 67. di Gropello, Kruse, and Tandon 2011. Thuy, and Long 2010). 68. OECD 2007b. 76. Evidence in Cambodia indicates that those 69. di Gropello 2006. with higher levels of head of household edu- 70. In China, participation and survival rates at cation, in the top three income quintiles, and the primary and secondary levels for ethnic living in urban areas have far higher educa- minorities are far lower than for Han Chinese tional attainment (Sakellariou 2010a). (Ma 2010). In Mongolia, Kazakh children 77. Linh, Thuy, and Long 2010. enjoy relatively high rates of access at the 78. Luan 2007. primary level but exhibit lower transition 79. Loyalka 2009. rates to the secondary level than the majority 80. Malaysia is a successful example of affi r- population by almost 10 percentage points mative policies in middle-income East Asia, (World Bank 2010d). for its Bumiputera ethnic group (Mukherjee 71. According to a recent study, the ethnic minority 2010). students demonstrating greater access to higher 81. It has eased access to political offi ce for these education most often live in urban areas and populations, loosened fertility restrictions, in households that are smaller, have a smaller and implemented several affi rmative policies. proportion of children, have larger shares of 100 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

82. Scholarships and boarding for primary and curricula and instituting more standardized secondary education have been offered to upper-secondary leaving examinations. most students in pastureland and mountainous 87. OECD 2008b. regions in China since 1995. This policy was in 88. di Gropello (2006) has shown that educa- place long before the national scheme of Two tional attainment and rural-urban disparities Exemptions and One Subsidy, which exempted in East Asia remain hampered by variables minorities from tuition fees and provided a such as distance to school, availability of subsidy to minority families whose children are basic services, school fees, cash and noncash enrolled in primary and secondary education. transfers, lack of learning materials, and lack 83. Information Offi ce of the State Council of the of alternative cost-effective delivery methods, People’s Republic of China 2005. such as distance learning. 84. Wu 2010. 89. They are weak for ethnic groups particularly. 85. In Korea serious attention to math and sci- 90. It was done through the creation of a ence at the secondary level started in 1973 Commission of Higher Education within the with the so-called movement to scientifi cize Ministry of Education (Fielden 2008). the whole people, performed in conjunction 91. Some policies involve actions to support fi rms with the government’s strong industrial push and research centers, for instance, which during the same period. Sorensen (1994) derive from broader economic and political shows that a similar emphasis on math and decisions. This is the case of fi scal, fi nancial, science soon followed at the tertiary level. and liberalization policies to encourage inno- 86. Several reports (World Bank 2009b, 2010d) vation in fi rms. Some others involve actions note the lack of links between secondary at other education levels. While not directly and tertiary curricula in these two countries. under the authority of higher education, They encourage more direct intervention by these policies need to be carefully consid- tertiary academics in designing secondary ered because they affect actors that are part of the higher education system and therefore the outcomes of higher education reforms. The combination of all these policies ultimately decides the success of higher education reform. 92. Rumbley, Pacheco, and Altbach 2008. Financing Higher Education 4

olid fi nancing is the backbone of a well- It starts with a broad review of financing functioning higher education system, needs, continues with a review of funding Sbut the systems in East Asia’s low- and strategies with emphasis on the role and effi - middle-income countries are not delivering ciency of public fi nancing and strategies to the skill and research outcomes they need, mobilize additional resources, and concludes as seen in the disconnects—often funding with a brief summary of policy options. related—discussed earlier. In part this is because public fi nancing goes to institutions Financing needs regardless of whether they are addressing public goods such as research, externalities, Many of the outcome gaps and related dis- or equity concerns. Public funding can then connects and constraints are related to fund- address the disconnects. For instance, it can ing. This section investigates how costly it help tackle the disconnect with early educa- would be to achieve better outcomes in a tion by supporting student transitions from sample of lower- and middle-income East secondary to tertiary education through Asian countries and, in light of the high scholarships and loans. Or it can tackle the costs, suggests priority areas for fund- disconnects between universities and fi rms ing. Before proceeding, a brief caveat is in in research and technology by supporting order: the aim of this section is to provide higher funding for research in universities. a broad order-of-magnitude estimate of The region thus needs to identify prior- fi nancing needs if countries were to ramp up ity areas for support and strategies to fund investments in quality and quantity, with- them. Public funding will have a critical out regard to current budget constraints. role to support research, science, technol- The estimates are not public investments ogy, education, and mathematics (STEM) but should be thought of as overall fi nanc- fields, and equity measures in lower- and ing needs that could be covered by myriad middle-income East Asia. But because public sources (public, private, philanthropic). This resources are scarce, countries will also need section highlights the trade-offs between to use them more effi ciently and effectively quality and quantity that countries will face and be innovative in mobilizing additional when fi nancing their systems. Pursuing both resources. This chapter has three parts. quality and quantity in equal measure would

101 102 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

be very costly without greater effi ciency of finance salaries, improving faculty qualifi- expenditures, a strong sense of priorities, cations through fellowships also constitutes and private fi nance leveraging. a robust share of recurrent expenditure at nearly 10 percent. Simulation analysis The fi nancing gaps that these projections imply for Indonesia over current expendi- A demographic-based model, in which cov- tures are large. Indeed, Indonesia will need erage targets and quality-improvement poli- to increase expenditure per student dra- cies are independent variables, can estimate matically over the near term, since in 2008 education expenditures in light of achieving it was spending only about 13 percent of these targets. In performing these simula- gross domestic product (GDP) per capita per tions, this chapter has adapted the United student (in public money). If public expendi- Nations Educational, Scientifi c and Cultural tures are maintained, to institute these inter- Organization’s Education Policy and Strat- ventions the financing gap will grow from egy Simulation Model to estimate fi nancing about 150 percent of GDP per capita to more needs for four regional countries. (Details of than 500 percent by 2019 (table 4.1). the architecture of the model are in appen- To widen access and raise quality, Viet- dix M.) nam also will have to mobilize significant This chapter presents the model’s results additional resources, mainly on increased for two scenarios: (a) expanding coverage and recurrent expenditures (about four-fi fths on improving quality (illustrative for countries salaries, followed by training, upgrading fac- with lower tertiary gross enrollment ratios, or ulty qualifi cations, and administration). Per GERs, using Indonesia and Vietnam) and (b) student expenditure will need to increase maintaining coverage and improving quality from US$1,500 to about US$4,000 over the (illustrative for countries with higher tertiary next decade. GERs, using the Philippines and Mongolia). If 2007 public expenditure levels are main- All scenarios have ambitious targets and tained, these projections imply a large fi nanc- interventions to provide an upper fi nancing ing gap. Tertiary expenditure per student, bound, though student-teacher ratios and measured as a percentage of GDP per capita, student-classroom ratios were maintained, will likely need to increase by three to four saving on costs. times its current levels by 2015, and by several orders of magnitude thereafter (table 4.2). Expanding coverage and improving quality: Indonesia and Vietnam Maintaining coverage and improving Financing simulations for Indonesia show quality: The Philippines and Mongolia that a steep ramp-up of tertiary investment The financing simulation for the Philip- will be required to expand coverage and pines similarly projects large and increasing improve quality. Interventions to increase recurrent expenditures. The model pro- faculty qualifi cations, salary and adminis- jects that expenditures need to grow from trative costs, and faculty training costs—as nearly US$6,000 per student per year to well as costs for central administration, cur- about US$10,000. More than 95 percent riculum development, and monitoring and of these expenditures are recurrent, and of evaluation—all imply larger fi nancing gaps those recurrent expenditures, salaries con- over current public expenditure. stitute by far the largest share (over 88 per- Recurrent expenditures make up more cent of recurrent expenditure), followed by than 95 percent of the projected quality administrative costs and costs for faculty improvement interventions. They rise over qualifi cation upgrading (both estimated at time from nearly US$2,000 per student per about 5 percent of recurrent expenditure). year to nearly US$8,000 over the decade. As with Indonesia and Vietnam, these fi g- While the majority of these recurrent costs ures represent expenditures several orders of FINANCING HIGHER EDUCATION 103

magnitude above what the Philippines cur- TABLE 4.1 Gap between projected per-student expenditure needed rently spends. If 2007 per student expendi- and current levels, Indonesia ture levels are maintained as a share of GDP percentage of GDP per capita per capita, the ﬁ nancing gap between needs Gap between and budgetary outlays is projected to reach projected levels about 300 percent of GDP per capita over 2008 tertiary Projected tertiary needed and the coming decade (table 4.3). Year expenditure expenditure needed 2009 levels Mongolia’s smaller volume of tertiary 2011 13.3 156.9 143.6 enrollments and higher student-teacher 2013 13.3 222.1 208.9 ratios place recurrent expenditure pro- 2015 13.3 325.9 314.6 jections slightly lower than those for the 2017 13.3 778.3 765.0 Philippines. In later years, the simulation 2019 13.3 579.8 518.1 projects recurrent expenditures per stu- Source: Authors’ calculations based on model presented in appendix M. dent to rise to more than US$7,000 per year, while capital expenditures remain TABLE 4.2 Gap between projected per-student expenditure needed low. Recurrent expenditures rise abruptly and current levels, Vietnam as more faculty complete fellowships and percentage of GDP per capita training. Overall, these figures imply a ﬁ nancing gap of 335 percent of GDP per Gap between capita by 2018 (table 4.4). projected levels 2007 tertiary Projected tertiary needed and Year expenditure expenditure needed 2009 levels Summary While these estimates are several orders 2011 61.7 151.8 90.1 of magnitude above what Indonesia, 2013 61.7 186.9 125.2 2015 61.7 236.3 174.6 Mongolia, the Philippines, Vietnam, and 2017 61.7 321.2 259.5 most other countries in the region are now 2019 61.7 579.8 518.1 spending on tertiary education, they should be put in context. The estimates are broadly Source: Authors’ calculations based on model presented in appendix M. in line with what neighboring countries and other middle-income countries are spend- TABLE 4.3 Gap between projected per-student expenditure needed ing on higher education. Per student tertiary and current levels, the Philippines spending is more than US$12,000 in Japan, percentage of GDP per capita nearly US$10,000 in Brazil, and nearly US$7,000 in Chile.1 The United States Gap between projected levels spends US$24,370 per tertiary student per 2007 tertiary Projected tertiary needed and year, and the Organisation for Economic Year expenditure expenditure needed 2009 levels Co-operation and Development (OECD) 2011 11.6 303.1 291.5 2 average is US$11,512. 2013 11.6 359.6 125.2 2015 11.6 430.7 174.6 Funding priorities 2017 11.6 536.4 259.5 2018 11.6 633.6 341.1

Attaining these spending targets appears Source: Authors’ calculations based on model presented in appendix M. very challenging in the short to medium term for all countries and should not be the cost norm for higher education in lower- and coverage increases in the short to medium middle-income East Asia. This suggests the term, notably Indonesia and the Philippines. need for greater selectivity in the targets and Increasing coverage is costly, raising quantity- in the activities to ﬁ nance. quality trade-offs. It may also be difﬁ cult to In line with the analysis of chapter 2, reduce student-faculty ratios across the board, it is clear that not all countries need broad implying that the higher differentiation of the 104 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

TABLE 4.4 Gap between projected per-student expenditure needed of research in higher education (acknowledg- and current levels, Mongolia ing that journals are not the best metric to percentage of GDP per capita capture the desirable effects of research) and Gap between that obtaining results has a cost (fi gure 4.1). Projected tertiary projected levels Chapter 1 showed that there is at least poten- 2007 tertiary expenditure needed and tial and need for better supporting small and Year expenditure needed 2009 levels medium enterprises in technological devel- 2011 17.1 117.5 100.4 opment in lower-income countries, thus jus- 2013 17.1 151.0 133.9 tifying higher focus on research. Without 2015 17.1 199.3 182.2 investment in tertiary research, disconnects 2017 17.1 280.2 263.1 between research and teaching in higher 2018 17.1 352.6 335.5 education will continue. Source: Authors’ calculations based on model presented in appendix M. Lower- and middle-income East Asia spend much less than upper-income East higher education system in colleges and other Asia in university research (figure 4.2). short-term institutions required by the labor Overall, the low research spending of market of some countries may be achievable most lower- and middle-income East Asian only gradually. And increasing faculty quali- countries points to a lack of priorities for fi cations may be possible only selectively, so research in higher education spending, low only a few universities will probably ever overall spending on research and develop- develop credible research capacity (given the ment (science and technology), and low high requirements for faculty qualifi cations). allocations of this spending for higher Realistic targets will also vary by income- education. They also spend less than other technology cluster. lower- and middle-income countries out- In line with being more selective in the tar- side the region. gets is setting priorities for what higher education activities countries should spend on. Cost constraints for poor and Chapter 3 suggested that activities with high disadvantaged groups and country externalities or market failures are probably responses underfunded in the region and thus need to With large education needs and little lee- be prioritized. This is further supported in way in public funding, most countries this chapter through an analysis of funding in the region are starting to rely more on for research, cost constraints, and equity- student fees to fi nance their institutions.3 enhancing interventions in the region. No sep- Although positive from the perspective of arate analysis is made for investment in STEM state budgeting, using this source brings fi elds, but the low enrollment shares in several inclusiveness challenges. countries of the lower and middle technology This section documents the fee structures clusters suggest that this is also an area with and related cost constraints that students high positive externalities (for its links with in the region face—as well as the coun- innovation) that is being underfunded. While try fi nancing policies to address these con- the focus here is on underfunded activities, straints. The aim is to show that countries spending also needs to continue or even be are still underspending on inclusiveness-en- boosted on gradually increasing coverage in hancing measures and to set the context for some countries, maintaining support to social further discussion of what approaches could science fi elds, and allowing for greater curric- tackle inclusiveness. ulum and degree diversifi cation when needed. Rich and poor countries operate with limited fiscal means and less than optimal Funding for research institutional capacity, using a familiar mix The tight relationship between journals and of funding sources to fi nance higher educa- tertiary research spending shows the benefi ts tion: tuition fees, government subsidies, and FINANCING HIGHER EDUCATION 105

income from other sources. Tuition fees are FIGURE 4.1 Research and development spending in tertiary equivalent to about a third or half of public education as a share of GDP and journals, latest available year university revenue in four of the fi ve countries, but only around a tenth in the Philippines .70 (table 4.5). The costs of tertiary education (tuition .60 Singapore fees and associated living costs) affect how .50 much a cash constraint can discourage oth- Japan .40 erwise talented students from enrolling and Korea, Rep. completing higher education. But financial .30 aid can reduce that constraint. .20 The variables infl uencing the decision to China pursue a tertiary education can be classifi ed .10 Thailand as monetary and nonmonetary, with three Malaysia R&D spending in tertiary (%) education 0 Vietnam types of monetary barriers (box 4.1). While 0 200 400 600 800 1,000 these barriers will be affected by many vari- journals per million people ables, fi nancial aid—in the form of assistance programs, scholarships, or loans—could Sources: EdStats database; UIS (UNESCO Institute for Statistics) Data Centre. address at least some of these constraints. It is the “net” costs (the costs less the fi nancial FIGURE 4.2 Research and development spending in tertiary aid) that really matter. education as a share of GDP, latest available year Several East Asian countries have fi nancial aid policies to help students overcome the cost constraint. It is important to assess the Japan full extent of the constraint. But the data do Singapore Hong Kong SAR, China not allow a calculation of “net costs” because Korea, Rep. few household surveys include information Malaysia on, say, scholarships. Secondary quantitative Thailand China information can help in assessing the likely Indonesia true costs. Philippines Mongolia Mongolia and Vietnam have used a fairly Vietnam effective combination of instruments to Lao PDR increase access among the poor, and Thailand 0 .10 .20 .30 .40 .50 .60 has been a pioneer in the use of student loans. R&D spending in tertiary education (% of GDP) But all lower- and middle-income countries in the region could implement more and Source: UIS Data Centre. better fi nancial aid policies. Beginning with a review of countries that have a combina- TABLE 4.5 Snapshot of public university revenue breakdown by tion of instruments to tackle inclusiveness proportion of funding source, selected economies (Vietnam, Mongolia, China), this section percent then reviews countries that have more spe- Government cific instruments (Indonesia, Cambodia, Economy subsidies Tuition fees Other income Thailand, Malaysia). China (2004) 47 30 23 Indonesia (2009) 56 38 6 Vietnam. In 2009, tertiary education costs Mongolia (2008)a 35 54 11 (tuition fees, extra-class fees, living and Philippines (2006) 73 11 16 accommodation costs) based on fees were Vietnam 50 45 5 estimated at 70 percent of household income Sources: China: OECD 2009a; Indonesia: World Bank 2010c (average of 82 universities); Mongolia: for the poorest quintile, and 30 percent for World Bank 2010d; Philippines: Orbeta 2008; Vietnam: World Bank 2010f. a. The government subsidy for Mongolia comprises funds from the state budget and the State 4 the richest quintile (table 4.6). Training Fund. 106 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 4.1 Determinants of higher education access

Both the academic literature and available data of higher education but simply cannot afford to (mainly household surveys) show that the decision attend university even after pooling internal (family to attend tertiary education has monetary and non- funds, savings, and wages) and external (grants and monetary variables. loans) funds. Financial aid, particularly the amount of aid, will increase liquidity. Monetary variables Debt-aversion barrier: Described by the econo- mist Richard Thaler as an “internalized liquidity Family income determines whether the student can constraint,” debt aversion arises when an individual afford the costs of the college. There are three mon- does not want to borrow even if he or she believes etary barriers to entry into tertiary education: the that the benefits of higher education outweigh cost-beneﬁ t barrier, the cash-constraint barrier, and the costs. Debt aversion occurs when this individ- the debt-aversion barrier. ual refuses to use the funds at his or her disposal Cost-benefit barrier: A certain segment of the because part of the funds might be loans, which at population (usually, lower-income or minority some point will have to be repaid. groups), when deciding on whether to attend university or not, performs a cost-beneﬁ t analysis of the costs of and expected returns to higher educa- Nonmonetary variables tion. The barrier arises when the group decides that the cost of attending university is greater than its Parental education, race and ethnicity, gender, and expected return to the education investment. geographical location—all play a role in the college Cash-constraint barrier: Also known as the decision-making process. So does prior academic “liquidity” constraint, it occurs when students who achievement, measured by the rigor of secondary have decided that the returns to education outweigh courses and the quality of that education. the costs still cannot put together the resources to obtain entry to universities. They believe in the value Sources: Johnstone 2004; Usher 2005.

TABLE 4.6 Vietnam: Tertiary education costs per month, 2009

Average monthly Average monthly Higher education costs Income quintile costs (US$) income (US$) (% of income) Poorest 67 95 70.1 Near poorest 83 152 54.2 Middle 95 212 44.8 Near richest 108 264 41.1 Richest 126 461 29.6

Source: Linh, Thuy, and Long 2010.

To address these constraints, Vietnam tuition fees in the total revenue of higher edu- has a comprehensive package of strategies cation institutions. While this strategy has and instruments, including the expansion helped Vietnam expand the subsector rapidly of the student loan scheme and the applica- without creating a heavy burden on the state tion of several other aid mechanisms.5 It has budget, it has also jeopardized the capacity of encouraged fee deductions and exemptions some students and their families to pay. This for the poor, revamped its aid scheme, and risk has been mitigated to an extent: access ﬁ ne-tuned its loan program. Cost-recovery to higher education by income quintile has mechanisms rapidly increased the share of become more equitable over time, as the gap FINANCING HIGHER EDUCATION 107

between the number enrolled from the poor- FIGURE 4.3 Change in ratio of tertiary enrollments and est quintile and the number enrolled from the completions for the poorest and richest quintiles, selected top income quintile has signiﬁ cantly declined, East Asian economies particularly in comparison to its middle-income neighbors (ﬁ gure 4.3). This is not the 60 result only of policies for higher education. Vietnam has also been active in supporting 40 more equitable access to higher-quality primary and secondary education, providing a 20 larger potential pool for tertiary education.6 0 But tertiary education policies have comple- mented these efforts by supporting the higher –20

transition of secondary graduates to tertiary percent education. –40 Since 1998 Vietnam has instituted policies to encourage fee deductions and exemptions, –60 benefi ting poor and ethnic minority students. In 2006 about 22 percent of disadvantaged –80 university students were benefi ting from sig- –100 nifi cant fee deductions of up to 50 percent of Cambodia Vietnam Indonesia Mongolia Thailand 7 tuition. low-income economies middle-income economies Vietnam has also had student aid schemes enrollment completion since the mid-1990s. Since 2006 students from ethnic minorities have received Source: Sakellariou 2010a. lump-sum assistance of about D470,000 Note: Cambodia data are from 2009 and 1999; Indonesia data are from 2009 and 1998; Vietnam data are from 2008 and 1998; Mongolia data are from 2007 and 1998; Thailand data are from 2006 and (US$25) a month—about a third of their 1996. Vietnam data include junior colleges and Indonesia data include tertiary diplomas. overall monthly higher education costs. Recent evidence from household surveys confi rms that the program is well targeted, with aid instruments to performance (or at least benefi ts accruing largely to poor and ethnic completion) could also be worth exploring, minority groups.8 Since 2007 scholarships as would focusing more on pretertiary educa- for poor students were also revised to cover tion and increasing the scholarship amount. the full tuition fees. In addition to scholarships, Vietnam Mongolia. Mongolian bachelor’s degree reformed its student loan scheme in 2006, students pay an average of US$270 per aca- increasing the amount of the loan by more demic year, roughly 16 percent of gross than 250 percent (from D300,000 per month national income per capita, higher than in to D800,000 per month, or US$15 to US$41 most OECD countries, on par with the per month) and lengthening the repayment Republic of Korea, and lower than in Chile. period. The scheme now supports 29 percent The fees in private tertiary institutions are of students enrolled in 103 universities. similar. When living expenses are included, Despite these positive steps, challenges an average student has to spend nearly remain, mainly to support greater equity in US$400 a year.9 the access and completion of ethnic minor- Like Vietnam, Mongolia has also relied on ity groups. Persistent inequity in comple- a fairly effective mix of loans and scholarships tion rates indicates that the mechanisms are to improve access to higher education. The better at supporting initial enrollment than State Training Fund provides about 28 per- retention, which may require an examina- cent of funding to tertiary education through tion of the way these mechanisms support grants and loans, with a need-based compo- students while enrolled. Tying some of these nent.10 On average, a recipient student would 108 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

receive 80 percent of tuition. These instru- between US$732 and US$879. In addition ments help offset the cost of education for to these fees, associated living costs are about 40 percent of students in the subsector, also high in China. which may explain the higher coverage and A combination of monetary instruments relatively lower inequity in Mongolia. While eases the barriers for the disadvantaged, but a signifi cant fraction of the grants are need- the outcomes have been mixed.12 The govern- based, they could be better targeted because ment has instituted both need-based and merit- of legislation that expanded eligibility to based scholarship programs to cover tuition the children of civil servants. Today, about and living expenses. A national school-based 40 percent of recipients are the children of loan program offers loans through commer- civil servants (table 4.7), though this propor- cial banks, to be repaid within six years of tion has been declining slightly. If loan recipi- graduation. And loans by the National Devel- ents are employed for eight consecutive years, opment Bank must be repaid within 10 years fi ve in a rural area, the loans are forgiven. after graduation. Universities have the autonomy to admin- China. Tuition fees in 2004 were ister fi nancial aid—based on need or merit. prominent sources of fi nancing and in the But researchers have noted shortcomings. absence of financial aid could present a Need-based aid, administered by the state large cash-constraint barrier to enrollment and by universities, is granted only to stu- and completion, particularly for the poor- dents majoring in the sciences or attending a est students. In Beijing average university fi rst- or second-tier university. And the aver- tuition fees per year range from US$615 to age scholarship amount, only about 15 per- US$806.11 Tuition fees are set differently cent of the average tuition fee, is too small to for different courses and programs. The really make a difference (table 4.8). The stu- tuition for science and engineering ranges dent loan program, not income contingent, from US$674 to US$806 in these univer- covers only a small proportion of students. sities, and for languages and medicine, A short repayment period is also taxing for

TABLE 4.7 Mongolia: State Training Fund recipients, by program area

2003 2004 2005 Grants Number Percentage Number Percentage Number Percentage Need-based grants 8,119 23.3 13,294 33.5 13,831 33.5 Disadvantaged group grants 2,216 6.4 2,454 6.2 2,149 5.2 Merit-based grants1530.41490.31260.3 Public employee family grants 15,915 45.7 16,335 41.2 16,428 39.8 Loans 8,409 24.2 7,390 18.6 8,696 21 Total 34,81210039,62210041,230100

Source: World Bank 2010d.

TABLE 4.8 China: Net payment in regular universities, 2008

Net payment Average Living cost Scholarship Work study Net payment (% of gross tuition (US$) (US$) (US$) Loan (US$) (US$) (US$) family income) 895 141 138 62 37 799 30

Source: Min et al. 2009 in Ma 2010. FINANCING HIGHER EDUCATION 109

many recipients. Overall, even after scholar- FIGURE 4.4 Indonesia: Tertiary education expenditure as a ships and loans, students still need to spend percentage of annual household income on average about 30 percent of their gross family income on higher education (fees and 50 living costs), a ratio that is certainly higher 45 45 for lower-income families. Other lower- and middle-income coun- 40 tries in the region have had a more piecemeal 35 32 approach to tackling inclusiveness issues. 30 26 Indonesia and Cambodia have relied mostly 25 on scholarships, and Thailand and Malaysia 20 mostly on loans, with mixed effects.13 15 % of household income 10 Indonesia. Private spending, primarily 5 for tuition fees and levies, constitutes the 0 14 bulk of financing for higher education. public education private education all education In 2009 the average spending per student per year was about US$2,200 in public institutions (tuition fees, student fees, and Source: World Bank 2010c. living expenses) and about US$1,200 in private institutions. Supporting one tertiary education student can cost up to a third of Cambodia. Scholarships are the primary yearly income (fi gure 4.4), a share likely to means of helping students overcome cost be much higher for the poorest quintile, constraints. In 2000 a circular from the and even higher if the student is in a public prime minister was issued to regulate the institution. proportion of fee-paying students in public To spur demand among the poor, the higher education institutions and stated that government introduced a full and partial a third of students enrolled in public higher scholarship scheme, but the scheme is tar- education institutions could be fee paying, geted to students already enrolled in tertiary with the remaining two-thirds on schol- education, ignoring high school leavers who arship.17 Scholarship students do not pay do not have the economic means to enroll. tuition fees, but the share of scholarships This may explain why enrollments in ter- is modest and declining: 15.3 percent of tiary education remain very inequitably dis- students held scholarships in 2004–05 and tributed in the country. Moreover, fi nancial 12.3 percent in 2007–08. aid awarded to students enrolled in tertiary Over the past few years, Cambodia’s education covers only 3 percent of the cost,15 Department of Higher Education has devel- even though, according to national law, stu- oped the selection processes to use the schol- dents are required to cover only up to 33 per- arship scheme to promote equitable access cent of tertiary educational costs. This may to higher education, improving the equity explain why Indonesia is the only country of access for the poorest (see figure 4.3). where completion gaps between the poorest But much more remains to be done. Of the and wealthiest population quintiles have been scholarships awarded, 60 percent are based increasing (fi gure 4.4). While 20 percent of on merit and 40 percent go to priority can- students from the poorest quintile are eligible didates. Of the priority scholarships, females for scholarships, these groups rarely receive receive 15 percent, the poor 15 percent, and scholarships because of narrow eligibility those from rural areas 10 percent. So the pro- criteria (scholarships are also merit based). portion of need-based scholarships remains Overall, scholarships cover only 5.6 percent low. Moreover, the financial support has of the student population.16 been signifi cantly less than the cost of study 110 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

and living (in effect, this is merely a tuition- has increased participation. To help lower- free mechanism). income students, government loans have to be repaid over 15 years at a 1 percent inter- Thailand. The poorest household spends on est rate. To faciliate administration, the average US$112 per month in higher educa- loans can be provided directly by universi- tion, about one-eighth of the spending of the ties (box 4.2). richest household.18 But for the poorest fam- Results have been encouraging: there is ilies, private spending on education is about evidence of significant effects on the par- 60 percent of their total income, and for the ticipation of the poorest to higher education. wealthiest, less than 1 percent (fi gure 4.5). These results could be improved with better Sending a child to school thus represents a targeting. There is also evidence suggest- signifi cant fi nancial burden for poor fami- ing that universities have extended loans to lies, not only because of high costs associ- underqualified applicants to boost enroll- ated with attending higher education but ment.20 The government is now attempting also because of the opportunity costs of for- again to introduce an income-contingent loan gone earnings. system. These fi gures do not include loans, and Thailand has increased access to higher edu- Malaysia. Tuition fees tend to be lower cation through student loan programs.19 than in other countries, but total living costs Thailand has also instituted a grant and are higher. Loans are the primary form of scholarship scheme to increase tertiary fi nancial aid for higher education, but the access among the poor, such as the One mechanism does not work equally well District Scholarship and scholarships for across fi elds and could be better targeted. A low-income students, but coverage has study to assess the effects of socioeconomic remained very limited. It is Thailand’s stu- status on university education and social dent loan program for needy students that and economic mobility at the University of

FIGURE 4.5 Thailand: Private expenditure in higher education, by income quintile

3,000 70

2,500 60

50 2,000 40 1,500 baht 30 1,000

20 expenditure % of total

500 10

0 0 poorest second third fourth richest income quintile monthly expenditure on education (left axis) educational expenditure (right axis)

Source: Socio-Economic Survey 2006, as cited in World Bank 2009a. FINANCING HIGHER EDUCATION 111

BOX 4.2 Evaluating the Student Loan Fund in Thailand

The Student Loan Fund, instituted by the Thai gov- nonpoor, compared with 19 percent of undergradu- ernment in 1996, was to increase access for low- ate student loan recipients. The success in targeting income upper-secondary, vocational, and under- upper-secondary students can be attributed partly to graduate students. The loans cover tuition fees, the closer relationships schools have with students education-related costs, and living expenses. Only and their families. students from households earning less than US$4,300 There were signiﬁ cant effects on the decision to a year are eligible for the need-based fund. Over the participate in higher education for those students ﬁ rst decade of implementation, US$5.7 billion was whose family income was close to the poverty line,a disbursed to more than 2.6 million students. but few effects on students with somewhat higher A recent evaluation examined the fund’s target- income, implying that the income limit for the loan ing and whether it increased access for low-income was set too high. students. Upper-secondary students were better targeted than undergraduate students. Only 7 percent Source: Tangkitvanich and Manasboonphempool 2010. of student loan recipients in upper secondary were a. The poverty line is approximately US$888 (B28,650) per year.

Malaya concluded that lower-income stu- Summary. First, countries should move dents were bound for certain disciplines, to need-based scholarships, because whereas upper-income students dominated merit-based ones do not promote inclusive- fi elds with traditionally higher returns.21 ness. Second, comprehensive packages— While benefiting both public and pri- including fee deductions for disadvantaged vate sector institutions, a positive feature groups, need-based scholarships, and stu- rather unusual for the region, Malaysia’s dent loans—have been more effective in student loan program could also be better addressing inclusiveness than more piece- targeted. The National Higher Education meal approaches. Third, the effectiveness of Fund Act 1997 provides for student loans to specifi c instruments (separate or in a wider faciliate access to public and private higher package) has varied significantly across education: full loans to students with fam- countries depending on design and imple- ily incomes below US$900, partial loans to mentation. Scholarships and loans have cover tuition fees and part of living expenses been more effective when covering a signifi - to students with family incomes between cant proportion of tuition costs and at least US$1,001 and US$1,380, and partial loans some living costs. Broad-based application to cover tuition fees for students with family across universities and fields has worked income greater than US$1,381.22 Together better than more selective application, when with affirmative action policies pursued carefully targeted to disadvantaged groups. since 1971, these loans were instrumental in increasing Bumiputera enrollment in univer- How to fund priority activities sities (reaching 60 percent of overall enrollment in 2006). But while the coverage of the How should underfunded activities be scheme and the volume of loans dispensed financed? There is a clear case for public have increased, it is not means tested, imply- financing to support research and STEM ing that there could be signifi cant leakages capacity—and to address inclusiveness. to the nonpoor. Very few loan recipients are Research and STEM are two areas with not Bumiputera.23 high positive externalities. While the initial 112 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

costs can be high, the social benefi ts are even spending signifi cantly more than Japan and higher, particularly in relation to their link Korea. with innovation. Private fi nancing tends to These trends are the result of how coun- underfund these activities. For inclusiveness tries deal with funding trade-offs and tight the case is based on the failures in capital budgets for the allocation of public funding markets: cost constraints are binding. So far, across levels of education. In a few high- most scholarships have indeed been fi nanced income economies in the region, Japan and with public money in the region. Korea, for example, the long-standing policy Countries thus need to assess their ability to has been to prioritize public funds for earlier allocate more public spending to higher edu- levels of education, a policy stance that has cation and, even more important, to fi nance had important implications for the strategies key activities. This will help higher educa- to expand and improve higher education over tion institutions address skill and research the years. Japan and Korea have relied heav- disconnects—providing students with better ily on private funding (both through public skills and increasing the talent pool entering and private delivery). Middle- and lower- tertiary education, and building the capacity income countries in the region have not mobi- of institutions for research. lized similar absolute levels of private sector The scarcity of public resources requires fi nance for their systems. While they should better targeted resources and performance- strive to do more, public funding will remain based allocations. More efficient financing critical. will also improve the match between skill Comparing public expenditure per tertiary provision and needs (first disconnect) and pupil as a share of GDP per capita with coun- research provision and firm needs (second tries outside East Asia, spending is generally disconnect) by tying funds to relevant skills on the low side in East Asia’s middle-income and research. countries even compared with other middle- To maximize the leverage of public funds, income countries in other regions: Brazil, it will be important to attract more private India, and Mexico all outspend them, apart funds and correct market failures by provid- from Malaysia (fi gure 4.7). ing student loans. Within a coherent fi nanc- In a troubling trend, expenditure mea- ing framework, private funding would not sured this way has also tended to decrease only complement public funding in fi nancing across the region, especially in Cambodia and some of the above activities but also focus on Lao PDR, where absolute per pupil tertiary system expansion and diversifi cation (target- expenditure was already low (figure 4.8). ing some other country priorities, such as Middle-income countries have also reduced increasing enrollment or service-related disci- such spending but, except for Malaysia, by plines) through public and private delivery. much smaller margins. These data suggest that several countries may have scope for further public spending Increasing and prioritizing in tertiary education, but do they have the public spending potential? A look at the ratios of tertiary Public spending ratios vary substantially in spending to total education spending and of East Asia and are not necessarily lower than total education spending to GDP casts light in high-income East Asia (fi gure 4.6). Ter- on this issue. According to these two indica- tiary spending in relation to GDP is lowest tors, Mongolia and Thailand are the coun- in the Lao People’s Democratic Republic, tries with most potential. Cambodia, and the Philippines. Ratios are Hong Kong SAR, China; Singapore; and higher in Vietnam, Indonesia, and particu- Malaysia spend the most on tertiary educa- larly Malaysia. Spending ratios also differ tion as a share of total education, at around substantially across high-income East Asia, 30 percent (figure 4.9). Japan and Korea with Hong Kong SAR, China, and Singapore allocate less than 20 percent of their public FINANCING HIGHER EDUCATION 113

education spending to tertiary education, as FIGURE 4.6 Public tertiary expenditure as a share of GDP, latest do the Philippines, Thailand, Mongolia, Lao available year PDR, and Cambodia. A comparison with lower- and middle-income countries outside the region can suggest some room for real- Japan 0.61 Singapore 1.07 location within education budgets. Hong Kong SAR, China 1.03 But the real room for intrasectoral reallo- Korea, Rep. 0.63 cation in favor of tertiary education is proba- Malaysia 1.69 bly not very large, given the many competing Thailand 0.71 24 needs in other parts of the education system Indonesia 1.2 and the fairly low overall education spending Philippines 0.34 as a share of GDP in several countries (fi gure Mongolia 0.93 4.10). This is particularly so for Lao PDR, Vietnam 1.18 Cambodia, and the Philippines. Thailand Lao PDR 0.21 and Mongolia may have more leeway. (Ter- Cambodia 0.05 tiary education shares have also been grow- 0 0.5 1.0 1.5 2.0 ing in most countries since the beginning share of GDP (%) of the decade, thus providing less scope for reallocation.) Source: UIS Data Centre. To increase spending in tertiary education, Lao PDR, Cambodia, and the Philippines should consider increasing their education and lower-income and lower technology clus- spending in relation to GDP. High tax shares ter countries clearly do not have the same to GDP would suggest more potential for room or even scope for increases in research public spending increases. as other countries. But further effort is Given the many competing needs, pri- needed. And in a broader sense, funding for oritizing public spending and improving research can also support the development of effi ciency in the allocation and use of pub- future researchers for both universities and lic funding will be critical. In cases such as the private sector. Scholarships for talented Vietnam, Indonesia, and Malaysia, which students in Thailand illustrate this option already spend above the average in tertiary well (box 4.3). education, a fi rst imperative will be to ensure higher public spending shares for research, Increasing effi ciency of public spending STEM, and scholarships. Further evidence comes from comparing overall public ter- Prioritizing public spending will be neither tiary spending in relation to GDP with ter- successful nor suffi cient without effi ciency tiary research spending in relation to GDP improvements. (figure 4.11) across East Asian economies. While these two indicators are not strictly Effi ciency gaps comparable (research and development funds Most countries could be more efficient include both public and private funds), they in the way they use and allocate public offer insights into the different priority given resources—as is also evident from the large to research within public budgets for higher spending gaps. For example, a very simple education. It is definitely much lower in comparison of higher education outcomes— lower- and middle-income economies than in the STEM share and the number of journal upper-income ones. In the middle technology articles—and spending indicators25 suggests cluster, China leads. that Cambodia, Malaysia, the Philippines, Lower- and middle-income East Asia need and Vietnam are less efficient than Korea to increase public spending for research in ter- in supplying STEM, and that Malaysia and tiary education. There is no ideal benchmark, Thailand are less efficient in translating 114 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 4.7 Public expenditure per tertiary pupil as a share of GDP per capita, 2006–07

a. Low-income economies, East Asia 70 60 50 40 30 20 10

share of GDP per capita (%) share 0 Cambodia Lao PDR Vietnam

b. Middle-income economies, East Asia c. Other middle-income economies 80 80

60 60

40 40

20 20 share of GDP per capita (%) share 0 of GDP per capita (%) share 0 Malaysia Mongolia Philippines Thailand Brazil Chile India Mexico

d. High-income economies, East Asia e. Other high-income economies 80 80

60 60

40 40

20 20

0 0 share of GDP per capita (%) share Hong Kong SAR, Japan Korea, Rep. of GDP per capita (%) share Finland United Kingdom United States China

Source: UIS Data Centre. Note: Although some countries have 2008 data, the 2006–07 range was maintained for comparability across countries.

funds into journal articles than Japan and within teaching on non-STEM fi elds. Malay- Korea (fi gures 4.12 and 4.13). sia’s high spending indicators in relation to Some of these differences may simply be the outcomes suggest some ineffi ciency. caused by different levels of private resources Beyond the poor targeting of their schol- invested into the system, either through fees arships, most countries spread their public in public institutions or private delivery. And resources for teaching and research too thinly they suggest different uses of funds as much among institutions, and few have competitive as ineffi cient use and allocation of these funds (or other performance-driven) funding to (for instance, a limited priority on STEM trigger systemic change. spending). But there is no obvious relation Governments across the world are turn- between private fees in public institutions ing to allocating resources to develop premier and results, and private delivery tends to be research (and teaching) universities. This more focused on teaching than research and move toward directing substantial resources FINANCING HIGHER EDUCATION 115

to a few premier institutions is partly a FIGURE 4.8 Public expenditure per tertiary pupil as a share response to the large resources universities of GDP per capita, 1998–2009 need to undertake high-level research (and a. Low-income economies teaching), which makes selectivity impor- 150 tant. Harvard University had an endowment of US$37 billion, and annual spending of US$3.2 billion, and per student spending of 120 US$105,041 before the crisis. Other top universities have expenditures in similar orders 90 of magnitude. Dedicating substantial resources to a few 60 premier institutions that attain international standards, in teaching or research, can be share of GDP per capita (%) share one way to reverse a trend of low-cost and 30 low-quality education. Pursuing this path, governments need to identify programs or 0 departments (not necessarily an entire insti- 1998 2000 2002 2004 2006 2008 2009 tution) that have good potential. Lao PDR Nepal Bangladesh Cambodia

Most high-income (and a few middle- b. Middle-income economies income) East Asian economies have started 150 moving this way. Through the World Premier International Research Center Initiative, 120 Japan intends to make Tokyo, Tohoku, Kyoto, and Osaka universities, as well as 90 the National Institute for Material Sciences, a public research institute, its core research 60 bodies.26 Korea is raising the caliber of its universities through the Brain Korea 21 share of GDP per capita (%) share 30 Project.27 The government in Hong Kong SAR, China, has set up ﬁ ve research centers 0 under the Hong Kong SAR, China, Research 1998 2000 2002 2004 2006 2008 2009 and Development Centers Program to lead India Mexico Malaysia Brazil Thailand and focus research on 13 priority sectors.28 In Mongolia Indonesia Chile Philippines Taiwan, China, the authorities have allocated c. High-income economies a budget of NT$65 billion for the Developing 150 Top Grade Universities and Research Centers to be spent between 2008 and 2015. 120 Some middle-income countries also have started to move in this direction. Malaysia 90 identified four universities as its main research universities. China identified 100 research universities for increased support 60 (about 6 percent of its universities). With these exceptions, however, most countries in of GDP per capita (%) share 30 the lower and middle technology cluster are still thinking of roughly equal allocation of 0 resources across institutions. Illustrating a 1998 2000 2002 2004 2006 2008 2009 new mindset, the Philippine government will Hong Kong SAR, China Finland Singapore United Kingdom spend US$70 million in 10 years to set up the United States Japan Korea, Rep. Engineering Research and Development for Source: EdStats database. 116 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 4.9 Public tertiary education expenditure as a share of total public education expenditure, 2006–07

a. Low-income economies, East Asia b. Other low-income economies 40 40

30 30

20 20

10 10 share of public share share of public share 0 0 education expenditure (%) expenditure education Cambodia Lao PDR Vietnam (%) expenditure education Bangladesh Ghana Nepal

c. Middle-income economies, East Asia d. Other middle-income economies 40 40 30 30 20 20 10

share of public share 10 share of public share 0 0 education expenditure (%) expenditure education Malaysia Mongolia Philippines Thailand Brazil Chile India Mexico South Africa education expenditure (%) expenditure education

e. High-income economies, East Asia f. Other high-income economies 40 40

30 30

20 20

10 10 share of total public of total share share of total public of total share 0 0 education expenditure (%) expenditure education education expenditure (%) expenditure education Hong Kong Japan Korea, Rep. Singapore Finland United Kingdom United States SAR, China

Source: UIS Data Centre. Note: Although some countries have 2008 data, the 2006–07 range was maintained for comparability across countries.

FIGURE 4.10 Overall public education expenditure as a share of GDP, latest year

World 4.53 Japan 3.46 Singapore 3.28 Hong Kong SAR, China 4.49 Korea, Rep. 4.23 Malaysia 4.13 Thailand 4.12 China 1.91 Indonesia 2.82 Philippines 2.80 Mongolia 5.60 Vietnam 5.34 Lao PDR 2.27 Cambodia 1.60 0123456 share of GDP (%)

Source: WDI database. FINANCING HIGHER EDUCATION 117

Technology Program, a consortium of eight FIGURE 4.11 Research and development spending and top engineering universities.29 total public spending in tertiary education, latest available Some gradualism for investing in centers year of excellence is warranted. It is important to create an environment that allows excellence Japan to emerge, and once identifi ed, to build on the comparative strengths of various institutions Singapore and their academic offerings. This requires Hong Kong SAR, China setting up a more autonomous and competitive governance framework for all universi- Korea, Rep. ties (discussed in the next chapters). Also Malaysia useful is a competitive (or other performance- driven) process to target the resources at the Thailand best research institutions. In Japan research China funding is allocated through competitive bid- Indonesia ding rather than being distributed in prede- termined amounts. More generally, perfor- Philippines mance-based allocations can trigger systemic Mongolia changes in quality—improving the effectiveness of both teaching and research (even in Vietnam fi elds that have less scope to be fi nanced with Lao PDR public funds). 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Lower- and middle-income East Asia still % of GDP have a long way to go to improve alloca- tertiary spending tertiary R&D spending tion mechanisms of public funds to improve skills and research. Most countries, usually in a centralized system, still use historically Source: UIS Data Centre. Note: Data on tertiary spending as a share of GDP for China are estimates. negotiated budgets (table 4.9), leading to

BOX 4.3 Overseas scholarships for outstanding students from Thailand

The Thai government has various scholarship and to secondary school graduates each year to continue loan programs. Some are explicitly targeted at the undergraduate study in foreign countries. The candi- poor to increase access among disadvantaged groups. dates are selected by academic performance, writing Others are merit based and aim to develop the techni- tests, and interviews. cal skills of the next generation of Thai leaders, both Anandamahidol Scholarships. The Anandama- in academia and government. Among the two most hidol Foundation was established by King Rama prestigious of these types of scholarships are the King IX to provide graduate-level study scholarships for Scholarships and the Anandamahidol Scholarships. students willing to continue postgraduate studies in King Scholarships. The King Scholarships were high-income countries. The scholarships’ main goal established in 1897 by King Rama V and are awarded is to support students pursuing degrees in one of the to outstanding students every year. The scholarship eight ﬁ elds identiﬁ ed as crucial for national develop- presentation was stopped in 1932 because of politi- ment, including science and technology. cal instability but was restored in 1964 by the current king, Rama IX. Several scholarships are given Source: World Bank 2009a. 118 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 4.12 Tertiary spending per student as a share of GDP per institutions, countries can foster higher capita and STEM enrollment share, latest available year quality and address critical research and skill disconnects. 50 Competitive funds, by promoting excel- 45 lence in research and teaching, can support Malaysia 40 Korea, Rep. more selective resource allocation. Higher education institutions submit their fund- 35 ing proposals along with their development 30 Philippines plans, with key performance indicators.30 25 Vietnam The funds are then used to fund equipment, 20 Cambodia facility improvement, and staff upgrading and 15 % of GDP per capita professional development. Part of the appeal 10 of competitive funds is that they can reduce 5 the incentives to use enrollment expansion to 0 0 10203040506070 sustain fi nancial viability—and they encour- STEM share (%) age faculty to devote time to teaching and research, improving quality across the system (box 4.4). Student fees can then be used by Sources: EdStats database; UIS Data Centre. institutions for supplemental investments. To improve relevance and research capacity, FIGURE 4.13 Tertiary spending per student and journals per funding could be awarded to the disciplines million people, latest available year and courses related to labor market needs. One of the most effective means to hold 900 institutions accountable is the government’s 800 power of the purse. In theory a government could simply remove or reduce fund- 700 ing to institutions that fail to comply with 600 approaches or fall behind on goals. But 500 Japan reducing funding for public universities, 400 Korea, Rep. particularly original budget allocations, is 300 extremely diffi cult. One way around this is 200 to have diverse fi nancing mechanisms beyond journals per million people the base fund, to incentivize institutions’ per- 100 Thailand Malaysia 0 formance. At a minimum a formula funding 0 2 4 6 8 10 12 14 allocating funds in proportion to the number tertiary spending per student, thousands of dollars of students is already a move in the right direction for negotiated budgets. Several mecha- Source: EdStats database; UIS Data Centre. nisms can enhance performance,31 beyond the competitive funds reviewed above: inefficiencies. Only China, Malaysia, and Performance contracts. Governments sign Thailand have moved to formula funding. regulatory agreements with institutions to set Indonesia has also been experimenting with mutual performance-based objectives. competitive funds and performance-based Performance set-asides. A portion of public grants. Mongolia and Vietnam started funding for universities is set aside to pay, on introducing some competitive funding for the basis of various performance measures. research. Payment for results. Output or outcome measures are used to determine all or a por- Moving forward tion of the funding formula. For example, By being more selective and performance universities are paid for the number of stu- based in resource allocations across dents that graduate, sometimes at higher FINANCING HIGHER EDUCATION 119

TABLE 4.9 Block grants in East Asia

Historically Formula Voucher-based Performance-based Economy negotiated budget funding allocation contracts High income Japan ● Singapore ● ● Hong Kong SAR, China ● Korea, Rep. ● ● Middle income Malaysia ●● Thailand ●● China ●● Indonesia ● a Philippines ● Low income Vietnam ● Lao PDR (National University of Laos) ● Cambodia ●

Sources: Raza 2010 on the basis of Expert Survey 2010, except Singapore, which was taken from Ministry of Education, Singapore 2005. Note: = noneրnot applicable; = partial. a. Formula funding incorporating performance indicators piloted in fi ve autonomous universities (and some nonautonomous universities) in Indonesia.

BOX 4.4 Competitive funds as an innovative fi nancing tool

A few countries—including, to a degree, Indonesia, they can handle complex indicators of impact and Mongolia, and Vietnam in East Asia—have com- success tailored to institutional characteristics and petitive funding schemes to allocate resources to needs. higher education institutions. Output oriented and Competitive funds over the past few years targeting a range of sector issues, they can increase have supported activities ranging from traditional cost-effectiveness and enhance educational quality investments to systemic reform. In Bangladesh and and relevance. Indonesia they have lifted quality by providing Under most competitive funds, institutions compete grants for equipment, libraries, laboratories, build- for investment on the basis of their own planning and ings, and university programs. choices. An independent body selects the best projects A critical component of a successful competitive based on potential, performance, and track record, fund is capacity: the government needs to set the encouraging a culture of fair competition. Its choices rules of eligibility, selection, and implementation, are based on transparency in objectives, eligibility, and institutions to supervise projects. and other criteria—and usually peer evaluation. On the downside, these funds offer less predict- Competitive funds offer several advantages. ability than formula funding, but they can stimulate They provide incentives for institutions to perform. systemic change. They encourage institutions to clarify their mission and strategy and plan for the medium term. And Source: World Bank staff reports.

prices for graduates in certain fi elds of study and equity, though they may also use them to or with specifi c skills. penalize the institution for underperforming. Many countries use payment for results The Indonesian government is using per- and performance contracts to improve insti- formance-based grants in the form of block tutional accountability for quality, effi ciency, grants for autonomous universities (and 120 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

some nonautonomous universities). This Performance-based funding requires cau- allows flexibility and some accountability. tion in its design, particularly performance Performance-based grants provide budget- set-asides, if there are no clear precedents. ary support, conditional on the universities’ The value of performance set-asides depends meeting certain performance targets at the on the amount and the indicators to assess department and unit level.32 institutional success. The set-aside can vary In Singapore, to accompany the cor- from 5 percent to 100 percent, and indica- poratization of the National University tors can also vary from 1 to 12 or more. But of Singapore and the National Techno- South Carolina provides a cautionary tale in logical University in 2006, the govern- the United States. South Carolina allocated ment adopted new ways of holding them 100 percent of its recurrent budget on numer- accountable. Besides requiring an external ous performance indicators. The program quality assurance process, the government failed simply because institutions could not mandates policy agreements and perfor- comprehend a clear vision of success because mance contracts. Policy agreements allow of the many indicators and standards built the Ministry of Education to provide stra- into the system.35 tegic direction to the higher education sec- Countries can also improve the targeting tor with clear goals for universities to guide of scholarships and loans to the poorest and them in formulating policy and ensure that disadvantaged groups. Targeting is a par- they are following the necessary conditions ticularly serious issue in Cambodia, Indo- to receive government funding. Performance nesia, and Mongolia. To improve targeting, contracts are established with each univer- governments need the capacity to identify sity for fi ve years. These set out the goals disadvantaged students, inform them of aid for teaching, research, service, and organi- opportunities, and monitor their perfor- zational development. Linked to each goal mance once enrolled. are clear targets and performance indica- Finally, efficiency gains can be made at tors. The ministry also sets workforce tar- the level of individual public institutions. gets linked to public fi nancing.33 Scholars have noted two main ways for Korea has also focused more on account- governments to promote internal effi ciency ability for public fi nancing. In 1995 all major and sustainability: by moderating costs to funding programs were restructured along conserve resources and by maintaining or the lines of performance contracts. The gov- increasing the rate at which students com- ernment evaluates the institutions’ achieve- plete their programs and receive degrees.36 ments against preset criteria. Under the Brain Korea 21 project and the New Universities Mobilizing private funding and for Regional Innovation project, contracts are developing student loans established between participating institutions and the government. If the institutions breach Recent research points to two interrelated the contract, they face government-imposed elements of successful pro-equity financ- penalties, usually fi nancial. ing for higher education: variable fees for These mechanisms may, however, have led students and income-contingent loans. to excessive regulation in Korea, highlight- Where these instruments have been imple- ing the diffi culty of achieving the right bal- mented well, higher education systems have ance between accountability and autonomy increased access for the poor and disadvan- (see chapter 5).34 More generally, there may taged while recovering costs. be some tension between performance contracts and higher accountability to university Variable fees boards if the contracts limit the capacity of Countries across the globe charge vary- the boards to freely fi x at least some institu- ing tuition fees, depending on their higher tional objectives. education structure. Japan, Korea, and the FINANCING HIGHER EDUCATION 121

United States, all countries with steep pri- systematically differentiated fee structures, vate sector participation in higher education, determined either by the government or by charge fairly high tuition fees as measured the universities. Vietnam has a partly dif- by percentage of GDP per capita. Canada ferentiated fee structure (through partial fee and the United Kingdom, by contrast, have deductions, and liberty to fi x higher fees for larger public sector financing and lower students enrolled outside the regular stu- tuition fees. dent quota), a strategy applied with some Variable (or liberalized) fees—set by uni- success. Combined with redistributive poli- versities—offer several benefi ts over a fl at fee. cies to promote access, variable fees can be They can increase the resources entering the progressive. Despite lack of formal evalu- higher education system by being open ended, ations, Mongolia and Vietnam argue for and they can increase competition among need-based scholarships. Supporting stron- universities, increasing quality and relevance, ger university-industry links and nontuition as well as the effi ciency of resource use.37 And private resources are other ways to leverage by being akin to income transfers to targeted private funds (box 4.5). income groups, they have the potential to be more equitable than other approaches to rev- Income-contingent loans enue generation, especially when they are set A second pillar of an effective financing at higher rates for those who can afford them scheme for higher education is income- and are combined with redistributive policies contingent loans. Traditionally, student loan to help poorer students pay those fees.38 programs have been either a conventional In most variable fee schemes, the govern- mortgage-type loan (the loan of choice in ment usually places a ceiling on the maxi- many countries, which usually requires mum39 and has most students make at least repayment after graduation with varying some contribution toward their education, repayment periods) or an income-contingent though exempting qualified poor students loan. Capital for a mortgage-type loan41 from fees based on need and equity helps may come from the government or a lend- ensure that they are not excluded or sent to ing institution. If the government is not the low-cost and possibly low-quality institu- lender underwriting the loan, it can provide tions. Otherwise, requiring students to pay interest subsidies and default assurance, at least some of the cost generally improves while allowing the lender to administer the their motivation and performance. Of course, loan. These types of loans can be means governments will differ in what the variable tested, and target students below a certain fee structures and ceilings should be to ensure income threshold, or they can be academic access, equity, and cost-recovery.40 merit loans, which target students who score East Asia has more scope for using high on entrance or exit exams. To ensure variable fee schemes, by applying more equity, loan designers also need to take into account the choice of higher education institution (public or private, accredited or not), TABLE 4.10 Average tertiary tuition fees location (in state, out of state, foreign), and percentage of GDP per capita status of study (full time or part time)—all Economy Tuition fees factors that may affect loan access. Australia 34.8 More governments recognize that income- Canada 20.6 contingent loans are better for access Japan 60.3 (box 4.6). Repayment is contingent on the Korea, Rep. 56.7 future income of the borrower: people with New Zealand 38.5 low earnings make low repayments, and peo- United Kingdom 18.5 ple with low lifetime earnings do not repay United States 36.7 the loan principal in full. Such a loan pro- Source: OECD 2008b. tects a student from excessive risk and can 122 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 4.5 Matching funds in Hong Kong SAR, China, and Singapore

One of the key reasons Hong Kong SAR, China, and each institution could access by raising donations to Singapore have mobilized public and private funding that amount. This structure gave smaller institutions for university research is that they both have effec- a fair chance to raise funds while encouraging com- tive government matching-fund programs and favor- petition among institutions and raising the profi le of able tax incentives. By using public-private partner- private philanthropy. ships, they have strengthened the capacity for raising Singapore has also mobilized large funds for uni- the independent income of colleges and universities versity research through its matching-fund program. and contributed to a philanthropic culture support- Although the government has traditionally invested a ing higher education. relatively larger share of public resources in university Starting in 2002, Hong Kong SAR, China, pro- research, it has augmented this investment by encour- vided nearly HK$7 million of seed money to 12 aging private participation. Starting in 1991, it began institutions to improve fund-raising capacity. To encouraging philanthropic support to research uni- encourage private donations to higher education, the versities with a matching ratio of 3 to 1. Private dona- government also raised the ceiling for tax-exempt tions were also eligible for double tax deductions. donations from 10 percent of income or profi ts to The success of both schemes points to strong 25 percent. And to support research in universities, institutional capabilities, as well as conducive legis- it created a national fund of HK$1 billion for match- lative climates and applicable tax laws. ing grants on a ratio of 1 to 1. Under the scheme, the government set a fl oor—a guaranteed minimum that Source: Sutton Trust 2004.

BOX 4.6 Advantages of income-contingent loans

Income-contingent loans have three main advan- of threshold economic means, whereas govern- tages over conventional mortgage-type loans. ment income-contingent loans are available to most Credit reputation. A bank- or government- students. guaranteed mortgage-type student loan protects the Repayment. Conventional mortgage-type loans bank or the government in case of default but offers are generally characterized by a repayment period no protection to the borrowers. If a student is unable where the borrower makes set payments over a spec- to fi nd employment that enables him or her to repay ifi ed period. Without the income-contingent part of the loan, he or she may have to declare bankruptcy, the loan, a borrower may have diffi culty in repaying impairing access to credit later in life. the borrowed amount plus interest. Access to loans. Conventional mortgage-type bank loans usually are available only to students Source: Johnstone 2004.

promote efﬁ ciency (by the protection from government’s cost of borrowing. For students risk) and access (fees ﬁ nanced by the loans (only while they are students), this means free resources for access). that their tertiary education is effectively free, How large should loan entitlement be? financed through taxation and an income- Experts argue that it should be enough to related government contribution. While they cover tuition fees and living expenses.42 And are students, they pay nothing at the time, it should carry an interest rate similar to the a fact compatible with government efforts FINANCING HIGHER EDUCATION 123

to improve access. Ministries of fi nance, of subsidy in the sense that the borrower is not course, bear the burden of upfront costs and required to pay back the full value of the loan receive repayments only later. Income-contin- received, thanks to below-market interest gent loans have been applied with varying suc- rates on the loan, periods when no interest cess across the world, starting with Australia is levied on outstanding debt (both during (box 4.7). Most other countries with income- study and in grace periods after study), and contingent loan programs are in Europe. repayments not linked to infl ation. But student loan schemes are not with- Some general steps can improve the out financial risk. They may differ in the fi nancial viability and cost recovery of loan underlying objectives and in organizational schemes. Governments can reduce the built-in structure, sources of initial funding, student subsidies (hidden grants). They can improve coverage, loan allocation procedures, and the effi ciency of loan schemes through con- collection methods. However, they almost taining administration costs. Or they can always share a common trait: they are highly reduce repayment leaks caused by default. subsidized by governments. Unlike commer- Some countries have taken steps in these cial loans, a sizable proportion of the total directions. The Canada Student Loan Pro- student loan outlay usually is not paid back. gram, for example, charges a zero nominal Experts note that this gap between total loan interest rate, subsidized by the government, disbursements and overall loan recovery is during the period of study, whereas the post- the result of two elements: (a) built-in inter- study repayment rate of interest is high (prime est rate subsidies, incorporated in the design plus 2.5 percent), resulting in an overall loan of the loan scheme, and (b) ineffi ciencies in system repayment ratio that nears 100 per- running the scheme, in substantial repayment cent.43 In the Czech Republic the interest rate default and high administration costs. Lend- charged throughout the period of the loan is ing conditions for almost all government- fairly high (above 12 percent), resulting in an sponsored loans are “softer” than those for overall system repayment ratio of more than regular commercial loans. This is a student 108 percent. Japan’s loan programs have

BOX 4.7 The higher education contribution scheme in Australia

Australia became the first country to adopt a structure was introduced. And the income threshold national income-contingent loan policy to ﬁ nance at which repayment was mandatory was decreased higher education, in 1989. Under the Higher Edu- to $A 23,000. cation Contribution Scheme, all Australian under- Apart from revenue generation, the main aim graduates pay a uniform tuition fee (in 1989, the of the scheme was to improve access, particularly fee was $A 2,250 a year), to be repaid in propor- among disadvantaged youth. It partly succeeded tion to future income. The fee could be either an because—though overall higher education participa- upfront fee, in which case students received a dis- tion increased when it was introduced, and though count, or a deferred fee, in which case repayment students from higher income levels were more likely was delayed until after graduation. The minimum to attend university—the participation rate for income threshold for repaying the loan was set at students from lower-income backgrounds did not $A 27,700 in 1989, with graduates paying 2 per- decrease. And the differential fee structure initiated cent of their taxable income; the rate was progres- in 1997 increased higher education enrollment for sive, rising to 4 percent for those at higher incomes. students from all backgrounds. In 1997, average charges for courses increased by 40 percent, and more important, a differential fee Source: Johnstone 2004. 124 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

charged higher interest rates and achieved an applicants who failed the means test by a slim overall recovery ratio of 68.3 percent (imply- margin; applicants less able to pay for their ing a 30 percent government subsidy). university education received more assistance. Clearly, these options may not be the most A separate non-means-tested loan was devel- attractive for countries confronting the ten- oped for other students in public universities sion between cost recovery and equitable who passed the means test. The interest rate inclusion. The key issue is cost-effectiveness. for these loans is the government’s no-gain- At what level should the built-in loan subsidy no-loss rate, 1.5 percentage points below be set for student groups, particularly the the average best lending rate of note-issuing poor and disadvantaged, to ensure adequate banks. This rate covers the government’s risk revenue, appropriate opportunity, and desired in disbursing unsecured loans. An adminis- outcomes? trative fee is charged annually to cover the Hong Kong SAR, China, and Korea offer full cost of processing and administering the useful lessons for how to reconcile fi scal sus- loans, further enhancing cost recovery. tainability with equity through tiered student If the region expanded such programs loans. Korea enacts no fewer than six differ- more widely, governments would have to ent types of student loan schemes—targeted upgrade the fi nancial management and fi du- to different segments of the population—to ciary capacity of the agencies charged with increase access and promote cost recovery.44 administering and monitoring. These pro- The separate loan schemes administered by grams would require governments to have Korea’s Ministry of Education, the Human systems for identifying qualifi ed individuals Resources Development Fund, and the Korea and calculating their repayment amounts— Research Foundation aim to increase access and ensuring collection. This could require of poor students to higher education. These coordination with the social security or tax loans target poor students, particularly from authority (together with interventions tar- farming and fishing villages; offer interest geted to earlier education levels). rates of less than 1 percent; are administered For most countries in East Asia, the com- through private commercial banks; and are bination of fees, scholarships, and loans46 can guaranteed by the government. Schemes increase equity and access to tertiary educa- administered by the Government Employ- tion. While fees are a necessary form of cost ees’ Pension Corporation, the Korea Teach- recovery, they should ensure the equality of ers’ Pension, and the Korea Labor Welfare opportunity for poorer and more vulnerable Corporation target the children of govern- groups. There is clear potential for scholar- ment employees, teachers and their children, ships when living costs are high and returns and industrial accident victims, respectively. are lower for some disadvantaged groups. But While still highly subsidized by the govern- fairly high higher education returns and cost- ment, they are offered at higher interest rates saving considerations make loans particularly and strongly enforce repayment after gradu- attractive. ation. With their emphasis on cost recovery, they cross-subsidize the equity-focused loan Summary of policy priorities schemes, helping ensure sustainability. Hong Kong SAR, China, is also well East Asia’s low- and middle-income coun- known for its multitiered student loan sys- tries are not delivering the skills and research tem. In the early 1990s, when Hong Kong outcomes they need. Many of the discon- SAR, China, decided to recover more higher nects are related to funding. The precise education costs through tuition fees, the challenges and related priorities differ for Local Student Finance Scheme was separated individual countries, but all countries face into two tracks.45 An Extended Loan Scheme some common imperatives. (not income contingent), charging a higher First, they need to be selective in deciding interest rate of 4–5 percent a year, targeted their targets and priorities. In line with the FINANCING HIGHER EDUCATION 125

analysis of chapter 2, it is clear that enroll- Lower- and middle-income countries face ment increases are not urgent or even advis- the challenge of improving and prioritizing able in all countries. Only a few universities their fi nancing policy for their higher educa- will ever be able to develop credible research tion systems. Meeting this critical challenge capacity (given the high requirements for will bear fruit only within a more flexible faculty qualifications). By contrast, most and competitive higher education system— research, STEM fi elds, and scholarships are the topic of the next two chapters. underfunded, making them a priority area. Second, countries need strategies to fund priority activities. There is a clear case for pub- Notes lic fi nancing to fi nance research, STEM fi elds, 1. OECD 2008b. and inclusiveness. (Private funding would 2. OECD 2008b. complement public funding in fi nancing some 3. As visible from the snapshot presented in activities and focus on increasing coverage table 4.5 and further evidence from Salmi and diversifi cation.) Countries could consider (2009). continuing to mobilize public resources, while 4. Linh, Thuy, and Long 2010. prioritizing them better. And because public 5. Linh, Thuy, and Long 2010. funds are scarce, countries should fi nd ways 6. For instance, fee exemptions introduced at to allocate them more effi ciently, attract more secondary education have been shown to private funds, and correct the source of mar- have increased enrollment and completion ket failures by offering student loans. of secondary school in Vietnam according to Mobilizing and prioritizing public funding. rigorous econometric analysis (World Bank 2010g). Countries should assess the scope for increas- 7. Sakellariou 2010b. ing public spending. The ratios of tertiary 8. World Bank 2008. spending to total education spending, of total 9. World Bank 2010d. education spending to GDP, and of taxes to 10. World Bank 2010d. GDP show the potential for increasing public 11. Ma 2010. The fi gures have been converted spending. On the fi rst two indicators, Mon- from renminbi (RMB) to U.S. dollars at a rate golia and Thailand have more potential than of US$1 = 6.83 RMB as of May 2010. the others. All countries should increase their 12. Loyalka 2009 as cited in Ma 2010; Ziderman shares of public spending for research, STEM 2004. fi elds, and scholarships. 13. The Philippines has done very little overall in Increasing the effi ciency of public fund- terms of equity (Orbeta 2008). 14. World Bank 2010c. ing. Because few countries have much fl ex- 15. Directorate General for Higher Education, ibility to ramp up their tertiary expenditures Ministry of National Education [Indonesia] quickly, more efficient use and allocation 2009. of public funds applies particularly in the 16. Moeliodihardjo 2010. short and medium terms. Greater effi ciency 17. World Bank 2010e. requires being more selective and perfor- 18. These fi gures are not directly comparable mance based in the way public funds for with the amounts in fi gure 4.5, which are cal- teaching and research are allocated across culated in Thai baht. institutions and better targeting of equity- 19. World Bank 2009a. enhancing measures. 20. Ziderman 2004. Leveraging private funds and correcting 21. Singh 1973 cited in Mukherjee 2010. 22. World Bank 2007a. the source of market failures. One way to 23. Mukherjee 2010. increase private funds is to design more effi - 24. For a richer discussion of the trade-offs cient and equitable fee structures. Combined involved in funding various levels of educa- with loan schemes, these policies can increase tion, see Mingat, Ledoux, and Rakotomalala access for the poor and disadvantaged while (2010). The authors provide fi nancing simu- helping to recover costs. lations for Sub-Saharan Africa under fi ve 126 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

varying scenarios for progress toward univer- 35. Salmi and Hauptman 2006. sal primary education under the Education 36. Salmi 2006. for All Fast-Track Initiative Framework, as 37. Barr 2008. well as tertiary education expansion. The 38. Barr 2008. authors draw attention to the implications of 39. This is true at least for public institutions. such scenarios, such as raising the share of Private institutions are generally—and should education in the national budget, reforming be—exempt from fee caps. the service delivery arrangements to manage 40. The use of fees can be maximized by charg- costs, diversifying the fl ow of students beyond ing variable fees at the institutional and course lower-secondary education, and enlarging level; higher-quality institutions and in-de- the role of private fi nance in postsecondary mand or higher-paying disciplines can charge education. greater fees. 25. Tertiary spending per student as a share of 41. This section draws heavily on a review on GDP per capita is used for the STEM share, student loan design by Johnstone (2004). given that the cost of providing these courses 42. Johnstone 2004. is likely to be quite dependent on local condi- 43. Ziderman 2006. tions; simple tertiary spending per student is 44. Ziderman 2003. used for the journal indicator, given the closer 45. Ping Chung 2007. correlation with international salary costs. 46. Programs to increase access to high-quality 26. Mok 2010. secondary schools and courses, particularly 27. Mok 2010. targeted at vulnerable groups, could have a par- 28. Mok 2010. ticularly high payoff, as shown by the example 29. Tan 2010. of the Urban Systemic Initiative in the United 30. These indicators include faculty strengths States. This initiative, which was designed to in course offerings, academic qualifi cation, provide opportunities for disadvantaged youth research and publications, student graduation to participate in high-level math and science rates, employment rates, and student evalua- courses, showed positive effects on access to tion results. advanced math and science courses, as well as 31. World Bank 2007. reductions in disparities between (a) African 32. Raza 2010. Americans and whites and (b) Latinos and 33. Ministry of Education, Singapore 2005. whites in science and mathematics course 34. Byun 2008. enrollment (Martinez and Klopott 2005). Managing Public Higher Education 5

inancial capacity forms an essential public sector can help tackle the disconnect pillar of effective higher education. between higher education institutions and FWhat helps tie it together is effec- skill and research users. tive management of public institutions and This chapter focuses on managing pub- stewardship of the higher education system. lic institutions through the optics of auton- This chapter deals with the management of omy and accountability. Higher education public institutions; chapter 6 addresses the worldwide in the past two decades has moved stewardship of the system, with emphasis to market approaches, attempting to achieve on private institutions and the interactions effi ciency gains. But markets, even for edu- between higher education institutions and cation, need regulators to oversee a correct skill and research users (the fi rms). Public balance between autonomy and account- tertiary institutions are critical in East Asia ability—a balance that many of East Asia’s because, notwithstanding significant dif- low- and middle-income countries have yet ferences across countries, 70 percent of all to fi nd. These countries should aim at com- students are enrolled in the public sector. prehensive autonomy, in both academic and Several information, capacity, and incentive procedural aspects. Their most important constraints and the related disconnects are accountability move will be to strengthen related to management. For instance, even if the mechanisms to nongovernment stake- institutions receive suffi cient funds for highly holders (communities, households, students, qualifi ed faculty, insuffi cient autonomy to and academic and other staff) to ensure that select staff and decide on academic pro- autonomy translates into more socially effi - grams makes it diffi cult for them to deliver cient outcomes. But governments remain what fi rms need. At the same time, lack of crucial in the process. And they need to sep- accountability of university management to arate long-term vision and policy direction, representative university boards may not be which they should direct, and quality assur- conducive to universities’ fulfi lling the needs ance, where they have a critical role, from of skill or research users. Examples abound operational management, which they should that poor management of public institutions grant to higher education institutions and has caused many of East Asia’s disconnects. intermediate “buffer” bodies. They must In this context, good management of the also generate program ownership, which

127 128 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

early, demonstrable success at a few top for monopolies. The government thus has a institutions can generate. clear role to play, but too strong a grip on How decision making is shared across higher education institutions or “protec- actors and how accountability is structured tive” behavior (such as guaranteed fi nanc- to translate autonomy into results are the two ing) undermines incentives to reform. A key critical decisions that regional policy mak- challenge is thus how to balance autonomy ers face in managing the public sector. The with accountability—and within account- following discussion reviews the case for ability, how to build on the different poten- providing higher education institutions with tial mechanisms available in a decentralized greater decision-making autonomy while sup- setting. porting strong accountability, then moves to Institutional autonomy increases the num- the main characteristics and issues related to ber of lines of accountability. But even as autonomy and accountability in the region. It institutions become more autonomous, gov- concludes with a summary of the main policy ernments typically continue to play an impor- implications. tant role in setting priorities and expecting results (system oversight), thus defi ning the Global moves to autonomy fi rst of three accountability relationships—a “compact” relationship (figure 5.1).1 In an Encouraged by governments, higher edu- institutional autonomy setting, front-line ser- cation sectors worldwide in the past two vice providers become, at least potentially, decades have increasingly adopted market- more transparently accountable to their like behavior in the hope of achieving effi - “clients” (local community, students, and ciency gains—often in the form of greater parents) through the second accountability institutional autonomy (box 5.1)—as they relationship—the “client power” relation- moved from state-controlled to state-steered ship.2 A third, the “internal management” systems. A crucial theme in this transition relationship, refers to the internal control and was structuring an alternative system of quality measures that need to be in place to accountability, for higher education is not ensure the accountability of the institution like other goods and is plagued with mar- management or of the faculty to the institu- ket failures that have traditionally justifi ed tion (box 5.2). a larger role for government. These include Research in the past few years3 has high- externalities, information asymmetry (at dif- lighted the importance of autonomy for devel- ferent levels of the sector), and the potential oping world-class universities and innovation.

BOX 5.1 Institutional autonomy defi ned

Institutional autonomy is the “degree of freedom notion of autonomy are efforts to encourage insti- of the university to steer itself”a or the “condition tutions to have the freedom to make choices about where academia determines how its work is carried their internal management and governance, given out.”b Allowing for increased institutional autonomy ideally existing market-driven incentives. means that governments increasingly exit from the day-to-day management of the tertiary sector, allow- a. Askling, Bauer, and Marton 1999, 177, and Marton 2000, 23f, both as quoted ing universities and other higher education institu- in Bladh 2007. tions to determine their own path. Underlying the b. Neave and van Vught 1994, 295, as quoted in Bladh 2007. MANAGING PUBLIC HIGHER EDUCATION 129

The key message is that autonomy combined FIGURE 5.1 Accountability relationships in an institutional with competition is conducive to innovation, autonomy setting especially for institutions closer to the technological frontier. Enhanced autonomy can also support a better match between output from tertiary politicians policy makers institutions and labor market needs.4 It also allows higher education systems to work with the community to develop programs for voice— compact developing skills needed by the local labor politics market (box 5.3). One study5 examined how different policies and institutions affect the number of new students management tertiary graduates6 as a proxy for investment parents client power community faculty, institutions in tertiary education in 19 countries of the managers Organisation for Economic Co-operation participation and/or exit and Development (OECD). The authors found that, on the supply side, the greater schooling responsiveness of supply of tertiary education (as measured by input fl exibility, output Source: Authors’ elaboration based on World Bank 2003. fl exibility, and accountability) had a positive effect on the number of graduates produced. On the demand side, they found that higher of skills and research between higher edu- internal rates of return also had a positive cation institutions and the labor market effect on graduate numbers, but that fi nanc- (fi rst and second disconnects) for differenti- ing systems (the ratio of education costs to ated local needs and for quality-enhancing available financing—an index for liquidity choices. Greater incentives are also positive constraints) had a negative effect. for resource mobilization, which is a basis for wider coverage and higher quality. Autonomy for low- and middle- Autonomy on its own, however, is income East Asia unlikely to achieve its potential without well- functioning accountability mechanisms. For Less affl uent East Asian countries face the example, universities may have the autonomy huge challenge of having to manage large to align their curriculum to what fi rms need and increasing public higher education sys- but may only really do so if a combination of tems—scaling up many times to thousands government, students, and parents holds them of institutions and to hundreds of millions accountable for graduates’ future employment of students—while making sure that these performance.7 Universities may also have the systems address the skill and innovation autonomy to hire the human resources to needs of their economies. Most do this using support an innovative research program for centralized management structures and lim- fi rms but may do so only if the government or ited institutional autonomy, but many are the fi rms hold them accountable for research failing—as seen in earlier chapters. commercialization.8 Adequate accountability Institutional autonomy, with account- structures have other benefi ts. Qualifi cation ability, could have huge benefits for skill frameworks have a direct effect on improv- development and innovation in low- and ing the incentives and information for coor- middle-income East Asia. This is largely dination among higher education institutions because greater incentives, arising from more (fourth disconnect), and governance arrange- fl exible governance structures, and greater use ments for the management of higher edu- of local information allow better matching cation (separating policy from operational 130 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 5.2 Translating autonomy into more socially effi cient outcomes

The main argument for higher institutions’ Similar reasoning holds for student involvement. autonomy lies in the potential gains that ﬂ exibil- And as important are possibilities for students to ity in substantive and procedural issues provides express preferences by voting with their feet (the in responding to changes in the labor market, in “exit” option). addressing differentiated local needs and in allow- Beyond client power, internal control and qual- ing quality-enhancing choices. None of these ben- ity measures (illustrating the internal management eﬁ ts will really materialize, however, if the client relationship) are also essential to support quality- power and internal management relationships do enhancing choices in higher education institutions. not work. While a university may know the needs Such measures typically include self-administered of its local community better than the central quality assurance (including self-evaluations), ten- government, those needs are most likely better ure systems, and other performance incentives for addressed if the community can directly express faculty, as well as university boards that can hold its preferences by, for example, participating in university management accountable for its decisions. the university board. More important, involving Institutional autonomy without a culture of internal members of the local community may be necessary quality—including holding staff and management to ensure that the university uses its knowledge to accountable for performance—runs the risk of sim- satisfy local needs. ply not producing the desired outcomes.

BOX 5.3 Addressing local labor market needs in the United States and Mexico

Clemson Universitya in South Carolina, United development projects as well as obtain professional States, by partnering with the carmaker BMW, skills.b It has wide representation of external mem- transformed itself from a predominantly agricul- bers on its board (external stakeholders control the ture and mechanical engineering school into a sig- direction of the institution) and grants a leading role niﬁ cant contributor to automotive and motor sports in decisions to faculty and students. The university research. The school made the switch because South has been particularly successful at setting up a wide Carolina has in recent years become a U.S. hub for network of business incubators, business accelera- the automotive industry. tors, and technology parks. In Mexico, curriculum reform has enhanced the prospects of the University of Monterrey. The a. Przirembel 2005, as cited in Salmi 2007. university requires students to take part in local b. Salmi 2007.

management) could even have positive conse- information is available, backs up the point quences for the interrelation between educa- that autonomy needs to be accompanied by tion levels (ﬁ fth disconnect). accountability. The evidence on the effects of autonomy in East Asia is less widespread than for OECD Eff ects of autonomy at countries, as real autonomy is still rare or, institutional level in some countries, very new.9 Still, the partial evidence now presented points to some There are a few positive examples of early successes of autonomy and, when this what autonomy can achieve in East Asian MANAGING PUBLIC HIGHER EDUCATION 131

institutions. In Thailand, a few universi- institutional ranking, patents, and jour- ties that have moved out of the govern- nals (see chapter 2) and the autonomy of its ment bureaucratic structure have begun to institutions, there is a positive relationship successfully position themselves as value- because autonomy increased at the same adding partners for several growth indus- time. Indeed, many reforms have been under tries. Prince of Songkla University, for way in Chinese higher education, including instance, has created new master’s and PhD priority of resource allocation to its fl agship programs in chemical engineering focused universities and very strict output moni- on bio-based chemicals and fuels that include toring, but the contemporaneous increase expertise in biomass agronomy, chemical in autonomy (box 5.4), extended to both engineering for biomass processing, saccha- teaching and research institutions, as well as rification, and fermentation. An emphasis further support of private higher education on biochemicals and polymers makes the providing some enhanced competition, has programs very useful to fi rms operating in played a role. The Chinese case appears to the rubber industry.10 illustrate the positive effects of a combina- The Hong Kong University of Technology tion of additional resources and governance became an internationally ranked univer- reforms aimed at increasing both account- sity within a decade of its founding in 1991, ability and autonomy. aided by two factors. First, its fi rst-tier fac- Another example, though more incipient, ulty recruitment policies were successful, is Singapore. The evidence suggests that fol- tapping into the potential of the Chinese lowing the move toward higher autonomy for diaspora, thanks to its level of academic its universities (see box 5.4), the National Uni- and administrative freedom (including free- versity of Singapore has seen its performance dom to fi x salaries). Second, the university pick up along several dimensions. In the mid- related well with local and regional firms 2000s, while active in publishing journals through a signifi cant presence of external and in patenting, the National University of members from businesses in the University Singapore in a context of increased competi- Council, and it had substantial freedom to tion fell from 18th in 2004 to 33rd in 2007 partner with fi rms both in and beyond Hong (based on 2006 results) of the Times Higher Kong SAR, China, and to launch regional Education Supplement (THES) ranking. initiatives.11 Such difficulties motivated the govern- Finally, Pohang University of Science and ment’s 2006 decision to corporatize the uni- Technology (POSTECH) in the Republic versity. As a result of greater flexibility in of Korea, which specializes in science and structuring recruitment offers, including the technology, is a private university that has provision of generous start-up research grants achieved world-class status over the past and reduced teaching loads in initial years for decade. The university has always had top researchers—and management fl exibility high levels of management autonomy and to support market adjustment allowances for used performance indicators. It also has faculty in fields with high market demand expanded its collaboration with many com- (such as medicine and fi nance)—the number panies in the electronics and mechanics of research collaboration agreements, inven- sector.12 tion disclosures, and patents granted to the National University of Singapore increased Eff ects of autonomy at country level significantly. In addition, the university developed relatively quickly a wide range of Some early country-level evidence shows new interdisciplinary educational programs the positive effects of autonomy along- such as nanotechnology and interactive digi- side accountability. For instance, while it tal media.13 While it is still early to appreci- is not possible to fl esh out the precise rela- ate the full effects of the reform, since 2006 tionship between recent Chinese trends in the National University of Singapore has 132 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 5.4 East Asian autonomy

The countries that began reform in the mid- to late Act of 2004, extended autonomy to all its national 1990s have generally followed a gradualist approach. universities, 87 of its total of 157 public universities.c The Korea Education Reform Plan of 1995 greatly In 2006 Singapore pushed through extensive auton- liberalized the private higher education institu- omy for its two premier public universities, the tions in Korea, which had previously had very little National University of Singapore and Nanyang Tech- autonomy.a Korea’s public higher education institu- nological University, after accepting the recommen- tions, however, remain highly centralized with lim- dations of the Steering Committee to Review Auton- ited autonomy. omy, Governance and Funding, set up in 2004. These Malaysia extended limited autonomy to all its 17 two universities were incorporated as not-for-proﬁ t public universities. But despite a strong blueprint for companies in separate acts, joining Singapore Man- reform, implementation was poor.b Indonesia and agement University, which already had this status. Thailand selected a few universities for autonomy. Elsewhere, recent attempts at reform in (In Indonesia, after some early reforms, momentum Cambodia are stuck in the legislature. In Vietnam slowed, then accelerated, then slowed again.) China the Higher Education Reform Agenda plans for pushed a more ambitious reform toward the end of greater autonomy and a comprehensive university the 1990s, decentralizing responsibilities to provincial charter. Already approved, implementation is only authorities and achieving fairly high levels of auton- just starting. omy for universities, though under strict regulations. In the mid-2000s, Japan and Singapore launched a. Byun 2008. reforms geared toward more extensive autonomy. b. World Bank 2007. Japan, through the National University Corporation c. OECD 2008b.

gained three places, ranking 30th in the 2009 Two fundamental issues THES. Finally, in Japan, while it is also quite early This section reviews the two main issues for a full assessment, since 2004, the year of with autonomy and accountability in low- governance reform (see box 5.4), the number and middle-income East Asia: (a) autonomy of world-class universities has doubled from is incomplete, and (b) the accountability 5 to 11 in the THES ranking (see chapter 2). structure, while including some elements Korea’s situation, in which it continues to supportive of strong accountability to the have highly centralized management struc- government, rarely hits the mark in develop- tures for higher education but still manages ing other lines of accountability. These two to have a number of high-quality institu- issues are worse among low-income coun- tions, shows that centralization may also tries, though they are present in middle- work. Given the huge private share, how- income countries as well. ever, the government has to directly manage only 20 percent of the higher education Incomplete autonomy system. Moreover, even then, that Korea’s institutions are not as successful as tech- The real push for decentralization of higher nological and innovation partners for fi rms education management in East Asia came in as they are in skill provision suggests that the mid to late 1990s, beginning with Korea the governance model may not suffi ciently and Malaysia and followed by Indonesia and encourage strategic vision, innovation, and Thailand. A second wave of reforms came in fl exibility.14 the mid-2000s, when Japan and Singapore MANAGING PUBLIC HIGHER EDUCATION 133

extended autonomy to their key higher edu- private universities. Even in Japan, where cation institutions (see box 5.4). the National University Corporation Act of Autonomy for higher education institu- 2004 increased substantive autonomy across tions can be divided into two types: sub- a range of areas, including human resource stantive and procedural15 (table 5.1). Lack decisions and enrollment size, Japanese uni- of alignment between the two is a problem versities are still required to consult with the because they need to work in a complemen- government when creating new departments tary fashion. For instance, innovation in sub- and faculties. They are also required to stantive areas may require additional ﬁ nancial secure government approval when modify- resources or higher-quality staff (often both), ing the number of students they enroll. and this requires procedural autonomy. In middle-income countries, governments Governments around the world vary in have extended some academic autonomy to their levels of intervention in substantive and higher education institutions in areas of aca- procedural issues, but they are generally let- demic structure and course content, but they ting autonomy rise (box 5.5). continue to make enrollment decisions. In In low- and middle-income Asian coun- China, for example, institutions are required tries, both areas of higher education auton- to get the approval of provincial and central omy are still limited. The two policy areas authorities to introduce any new program. where governments have extended autonomy Certain course elements are mandatory, but furthest are (a) institutions’ control over aca- institutions have some room to modify con- demic content and structure of programs tent. Admissions policy is based on a national and (b) at least in middle-income countries, exam and Chinese institutions are not the introduction of block grants. Signiﬁ cant allowed to accept students outside a set quota. limitations remain in other areas (table 5.2).16 A recent pilot project, however, is allowing Box 5.6 highlights Japan’s recent shift toward some institutions to take in up to 5 percent of more substantive and procedural autonomy enrollment at their own discretion.17 for its national universities. Higher education institutions in Indonesia also have to meet certain mandatory require- Substantive autonomy ments for course content, but beyond that, Academic autonomy is one area where sev- autonomous institutions have the freedom eral economies in the region have made to develop content. In Malaysia autonomy some inroads, but this trend is clearer in over content is quite restricted. A university is upper-income economies, such as Hong allowed to modify up to 30 percent, but must Kong SAR, China, and Singapore, much of gain approval from the Ministry of Higher whose widespread academic autonomy pre- Education for anything beyond this.18 dates the recent push for reform. Historically Low-income countries have some auton- in East Asia, academic autonomy has been omy in areas of academic structure and limited, and governments have often con- course content but little else. In the Lao trolled academic content for both public and People’s Democratic Republic, the National

TABLE 5.1 Substantive and procedural autonomy

Substantive (academic) Procedural (nonacademic) Curriculum design Budgeting Research policy Financing management Entrance standards Nonacademic staﬀ appointments Academic staﬀ appointments Purchasing Awarding degrees Entering into contracts

Source: Raza 2010. 134 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 5.5 Higher education autonomy outside East Asia

The United States has always had substantive auton- In centralized systems, government authorities hire omy for higher education institutions, but individual academics, often treating them as civil servants, states vary on procedural autonomy. In Europe such and set salaries based on a government scale. In institutions have been given substantial freedom to some systems, salaries are fully determined by the develop their own academic programs, with gov- government.b ernments continuing to steer academic provision Many aspects of procedural autonomy are on the through a range of secondary tools, specifically, rise. A recent study shows that Europe is marking a different forms of accreditation, licensing, or nego- trend toward allocating public funds in the form of tiating procedures. Areas in Europe that are still block grants, usually attached to certain performance restricted include selection of students and, in some criteria.c European universities are increasingly countries, stafﬁ ng. Salaries for staff in most coun- allowed to collect fees from some of their student tries are set by institutions (with or without govern- population, but the rules vary hugely. ment intervention), and only in a few countries do a governments directly determine salaries. a. Estermann and Nokkala 2009; De Weert 2001, cited in OECD 2008b. In decentralized systems, faculty members are b. Australia, Chile, and the United Kingdom are among many countries follow- hired by the institution, which also sets salaries ing a decentralized approach; countries such as Greece and Portugal follow a centralized approach. and hires under general contractual agreements. c. Estermann and Nokkala 2009.

University of Laos is the only university that are self-fi nanced can hire and fi re fac- that has acquired some autonomy. A decree ulty (which excludes most university staff, in June 2000 stipulated its academic and who continue to be paid by public funds). administrative autonomy, specifically an The Thai government no longer requires the increase in students admitted and changes 11 autonomous universities to offer faculty to the curriculum. In Vietnam some auton- lifetime contracts and encourages pay to be omy over academic content is allowed. incentivized through rewards.19 It has made Higher education institutions there have some attempt to convert university posts to also been given some autonomy over non–civil service positions, case by case. For enrollment: they can take in fee-paying stu- the most part, however, “public university dents once they have met the government’s employees are currently civil servants, which directives. impose[s] higher costs and less fl exibility in The other area where higher education terms of hiring and fi ring staff.”20 institutions have some autonomy is staffi ng. China and Malaysia seem to have more In Hong Kong SAR, China, they have histori- room to actually hire and fi re. It is subject cally had such autonomy and can hire and to contract in China, and in Malaysia, given fi re faculty as well as set salaries. Japanese that university staff are still civil servants, institutions, after the 2004 reforms, have effective autonomy is not a given. been able to convert the status of university Low-income countries in the region have employees to non–civil servants and have no or very limited staffi ng autonomy. committed to introduce fi xed-term contracts. Korea, by contrast, continues to limit staffi ng Procedural autonomy autonomy in these institutions. Even more than for substantive autonomy, Middle-income countries’ institutions seem procedural—especially fi nancial—autonomy to have partial autonomy on staffi ng, though differs widely between high-income econo- it is quite restrictive. Institutions in Indonesia mies and other categories. MANAGING PUBLIC HIGHER EDUCATION 135

TABLE 5.2 Autonomy among higher education institutions, East Asia

Substantive autonomy Procedural autonomy Set academic Employ Spend Decide structure/ and dismiss Decide size Own budgets level of course academic of student building and Borrow to achieve tuition Set Economy Institution type content staff enrollment equipment funds objectives fees salaries High income Japan National ●● ●●● ● Singapore National University of Singapore, Nanyang Technological University, and Singapore Management University ●● ●●● Hong Kong Public SAR, China ●● ●● ● Korea, Rep. National/public ● Middle income Malaysia Public ● ● Thailand Autonomous ● ● China National and regional ● Indonesia Autonomous ●●● Philippines Public ● ● ● Low income Vietnam Public Lao PDR National University of Laos Cambodia Public

Source: Raza 2010, based on Expert Survey for this study. Note: has no autonomy; ● has autonomy; has autonomy in some respects.

In Japan after the 2004 reforms, national After 2006 in Singapore, though the gov- universities for the fi rst time received none- ernment remains committed to being the armarked block grants to spend as they see major funder in the sector, newly autonomous fi t. And although the Ministry of Education universities have been encouraged to seek out and Training sets the standard annual tuition other sources of funding, particularly from fee, it allowed higher education institutions industry. These universities have been given to increase fees 20 percent in 2007 (if they the freedom to set tuition fees and have been wanted to do so).21 given full autonomy over human resources, Hong Kong SAR, China, has some fi nan- including setting remuneration packages. cial autonomy across a range of areas. For Korea, again, is different. Public univer- example, its institutions can own and sell sities remain constrained in areas of proce- buildings that have been donated or have dural autonomy despite a series of reforms been self-fi nanced. Universities can borrow after 1995.22 Paradoxically, the reform pro- funds from commercial banks and fi nancial cess has led to an excessive focus on fi nancial markets but cannot be publicly listed. Uni- accountability in a negative way. Funding has versities can set fees only for those programs become much more results focused and has that are self-funded. produced a culture of excessive regulation 136 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 5.6 Autonomy in Japan’s higher education before and after the National University Corporation Act of 2004

Japan’s National University Corporation Act of time, in admissions policies, building and equipment, 2004 greatly increased the institutional autonomy long-term borrowing, and spending block grants. of its higher education institutions. The act incorpo- Faculty members were no longer civil servants, and rated national universities with their own governing universities had autonomy to set faculty salaries. But boards. Under government supervision, universities these institutions still had limited autonomy in set- moved from limited to extensive autonomy in both ting their academic structure and content and had substantive and procedural autonomy (See table less autonomy in deciding tuition fees than they did B5.6). These institutions were given more autonomy before the act. in managing their human resources and, for the ﬁ rst

TABLE B5.6 Institutional autonomy of higher education institutions in Japan, 2003 and 2007

Substantive autonomy Procedural autonomy

Set academic Employ and Decide size Own their Spend budgets Decide level structure/course dismiss academic of student building and Borrow to achieve of tuition Set content staff enrollment equipment funds objectives fees salaries 2003 ● 2007 ●●●●● ●

Source: Raza 2010. Note: has no autonomy; ● has autonomy; has autonomy in some respects.

Sources: Byun 2008; Newby and others 2009.

rather than “steering.”23 By contrast, private Indonesia’s autonomous universities have universities have seen their financial (and been given signifi cant procedural autonomy, other) autonomy increase. though this has not always translated into Governments in middle-income countries effective fi nancial autonomy.25 But Indonesia have granted some financial autonomy to has been successful in introducing different selected universities. This has taken the form types of competitive funds, going beyond its of transferring public funds as block grants, autonomous universities. Malaysian higher allowing some flexibility to set fees for education institutions, too, receive their pub- selected programs, and in a few cases granting lic funds through block grants. the ability to add to basic staff-remuneration Higher education institutions in low- packages. But even autonomous institutions income countries still have little procedural are still restricted in borrowing funds com- autonomy, but it is growing. In Vietnam mercially and in owning property. they are shifting from relying only on the In Thailand, the autonomous universities state budget and are being encouraged by receive public funds through block grants the government to seek other sources of and have autonomy to establish their own funding.26 The number of fee-paying stu- administrative structures or formulate rules dents has exceeded the number of students on personnel and staffi ng.24 These universi- sponsored by government.27 In Lao PDR the ties also have the authority to manage and National University of Laos has been given use state property. Autonomous universities some financial autonomy. A financial sys- in Indonesia can also do that. Legislatively, tem allows the university to manage its own MANAGING PUBLIC HIGHER EDUCATION 137

revenue under the supervision of the univer- stakeholders participate in boards, this is lim- sity council.28 ited for students. With the exception of Hong Kong SAR, In Thailand the role of the university coun- China, university branches or satellites cil has been strengthened in the autonomous (domestic or foreign) have very little academic universities,30 though they remain heavily and procedural autonomy.29 influenced by the Commission of Higher Education.31 In Japan the authority of the Incomplete accountability university councils is second to that of the president of the university, which is unusual Although decentralization is only incipi- for both the region and world. The National ent or incomplete in most low- and middle- University Corporation Act of 2004 central- income East Asian countries, the trend is still ized power under the presidents, far greater under way, and all countries therefore have than in other OECD countries. In Indonesia to align their accountability systems with it. autonomous universities are accountable to Current systems still fall short because of a board of trustees,32 comprising representa- the limited development of accountability tives of ministries, the academic senate, and lines to nongovernment stakeholders (com- broader society, for example. munity, households, students, and academic On other accountability mechanisms, and other staff) and some defi ciencies in the most countries have a national career struc- design and implementation of accountability ture for academic staff, usually organized toward the central (government) level. by career ranks.33 The criteria for career How does East Asia fare on the vari- advancement and tenure nearly always ous mechanisms, relative to the rest of the include qualifi cations and achievements in world? The answer may be broken down into teaching and research,34 though the impor- accountability to nongovernment stakehold- tance of these criteria varies by country. ers and to the government. (Seniority and personal connections are often more important than achievement.) In some Accountability to nongovernment countries, such as China,35 achievement stakeholders in research is more valued than teaching As higher education institutions receive skills. But many institutions in Cambodia, more decision-making power, they need to Indonesia, and Mongolia, for example, have increase their own institutional and man- a weighting toward teaching.36 Performance- agement capacity. One important area is the pay management is still little used in the governing board (box 5.7). region, though growing. Box 5.8 provides Like the rest of the world, East Asia is examples of faculty performance programs showing a trend toward establishing gov- in China. erning boards or university councils, but Competition, the main mechanism to their powers and representativeness are still ensure exit, is constrained by lack of systemic limited, even in upper-income East Asia financing and information mechanisms to (table 5.3). All economies except Korea have support mobility across institutions. Com- governing boards, but their functions are petition is an effective tool to ensure client fairly limited in selecting the university lead- power (see fi gure 5.1), by allowing consum- ership, and only in Singapore does the board ers who are unhappy with the provision of select vice chancellors, presidents, and rectors. services to exit and opt for another provider. East Asian board heads are selected by the The fear of exit forces providers to be more government, internal university bodies, the accountable to consumers. Key elements for boards themselves, or a by a mixed approach ensuring that this mechanism works include (boards’ members are generally selected by a strong and comparable alternative to exist- either the government or a mixed approach). ing higher education institutions (public or Although most East Asian countries’ external private), the option to exit, and information 138 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 5.7 Governing boards worldwide

Globally, there has been a trend toward establishing informed understanding of its programs and activi- boards that favor a managerial model with a small ties. In setting the strategy and direction of the insti- number of external representatives.a tution, the board is a key actor in translating public Universities and other higher education institu- policies and orientation in institutional practice tions in most countries worldwide are free to set and policy implementation. The board’s functions their own internal academic structure, within a tend to be strong on the role of presidents or rec- basic framework stipulated by the law. Both dual tors, who are usually appointed by the board and (board or council and senate) and unitary gover- are thus accountable to it. The board’s head tends nance structures have been adopted, with a bit of to be elected (or appointed) by the board itself or by an edge for dual structures. Alongside the external another internal university body. members who feature in a heavy majority of coun- These arrangements allow for some accountabil- tries, boards are largely composed of academic staff, ity of the institution to external stakeholders (such nonacademic staff, and students. as private sector representatives) and for account- Typically, in OECD countries the governing ability of university leaders to their institution board has responsibility for the mission and goals (through boards appointing the presidents or rectors of the institution; approval of its policies and pro- and board heads appointed by the university). cedures; the appointment, review, and support of its president; oversight of its resources; and having an a. Fielden and LaRocque 2008.

about the quality of provision so that con- In East Asia, most upper-income econo- sumers can make choices. mies have a strategic vision as well as the It is unclear how much private and public necessary accompanying legislation.37 delivery compete, however. Most countries do Middle-income countries such as Indonesia, not have academic credit transfers allowing Malaysia, and Thailand are still transitioning students to move easily between institutions, and setting up the new systems, as are low- and public information on the quality and rel- income countries.38 evance of institutions and their programs is While academic autonomy is important patchy (table 5.4). Narrow use of demand-side to support a better fi t between supply and fi nancing, such as voucher schemes and gov- demand for skills, core curriculum guidelines ernment scholarships, for public and private have to be set by the center. As part of their sectors, also makes such mobility diffi cult. strategic vision for higher education, countries in lower- and middle-income East Asia Accountability to the government will need to incorporate elements of a curric- While less involved in operational manage- ulum reform. Chapter 3 has pointed out that ment, governments worldwide still have a curricula are currently often outdated. Hold- leading role in strategic vision, monitoring ing faculty and institutions accountable for and evaluation, and fi nancing of the public better skill delivery will be greatly facilitated sector. Various instruments governments by a new curriculum approach (box 5.9). need to manage a more autonomous pub- All countries in the region have a quality lic system are a strategic vision and higher assurance body and seem to be moving to a education legislation, a quality assurance more outcome-based system, as seen in the system, education management information widespread use of accreditation, audit, and systems, and performance-based fi nancing assessments (see table 5.4). Upper-income (discussed in chapter 4). economies have quality assurance approaches MANAGING PUBLIC HIGHER EDUCATION 139

TABLE 5.3 University governance, East Asia

Vice chancellors, Leadership of board Members of board presidents, rectors Senior management of Economy selected by selected by selected by universities selected by Composition of board High-income Japan Governing board Governing board Internal selection Appointed by vice Academic staff , nonacademic (public universities) chancellors, presidents, staff , external stakeholders Presidential selection or rectors and internal committee (national selection universities) Singapore Governing board Government Governing board — Academic staff , nonacademic staff , external stakeholders Hong Kong Government Mixeda Professional selection Professional selection Academic staff , nonacademic SAR, China and appointed by vice- staff , external stakeholders (2:1 chancellors, presidents, ratio of external stakeholders or rectors to university members) Korea, Rep. Boards not allowed n.a. n.a. n.a. n.a. by law (in public universities) Middle-income Malaysia — Mixedb Government Vice chancellors, Academic staff , nonacademic presidents, rectors staff , external stakeholders Thailand Governing board Mixedc Professional selection Professional selection Academic staff , nonacademic staff , external stakeholders China Internal university Internal university Government Professional selection Academic staff , nonacademic bodies bodies or staff , external stakeholders government Indonesia Governing board University senate Internal selectiond Vice chancellors, Academic staff , nonacademic presidents, rectors staff , external stakeholders Philippines Government Mixede Internal selection Vice chancellors, Mix of government offi cials presidents, rectors and private citizens appointed by the president; students, and faculty Low-income Vietnam Government Government Professional selection Professional selection — Lao PDR Government Government Government Government — (National University of Laos) Cambodia Mixedf Government Mixed Mixed Academic staff , nonacademic staff , external stakeholders

Sources: Raza 2010 based on Expert Survey; OECD 2008b. Note: Most institutions are public. — = not available; n.a. = not applicable. a. Some members are selected by the chief executive (often the chancellor) of the university, and others are elected. b. Members of the governing board are appointed by the government and the governing board. c. Diff ers by institution. d. Internal election involving the entire university community. e. Members of the governing board are appointed by the government and university. f. Appointed by the government (public) and the university owner (private). that include accreditation, audit, and assess- quality assurance process, and apart from ment, while low- and middle-income coun- Hong Kong SAR, China, the same for private tries mainly use accreditation or audit.39 institutions (more on private institutions in Incentives for compliance appear to be gen- the next chapter). erally strong: governments require public It is not clear yet, however, to what extent higher education institutions to undergo the quality assurance is ready to take up the 140 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 5.8 Evaluating faculty performance in China

China has introduced over the years a number of System Reform Plan initiated at Beida (Beijing Uni- faculty performance programs with positive effects versity) has been the precursor to personnel reform that may be useful to other lower- and middle- plans in other universities including Tianjin. The income countries. Beida reforms aim to make university hiring more Teaching evaluation. The Chinese government competitive by giving existing lecturers a set number formally established a teaching evaluation program of years to be promoted; if their performances are of all regular higher education institutes in 2000. not up to par, their contracts will not be renewed. This evaluation involves a ﬁ ve-year cycle with the As part of the reform, Beida institutions and depart- ﬁ rst round of outcomes made available to parents ments are encouraged to not hire their own gradu- and the general public in 2003. ates, but instead to look for talent both domestically Under the auspices of the program, some institu- and abroad. Research outputs have increased in both tions are making serious efforts to better assess their Tianjin and Beijing universities. instructors’ teaching effectiveness, using several Graduate employability. China faced both a seri- tools. In some cases the teaching evaluation done ous decrease in teaching quality and an increase in at the university level consists of three parts, each unemployment rate among university graduates. conducted by a different group. Part 1 is teaching While the rapid drop in educational quality was observation conducted by a “committee on teaching the focus of most faculty concern, the government supervision” composed of senior teaching colleagues was more concerned about graduates’ unemploy- in an instructor’s subject area (such as deans and pro- ment. The government’s response included graduate fessors). These supervisors have the right to attend employment rate as a major indicator of program any lecture for the purpose of assessing the instruc- quality on the national “Assessment on the teaching tor’s teaching ability. Part 2 consists of teaching standard of undergraduate programs in higher insti- observations conducted by fellow colleagues. Part 3 tutes.” This focus on employment placed enormous is the use of student course evaluations, perceived by pressure on university faculties to quickly address many instructors to be the most effective indicator of the problem or otherwise face consequences. If a spe- teaching performance, because student assessments cialization could not reach a graduate employment affect the teachers’ prestige in the faculty. Institutes rate of 60 percent or above for a certain number of applying these measures have improved their teach- years, the specialization would then be eliminated. ing facilities, increased their educational spending, The government felt its actions had been largely suc- and put extra emphasis on teaching quality. cessful when the Ministry of Education announced Research and performance. Tianjin University in 2006 that the new reforms resulted in an annual considers papers published abroad and the frequency graduate employment rate of 70 percent. The recent with which those papers are cited and quoted abroad. crisis brought back unemployment issues but to a It has also moved to a system similar to merit pay in lesser extent. which instructors receive “work post subsidies” on the basis of their performance. Moreover, the Beijing University Teachers’ Engagement and Promotions Sources: Lai and Lo 2007; Li 2005; Y. Lin and others 2005; Postiglione 2006.

challenges of internationalization and dis- this is not the case in Cambodia, Lao PDR, tance education.40 In particular, the inter- Vietnam, and Singapore. In the first three nationalization of higher education offers countries this underlines the early stages of tremendous opportunities but also some the decentralization process; in Singapore the challenges (discussed in chapter 6). longer-term goal is to set up an independent Also noticeable is the fact that while the quality assurance framework (the quality majority of countries’ quality assurance bod- assurance body is currently based in the min- ies are independent or semi-independent, istry of education). In Vietnam the lack of X X ed None X X X X Disclosure X X X X X X X a X X X X X X X X X X X Requirement Mandatory Voluntary Complete Limit X X X Body funding source X X X X X X X X X XX X 05. t Government Institutions Others X X X X es. X X X X X X X X Type of system X X X X X X X X X X X X X X body Accreditation Audit Assessmen ch information taken from Ministry of Education, Singapore 20 represented represented Government- nless otherwise stated. atory for public universities and voluntary for private universiti X X XX body Semi- autonomous Type of body X X X X X X body Independent Independent b External quality assurance, East Asia Raza 2010 basedRaza on 2010 Expert Survey, except Singapore, for whi Economy For Hong Kong SAR, China, quality assurance is mand This requirement is for public and private institutions, u

TABLE 5.4 China Indonesia Philippines Low-income Vietnam Thailand PDR Lao Cambodia Korea, Rep. Middle-income Malaysia b. a. Sources: Sources: High-income Japan China SAR, Kong Hong Singapore 141 142 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 5.9 Curriculum reform for East Asia

For higher education systems in the region to be student independence and creativity, minimizing more relevant to labor market needs, curricula faculty supervision. These approaches could help across disciplines must serve broader objectives than improve the quality of teaching even within the cur- simply transmitting academic knowledge. Previ- rent faculty constraints. ous evidence has made clear that countries should Universities in Europe, the United States, and place more emphasis on core behavioral skills that Australia are instituting more of these courses. Fol- are particularly applicable to services, such as deci- lowing core national guidelines, the University of sion making, communication, and client-orientation South Australia, for instance, has enumerated seven skills, do a better job of incorporating in their cur- skills that it seeks to cultivate in its graduates and has ricula problem solving and creative thinking, and reformed its pedagogical practices to achieve them: provide more relevant and practical technical and • Technical skills within a given discipline in sufﬁ - business skills for managers and professionals. cient depth to begin professional practice As a general pedagogical trend, multidisciplinary • The ability to undertake lifelong learning and transdisciplinary courses have become popular • Problem-solving skills and critical thinking for equipping graduates to learn and think across • Teamwork skills a broad range of fields while also developing in- • Ethics and social responsibility depth academic skills. These courses often follow • Communication skills a problem-centered approach and use case stud- • Skills to demonstrate international perspectives as ies to understand complex systems. They do not a citizen and a professional require higher-qualiﬁ ed or more faculty. Indeed, if well developed and accompanied by adequate faculty training, problem-centered approaches enhance Sources: Hicks and George 1998; Kennedy and Lee 2008.

independence of the quality assurance body standards review, and budget allocation has also hampered the transition from a more (that is, operational functions). The central traditional input approach to an outcome- government maintains responsibility for based system of quality assurance, given mission, strategic planning, and the public that input control is the routine monitoring budget. Buffer bodies have advantages and strategy.41 disadvantages. Ministries of education are Few economies in East Asia have fol- wary that buffer bodies will become too lowed the international trend of separat- powerful and will stand in the way of more ing operational management from national far-reaching reforms. For their part, buffer policy (including determining the scale and bodies worry that the ministry may begin to scope of the sector) at the central (govern- interfere in the daily management of higher ment) level. Most governments continue education, particularly when subjected to to manage their higher education sectors political pressure from lobbying groups.42 through their ministries of education. The Still, they generally help ensure less public only exceptions are China, which divides interference in operational management, the management of higher education higher standards, and more time for strategic between national and provincial authori- planning for the ministry.43 ties, and Hong Kong SAR, China, which Finally, the lack of well-functioning and has a buffer body. comprehensive education management infor- Key functions are often delegated to mation systems (including information not such bodies, including quality assurance, only on institutions but also on graduate MANAGING PUBLIC HIGHER EDUCATION 143

employment and skills) in all countries is staffi ng may be diffi cult because of the civil likely to weaken quality assurance by limit- service system). ing information on outputs and outcomes.44 Aligning substantive and procedural This may also be why performance-based autonomy. Governments worldwide used to fi nancing is still in its infancy. be more generous with substantive autonomy than procedural autonomy, though this is Moving forward now changing. In East Asia, too, governments prioritize autonomy in substantive areas more All indicators suggest that the decision- than procedural areas. Governments need to making autonomy of higher education insti- conceive both types of autonomy as a whole, tutions is still limited in low- and middle- because many aspects of substantive auton- income East Asian countries. This explains omy can be undermined by the lack of proce- at least part of the disconnect between dural autonomy. institutions and fi rms in skills and research. For example, within staffi ng policies, hir- Accountability lines to nongovernment ing and fi ring faculty will depend critically on stakeholders, including the role of boards, the ability of institutions to set salaries. Diver- remain underdeveloped. Internal manage- sifying funding is also an essential means of ment processes are generally weak, with increasing autonomy.45 Securing funding boards struggling (or even not having the from the private sector is particularly impor- power) to fulfi ll their fi duciary role. tant, because it is a critical mechanism of The compact relationship (see fi gure 5.1) external effi ciency. As higher education insti- tends to be more developed, at least in terms tutions continue to be mostly public bodies of regulation and existence and the manda- reliant on base funding, diversifying funding tory nature of quality-assurance mechanisms, offers these institutions greater autonomy. and has been moving to some extent from an input-based to an output-based approach Accountability and system oversight (from regulation to quality assurance). Yet performance measures on graduation, labor Strengthening accountability mechanisms. As market outcomes, and research are rarely they move to higher autonomy, all countries used for quality assurance and funding allo- will need to align their accountability frame- cation, and performance standards are not work to this new setting. The most impor- set high. tant step will be to strengthen accountability Some of the main implications from this mechanisms to nongovernment stakeholders chapter are now given. to ensure that autonomy translates into more socially effi cient outcomes. This will require Autonomy two main types of measures, both of which should be supported by the governments. Increasing autonomy across income groups. The fi rst is strengthening and empower- While governments in middle-income coun- ing governing boards. This should include tries may feel more urgency to increase the appointment of the board head by the autonomy than low-income ones (par- board itself or internal bodies of universi- ticularly China, which is the closest to the ties as well as the capacity of the board to technology frontier), low-income countries appoint the president or rector (as in most should not lose time either, and they should countries). Boards may need to receive exten- aim at comprehensive and not piecemeal sive training, particularly in low-income autonomy. countries. This move may also require keep- Increasing autonomy in staffi ng. Institu- ing a limited number of members but pre- tions should have full autonomy on hiring serving broad representativeness, which and fi ring (though adapting a decentralized is essential to strengthen the client power hiring system in countries with centralized relationship.46 144 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

The second is strengthening the exit would allow governments to maintain a dis- option for students. This will entail steps to tance from the regular lobbying in the sector increase mobility and competition, including to focus on policy47 and help them articulate national qualifi cation frameworks (box 5.10), the sector’s priorities with the broader over- disclosure and publication of information on lapping labor market and education system. institutional and graduate performance, and In most cases, East Asian governments have some demand-side fi nancing. established separate departments within the More widely, the role of government, ministry of education to be responsible for though changed, remains critical. As the the sector, which may be better than entirely power (and capacity) of boards increases, separate bodies for policy coordination but is priority setting as well as monitoring and insuffi cient to grant full policy coherence if evaluation will become a shared responsibil- the departments are too involved in the oper- ity. The challenge is how to make the vari- ational management of the sector. ous accountability relationships work in a Strengthening education management complementary way, building on relative information systems. Alongside perfor- strengths. mance-based fi nancing, governments should Implementing successfully the compact improve information to help enforce quality relationship will require the capacity to hold standards (with quality oversight provided by institutions accountable on broad clear goals, boards). while minimizing intrusions in daily management and protecting the capacity of institu- Creating an enabling environment tions to fi x many of their priorities. This will entail setting clear goals for the system as Several factors are important in creating an well as an effective quality assurance system enabling environment for successful reform. focused on initial accreditation, audits, and Generating ownership of the reform pro- outcome assessments. A semi-independent cess. Successful outcomes are usually driven or independent system will be more cred- by domestic ownership of the reform pro- ible. And the quality assurance system must cess and by a political consensus that decen- address the challenges of internationalization tralizing higher education management is (discussed in chapter 6). better for the economic needs of the coun- Separating policy and operational man- try.48 More emphasis needs to be placed on agement. Such separation at the central level undertaking dialogue with key stakeholders,

BOX 5.10 National qualifi cations frameworks

Part of the broad quality assurance system, national modular (competency-based) curricula, and design- qualifi cations frameworks help provide nationally ing assessment methods and new performance tests. consistent recognition of outcomes in postcompul- These shifts have rarely been adopted in the region. sory education and thus are essential for student Australia’s national qualifi cations framework is the mobility between education and training institu- most accomplished. It was introduced in 1995 on a tions. While they can potentially address discon- nationwide basis and was phased in over a period nects among skill providers (whether in a different of fi ve years, achieving full implementation by year or even in the same education level) and should 2000. therefore be supported, their design and implementation are complex. Challenges include developing Sources: Australian Qualifi cations Framework Web site, www.aqf.edu.au; Johanson standards-based on-the-job analysis, preparing new and Adams 2004. MANAGING PUBLIC HIGHER EDUCATION 145 particularly higher education institutions Notes themselves, to ensure that they see the merits of how a decentralized system can poten- 1. Defi nition of these relationships follows the World Development Report (WDR) 2004 tially benefi t them. Leadership at the govern- terminology (World Bank 2003). ment and higher education institution levels 2. Again, this follows WDR 2004 terminology. is also critical. The client power relationship can build on Building institutional capacity. Reforms two main options: “participation,” where often stumble because capacity is lacking communities have the mechanisms (such as within either government or higher educa- councils and stakeholder associations) to tion institutions, and this lack also makes express their preferences and hold the man- governments wary of pursuing reform. Build- agement of the institution accountable for ing capacity in areas such as fi nancial man- results, or “exit,” where students can leave agement in higher education institutions or nonperforming institutions (which requires establishing an education management infor- competition among institutions). 3. Aghion and others 2007, 2008, and 2009. mation system, for example, even before the 4. OECD 2008b. reform formally starts, is essential. Given the 5. Raza 2010. limited capacity in institutions in low- and 6. This was measured as the share of graduates middle-income East Asian countries, the stag- in the 20- to 29-year-old age cohort. gered approach to introducing autonomy that 7. Such accountability may occur, for example, many countries have pursued makes some through performance-based fi nancing or by sense to address the issue of limited institu- requiring the publication of employment tional capacity. By contrast, this slow pace information. has often led to reform stagnation or even 8. They can do so through competitive fi nanc- reversal. A better option may therefore be to ing for research and broad-based university ensure that a minimum set of preconditions boards, for example. This latter point is discussed further below. is in place and then start with far-reaching 9. There is also a nearly complete lack of rigor- reform. ous studies assessing the effects of autonomy Prioritizing the legislative framework. Fail- on higher education in the region. ure to adequately consider reform sequencing 10. Doner, Intarakumnerd, and Ritchie 2010. can slow the process, and prioritizing and get- 11. Postiglione 2011, as cited in Altbach and ting the legislative framework right is key.49 Salmi forthcoming. The legal focus in the region, rightly so, has 12. Shik Rhee 2011, as cited in Altbach and Salmi prioritized the conversion of public higher forthcoming. education institutions into autonomous inde- 13. Altbach and Salmi forthcoming. pendent entities. What is equally important, 14. These are critical characteristics of excellence, however, is to ensure that the overall legis- according to Salmi 2009. 15. Berdahl 1971. lative framework is adequately reformed to 16. This information comes largely from an insti- accommodate the new autonomous role of tutional survey applied to a group of regional 50 these publicly owned institutions. experts (The Expert Group on New Skills for Demonstrating success. One way to New Jobs) in 2010. Mongolia, not included in increase ownership and address opposition the survey, has a very centralized structure. is providing autonomy only to top univer- 17. Raza 2010. sities to generate a “demonstration” effect. 18. World Bank 2007. (This is a stronger justifi cation than lack of 19. Raza 2010. capacity for staggering reform.) This strat- 20. World Bank 2009a, 84. egy could be useful in low-income countries 21. Byun 2008; Newby and others 2009. where the domestic climate is yet not ready 22. Byun 2008. 23. Byun 2008. for reform or in middle-income countries 24. World Bank 2009a. where some top-performing universities are 25. In the areas of fi nancing and staffi ng, effec- 51 ready to take off. tive autonomy has been undermined by the 146 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

lack of comprehensive reform in legislation, represents between a fi fth and a third of total though a new law was passed in 2009, which higher education enrollment across China, seeks to cover some previous gaps in auton- Indonesia, Korea, and Thailand. But the omy (World Bank 2010c). emergence of distance learning in the region 26. World Bank 2009b. does pose new challenges that can threaten 27. UNESCO 2006. educational quality in countries with weaker 28. UNESCO 2006. regulatory capacity. The need to strengthen 29. Mok 2009. the capacity of national quality assurance sys- 30. Raza 2010. tems must accompany the region’s efforts to 31. World Bank 2009a. take advantage of these new opportunities. 32. World Bank 2005. 41. World Bank 2008. 33. OECD 2008b. 42. Fielden 2008. 34. Chapman 2010. 43. By the same token, this may help address 35. Wu 2010. the disconnect between education levels by 36. World Bank 2010c, 2010d, 2010e. enhancing time and attention for strategic 37. Raza 2010. planning and public budget development 38. In Indonesia, for example, although the deci- across education levels within the ministry. sion to make universities autonomous goes 44. World Bank 2009a, 2009b. back to 1991, the strategic vision was only 45. Fielden and LaRocque 2008. issued in 2003, the quality assurance system 46. Empowering governing boards would also be was put in place in 2008, and the legal frame- instrumental in providing higher autonomy work for higher education was promulgated for satellites, by making them accountable to in 2009. In Malaysia the legal framework the boards. was in place as far back as 1996–97, but the 47. Fielden and LaRocque 2008. strategic vision and quality assurance body 48. One reason why the Mongolia decentraliza- were established only in 2007. tion process did not succeed in the 1990s 39. The push for more outcome-based account- was that the process was imposed by inter- ability has evolved not only because of the national donors and lacked strong domestic changing relationship between government roots (Steiner-Khamasi and Stolpe 2004). and universities, but also because of the 49. An example is Indonesia, where the lack of a increased focus on effi ciency, value for money, legislative framework has delayed the effec- and the globalization of education (Huisman tive autonomy of institutions. and Currie 2004). 50. The recent approval of a university char- 40. Distance education is a cheap and effective ter providing a clear framework for higher way to enroll students who under traditional autonomy and the role of university councils modes would be unable to participate in the is a good fi rst step in Vietnam. tertiary education system. The initial costs of 51. Vietnam is following this strategy with its distance education are high. Once the system special autonomy to some of its new model is established, however, it can grow to scale universities. It has worked in a few other coun- at a relatively low cost. Distance education tries, such as South Africa (Cloete 2002). Providing Stewardship for Higher Education 6

ffective management of public insti- other important areas for support, briefly tutions will go a long way to address touched upon in this chapter.) Finally, bet- Emany of the disconnects—both in ter coordination among government bodies skills and research. But at the core of higher can go a long way to address several of the education systems are actors that do not disconnects. operate directly under the responsibility of This chapter addresses the four critical higher education departments: private higher challenges of providing effective coordina- education institutions (PHEIs; as skill and tion between higher education departments research providers), international skill pro- (or ministries) and the other education viders, and fi rms (as skill and research users departments (or ministries), ministries of and providers), among others. This chapter science and technology, and ministries of discusses how countries can coordinate and fi nance and labor; effective stewardship for handle these actors in ways that enhance private and international institutions with a skill and innovation outcomes (in other focus on regulation and fi nancing; and man- words, exercise appropriate stewardship of aging university-industry links. the overall system). The lack of an integrated and coordinated As seen previously, many disconnects are stewardship structure to guide the strategic related to stewardship. While governments development of tertiary education must be have less authority over PHEIs and inter- addressed. Countries will need to develop national skill providers, they can go a long stronger capacity of coordination within their way to ensure that they fulfi ll their potential governments. through effective policy decisions, helping Private higher education (PHE) in all coun- address skill-related disconnects. And while tries will play a large—or larger—role, given fi rms are outside their sphere of day-to-day constraints on public funds. As the sector control, encouragement of effective university- expands, though, the critical challenge will industry links can provide the incentives for be to avoid sacrificing quality to quantity. fruitful collaboration and interaction, help- Put another way, the quality of too many of ing address the disconnect between fi rms and the affordable demand-absorbing institutions providers in skills and research. (Interactions is low in some countries, while the share between institutions and research centers are of “semi-elite” institutions (usually more

147 148 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

expensive but with stronger potential for related to the ambiguity of content, feasibility innovation) is also low. The policy environ- for implementation, comprehensiveness, and ment thus needs to encourage a move to both consistency between different regulations more and higher-quality institutions. But because of the lack of coordination between because policy makers and the public seem to agencies responsible for administering higher have ruled out direct subsidies to PHE, policy education, science and technology, and uni- makers and academics should consider the versity relationships with industry.1 potential of indirect subsidies—from student In Mongolia, different agencies responsible loans to tax subsidies to competitive funds for different roles in the technical and voca- for research—in greater detail. tional education and training (TVET) system In managing their university-industry are working in silos without much coordina- links, countries need to continue building tion, interaction, or synergy.2 The National capacity. As this will take many years, pol- Vocational Education and Training Method- icy makers should start encouraging stron- ology Center has 70 skill standards, but the ger links by assessing and gradually helping National Council for Education Accredita- develop selected modalities. Approaches to tion (located separately in the Ministry of be considered include better aligning uni- Education) has yet to accredit any of these versity teaching with firms’ skill demands programs. And although there is evidence of through collaboration in curriculum devel- collaboration with industry, such collabora- opment, encouraging entrepreneurship, set- tion is still very limited to students’ attach- ting up university incubators, and establish- ments, which range from 30 to 45 days. Any ing technology licensing offi ces (TLOs) and attachment beyond the stated duration is at spin-offs. Governments need to provide the the discretion of school directors, and the incentives required to make such modalities Ministry of Education has little initiative to work, including bringing in intermediaries forge closer collaboration with industry or and offering matching funds. employers. Recently, the Ministry of For- Finally, the globalization of higher educa- eign Affairs has been considering a special tion has the potential to address some of the economic zone for small and medium enter- disconnects constraining higher education. prises. But neither the Ministry of Education The fi nal section of this chapter focuses on nor the Ministry of Science and Technology providing adequate stewardship of interna- has been part of this effort. tional skill providers. Governments need to Most countries need to do a better job manage the growing trends of international clearly articulating the direction and general education in ways that promote quality action map for their higher education systems. and recognition, increase access and equity, Many governments have ambitious objectives encourage cost-effectiveness, and build for their higher education systems, but if deci- capacity. sion making is fragmented and does not consider related issues and challenges, progress Providing eff ective coordination will be diffi cult. Without a bold, systematic, among government bodies and comprehensive vision—and processes for implementing its reform—progress in most Some of the disconnects described in chapter of East Asia will be much less substantial 3 have been aggravated by poor coordina- and systematic than expected. Indeed, over- tion. Indeed, institutional segmentation and coming the fi ve disconnects will most likely fragmentation appear to characterize many involve coordination among entities spread of the region’s higher education systems. over multiple places of decision making (min- In Vietnam, the planned transfer of deci- istries of education, science and technology, sion-making authority to higher education and labor, to name three). How can govern- institutes in recent years has been welcome, ments do a better job at coordinating across but it has also been accompanied by challenges these agencies? PROVIDING STEWARDSHIP FOR HIGHER EDUCATION 149

There are no obvious solutions for Commission, Te Amorangi Ma¯tauranga addressing this challenge. For instance, the Matua, steers the tertiary education sys- integration of higher education within min- tem to achieve national objectives.4 Its main istries of education can help solve the dis- function is to put into effect the tertiary connects between education levels but may education strategy and coordinate its imple- make the connections with science and tech- mentation, which may be suboptimal for nology more distant. And even within min- separating policy and operations in higher istries, coordination between departments education. Still, in pursuing its directive, it cannot be taken for granted. Overall, coun- works with several key government agencies, tries will need to dedicate more attention to including the Ministry of Education; the New these links, including developing stronger Zealand Qualifications Authority; and the mechanisms for coordination within their Ministry of Research, Science, and Technol- governments. ogy, as well as a range of private, education While there are many different ways to sector, and industry representative groups—a coordinate the myriad actors in the higher very positive feature. education system, some countries may fi nd it desirable to set up a national commis- Steering private delivery sion for tertiary education or similar body with broad representation from industry, East Asian countries face the challenge of key professions, and national and interna- managing their growing PHE sector in a tional academics. The commissions could way that maximizes its potential to deliver be charged with the responsibility to set the key outcomes. If handled well, PHEIs can strategic direction for tertiary education, increase the capacity of the higher educa- coordinate the various actors, and monitor tion system to expand coverage, diversify, how higher education facilitates national and provide many useful intermediate skills and regional development. The commis- through quality college-level education (fi rst sions could also sponsor policy analysis and disconnect). Adequate steering of private conduct strategic planning, but they have to institutions can also improve the interre- be in very close contact with the ministry lationships among higher education insti- of education and its different departments tutions (fourth disconnect). This section (they could even be part of ministries of reviews the main characteristics of private education). The Republic of Korea’s Min- delivery in East Asia, assesses some of the istry of Education and Human Resources critical issues, and concludes with some pol- Development and the Korean Council for icy recommendations. University Education have partnered to serve this function and mediate between Characteristics of private delivery the central government (including its differ- in East Asia ent agencies), fi rms, and higher education institutions.3 Across the world, PHE has developed as Other countries outside the region have an alternative to public education with sig- formal stewardship arrangements. In the nificant potential to increase coverage at United Kingdom the Commission on Higher low public cost and even provide skills and Education helps set strategic direction by innovation that the public sector may not be articulating fi ve aims: enhancing excellence providing (through, for instance, state-of- in learning and teaching; widening partici- the-art pedagogy and facilities). pation and fair access; promoting employer A review of the characteristics and poten- engagement and skills; enhancing excellence tial of PHE is particularly relevant in the con- in research; and enhancing the contribu- text of East Asia because it is not yet used to tion of higher education to the economy and its full potential, either (more commonly) for society. New Zealand’s Tertiary Education quality or (less commonly) for quantity. 150 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

Private enrollment shares Vietnam have some of the lowest enroll- PHE accounts for almost 30 percent of ment ratios, but Cambodia has a much higher education enrollment in East Asia. higher share. Indonesia and the Philippines This places East Asia behind only Latin have shares coming close to those of Japan America among the world’s regions, ahead and Korea, while Malaysia and Thailand of South Asia and the United States and far have lower or much lower shares. In general, ahead of Europe and Africa (figure 6.1).5 East Asian countries with early PHE sectors Figure 6.2 provides the private share in (Japan, Korea, Indonesia, and the Philippines) terms of enrollment and institutions by econ- have larger PHE shares than countries with omy. The enrollment share of PHE exceeds middle (Malaysia, Thailand) and late (China 50 percent in eight economies of the region and Southeast Asia) PHE sectors. and reaches 70–80 percent in Korea; Japan; Taiwan, China; and Indonesia. Demand-absorbing institutions Beyond the generally high private share, The lack of a tight income relationship is in two characteristics are notable: (a) mainly part because the nonelite, demand-absorb- higher shares for institutions than enrolling institution type is by far the largest in ment, highlighting the relatively small size East Asia.6 (As the name suggests, this type of many private institutions, and (b) some of institution has the potential to address substantial differences in shares across econ- unfulﬁ lled overall demand and thus contrib- omies for enrollment. Enrollment shares ute to increased coverage, with minimum vary from about 10 percent in Thailand and cost to governments—assuming it is entirely Vietnam to about 80 percent in Korea. ﬁ nanced with private funds, the most com- Upper-income economies have higher pri- mon pattern.) vate enrollment shares, but there is no linear Strong, large public sectors long antedat- relationship between the private enrollment ing PHE were rarely a reality at the time East share and the country income level in low- Asian governments approved PHE. This may and middle-income East Asia (table 6.1). explain why in many countries the private The Lao People’s Democratic Republic and sector expanded so swiftly with broad overall government support. Demand for higher edu- FIGURE 6.1 Private higher education enrollment share, by cation became powerful with postwar eco- region or country nomic development and middle-class growth, and appeared all the more explosive because it came in the historic context of low cohort Latin America enrollment in higher education. Postwar development prioritized rural, primary, and then secondary education rather than higher Asia education in the public sector. For higher education, the vehicle for growth in several large United States East Asian nations would be PHE, an important feature in the following assessment. Europe The demand-absorbing institution is also the basic type for forms of PHE that have Africa other characteristics. For instance, Thai- land shows a powerful correlation between 0204060 80 100 percent family-owned PHEIs and demand-absorbing 7 private higher education enrollment institutions. Much of Chinese PHE involves total enrollment institutions for adult studies, part-time status, and “self-study” to prepare for a national

Source: PROPHE (Program for Research on Private Higher Education) International Databases, exam, and most of them are at the prebach- latest available year (2001–09). elor’s level.8 Globally, the private share is PROVIDING STEWARDSHIP FOR HIGHER EDUCATION 151

FIGURE 6.2 Private higher education shares

120

100

60 percent

Japan China Lao PDR Vietnam Indonesia Malaysia Mongolia Thailand Korea, Rep. Philippines Cambodia Taiwan, China

Hong Kong SAR, China share of total private higher education enrollment share of total private higher education institutions

Sources: Based on Levy 2010; PROPHE International Databases, latest year available (2002–08); and WDI (World Development Indicators) database, 2007 data. Note: Although data come from the most reliable (usually offi cial) sources, criteria and inclusiveness vary greatly across countries, so comparisons should be made with caution. For example, the meaning of higher education, university, and tertiary education varies. In some databases, only accredited or at least licensed institutions may be counted; in others, the fi gures are more broadly inclusive. There are also diff erences in how enrollment and other variables are measured. Data years are the most recently available by country (generally 2006, 2007, or 2008).

TABLE 6.1 Private higher education shares and income groups

Private higher Low-income Middle-income education share economies economies High-income economies <33% Lao PDR, Vietnam China, Mongolia, Thailand n.a. 33%–65% Cambodia Malaysia n.a. Hong Kong SAR, China; Japan; >65% n.a. Indonesia, Philippines Korea, Rep.; Taiwan, China

Source: Authors’ elaboration. Note: n.a. = not applicable.

usually higher than the public share for shares of nonuniversity enrollment in PHE nonuniversity enrollment, unlike university than public higher education. enrollment. In East Asia, this is particularly Demand-absorbing institutions make a true for Malaysia, where the nonuniversity signiﬁ cant contribution to the labor market enrollment share is much larger than uni- particularly in the private market and can versity enrollment for PHE, but all countries even make inroads into ﬁ elds more expensive (including Japan and Thailand) show higher to offer, such as engineering and information 152 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

technology (as recently documented by the below average and that public technical insti- Philippines9 or even by some cases in China tutes perform better than private ones (fi gure and Thailand).10 However, the main demand- 6.3). Other sources confirm this evidence. absorbing fi eld remains business administra- BAN-PT’s assessments generally show lower tion, with related studies in management, quality for private institutions.12 Buchori and tourism, and the like. Malik (2004), too, noted that most PHEIs are of poor quality. They also state that pri- Quantity-quality trade-off vate schools are often a prospective student’s The lack of a tight relationship with income second choice. levels also suggests gaps in the quality of Similarly, in the Philippines graduates from some institutions, pointing to a quantity- private institutions are less likely to be ranked quality trade-off in some countries: too “good” than those attending public universi- much quantity can mean lower quality. ties or polytechnics (fi gure 6.3). Care needs While data to compare the performance to be taken in interpreting too narrowly these between public and private institutions are comparisons since, as is the case in Indone- extremely scarce—starting with the impossi- sia, “student sorting” (selection bias) between bility of distinguishing people who have stud- institutions may be common, with the best ied in public or private higher institutions in students preferring public to private institu- household and labor force surveys—skill sur- tions, which would help explain differences veys in Indonesia and the Philippines offer at graduation (though such sorting already some clues. suggests issues with PHE). Beyond these two In Indonesia graduates from public rather countries, evidence from Cambodia indicates than private institutions seem to fare better that private universities there may be of lower across the board. The 2008 employer skill quality than public ones.13 Some evidence also survey11 shows that there are clearly more confi rms that the research output is lower at private than public universities performing private than at public institutions.14

FIGURE 6.3 Rating of graduates from public and private tertiary institutions, Indonesia and the Philippines, 2008

100 90 80 70 60 50 40

% of graduates 30 20 10 0 poor average good poor average good poor average good poor average good public private public private tertiary vocational institutions universities rating of graduates

Indonesia Philippines

Sources: Indonesia and Philippines Skill Surveys 2008, as cited in di Gropello, Kruse, and Tandon 2011 and di Gropello, Tan, and Tandon 2010, respectively. Note: Tertiary vocational institutions cover polytechnics only in the Philippines. PROVIDING STEWARDSHIP FOR HIGHER EDUCATION 153

In some other countries, such as Korea, such as science, technology, engineering, however, this quantity-quality trade-off is and mathematics. not visible, and, in fact, in East Asia there Countries of late PHE development in East are many successful examples of private sec- Asia tend to have a higher share of semi-elite tor institutions, illustrating that there does institutions. Leading the way is, however, not need to be a trade-off. In Malaysia, Korea, which has equal representation of pri- for example, PHEIs have pioneered in job- vate and public higher education institutions relevant fields, especially English, rarely in the country’s top-10 ranking and where available before, and some reports indicate semi-elite institutions have a strong presence that private institutions have better job- in research and the sciences. Thailand’s fi ve placement records.15 A number of demand- semi-elite universities aim to produce gradu- absorbing PHEIs in China are accredited ates for business fields and technologically technical and vocational colleges that have related industries networking with them. made major efforts investing in infrastruc- All these universities in Thailand are com- ture and equipment, enhancing curricula, prehensive in their programs but originally establishing useful niche programs, and emphasized business-related fi elds, according building links between graduates and to their founders.17 Other countries also have employers. Similar to those in China, sev- their private semi-elite institutions: Japan’s eral demand-absorbing PHEIs in Thailand private sector is academically led by Waseda are fast-growing and successful specialized and Keio; the Philippines has its Ateneo de private institutions. Manila, La Salle, and Santo Tomas;18 Indo- nesia, Santa Dharma; and Vietnam, Phang Semi-elite private universities Dong university. Like much of the world, East Asia is While on the rise, semi-elite institutions are showing a rise in semi-elite universities.16 still relatively rare in most countries. Thailand, These institutions, while not world-class with as much as 40 percent of PHE enroll- 19 in rankings or research, are high-status ment in the semi-elite subsector, appears to institutions within their countries with an be an exception in middle-income East Asia. objective of high-quality provision. They Even in the Philippines, the fact that only often compete with good public universi- about 30 percent of private sector university ties below the academic peak. They also graduates are considered good (against about sometimes compete with top-tier public 40 percent for the public sector) indicates that universities in certain niche fi elds, above all either the average quality of demand-absorb- business. They may also compete with the ing institutions is still not good enough or that elite public universities for top students by there is scope for more semi-elite institutions paying more attention to teaching, though (or more enrollees in semi-elite institutions). they lack the research capacity of top pub- In Indonesia the data suggest paucity of semi- lic universities. elite private tertiary institutions. In other words, they have the potential of providing very good academic, technical, Potential to expand and improve and generic skills. Indeed, they can offer private delivery tough labor market competition to pub- The clearest case of higher education expan- lic institutions and, being entrepreneurial, sion led by the private sector is in Korea, can take the lead in innovative fi elds and where the private gross enrollment ratio fashion ties with businesses, nationally (GER) increased from less than 10 percent and internationally. They can even provide in the 1970s to almost 100 percent in 2008, greater breadth of fi elds of study than pub- alongside a massive expansion of PHE lic institutions by focusing on more costly (figure 6.4). Indeed, Korea has built a but nonetheless essential fields of study, strong higher education system based on 154 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

FIGURE 6.4 Higher education GER and enrollments, Republic of their current level of supplying more than Korea, 1971–2008 70 percent of higher education enrollment. Further down the income gradient, several 120 3.0 low- and middle-income East Asian countries with low expansion or low quality of the pri- 100 2.5 vate sector seem not to be using private delivery to its full potential. This failure has impli- 80 2.0 cations for their capacity to deliver skills and innovation. Some countries have much scope 60 1.5 for further expansion, especially China, Lao PDR, Mongolia, and Vietnam. 40 1.0 enrollment (millions) In countries such as China and Vietnam, gross enrollment ratio (%) with less than 30 percent PHE enrollment, it 20 0.5 is natural to assume that a further expansion of PHEIs would greatly help address coverage 0 0 constraints or, conversely, that coverage may 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 enrollment in private higher education (right axis) be currently constrained by low PHE shares. higher education gross enrollment ratio (left axis) Indeed, in Malaysia the expansion of PHEIs enrollment in public higher education (right axis) was a particularly powerful complement to public institutions in rapidly expanding Sources: PROPHE International Databases; WDI database. access. Malaysia had no PHEIs until 1992, when the number skyrocketed from 156 to 632 in 2000. This surge translated into an some excellent PHEIs. Box 6.1 maps the increase in the private enrollment share from growth of Korea’s higher education sys- about 10 percent in 1994 to about 40 percent tem and the rise of private institutions to in 2007 and, at the same time, an increase meet the demand for more access to higher in the overall tertiary GER from about education. 15 percent in the mid-1990s to 30 percent in The growth of Japan’s PHE is an early 2006.21 By comparison, the private enroll- instance of demand absorption that led to ment share (and the overall tertiary GER) high-quality private institutions.20 Near the has remained fairly stagnant in Thailand and end of the 19th century, demand for higher Vietnam (fi gure 6.5). education and the government’s insuffi- Private delivery also has the potential to cient supply of public universities provided lead to system diversifi cation and better pro- a market for PHE to fi ll the shortage. Pri- vision of intermediate skills. In Mongolia, for vate universities soon gained public approval instance, a stronger role for PHE may help because of their high-quality education. At support higher diversifi cation of an otherwise the beginning of the 20th century, the gov- homogeneous higher education sector, satis- ernment began to support PHE because of fying many skill needs. For its part, Thailand a desire to better link primary, secondary, may have the potential to increase its PHE and higher education; demand for more share to tackle issues of adult learning, for skilled workers in science and technology; example. support for a merit-based education sys- Other countries have much scope for tem; and a new view that education could improving the quality of the private sector. increase social opportunity. These factors The distinction between nonserious, demand- led Japan to establish its University Code absorbing institutions (often called diploma of 1918 to dramatically expand PHE and mills) and serious demand-absorbing institu- offi cially recognize current private institutions is crucial. The former often accept stu- tions as accredited colleges and universi- dents least prepared for higher education and ties. By 1938 private institutions reached provide them with little help or direction. The PROVIDING STEWARDSHIP FOR HIGHER EDUCATION 155

BOX 6.1 Expanding private higher education in the Republic of Korea

In 1949 the new Republic of Korea announced a FIGURE B6.1 Enrollment by education level, Republic vision of universal primary education. The sub- of Korea, 1971–2008 sequent rise in primary-educated students led to increases in secondary enrollment, and eventually 120 increasing demand for higher education (see ﬁ gure B6.1). In 1995 a new deregulation policy eliminated 100 higher education enrollment quotas, resulting in a boom of new universities and branch campuses. 80

Altogether, the rising demand for higher education, 60 the insufﬁ cient public higher education system, and the government’s focus on primary and secondary 40 education created a market for the private sector.a During the 1990s the higher education system 20

experienced another boom, following the govern- 0 ment’s drive to build a market economy through 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 gross enrollment rate of school-aged population (%) privatization and competition. The government also primary secondary tertiary implemented its Brain Korea 21 (BK21) project to increase innovation and promote a knowledge-based universities receiving BK21 funds being private society focused on research and quality education universities.b Today, Korea’s tertiary enrollment rate for the 21st century. Among other initiatives, BK21 stands at 98 percent, with the private sector accom- emphasized privatization, with more than half the modating 80 percent of the students.c

Source: EdStats database. a. Kim and Lee 2006. b. Kim and Lee 2006; Ministry of Education and Human Resources Development, Republic of Korea 2010; Moon and Kim 2001. c. EdStats database, 2008 data.

latter, by contrast, tend to orient themselves FIGURE 6.5 PHE enrollment share, Malaysia, Thailand, and to the labor market, including tracking job- Vietnam, 1994–2009 demand signals, providing counseling, engaging in joint initiatives with local businesses, 45 and seeking feedback on job placements. 40 More generally, one can broadly differ- 35 entiate institutions into those having prob- 30 lematic characteristics and of a poor quality 25 and reputation (“garage”22) and those show- 20 percent ing perseverance in job training (“serious”). 15 Garage institutions typically are concerned 10 with ﬁ nancial gain and focus on the short- 5 cycle labor market (emphasizing low-cost 0 fields), often through inferior resources, a 1994 1996 1998 2000 2002 2004 2006 2008 2009 lack of transparency, and overall low qual- Malaysia Thailand Vietnam ity. Indonesia’s ratings of PHE graduates appear to suggest that such garage institu- Sources: PROPHE International Databases; WDI database. tions are quite common. Characteristics of 156 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

serious institutions include responsiveness countries with early PHE sectors (Indonesia, to changing demands in the labor market, Japan, Korea, and the Philippines). Private attraction of nontraditional students, effec- demand absorption has, for instance, been tiveness in management, and entrepreneurial very clear in Indonesia and the Philippines. orientation. Government stances were less clear-cut with Looking ahead, the largest share of PHEIs private sectors that developed later (Malaysia will have to remain demand absorbing, given and Thailand). the higher cost of semi-elite institutions. But Malaysia’s well-documented stance the current curriculum disconnects between toward a more pro-private policy start- institutions and the labor market, alongside ing in the 1990s—with its impact visible in the potential of semi-elite institutions, suggests fi gure 6.5—is striking because it comes after that quality could also be improved by sup- the 1970s policy of “constraint” and then a porting a larger role for semi-elite institutions. 1980s policy of “controlled development.”24 The aim of access without public money Main issues for private delivery became coupled with an approach to cut in East Asia the brain drain of domestic students going abroad and to attract thousands of foreign Two main higher education issues face low- students to Malaysia, specifi cally to Malay- and middle-income East Asian countries: sian PHEIs. More long-standing, since the expansion (mainly in low-income and lower 1980s, have been twinning and related poli- technology cluster countries but also in cies that have brought foreign institutions to China and Thailand), and a quantity-quality Malaysia or sent Malaysian students abroad trade-off (more visible in middle-income and for part of their studies, though much of the middle technology cluster countries, nota- thrust is now to avoid study abroad as too bly Indonesia and the Philippines, but also expensive.25 The government is not fully pro- in countries of the lower technology cluster private, however, because it involves PHE in where the private sector has been growing policy planning only to a limited extent. fast, such as Cambodia). Low PHE shares Thailand is a slight aberration for its low appear to stem largely from an earlier unfa- and recently fallen share of PHE. The govern- vorable policy stance or too restrictive regu- ment does not proclaim goals for PHE shares lation (or both), with insuffi cient resources and is attracted to expanding enrollment playing a role as well. The quantity-quality without a greater cost burden, but regula- trade-off may be an outcome of delayed tions remain quite restrictive. As a result, the or no regulation, leading to low-quality, private enrollment share has been stagnant demand-absorbing institutions, and limited since the early 1990s (see figure 6.5). The resources and overregulation have no doubt government itself seems to be divided at times hampered semi-elite institutions from gaining on the issue, leaving it to the Commission on a higher share. The following review of policy Higher Education to largely determine policy. stance and regulation, and of the fi nancing PHE shares do not rise easily in the absence framework, shows how these aspects affected of public policy limits on public expansion. quantity and quality of private delivery. The situation is different again in low- income or socialist-leaning economies where Policy stance and regulation the private sector has been traditionally seen East Asian government policy has been in less favorable terms, but the policy in these mostly pro-private on enrollment expan- countries has been changing. China’s policy sion, though with notable differences among stance, for instance, was for a long time unfa- countries.23 Government promotion follow- vorable to the private sector. By 1994 it offi - ing the demand-absorbing logic of encour- cially encouraged private growth (though less aging more private supply to meet soaring than wholeheartedly), and the stance really demand has been particularly evident in became unambiguous only in 2002. Since PROVIDING STEWARDSHIP FOR HIGHER EDUCATION 157

then the government has set goals for a grow- with laws to regulate PHE.29 All these coun- ing private share (40 percent by 2020); has tries appear to have fairly tight regulations invited money from abroad and Hong Kong on PHE, but they were developed in different SAR, China; and is working on a long-term periods. Some evidence suggests that Chinese plan for PHE.26 regulation is not extensive and instead allows Vietnam, too, has lately shown a more ample competition.30 In Indonesia autonomy favorable stance, setting a goal in its Higher from government seems to be high in prac- Education Agenda of higher education being tice, despite rules.31 40 percent private by 2020. Its earlier goal of This overview allows a few suggestions to 30 percent by 2010 has, however, stalled, and be made on the role of the policy stance and regulations on private sector participation are regulation in East Asia in relation to expan- ambiguous.27 The private enrollment share is sion and the quantity-quality trade-off. struggling to pick up (see ﬁ gure 6.5). The relationship between the policy Beyond the policy stance, regulations vary stance and the regulatory framework, and in nature, extent, and time of implementation the growth of the PHE sector, is strong. The across countries (table 6.2).28 Cambodia, policy stance and regulatory framework have Malaysia, and Thailand are among countries generally closely affected the development of

TABLE 6.2 Government regulations on private higher education, East Asia

Country Policy or approach Signifi cance of the policy or approach Malaysia Private higher education Regulates PHE (tight control) act (1996) For profi t Are explicitly permitted and extensive Purchasing power parity Are mainly foreign universities and domestic colleges, with the former having some regulatory role over the latter, such as quality assurance Quality assurance PHEIs have to obtain accreditation from the National Board of Accreditation; public universities have to carry out regular audits Thailand Private higher education Regulates PHE (comparatively restrictive) act (1969) Ministerial regulations Regulations and guidelines support the PHE act for procedural implementation of PHEIs (comparatively restrictive and detailed) Ambiguous for profi t For-profi t institutions not allowed in higher education but “30 percent returns allowed” for licensees and investment encouraged Quality assurance and PHEIs must be externally assessed and accredited by the Offi ce for National Education accreditation Standards and Quality Assessment every fi ve years; PHEIs and public universities use similar standards China Regulation (overall) Not extensive and allows ample competition Varies by province and period Accreditation Enhances autonomy and legitimacy (as well as quality), though most private higher education institutions have not earned it (620 have, 866 have not) Ambiguous for-profi t For-profi t institutions not allowed in higher education but “reasonable returns” allowed and investment encouraged; for-profi ts allowed in training institutions Purchasing power parities Are mainly public universities and private colleges, with the former having some regulatory role over the latter, such as quality assurance Governance: the Private Requires a board of trustees for every private institution; stipulates that it consist of at Education Law (2002) least fi ve people and that the institution’s legal identity be with the chairperson of the board or the institution’s president Indonesia Rules (overall) Grant high autonomy from government in practice For profi t Are explicitly permitted and extensive Philippines For profi t Are explicitly permitted and extensive

Source: Levy 2010. Note: The table is selective, and a policy shown in one country does not imply its lack in another. 158 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

the private sector in East Asia. In Indonesia, consistently, to a lower level, given the more for instance, the expansion of the PHE sector ambiguous policy stance. was facilitated by a favorable policy stance By contrast, there is also a strong case combined with lax application of laws. against overregulation. While the pre-1995 China’s PHE share is still low due to late regulations on Korea’s PHE did not hamper take-off but has been increasing strongly32 growth, they may have hampered creativ- with the latest policy and regulatory steps. ity and innovation because of their intrusive Vietnam’s low share is due to the ambigui- nature. This led to a relaxation of regulations ties of the regulatory framework (and policy in 1995, which then supported the growth stance), and Thailand’s PHE share is ham- and development of, for example, institutions pered by the very tight legislation and gener- such as Pohang University of Science and ally unfavorable policy stance. Technology (POSTECH). There is also a strong relationship between It is also striking that specifi c regulations too laxly applied or “delayed” regulation and are sometimes tougher on private than pub- quality. Too lax or delayed regulation may lic institutions. This is, for instance, the case have hampered (or be hampering) PHE qual- in Malaysia, where only PHE needs to be ity in some countries with a high or grow- accredited (and this decision may have con- ing PHE share. The regulations in Malaysia tributed to a slowdown of private enrollment are likely to have helped maintain a mini- growth at the beginning of the new century— mum quality threshold in private institutions, see fi gure 6.5). This duality can greatly con- thereby alleviating (or at least addressing) the strain expansion. quantity-quality trade-off. Against this, the lax application of regula- Financing framework tions in Indonesia is most likely a contribu- The financing framework for PHE is gen- tor to the poor quality of its PHE. Similarly, erally sound in low- and middle-income delayed regulation in the Philippines may also East Asia, though there is scope to extend have played a role in the low quality of some student loans to the PHE sector to support institutions. While moving in the right direc- its demand-absorbing role and to further tion, China needs to be careful to combine improve incentives to nontuition private the higher level of autonomy given to private funds to help establish more semi-elite insti- education with mechanisms of accountability. tutions (and better demand-absorbing insti- The fact that all Chinese private institutions tutions). There may also be some scope to need to have a board of trustees is, however, a extend competitive research funding to the fairly reassuring measure of accountability.33 extent that competition between public and Lack of disclosure of higher education results private institutions at the top can help spur in several countries has certainly contributed innovation. Below is a brief review of the to some quality gaps. fi nancing framework in East Asia, followed These fi ndings thus point to the need for by an assessment of its soundness to address the right regulation coupled with a favorable coverage and quality needs. policy stance. The policy stance is critical East Asia has for the most part followed because as long as it has been favorable, regu- the global rule: PHE is overwhelmingly lation has not necessarily hampered private funded through tuition. If anything, tuition sector growth in East Asia. The regulation dependence is especially marked in East Asia applied in Korea (both before and after 1995), because PHE is concentrated in the demand- for instance, did not hamper the growth absorbing form, which, almost everywhere, is of the sector very much in the presence of a tuition dependent.34 PHEIs in the Philippines favorable general policy stance. A similar get 85 percent of their income from tuition story holds for Malaysia (where the private and in Vietnam, 82 percent,35 and even in share has continued to increase after the Japan36 about 70 percent.37 Overall, East 1996 private higher education act), though, Asia’s tuition share accounts for more than PROVIDING STEWARDSHIP FOR HIGHER EDUCATION 159

90 percent of operating costs, sometimes institutions in the Netherlands.43 The Chi- nearly the full load.38 There are in most cases nese government has been wary of foreign no fee caps, allowing fees to be higher— ownership but has encouraged fi nancial con- sometimes signifi cantly so—than the public tributions, while Malaysia’s twinning setups sector’s. became substantial quickly.44 Box 6.2 offers There are also some nontuition sources examples of resource diversifi cation in China of PHE income. Religious institutions may and Malaysia. get funding from sponsoring or ownership Direct government subsidies to PHEIs are organizations, or donated services. East very rare in East Asia, partly because of the Asian academics, sometimes retired, may numerical weight of the demand-absorbing offer their services. More important, wealth subsector, globally the least likely to get accumulation in East Asia has led to philan- such subsidies. Similarly, the recent nature thropy (which shades into entrepreneurial of PHE in many East Asian countries plays investment), playing a larger role in PHE, as a role because, internationally, older PHEIs in the Philippines,39 Singapore (the govern- are more likely to get government funding. ment matches private donations to higher Even in Japan, the share of private univer- education),40 Japan (some families have used sity income from that source is now only gains from their lower-education enterprises 10.5 percent. Nor does Korea offer a model to donate to education corporations),41 or for public subsidization of PHEIs: they get Malaysia (businesses opened institutions or less than 2 percent of their income from the bought some set up by public university pro- government.45 fessor–entrepreneurs in the 1980s).42 Indirect subsidies are more frequent, Funding from abroad is another source though not systematic. The three main indi- of income, and of course arrangements vary rect funding streams are student assistance, greatly. In Vietnam, PHE started with foreign justifi ed mostly on access and equity grounds; money, with domestic businesses following. competitive research awards; and PHE hir- Some Indonesian PHEIs maintain ties with ing of public university teachers. (Donations

BOX 6.2 Resource diversifi cation in China and Malaysia

China Malaysia • Funding comes from overseas Chinese and • Foreign funding is permitted for PHEIs. domestic businesses, with much initiative by • Business fi rms open their own institutions or buy professors and administrators from public institutions set up by individual proprietors. universities. • Various sorts of foreign fi nancial and other participa- • Family-owned and social group undertakings tion are permitted for both higher education sectors. provide funds. • Foreign funds are prominent in private semi-elite • Public universities provide initiatives for affi liated institutions. colleges. • Government pushes higher education to be a • Partnerships exist between foreign institutions and regional hub with a large, foreign student body Chinese private (or public) institutions. paying tuition. • Recent change allows government loans for • Students enrolled in accredited programs in PHEIs students in private institutions. are entitled to apply for government student loans. • Donations of land or buildings provide in-kind resources. Source: Levy 2010. 160 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

of land or buildings and favorable tax poli- Some arguments favor sector neutrality (sub- cies are other examples of indirect fi nancing sidies provided to both public and private instruments.) institutions);48 others claim that it can dilute Student loans are still not systematically already scarce public funding and overburden available to both the public and private sec- the private sector with attached government tors in East Asia, but policies are changing. regulation.49 More important, private returns Thailand has gone far enough that many to tertiary education would seem to be high of its PHEIs could be called “government enough—overall and in the service sector— dependent,” given that student loans are their to provide an adequate incentive for private major income source.46 Demand-absorbing sector delivery and financing, especially in Thai PHEIs rely on student loans the most highly demanded “soft” sectors,50 without because they need to keep fees reasonably public subsidies. affordable and tend to enroll poorer students A case for direct subsidization can be than those at the semi-elite private universi- made for semi-elite institutions—for instance, ties (or public institutions). The semi-elite in science, technology, engineering, and universities bank on student loans the least mathematics—because of higher costs and among PHEIs because they charge very high (possibly) lower private returns, or larger tuition and fees and enroll students with externalities. But the evidence is missing that privileged backgrounds. They also get other the private sector is more effi cient than the sources of income. Until recently, China did public sector in providing these sorts of skills not permit private students to access govern- (and the private sector has shown itself to be ment loans—but that it does now is a signifi - capable of mobilizing suffi cient funds to pro- cant change. vide these types of skills). While it may there- Competitive research funding is not yet fore be useful to undertake further analysis systematically available to semi-elite uni- to assess the cost-saving potential of subsidies versities. Very little research is undertaken for these “hard” fi elds, no strong case can be in demand-absorbing institutions, making made for directly subsidizing the PHE sector this subsidy route unlikely for this type of generally.51 institution. One example of such funding Indirect subsidies may still not be used for semi-elite institutions is Korea, where suffi ciently to support private sector expan- more than half the relevant government sion and quality. This is the case of student money goes to the top 10 of the 124 private loans. Student loans have many rationales institutions.47 (reviewed in chapter 4). In Thailand student Long seen in Latin America and now loans may explain why the disadvantaged increasingly elsewhere is PHE hiring of pro- have fairly good access to higher education, fessors from public universities who receive unlike in many other low- and middle-income their main salary and benefi ts from the public countries where traditionally limited access institution. These professors are thus a public to such loans in the private sector may have subsidy for PHE. This appears very common hampered equity and expansion. Middle- in East Asia. In China PHE use of public sec- income countries such as China and Malaysia, tor teachers has been facilitated by slack in having expanded loans to the private sector, the system, where some public professors had may now be seeing the benefi ts of such a policy good benefi ts but did little research. in coverage. On the basis of the preceding review, the Competitive research funding could also following assessment of the fi nancing frame- be made more systematically available to work is offered. semi-elite universities. Allowing them to The basic fi nancing approach to PHE is compete for these grants may both strengthen reasonably sound. This is largely because them and spur healthy competition with the there appear to be no strong grounds to sup- public sector. The Korean example is a case port direct subsidies to the private sector. in point. PROVIDING STEWARDSHIP FOR HIGHER EDUCATION 161

Finally, favorable tax policies, while supporting a continued expansion of the they remain rare in the region (thus not dis- PHE sector,53 the policy environment thus cussed above), could have a positive effect on also needs to encourage a move away from resource mobilization. The case for such poli- diploma mills to high-quality institutions to cies is strong. They make sense for both pri- improve the quality and relevance of the private and public institutions. On the one side, vate and overall higher education sector, nota- there can be exemptions or partial exemp- bly through these two types of institutions. tions from certain taxes; on the other, there Regulation (and information). Clear and can be tax breaks for individuals and busi- effi cient regulation is needed to provide clar- nesses to donate, thus boosting philanthropy, ity for private sector participation while which would have the potential to expand restraining the development of low-quality the semi-elite system. demand-absorbing institutions. But excessive and undifferentiated regulation can greatly Moving forward constrain participation—and even quality. Key aspects to be regulated are rules for Private education is clearly an integral part use of public professors, the minimum cur- of the higher education system in East Asia, riculum, degree granting, licensing, examina- with about a third of the enrollment and with tions, and possibly board structures. Some scope to continue growing.52 In all countries pragmatism is needed to support the devel- PHE will continue playing or will have to opment of both serious demand-absorbing play a large role, given constraints on pub- and semi-elite (or even elite) private institu- lic funds. This is particularly the case for tions, which have different strengths and thus demand-absorbing institutions that have a are both important to protect. This suggests critical role in providing relatively low-cost that using the same standards (for instance, but highly demanded fields. To the extent within an accreditation process) for public that more resources can be mobilized, semi- and private institutions may not be desirable, elite institutions can also have an important particularly when the private institutions are role in supporting innovative management demand absorbing. Other levels of differen- and academic practices and in providing the tiation may include greater or lesser regula- public sector with high-level competition. tions for merely licensed institutions and for Expansion will be all the more necessary in accredited ones, and for institutions receiv- low-income countries where overall higher ing or not receiving public fi nance. Finally, education coverage is still relatively low, and the PHEI practice of voluntary accreditation in countries with still low PHE participation. should be supported. As the sector expands, though, the critical Concrete incentives to performance will be challenge will be to avoid the quantity-quality more effective in supporting quality and rel- trade-off dilemma that has appeared in some, evance than strict and detailed bureaucratic mostly middle-income, countries. Addressing rules. A clear example is allowing public and this trade-off and expansion requires policy private higher education institutions to com- and regulation, as well as a fi nancing frame- pete for research funds and to receive student work, to ensure that PHE is used to its full loans. For both public and private institu- potential. tions, regulation should as much as possible In low- and middle-income East Asia a be outcome based rather than input based. proportion of demand-absorbing institutions Similarly, various strategies should be (with high potential to provide affordable devised to increase accountability to different access to fi elds of study in high demand by the stakeholders. Publication and dissemination labor market) is of low quality, and the share of student performance in fi nal exams, pro- of semi-elite institutions (more expensive but fessional board licenses, and rules on board with higher potential for innovation and, to composition and responsibilities are all ways some extent, even hard sciences) is low. While to ensure higher accountability for quality 162 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

and relevance. Provision of information to Well-handled university-industry links can all relevant stakeholders, including prospec- improve information flow between higher tive students to allow them to make informed education and firms and the incentives to choices, will be particularly important. Infor- interact, enabling higher education institu- mation should include indicators of academic, tions to meet skill and research demands of research, and labor market performance of fi rms (fi rst and second disconnects), and col- private sector institutions. laborate with other research providers (third Financing. A strong case exists for public disconnect). This section reviews the avail- policy to encourage a variety of nontuition able modalities to build stronger links. funding. This can help diversify the private Although managing public and private sector (including supporting semi-elite insti- institutions very much revolves around issues tutions) and build a broader, more secure, of autonomy and accountability, “reconnect- and fl exible base, from which welcome inno- ing” the system—particularly the university- vation springs. Tax policies more supportive industry relationship—will also require other to philanthropic and entrepreneurial funding governance-related decisions to translate with regulations favorable to foreign funds improved governance at the institutional level would help support a larger semi-elite sector. into stronger links for skills and research. At the same time, as tuition will remain For university-industry links (also cov- the main income source for all types of East ering tertiary institutions generally), these Asian PHE, and the overwhelming source governance-related elements include decisions for most PHEIs, public policy should be on the modality of the links and the fi nanc- restrained in setting regulatory limits on it. ing, fiscal, legal, and information-related Similar logic applies to restraint on imposing incentives that may be needed to support caps on enrollment at PHEIs.54 The way to them. These decisions will be all the more address equity issues and make PHE more critical in the East Asian context where con- affordable is rather through student loans straints to stronger university-industry links and scholarships, offered by the institutions in research and technology are multiple (dis- themselves or the government. cussed in chapter 3). Direct subsidies of the government to The major hurdles in managing university- PHEIs in East Asia are uncommon and industry links are that fi rms in East Asia do should probably remain so given the limited not rank universities high either as sources available public funding and the associated of commercializable ideas and technical risks of overregulation. Indirect subsidies— and commercial information or as fruitful from competitive funds for research and stu- collaborators. Additionally, risk-averse fi rms dent loans to tax subsidies—have much more with little exposure to new technologies potential, and low- and middle-income East and a lack of technically skilled labor are Asia should make more use of them. often unwilling to bear the risks or put up the financing to bring new technologies to Encouraging eff ective market. university-industry links Moving forward While improved institutional management and coordination at the central level will Some of the deeper constraints related to go a long way toward strengthening links limited research capability and lack of with firms, it will be insufficient unless demand from the private sector will be it addresses other constraints. Some are addressed only with time, investment, smart related to capacity, but others can be tack- governance, and economic reforms. Indeed, led by smart choices in managing university- the success of university entrepreneurship industry links. Policy makers across the region will depend ultimately on the generation are looking at ways to make this link work. of significant ideas and findings and its PROVIDING STEWARDSHIP FOR HIGHER EDUCATION 163

reputation in providing high-quality educa- relevance of curricula, fi nancially supported tion, so countries need to continue building student work placements, and developed a capacity, step by step. Business Links Fund to formalize industry Not all countries have the same potential input into curricular design.55 France has for university-industry links in research and enacted legislation (Law of August 2007 on technology. Countries from the lower tech- the Freedom and Responsibility of Univer- nology cluster in particular see their potential sities) to mandate that two representatives very constrained by capacity gaps. They there- from industry or industrial associations serve fore need to continue strengthening their union university governing boards. versities. Demand for innovation from their Persuading tertiary institutions to play private sectors is also more limited. More a more active role in training fi rms’ current selectivity, too, will be required for university- workers. Handled well, this modality has industry links: only a few fi rst-tier universities positive consequences for fi rms and curricula. in a country will ever have the potential to For example, the University of the Philippines, have strong formal interaction with fi rms. Diliman, has partnered with the Department While capacity is being built, policy mak- of Science and Technology to have engineer- ers can move to support stronger links (or at ing students and industrial plant workers least put them in the right direction) in both undertake training at the university and at skills and research. plant sites, partly under the direction of uni- The main modalities relevant to East versity instructors. Lessons learned during Asia involve, in roughly descending order of training are analyzed in the classroom, in line importance, better aligning the teaching at with theoretical and academic lessons, and university and fi rms’ skill demands through then fed back to industry.56 In Vietnam the collaboration in curriculum development, Ho Chi Minh University of Technology has training fi rms’ current workers, encouraging been active in training and skill development entrepreneurship, setting up university incu- for local businesses.57 bators, establishing TLOs and spin-offs, and Encouraging entrepreneurship in ter- developing extension and product develop- tiary institutions. Entrepreneurship training ment services. Bringing in intermediaries, pro- for faculty, students, managers, and work- viding matching funds, and pursuing greater ers is a very promising university-industry sectoral and subsectoral selectivity will be link for all countries. This will help boost instrumental in these modalities’ success. weak management and leadership skills in Encouraging university-industry collabo- low- and middle-income East Asia, indi- ration in curriculum development. Aligning rectly lifting productivity. It should also help the skill demand of firms and instruction overcome the reluctance of some faculty at university is probably the most urgent members to get in touch with the business imperative, and stronger collaboration in world—possibly because of a lack of business curriculum development is necessary for acumen and entrepreneurial drive—even if meeting it. Among the most widely used (and they have promising fi ndings and new com- most promising) modalities are appointing mercializable ideas. Deepening the pool of industrial practitioners on the staff of higher entrepreneurs can also complement research- education institutions, establishing consulta- promotion efforts.58 Potter (2008) suggests tion mechanisms with industry, prioritizing various routes,59 and the National University student internships to encourage on-the-job of Singapore has introduced several of these learning and retrofit the curriculum, and approaches in its entrepreneurship courses. including fi rm representatives on university Setting up university incubators. This can boards. help budding entrepreneurs commercialize In New Zealand, for example, the Tertiary ideas produced in a university. Universities in Education Commission has funded “experts Taiwan, China, for example, are setting up in residence” from industry to increase the incubation services to stimulate interactions 164 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

with fi rms;60 universities there run 81 percent owned in other countries and just happen to of all incubators. But for incubators to launch be owned by universities in China.66 viable start-ups, they should meet several Start-up risks can be high. Even in more conditions, in descending order: the univer- promising fi elds, such as biopharmaceuticals sity is a hotbed of ideas and a supportive pro- and nanotechnology, the payoff is uncer- moter;61 incubation facilities are networked tain.67 And start-ups initiated by faculty with other incubators and fi rms;62 and the “may lead productive faculty to leave the university provides services facilitating the university, ultimately undermining the uni- exit of fi rms from the university orbit.63 versity’s underlying capacity to generate new Establishing TLOs and spin-offs. Commer- knowledge.”68 Spin-offs can, however, fulfi ll cializing research and transferring technology a role of knowledge and technology transfer, by setting up TLOs and university-supported including joint research projects and assimi- or sponsored spin-offs have good potential lation of existing technology, and so offer (box 6.3) but need to be well managed. some potential in countries of the middle Among other things, TLOs can help uni- and possibly lower technology clusters. versities address legal issues and deal with Reinforcing research and education policies the “nitty-gritty” in reaching out to firms. through extension and product development Yet experience with university TLOs in the services. Particularly to small and medium United States and Europe suggests that they enterprises, such services can offer several often struggle to recruit and retain staff with benefits: providing a means of transferring the needed skills and that their incentive is to valuable technical and problem-solving skills maximize revenue (inducing them to focus on to industry; increasing the skill intensity of the the few most promising areas of research and subsector and encouraging research and devel- to neglect other technologies with longer- opment activity in fi rms that rarely engage in term potential). Skills and university leader- it; giving university graduates an opportunity ship are thus needed to support results. Sepa- to acquire practical experience and provid- rately, income generation through TLOs will ing job opportunities; and partly neutralizing probably remain small relative to total uni- the disincentive effects of the recent global versity income.64 economic downturn for students contemplat- The potential of spin-offs (start-ups) is ing a future in science and engineering or in possibly more for knowledge transfer than research and development.69 commercializing research. Spin-offs have The Fraunhofer Institutes in Germany,70 become a lucrative source of revenue for the Technology Transfer from Research many research institutions in China,65 but Institutes to Small and Medium Enterprises only a handful are truly innovative; the bulk program in Sweden, the Advanced Technol- are regular firms that would be privately ogy Program in the United States,71 and the

BOX 6.3 Technology licensing offi ces in three economies

Universities in Taiwan, China,a and in Hong Kong university-afﬁ liated enterprises (unique to China). SAR, China, have set up TLOs to facilitate technol- The two leading Beijing universities for commercial- ogy transfer and research collaboration.b Chinese izing ideas have established internal TLOs. universities are also collaborating increasingly with industry through contracts for technology services, a. Mok 2010. patent licensing, and sales, and working through b. Mok 2010. PROVIDING STEWARDSHIP FOR HIGHER EDUCATION 165

Technology Transfer Initiative (TTI) in Ire- by providing matching funds to fi rms or uni- land (box 6.4) are successful models. versities to collaborate in adapting or devel- Bringing in intermediaries. Firms them- oping technology. Such funds typically go selves should take the lead in initiating and further than competitive research funding by organizing collaboration,72 but this rarely supporting the commercialization of research happens because of credibility and communi- and tying up universities, research centers, cation gaps. Yet examples abound that inter- and fi rms in collaborative research and tech- mediary groups or associations can help to nology projects. In a skill-provision role, bridge the university-industry gap. (Box 6.5 applying the different university-industry link provides two—the fi rst suitable to all coun- approaches may require some fi nancial incen- tries, and the second more suitable to coun- tives. Additionally, training funds and dual- tries of the top technology cluster.) training programs to support internships and While more unusual, local governments postemployment training could be explored could also have an important role in relating further.74 universities to fi rms for both skill and techno- Adopting greater selectivity. Skill-related logical development, as shown by the Ho Chi university-industry links make sense in all Minh City government in Vietnam. It encour- sectors, but only a few fi rst-tier universities aged links with local businesses by establish- will ever have the potential to have strong ing a university council, which advised the formal interactions with fi rms in research and government on how to promote training and development, and within these universities innovation in universities linked to the city’s only a few departments representing specifi c development.73 sectors and subsectors may be able to pursue Providing matching funds. Along the lines university-industry links. The priority sectors of providing more effective incentives, inno- or subsectors should be the most dynamic vation funds may also support stronger links and innovative parts of the economy.

BOX 6.4 Creating university-industry links in Ireland

The Technology Transfer Initiative is an innova- Companies request visits from TTI personnel tive support structure for small and medium enter- who help identify potential research projects and prises in some regions of Ireland. It emerged from topics to be addressed at specialist seminars with an alliance of three regional universities. The TTI staff at universities. The TTI also organizes various is cofunded by them and Enterprise Ireland, a innovation clubs between industry and academics. state development agency focused on developing Companies meet regularly to discuss, present, and industry. brainstorm new research areas, to share experi- The role of the TTI is to act as a gateway for ence in new innovation, and to meet with academic companies by facilitating access to the expertise and researchers. resources of the three universities. Its core aim is Feedback on the TTI from industry represen- to encourage and assist Irish companies to become tatives suggests that it has been effective in giving more innovative and thus more competitive and small and medium enterprises access to in-depth profitable. Essentially, the TTI aims to enhance knowledge of universities and that it has given uni- technology transfer on an interindustry and inter- versity researchers a more practical look at the needs regional basis, increase innovation through research of industry. and development, and act as a single point of contact between expertise in the participating universi- Source: Technology Transfer Initiative; see http://www.biotechnologyireland ties and local industry. .com/pooled/profi les/BF_COMP/view.asp?Q=BF_COMP_9249. 166 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 6.5 Intermediary organizations

TAMA association in Japana between the University of Cambridge, United Kingdom, and the Massachusetts Institute of Tech- The aim of the TAMA (Technology Advanced Met- nology, United States) may be a good model for ropolitan Area) association is to improve the com- government-industry-university collaboration. petitiveness of small and medium enterprises in the One example is the Silent Aircraft knowledge- Tama region (northwest of Tokyo). The creation of integrating community of the U.K. government. the association was initially supported by the Kanto Because of the nature of the problem and the indus- regional bureau of the Ministry of Economy, Trade, try, collaboration among these three parties was and Industry. Members of the association include seen as essential. A government body was needed small and medium enterprises, regional tertiary to ensure compliance with existing regulations and institutes, and banks. potentially amending some of these. Industry part- The benefits of belonging to the association ners (airlines) were involved because they are the end accrue mostly to those enterprises engaged in prod- users. The analytical and research part was under- uct innovation, rather than process innovation. So taken by a research institute. The involvement of the ﬁ rst condition for this model (and other interme- banks helps address issues of capital. diary models) to have some utility is the existence of ﬁ rms that are innovation driven, regardless of size.

Knowledge-integrating communityb

A knowledge-integrating community supported a. T. Kodama 2008. by the Cambridge–MIT Institute (a collaboration b. Acworth 2008.

High-tech subsectors offer potential for international: students enroll in full-time university-industry links—and in China, programs in foreign countries, faculty take Thailand, and Vietnam particularly, in up fellowships and posts in countries other information technology and electronics. Yet than their own, and more and more students other more medium-tech parts of an econ- are using distance learning opportunities to omy may offer scope for these links, such take courses online in other countries. Pro- as the rubber industry in Thailand, which viding effective stewardship of their higher has shown promising collaboration, nota- education systems will mean that govern- bly in testing and standards.75 Some lower- ments will have to confront the complex value-added sectors dominated by small and growing trends of people mobility, pro- and medium enterprises may also offer such gram mobility, and institutional mobility in scope, to the extent that they are upgrad- ways that promote quality and recognition, ing their technology, or some subsectors in increase access and equity, encourage cost- agriculture.76 effectiveness, and build capacity. Effective stewardship of the international dimensions Stewardship of the of higher education can help higher educa- internationalization of tion institutions meet skill and research higher education demands of ﬁ rms (the ﬁ rst and second disconnects) and collaborate with other skill Over the past decade, higher educa- and research providers (the third and fourth tion around the world has become more disconnects). PROVIDING STEWARDSHIP FOR HIGHER EDUCATION 167

Characteristics of internationalized FIGURE 6.6 Tertiary students studying abroad, 2007 higher education in East Asia Cambodia Perhaps the biggest growing trend in internationalized higher education in East Asia is China that of international “people mobility.” Of all Hong Kong SAR, China East Asian countries, China sends by far the Indonesia greatest number of students to study in foreign Japan countries. In 2007 more than 421,000 Chi- Lao PDR nese students were enrolled in tertiary educa- Malaysia tion classes in a foreign country (fi gure 6.6). Mongolia While the large disparity between China and Philippines other countries in the region on this indica- Singapore tor can likely be explained by population Thailand differences, the number of tertiary students Vietnam studying abroad from other less populated 0 50 100 150 200 250 300 350 400 450 countries in the region is not insignifi cant. thousands More than 54,000 Japanese tertiary students were studying abroad in 2007, 45,000 Malay- Source: UIS (UNESCO Institute for Statistics) Data Centre. sian students, and more than 32,000 students from Hong Kong SAR, China. Among lower- income countries in the region, Vietnam FIGURE 6.7 Asian students studying abroad in tertiary schools sent about 27,000 tertiary students to study in the top-fi ve receiving countries, 2007 abroad in 2007. The highest proportion of Asian tertiary students studying abroad in 2007 chose 35 to study in the United States (33 percent; 30 fi gure 6.7). About 13 percent chose to study in the United Kingdom, 12 percent in 25 Australia, 10 percent in Japan, and 8 percent 20 in Germany. This list of receiving countries is

percent 15 not surprising: the United States, the United Kingdom, Australia, Japan, and Germany 10 account for almost 80 percent of foreign stu- 5 dents worldwide and more than 76 percent of 77 0 Asian students. United United Australia Japan Germany Program mobility (which involves States Kingdom e-learning and face-to-face teaching in local partner institutions) is the second most popu- Source: UIS Data Centre. lar form of cross-border higher education (box 6.6). Institutional mobility, where universities open campuses in foreign countries, institutions. Some countries see cross-border is also an increasingly important, if novel to exchanges as skilled migration: the recruit- East Asia, feature of cross-border education. ment and retention of talented international Cross-border higher education can address students and faculty can pay large dividends several disconnects identified in chapter 3, for a host country’s knowledge economy and and countries are pursuing opportunities to make its higher education system more com- improve their higher education systems. The petitive and innovative. In this way, interna- mobility of domestic and foreign students tionalization of higher education can help and staff often encourage political, cultural, address the first and second disconnects. and academic mutual understanding between Moreover, giving more foreign students access 168 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

BOX 6.6 E-learning and virtual universities as instruments of internationalization

To increase accessibility to tertiary education, coun- Open University Catalunya’s enrollment is more than tries are turning with greater frequency to distance- 20 percent international. German and Mexican vir- learning centers and virtual universities. For many tual universities reported similar international enroll- countries, this trend is attractive for three main rea- ment ratios. sons: it allows the sharing of resources, costs, and As virtual universities become more popular, infrastructure to deliver e-learning; it provides a a central challenge will be to ensure quality and platform for competing with international provid- consumer protection for students and parents. ers; and it can reduce duplication among existing National regulatory frameworks will have to take universities. into account international benchmarks to ensure In most of the world, particularly East Asia, vir- the comparability of degrees and easier comparison tual universities are a novel activity. While enroll- against international standards. With these efforts ments are difﬁ cult to track, particularly in the region, to enhance the international comparability of virtual the Organisation for Economic Co-operation and learning, countries will need to improve the infor- Development’s Center for Educational Research and mation available to prospective students so that they Innovation recently surveyed tertiary institutions can make informed decisions about where to enroll. offering e-learning modules to learn more about their Institutional autonomy will also be important. Being activities. It found that many virtual universities are able to partner with information technology institu- being used as instruments of internationalization. tions and providers to keep costs manageable will be For example, the United Kingdom’s Open University a continuing challenge for virtual universities. enrolls more than 25,000 international students, more than 15 percent of the total enrollment. Similarly, the Source: OECD 2004.

to domestic higher education can lead to (as when foreign students generate income for increased revenue generation for institutions, institutions through entrepreneurship), and through charging full tuition fees. And cross- the need to develop a more skilled workforce border education can offer opportunities (such as scholarship programs for domes- for capacity building for sending countries: tic civil servants and academics to facilitate fellowship programs for civil servants, aca- knowledge transfers between host and local demics, and students can be policy-building institutions). Malaysia; Hong Kong SAR, instruments, in addition to developing teach- China; China; and Singapore have been paring and research skills. These opportunities ticularly active in cross-border education dur- can help address the ﬁ rst and second discon- ing the past decade. nects, as well as the disconnects between Recent research confirms that the deci- higher education institutions themselves and sions to study abroad are often rooted in higher education institutions and other skill the labor market: students decide to enroll providers (fourth disconnect). in tertiary education abroad to obtain a job In East Asia, the increasing cross-border in that country.78 Indeed, greater numbers of participation appears to be demand driven: students decide to pursue higher education the growing internationalization of higher in the United States from “lower skill-price” education is the result of skilled-worker countries than from “higher skill-price” coun- migration (such as the recruitment of talented tries. Moreover, “higher skill-price” countries international students to work in the host attract more foreign students than “lower country), revenue generation opportunities skill-price countries.” PROVIDING STEWARDSHIP FOR HIGHER EDUCATION 169

Increasing the number of tertiary insti- educational backgrounds are less likely tutions in a sending country with a low to participate. Financial support through skill-price increases outbound migration of targeted or means-tested scholarship pro- tertiary students, whereas improving the grams and increased information about quality of domestic tertiary institutions cross-border programs could narrow this decreases student migration.79 This is likely equity gap. To help universities become the result of the higher number of college more relevant, demand driven, entrepre- graduates increasing the number of work- neurial, and profi table, they need fi nan- ers who would benefit from migrating to cial autonomy so that they can control the high skill-price countries. But higher quality revenue they generate. retains students in country. • Fourth, to build capacity, cross-border In East Asia, students who study abroad at initiatives need to meet quality and skill the tertiary level tend to return to their home needs of countries in ways that promote country in larger numbers than do students lifelong learning, while managing the risk who study abroad from other regions.80 But of brain drain. foreign-trained students from low-wage countries are less likely to return than foreign- *** trained students from high-income countries. The previous parts of this study documented some of the main constraints that Moving forward East Asia’s higher education systems are facing. Five fundamental disconnects prevent While the potential of the internationaliza- the region’s higher education systems from tion of higher education for East Asia is contributing to innovation and technologi- great, it must be managed carefully. How cal capability and from addressing the mis- can it be handled in ways that promote matches between the demand for and supply quality and recognition, increase access and of skills. equity, encourage cost-effectiveness, build These structural deficits result largely capacity, and best address the disconnects from country failures to implement coherent identifi ed in chapter 3? sets of policies to address the challenges and Lessons from the international experience shape the directions of their higher education suggest the following:81 systems holistically. This lack of vision has • First, higher education policies in low- led to ineffi ciencies throughout the system. income countries and countries with There is no magic formula for articu- low skill-prices need to be informed by lating and implementing a fully fledged the causes and consequences of the out- vision for higher education, but success migration of students seeking higher- will depend on two key factors the previous paying jobs. chapters have examined: (a) setting the right • Second, as the global competition for skills incentives and (b) mobilizing and allocating intensifi es, promoting consumer protec- resources effi ciently and effectively. Effec- tion is paramount: students need trans- tive fi nancing policies will need to include parent information about the quality and a strategic framework that allows further international validity of programs they resource mobilization while maximizing wish to enroll in. This implies that quality the effectiveness and effi ciency of fund use assurance and accreditation systems need and allocation. Improving the manage- to cooperate at the international level and ment of public institutions will require that reference international benchmarks. countries reassess autonomy and account- • Third, because student participation in ability within their systems. And providing cross-border education is largely fi nanced effective stewardship will require countries by students, those of lower income and to better coordinate different government 170 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

agencies, steer private higher education, Commerce are individually much larger than encourage university-industry links, and the country’s demand-absorbing institutions navigate the increasing trend of cross- (Praphamontripong 2010). border education. 20. Okada 2005. Two steps forward, one step back. East 21. Levy 2010. 22. Levy 2010. Asia’s higher-income economies have made 23. The government policy stance mostly includes spectacular gains in higher education delivery the existence of a supportive legal, strategic, over the past few decades, but the low- and and fi scal framework for PHE (PHE promo- middle-income economies are struggling to tion laws and decrees, PHE growth targets, replicate the standard bearers’ success. These fi scal incentives) and decisions on public sec- final chapters have offered suggestions for tor expansion. how the higher education systems in low- and 24. M. Lee 1999. middle-income countries can close the gap 25. Mei 2002. with global leaders—some of which are their 26. Cao 2007; Deng 1997; J. Lin 1999. neighbors. 27. World Bank 2009b. 28. The diagnostic of the regulatory environment across countries is constrained, however, Notes by lack of systematic and fully comparable information across countries. 1. World Bank 2010f. 29. UNESCO 2006. 2. World Bank 2010d. 30. Cai and Yan 2009. 3. Mok 2010. 31. Sukamoto 2002. 4. Tertiary Education Commission of New 32. The share of the private sector has been Zealand 2009. increasing rapidly in China from less than 9 5. This omits some Asian countries with older percent in 2002 to about 20 percent now. numbers. 33. Lax regulation is more likely to be an issue 6. Levy 2008b, 2010. where the internal management structure of a 7. Praphamontripong 2010. PHEI does not support accountability to the 8. Although many Chinese PHEIs carry “univer- broader community. sity” in their name, there are offi cially no pri- 34. Alternative (nontuition) private funding in vate universities (Yan 2010). East Asia, as elsewhere, is more common 9. Valisno 2002. in semi-elite and religious or cultural subsec- 10. Levy 2010. tors. 11. Di Gropello, Kruse, and Tandon 2011. 35. Valisno 2002; World Bank 2008. 12. BAN-PT (Indonesia’s National Higher Edu- 36. Japan is often taken as an example and bench- cation Accreditation Board) was established mark in the advocacy of East Asian govern- to introduce a quality-awareness and self- ment subsidies to PHEIs. assessment tradition among public and pri- 37. Yonezawa and Baba 1998. vate sector providers. It has also recently 38. Levy 2010. begun to reference international benchmarks 39. LaRocque 2002. in its accreditation processes. 40. Bjarnason and others 2009. 13. World Bank 2010e. 41. Yonezawa and Baba 1998. 14. In Vietnam in 2006, for example, average 42. Mei 2002. published articles per academic staff were 43. Bastiaens 2009. only 0.01 percent for private universities but 44. M. Lee 2004. 0.45 percent for public universities (World 45. Postiglione and Mak 1997. Bank 2008). 46. Ziderman 2003, 2006. 15. Orbeta 2008; Yilmaz 2009. 47. S. H. Lee 1998. 16. Levy 2008a. 48. Arguments include fairness to the private sec- 17. Doner, Intarakumnerd, and Ritchie 2010. tor, expansion of PHE, improvement of PHE, 18. LaRocque 2002. increased competition between the public 19. Semi-elite universities such as Bangkok and private sectors, that higher education University, Dhurakij Pundit University, provides externalities and so merits public and the University of the Thai Chamber of subsidies, and that a vulnerable private sector PROVIDING STEWARDSHIP FOR HIGHER EDUCATION 171

can lead to a proportional shift that burdens 62. Hansen and others 2000. the public sector. 63. Potter 2008. 49. Chapman and Drysdale 2008. 64. Perhaps no more than fi ve universities in the 50. These are typically business, fi nance, lan- United States derive signifi cant income from guages, and various social sciences. licensing of research fi ndings or from royal- 51. Limited subsidies have not prevented the ties (Geiger and Sa 2008). sector from expanding (or moving to higher 65. Wu 2010. quality) in Korea, for example. 66. Chen and Kenney 2007. 52. Notwithstanding the proportional decline in 67. Chinese universities are also discovering a couple of countries. the downside of start-ups and are distanc- 53. In this respect, Malaysia, Thailand, and ing themselves from direct ownership and Vietnam may want to change their policy responsibilities (Zhou 2008). stance. They might also consider making the 68. Miner and others 2001. impact on PHE one element in their delibera- 69. Bramwell and Wolfe 2008; Lundvall 2007. tions on public expansion. 70. Yusuf and Nabeshima 2009. 54. As seen in Taiwan, China. 71. Darby, Zucker, and Wang 2004. 55. OECD 2008b. 72. As Seagate did in Thailand in the late 1990s 56. Tansisin 2007. (Doner, Intarakumnerd, and Ritchie 2010). 57. Ca 2006. 73. Ca 2006. 58. Audretsch 2008. 74. In East Asia the case of the dual-training pro- 59. These approaches include classroom lectures, gram in the Philippines is a good example of business plans, case studies, entrepreneurs as government, fi rm, and institution collabora- guest speakers, student business start-ups, busi- tion in training (di Gropello, Tan, and Tandon ness games, student entrepreneur clubs and net- 2010). works, placements with small fi rms, feasibility 75. Doner, Intarakumnerd, and Ritchie 2010. studies, communication training, consulting 76. See, for example, Brimble and Doner (2007) for small and medium enterprises, support for and Kruss and Lorentzen (2007) on the graduate student start-ups following the course, shrimp industry in Thailand and the wine universitywide entrepreneurship education, industry in South Africa, respectively. specialist entrepreneurship degrees, distance 77. Rosenzweig 2009. education programs, external partnerships, and 78. Rosenzweig 2009. courses for entrepreneurship teachers. 79. Rosenzweig 2009. 60. Mok 2010. 80. Rosenzweig 2009. 61. Patton, Warren, and Bream 2009. 81. OECD 2004; UNESCO 2006.

Appendixes

A Number and Type of Higher Education Institutions in East Asia ...... 174

B Economic Indicators ...... 175

C Trends in Returns to Skill and Share of Skilled Workers, by Sector . . . . . 179

D Changes in Wage and Industry Education Premiums, by Sub-Sector . . . . 182

E Openness, Technology, and Demand for Tertiary Graduates, Regression Tables ...... 186

F Demand for Job-Speciﬁ c and Generic Skills in East Asia ...... 189

G Determinants of Skill Gap Indicators ...... 194

H Reasons for Skill Shortages in East Asia ...... 200

I Comparative Demand- and Supply-Side Indicators ...... 202

J Skill Gaps in East Asia ...... 204

K Skill Gaps across Professionals and Skilled Workers in Indonesia and the Philippines ...... 208

L Doctoral Degrees Earned in Science and Engineering ...... 210

M Simulations for Financing Higher Education ...... 211

173 174 APPENDIX A

Appendix A Number and Type of Higher Education Institutions in East Asia

TABLE A.1 Number and type of higher education institutions in East Asia

Economy Types of higher education institutions Cambodia Public: 34 institutions Private: 57 institutions Other: higher education includes universities, institutes or technical institutes, and a royal academy China Public: 1,079 four-year institutions, 1,184 three-year vocational or technical institutions, 400 adult higher education institutions Private: 218 Minban higher education, 322 independent colleges and universities Other: self-study programs provided by distance-education universities Indonesia Public: 48 universities, 7 institutes, 2 colleges (Sekolah Tinggi), 26 polytechnics Private: 372 universities, 42 institutes, 1,249 colleges (Sekolah Tinggi), 985 academies, 118 polytechnics Other: 52 Islamic universities Korea, Rep. Public: 26 universities, 15 junior colleges, 11 University of Education, 8 industrial universities, 1 open university Private: 145 universities, 143 junior colleges, 10 industrial universities, 1 technical university, 17 cyber universities, 1 corporate university, 28 grad school universities, 5 miscellaneous schools Lao PDR Public: 22 institutions Private: 77 institutions Others: higher education includes universities, institutes, or technical institutes Malaysia Public: 20 universities Private: 487 total (20 universities, 21 university colleges, 398 colleges, 5 foreign branch campuses, open universities, virtual universities) Other: 24 polytechnics, 37 community colleges Mongolia Public: 49 institutions Private: 131 institutions Other: colleges award diplomas (three years) and bachelor’s degrees (four years); institutions may also provide master’s (1.5 to 2 years); universities off er all degrees including doctoral (three to fi ve years) Philippines Public: 110 state universities and colleges with 424 satellite campuses,a 1 Commission on Higher Education–supervised institutions, 93 local universities and colleges, and 15 other government schools and special higher education institutions Private: 334 sectarian institutions (religious, private nonprofi t) and 1,270 nonsectarian institutions (private for profi t) Singapore Public: 5 autonomous universities Private: 9 universities/institutions Other: 13 junior colleges (two to three years), 1 centralized institute (three years), 5 polytechnics, 1 Institute of Technical Education (one to two years), 2 arts institutions Thailand Public: 66 selective-admissions universities, 2 open-admissions universities, 11 autonomous universities, 19 two-year community colleges Private: 39 universities, 25 colleges, 7 private institutions Other: 1–4 years for vocational diploma with two to three years for bachelor’s degree, 2 autonomous Buddhist universities Vietnam Public: 306 institutions Private: 80 nonpublic institutions Other: four- to six-year universities, three-year colleges, one- to three-year master’s degrees, two- to four-year doctorate (after master’s), vocational college, vocational secondary, vocational training center, in-service higher education

Source: Authors’ elaboration. a. Satellite campuses include state universities and college extension campuses and external study center. ECONOMIC INDICATORS 175

Appendix B Economic Indicators

TABLE B.1 GDP growth of upper-income East Asian economies average annual percent

Economy 1990–99 2000–09 Singapore 7.6 5.6 Hong Kong SAR, China 3.6 5.0a Korea, Rep. 6.3 4.4 Taiwan, China 6.5 3.6a Japan 1.5 0.7

Source: World Development Indicators (WDI) database. a. Average for 2000–08.

TABLE B.2 Savings and investment of upper-income East Asian economies average percentage of GDP

Gross domestic savings Gross capital formation Economy 1990–99 2000–09 1990–99 2000–09 Singapore 48.3 45.7a 35.2 23.2a Korea, Rep. 36.3 31.6 35.4 29.5 Hong Kong SAR, China 31.9 31.5a 29.4 22.6a Taiwan, China 26.8 25.9a 24.4 20.7a Japan 30.6 24.6 29.0 23.4

Source: WDI database. a. Average for 2000–08.

TABLE B.3 Exports of upper-income East Asian economies average percentage of GDP

Economy 1990–99 2000–09 Singapore — 215.0a Hong Kong SAR, China 133.3 179.7b Taiwan, China 45.7 62.1b Korea, Rep. 30.8 40.7 Japan 9.9 13.6

Source: WDI database. Note: — not available. a. Average for 2001–08. b. Average for 2000–08. 176 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

TABLE B.4 Leading export sectors, upper-income East Asian economies percentage of total exports

Economy 1995 2007 Japan Engineering products 26.5 25.7 Automotive products 18.0 23.3 Electronic and electrical products 28.1 17.6 Process industry 7.1 9.0 Korea, Rep. Electronic and electrical products 29.5 26.9 Engineering products 15.1 18.3 Automotive products 7.4 13.2 Other resource-based products 4.4 10.5 Taiwan, Chinaa Electronic and electrical products 36.0 35.5 Other low-technology products 16.1 13.5 Process industry 8.6 12.0 Engineering products 13.7 11.3 Singapore Electronic and electrical products 52.5 38.4 Other resource-based products 9.7 19.4 Engineering products 14.5 18.7 Process industry 4.6 6.2

Source: World Integrated Trade Solution (WITS) database. Note: Data are not available for Hong Kong SAR, China. a. Data for initial year are from 1997.

TABLE B.5 GDP growth of middle-income East Asian economies average annual percentage

Economy 1990–99 2000–09 China 10.0 10.3 Mongolia −0.3 5.9 Indonesia 4.8 5.1 Malaysia 7.2 4.8 Philippines 2.8 4.6 Thailand 5.3 4.1

Source: WDI database.

TABLE B.6 Exports of middle-income East Asian economies average percentage of GDP

Economy 1990–99 2000–09 Malaysia 91.2 110.5 Thailand 43.0 69.9 Mongolia 50.5 59.8 Philippines 38.1 46.1 Indonesia 30.1 32.4 China 19.6 31.1

Source: WDI database. ECONOMIC INDICATORS 177

TABLE B.7 Savings and investment of middle-income East Asian economies average percentage of GDP

Gross domestic savings Gross capital formation Economy 1990–99 2000–09 1990–99 2000–09 China 41.2 45.8 39.1 41.2 Malaysia 40.7 42.2 36.3 21.8 Thailand 35.3 31.7 36.3 25.9 Indonesia 30.2 30.5 27.6 25.0 Mongolia 18.7 25.4 31.2 35.3 Philippines 16.2 14.9 22.4 16.6

Source: WDI database.

TABLE B.8 Leading export sectors, middle-income East Asian economies percentage of total exports

Economy 1995 2007 China Electronic and electrical products 10.8 30.1 Textiles, garments, and footwear 30.9 16.5 Engineering 10.6 15.3 Other low-technology products 15.8 14.8 Indonesia Primary products 36.0 35.4 Agro-based products 18.6 17.6 Other resource-based products 8.3 12.5 Textiles, garments, and footwear 15.9 9.4 Malaysia Electronic and electrical products 38.7 39.3 Primary products 10.9 15.3 Agro-based products 15.2 11.4 Engineering products 15.4 9.6 Philippines Electronic and electrical products 49.0 47.9 Primary products 7.3 9.9 Agro-based products 7.3 9.0 Other resource-based products 3.0 7.3 Thailand Electronic and electrical products 23.0 23.2 Engineering products 10.1 13.6 Primary products 16.9 11.4 Other low-technology products 11.1 10.2

Source: WITS database. Note: Data are not available for Mongolia.

TABLE B.9 GDP growth of lower-income East Asian economies average annual percentage

Economy 1990–99 2000–09 Cambodia 7.2a 8.1 Vietnam 7.4 7.3 Lao PDR 6.3 6.7

Source: WDI database. a. Cambodia average for 1994–99. 178 PUTTING HIGHER EDUCATION TO WORK: SKILLS AND RESEARCH FOR GROWTH IN EAST ASIA

TABLE B.10 Exports of lower-income East Asian economies average percentage of GDP

Economy 1990–99 2000–09 Vietnam 37.6 65.8 Cambodia 29.1a 60.1 Lao PDR 23.0 31.4b

Source: WDI database. a. Cambodia average for 1993–99. b. Lao PDR average for 2000–08.

TABLE B.11 Savings and investment of lower-income East Asian economies average percentage of GDP

Gross domestic savings Gross capital formation Economy 1990–99 2000–09 1990–99 2000–09 Vietnam 16.0 28.3 23.5 35.8 Lao PDR — 19.9a —31.4a Cambodia −0.6b 11.5 13.8 b 19.1

Source: WDI database. Note: — not available. a. Lao PDR averages for 2000–08. b. Cambodia averages for 1993–99.

TABLE B.12 Leading export sectors, lower-income East Asian economies percentage of total exports

Economy 2000 2007 Vietnam Primary products 51.6 40.6 Textiles, garments, and footwear 26.6 27.3 Other low-tech products 4.6 10.4 Engineering products 2.9 5.6 Cambodia Textiles, garments, and footwear 72.8 72.9 Other low-tech products 20.3 23.1 Primary products 1.3 2.3 Agro-based products 4.5 0.6

Source: WITS database. Note: Data are not available for Lao PDR. TRENDS IN RETURNS TO SKILL AND SHARE OF SKILLED WORKERS, BY SECTOR 179

Appendix C Trends in Returns to Skill and Share of Skilled Workers, by Sector

FIGURE C.1 Return to skill by sector over time FIGURE C.2 Proportion of skilled workers by (tertiary and above), Indonesia sector over time (tertiary and above), Indonesia

20 40

15 30

10 20 percent percent 5 10

0 0 1994 1999 2003 2007 1994 1999 2003 2007 agriculture/mining manufacturing agriculture/mining manufacturing trade/services trade/services

Source: di Gropello and Sakellariou 2010, based on labor and household Source: di Gropello and Sakellariou 2010, based on labor and household surveys (various years). surveys (various years).

FIGURE C.3 Return to education by sector over FIGURE C.4 Proportion of educated workers by time (tertiary and above), Philippines sector over time (tertiary and above), Philippines

10 50

8 40

6 30 percent percent 4 20

2 10

0 0 1988 1994 2001 2006 1988 1994 2001 2006 agriculture/mining manufacturing agriculture/mining manufacturing trade/services trade/services

Source: di Gropello and Sakellariou 2010, based on labor and household Source: di Gropello and Sakellariou 2010, based on labor and household surveys (various years). surveys (various years). 180 APPENDIX C

FIGURE C.5 Return to skill by sector over time FIGURE C.6 Proportion of skilled workers by (tertiary and above), Thailand sector over time (tertiary and above), Thailand

30 25 25 20 20 15 15

percent 10 10 percent 5 5 0 0 1990 1994 2000 2004 1990 1994 2000 2004 agriculture/mining manufacturing agriculture/mining manufacturing trade/services trade/services

Source: di Gropello and Sakellariou 2010, based on labor and household Source: di Gropello and Sakellariou 2010, based on labor and household surveys (various years). surveys (various years).

FIGURE C.7 Return to education by sector over FIGURE C.8 Proportion of educated workers by time (tertiary and above), Vietnam sector over time (tertiary and above), Vietnam

10 25 20 5 15

percent 10 percent 0 5 0 –0.5 1990 1998 2004 2006 1992 1998 2004 2006 agriculture/mining manufacturing agriculture/mining manufacturing trade/services trade/services

Source: di Gropello and Sakellariou 2010, based on labor and household Source: di Gropello and Sakellariou 2010, based on labor and household surveys (various years). surveys (various years).

FIGURE C.9 Return to skill by sector over time FIGURE C.10 Proportion of skilled workers by (tertiary and above), Cambodia sector over time (tertiary and above), Cambodia

15 15

10 10 5 percent percent 0 5 –5

–10 0 1997 2003–04 2007 1997 2003–04 2007 manufacturing trade/services manufacturing trade/services

FIGURE C.11 Return to skill by sector over time FIGURE C.12 Proportion of skilled labor by sector (tertiary and above), China over time (tertiary and above), China

8 40 35 6 30 25 4 20 percent

percent 15 2 10 5 0 0 1999 2005 1999 2005 agriculture/mining manufacturing agriculture/mining manufacturing trade/services trade/services

Source: di Gropello and Sakellariou 2010, based on labor and household Source: di Gropello and Sakellariou 2010, based on labor and household surveys (various years). surveys (various years).

FIGURE C.13 Return to skill by sector over time FIGURE C.14 Proportion of skilled labor by sector (tertiary and above), Mongolia over time (tertiary and above), Mongolia

10 60 8 50 40 6 30 4 percent percent 20 2 10 0 0 1998 2002 2007 1998 2002 2007 agriculture/mining manufacturing agriculture/mining manufacturing trade/services trade/services

Appendix D Changes in Wage and Industry Education Premiums, by Subsector

FIGURE D.1 Change in unstandardized weighted industry and skill premiums by sector, Indonesia, 1996–2007

4 3 2 1 0 –1

% change –2 –3 –4 –5 –6 1234567891011121314151617181920 industry code change in industry premium change in skill premium

Source: di Gropello and Sakellariou 2010. Note: The numbers on the x axis refer to these industries: 1. agriculture; 2. forestry; 3. fi shery; 4. mining/minerals; 5. metal ore/other mining; 6. food/ beverage/tobacco; 7. textile; 8. timber and furniture; 9. paper and printing; 10. chemicals; 11. nonmetallic production; 12. metal industry; 13. other industry; 14. utilities; 15. construction; 16. wholesale trade; 17. retail trade; 18. transportation/communications; 19. fi nance/real estate; 20. public administration/ health/other services. CHANGES IN WAGE AND INDUSTRY EDUCATION PREMIUMS, BY SUBSECTOR 183

FIGURE D.2 Change in unstandardized weighted industry and skill premiums by sector, Philippines, 1988–2006

–10 % change –20

–30

–40

–50 123579114 6 8 10121416182013 15 17 19 industry code change in industry premium change in skill premium

Source: di Gropello and Sakellariou 2010. Note: The numbers on the x axis refer to these industries: 1. growing crops; 2. fi shing; 3. nonmetal mining; 4. food/beverages/tobacco; 5. textiles; 6. wood/furniture; 7. paper products; 8. chemicals; 9. utilities; 10. construction; 11. wholesale trade; 12. retail trade; 13. hotels/restaurants; 14. transportation; 15. fi nancial services; 16. business services; 17. public administration; 18. education; 19. health and social work; 20. private household employment.

FIGURE D.3 Change in unstandardized weighted industry and skill premiums by sector, Indonesia, 1990–2004

–5 % change –10

–15

–20

–25 1 2 345678 industry code change in industry premium change in skill premium

Source: di Gropello and Sakellariou 2010. Note: The numbers on the x axis refer to these industries: 1. agriculture; 2. mining; 3. manufacturing; 4. utilities; 5. construction; 6. trade; 7. transportation; 8. other services. 184 APPENDIX D

FIGURE D.4 Change in unstandardized weighted industry and skill premiums by sector, Vietnam, 1992–2006

4 % change 2

–2 12345678910111213141516 industry code change in industry premium change in skill premium

Source: di Gropello and Sakellariou 2010. Note: The numbers on the x axis refer to these industries: 1. agriculture; 2. mining; 3. food/beverage/tobacco; 4. textiles; 5. wood/furniture; 6. paper; 7. chemicals; 8. nonmetal mining products; 9. metal; 10. other manufacturing; 11. utilities; 12. construction; 13. trade; 14. transport/communications; 15. fi nance/business; 16. other services.

FIGURE D.5 Change in unstandardized weighted industry and skill premiums by sector, Cambodia, 1997–2007

% change 0

–5

–10 123456789 industry code change in industry premium change in skill premium

Source: di Gropello and Sakellariou 2010. Note: The numbers on the x axis refer to these industries: 1. agriculture/mining; 2. manufacturing; 3. utilities; 4. construction; 5. trade; 6. transportation/ communications; 7. fi nance/business; 8. public administration; 9. other services. CHANGES IN WAGE AND INDUSTRY EDUCATION PREMIUMS, BY SUBSECTOR 185

FIGURE D.6 Change in unstandardized weighted industry and skill premiums by sector, China, 1995–2005

–5 % change –10

–15

–20 12345678 industry code change in industry premium change in skill premium

Source: di Gropello and Sakellariou 2010. Note: The numbers on the x axis refer to these industries: 1. agriculture/mining; 2. manufacturing; 3. utilities; 4. construction; 5. transportation/ communications; 6. trade; 7. public administration; 8. other services (including fi nance and business and others).

FIGURE D.7 Change in unstandardized weighted industry and skill premiums by sector, Mongolia, 1998–2007

% change 0

–5

–10 12345678 industry code change in industry premium change in skill premium

Source: di Gropello and Sakellariou 2010. Note: The numbers on the x axis refer to these industries: 1. agriculture/mining; 2. manufacturing; 3. utilities; 4. construction; 5. transportation/communications; 6. trade; 7. public administration; 8. other services (including fi nance and business and others). 186 APPENDIX E

Appendix E Openness, Technology, and Demand for Tertiary Graduates, Regression Tables

TABLE E.1 Openness, technological innovation, and the demand for skills

Variable (1) (2) (3) (4) (5) (6) (7) Exporter −0.177*** 0.043 0.041 −0.017 −0.019 −0.029 −0.058 [0.0350] [0.0350] [0.0359] [0.0389] [0.0388] [0.0387] [0.0388] Foreign ownership 0.429*** 0.358*** 0.354*** 0.283*** 0.278*** 0.262*** 0.211*** [0.0386] [0.0374] [0.0381] [0.0395] [0.0394] [0.0393] [0.0394] Technological innnovation 0.246*** 0.230*** 0.226*** [0.0316] [0.0316] [0.0318] Age of fi rm −0.00907*** −0.00320*** −0.00377*** −0.00719*** −0.00706*** −0.00713*** −0.00672*** [0.00130] [0.00121] [0.00121] [0.00127] [0.00127] [0.00125] [0.00124] Public ownership 0.261*** 0.268*** 0.291*** 0.274*** 0.267*** 0.260*** 0.256*** [0.0486] [0.0449] [0.0454] [0.0486] [0.0484] [0.0482] [0.0483] Managerial postsecondary education 0.745*** 0.696*** [0.0877] [0.0911] Sales per employee (log) 0.0447*** [0.00814] Industry fi xed eff ects? No Yes No No No No No Country fi xed eff ects? Yes Yes No No No No No Country-industry fi xed eff ects? No No Yes No No No No Country-industry-size fi xed eff ects? No No No Yes Yes Yes Yes Observations 8,087 8,087 8,087 8,087 8,047 8,047 7,746 R-squared 0.09 0.23 0.25 0.32 0.33 0.34 0.34

TABLE E.2 Openness, technological innovation, and the demand for skills: Robustness to technological variables

Variable (1) (2) (4) (5) (6) Exporter −0.0634 −0.135*** −0.137*** −0.228*** 0.00842 [0.0415] [0.0467] [0.0469] [0.0647] [0.0429] Foreign ownership 0.208*** 0.247*** 0.232*** 0.0691 0.142*** [0.0411] [0.0473] [0.0474] [0.0606] [0.0442] Technological innovation 0.199*** 0.204*** 0.244*** 0.264*** 0.0616 [0.0341] [0.0410] [0.0405] [0.0469] [0.0376] ISO certiﬁ cate 0.267*** [0.0381] R&D 0.295*** [0.0424] R&D/sales 0.00603*** [0.00214]

(continued next page) OPENNESS, TECHNOLOGY, AND DEMAND FOR TERTIARY GRADUATES, REGRESSION TABLES 187

TABLE E.2 (continued)

Variable (1) (2) (4) (5) (6) Use of computers 0.642*** [0.0543] Use of e-mail and Internet 0.558*** [0.0434] Baseline ﬁ rm characteristics? Yes Yes Yes Yes Yes Country-sector-size eﬀ ects? Yes Yes Yes Yes Yes Observations 7,208 5,069 5,069 4,059 4,928 R-squared 0.346 0.302 0.297 0.394 0.385

TABLE E.3 Openness, technological innovation, and the demand for skills: Robustness to alternative samples

Cambodia China Indonesia Korea, Rep. Malaysia Philippines Thailand Vietnam Variable (1) (2) (3) (6) (4) (5) (7) (8) Panel A: Excluding one country at a time Exporter −0.0683* 0.118*** −0.0888** −0.061 −0.0725* −0.056 −0.130*** −0.0776* [0.0386] [0.0426] [0.0397] [0.0393] [0.0393] [0.0395] [0.0475] [0.0455] Foreign ownership 0.201*** 0.186*** 0.234*** 0.208*** 0.216*** 0.229*** 0.255*** 0.0969** [0.0393] [0.0448] [0.0405] [0.0401] [0.0411] [0.0402] [0.0474] [0.0441] Technological innovation 0.240*** 0.123*** 0.227*** 0.232*** 0.234*** 0.205*** 0.251*** 0.250*** [0.0322] [0.0379] [0.0326] [0.0322] [0.0326] [0.0324] [0.0363] [0.0364] Baseline fi rm characteristics? Yes Yes Yes Yes Yes Yes Yes Yes Country-sector-size eff ects? Yes Yes Yes Yes Yes Yes Yes Yes Obser vations 7,4 43 4,931 7,270 7,261 7,427 7,255 6,451 6,184 R-squared 0.313 0.36 0.333 0.332 0.343 0.351 0.344 0.355 Panel B: Including one country at a time Exporter 1.531*** −0.406*** 0.388** 0.015 0.336 −0.028 0.142** −0.050 [0.542] [0.0755] [0.159] [0.210] [0.228] [0.166] [0.0562] [0.0665] Foreign ownership 0.628 0.191*** −0.118 0.419*** 0.131 −0.231 0.012 0.400*** [0.422] [0.0705] [0.178] [0.156] [0.138] [0.198] [0.0618] [0.0797] Technological innovation −0.161 0.342*** 0.247* 0.051 0.007 0.497*** 0.134** 0.055 [0.179] [0.0555] [0.139] [0.168] [0.135] [0.142] [0.0564] [0.0592] Baseline fi rm characteristics? Yes Yes Yes Yes Yes Yes Yes Yes Country-sector-size eff ects? Yes Yes Yes Yes Yes Yes Yes Yes Observations 303 2,815 476 485 319 491 1,295 1,562 R-squared 0.366 0.344 0.303 0.383 0.172 0.246 0.22 0.259

Source: Almeida 2009b, based on World Bank Investment Climate Survey database. Note: Dependent variable is share of workers in the fi rm with more than secondary education (12 years of schooling). Panel A excludes from the sample one country at a time. Panel B restricts the sample to one country at a time. Standard deviation shown in brackets. Signifi cance level: * = 10 percent; ** = 5 percent; *** = 1 percent. 188 APPENDIX E

TABLE E.4 Openness, technological innovation, and the demand for skills: Robustness to alternative samples

Low-income Low-middle-income Middle-income Excluding capital countries countries countries city Variable (1) (2) (3) (4) Exporter 0.00159 −0.115** −0.0958** −0.0728* [0.0746] [0.0489] [0.0474] [0.0425] Foreign ownership 0.459*** 0.0721 0.0785 0.182*** [0.0933] [0.0512] [0.0486] [0.0426] Technological innovation 0.015 0.297*** 0.280*** 0.225*** [0.0589] [0.0395] [0.0380] [0.0345] Baseline ﬁ rm characteristics? Yes Yes Yes Yes Country-sector-size eﬀ ects? Yes Yes Yes Yes Observations 1,865 5,077 5,396 6,635

Source: Almeida 2009b, based on World Bank Investment Climate Survey database. Note: Dependent variable is share of workers in the fi rm with more than secondary education (12 years of schooling). Standard deviation shown in brackets. Signifi cance level: * = 10 percent; ** = 5 percent; *** = 1 percent. DEMAND FOR JOB-SPECIFIC AND GENERIC SKILLS IN EAST ASIA 189

Appendix F Demand for Job-Specifi c and Generic Skills in East Asia

FIGURE F.1 Key job-specifi c and generic skills in a sample of East Asian economies (according to employers)

a. Philippines: Five most important job- b. Philippines: Five most important generic specific (and source of job-specific) skills for skills for managers, professionals, and managers, professionals, and skilled skilled workers, according to employers workers, according to employers

foreign degree time management technical qualification initiative negotiation voc/tech qualification leadership experience in different field problem solving experience in same field creativity teamwork practical skills independent work general experience communication theory math grades computer writing secondary diploma literacy local degree language 051015 20 25 30 0 2 4 6 8 10121416 percent percent managers skilled workers managers skilled workers

c. Indonesia: Share of firms rating job- d. Indonesia: Share of firms rating core specific (and source of job-specific) skills as skills as “very important” for managers, “very important” for managers, professionals, and skilled workers professionals, and skilled workers primary education basic skills SMU (math, literacy) SMK local university thinking skills diploma/vocational university foreign university behavioral grades and transcripts theoretical knowledge practical knowledge computer skills experience in same field experience in different field English general experience 0102030 40 50 60 0 10 20 30 40 50 60 70 80 percent percent managers skilled workers managers skilled workers

(continued next page) 190 APPENDIX F

FIGURE F.1 (continued)

e. Thailand: Share of ﬁrms indicating most important f. Malaysia: Share of ﬁrms indicating most important considerations used in recruiting (manufacturing) considerations used in recruiting (manufacturing)

education

education technical skills

experience

loyalty technical skills ethnic quotas

interpersonal skills language experience proﬁciency

othera

051015 20 25 30 35 40 45 0 5 10 15 20 25 30 35 40 percent percent 2004 2006 2002 2007

g. Vietnam: Share of firms indicating which h. Mongolia: Share of firms indicating which workers’ skills are among the three most workers’ skills are among the three most important, by education level important (manufacturing) job-related job-related theory technical skills job-related creativity practical skills theoretical communication knowledge good writing/ computer skills verbal skills foreign language English proficiency

punctuality leadership

independence problem solving general experience teamwork 0510 15 20 25 30 0 10 20 30 40 50 60 70 80 90 percent percent technical school professional school college

(continued next page) DEMAND FOR JOB-SPECIFIC AND GENERIC SKILLS IN EAST ASIA 191

FIGURE F.1 (continued)

i. Cambodia: Firm responses to skills and attributes that are the most important for professionals

analytical skills

decision-making skills

experience

foreign language proﬁciency

work attitude

communication

technical skills

education

computer skills

0 10203040 50 60 percent

Sources: di Gropello, Tan, and Tandon 2010, based on 2008 Philippines Skills Survey; di Gropello, Kruse, and Tandon 2011, based on 2008 Indonesia Skills Survey; Thailand World Bank Investment Climate Survey 2004 and 2006; Malaysia World Bank Investment Climate Survey 2002 and 2007; World Bank 2008, based on 2003 Vietnam MOLISA-ADB survey on labor market; Mongolia World Bank Investment Climate Survey 2010; HRINC. 2010, based on Cambodian Federation of Employers and Business Associations Youth and Employment Study (BDLINK Cambodia Co. 2008). a. “Other” includes thinking and problem-solving skills among other factors. 192 APPENDIX F

FIGURE F.2 Key skills in a sample of East Asian economies (according to employees)

a. Indonesia: Share of workers rating b. Indonesia: Share of workers indicating these skills as “important” or “very which skills are among the three most important” for job performance important for job performance writing dealing with people 68 English communications team skills leadership creative thinking 66 time management independence adaptability basic reading and math 60 creativity numerical skills problem solving computer skills 61 computer skills technical skills 54 56 58 60 62 64 66 68 70 0246 8101214 percent percent

c. Thailand: Share of workers rating these skills as d. Thailand: Share of workers indicating which skills are among “important” or “very important” for job performance the three most important for job performance (manufacturing professionals and skilled workers) (manufacturing professionals and skilled workers)

English 27

dealing with people 71 professional 4 social skills 3

teamwork 3

analysis and planning 52 leadership skills 6 time management 4

adaptability 2

caring for others 80 creativity/innovation 5 numerical skills 11

problem solving 6

computer skills 30 information technology skills 18

technical/professional skills 11 0 20406080100 0 5 10 15 20 25 30 percent percent

(continued next page) DEMAND FOR JOB-SPECIFIC AND GENERIC SKILLS IN EAST ASIA 193

FIGURE F.2 (continued)

e. Malaysia: Share of workers rating these skills as f. Malaysia: Share of workers indicating which skills are among “important” or “very important” for job performance the three most important for job performance (manufacturing professionals and skilled workers) (manufacturing professionals and skilled workers)

English 18 dealing with people 80 professional 12 social skills 7 teamwork 5 analysis and planning 71 leadership 7 time management 6 adaptability 4 caring for others 90 creativity/innovation 7 numerical skills 3 problem solving 5 computer skills 85 information technology skills 15 technical/professional skills 12 020406080 100 02468101214161820 percent percent

g. Mongolia: Share of workers rating these skills as “important” or “very important” for job performance (professionals and skilled workers)

job-related skills 33

creativity 12

communication 28

computer skills 34

English 57

leadership 27

problem solving 7

teamwork 10 0 10 20 30 40 50 60 percent

Source: di Gropello, Kruse, and Tandon 2011, based on 2008 Indonesia Skills Survey; Thailand World Bank Investment Climate Survey 2004; Malaysia World Bank Investment Climate Survey 2002; Mongolia World Bank Investment Climate Survey 2010. 194 APPENDIX G

Appendix G Determinants of Skill Gap Indicators

TABLE G.1 Determinants of time to fi ll skilled vacancies

Variable (1) (2) (3) (4) (5) Technological innovation 0.114 0.083 0.077 0.082 0.073 0.025* 0.026*** 0.027*** 0.026*** 0.026*** Open 0.129 0.084 0.082 0.084 0.069 0.026* 0.028*** 0.029*** 0.028*** 0.028** Small enterprise 0.063 0.119 0.074 0.078 0.050 0.047** 0.049 0.048 Medium enterprise 0.138 0.189 0.155 0.158 0.057** 0.055*** 0.056*** 0.055*** Large enterprise 0.155 0.190 0.175 0.195 0.057*** 0.055*** 0.056*** 0.056*** Very large enterprise 0.231 0.241 0.256 0.261 0.057*** 0.054*** 0.056*** 0.055*** Public ownership 0.011 0.002 0.010 0.030 0.051 0.054 0.051 0.051 Age of fi rm > 2 years and < 4 years −0.142 −0.136 −0.137 −0.094 0.794 0.792 0.784 0.792 Age of fi rm > 4 years and < 6 years −0.192 −0.187 −0.191 −0.144 0.793 0.792 0.784 0.792 Age of fi rm > 6 years −0.247 −0.242 −0.245 −0.196 0.793 0.7912 0.783 0.791 Average years of schooling of workforce 0.005

Share of skilled workers 0.158 0.142 0.060*** 0.058** Basic fi rm-level controls included? Yes Yes Yes Yes Yes Industry fi xed eff ects? Yes Yes Yes Yes No Country fi xed eff ects? Yes Yes Yes Yes Yes Country-industry fi xed eff ects? No No No No Yes Observations 4,351 4,226 3,705 4,214 4,214 R-squared 0.098 0.102 0.094 0.103 0.107

Source: Almeida 2009a, based on World Bank Investment Climate (IC) Surveys database. Note: Dependent variable is a dummy variable that assumes the value 1 if the fi rm fi lls skilled vacancies. Table reports the marginal eff ects (at mean values) on the fi rm’s propensity to fi ll skilled vacancies from probit regressions. Microfi rms (with fewer than than 10 employees) are the omitted size group. Signifi cance level: * = 10 percent; ** = 5 percent; *** = 1 percent.

TABLE G.2 Determinants of managerial perceptions of skills and education of East Asian workforce

Variable (1) (2) (3) (4) (5) Technological innovation 0.046 0.036 0.033 0.035 0.033 0.010*** 0.010*** 0.011*** 0.010*** 0.010*** Open 0.043 0.028 0.026 0.030 0.032 0.010*** 0.010*** 0.011** 0.010*** 0.011*** Small enterprise (10–49 employees) 0.038 0.026 0.048 0.051 0.021* 0.025 0.022** 0.023** Medium enterprise (50–99 employees) 0.090 0.077 0.105 0.107 0.026*** 0.030*** 0.028*** 0.029***

(continued next page) DETERMINANTS OF SKILL GAP INDICATORS 195

TABLE G.2 (continued)

Variable (1) (2) (3) (4) (5) Large enterprise (100–249 employees) 0.094 0.076 0.109 0.112 0.027*** 0.030*** 0.028*** 0.029*** Very large enterprise (+250 employees) 0.087 0.066 0.104 0.109 0.026*** 0.029** 0.027*** 0.029*** Public ownership 0.021 0.036 0.017 0.018 0.017 0.018** 0.017 0.017 Age of fi rm > 2 years and < 4 years −0.065 −0.064 −0.068 −0.069 0.036 0.043 0.036* 0.036* Age of fi rm > 4 years and < 6 years −0.074 −0.074 −0.077 −0.078 0.035* 0.042 0.035** 0.035* Age of fi rm > 6 years −0.103 −0.099 −0.105 −0.106 0.048** 0.057* 0.048** 0.049** Average years of schooling of workforce 0.001 0.002 Share of skilled workers 0.068 0.079 0.022*** 0.023*** Industry fi xed eff ects? Yes Yes Yes Yes No Country fi xed eff ects? Yes Yes Yes Yes Yes Country-industry fi xed eff ects? No No No No Yes Observations 8,099 7,961 6,737 7,864 7,726

Source: Almeida 2009a, based on World Bank IC Surveys database. Note: Dependent variable is a dummy variable equal to 1 where fi rms rate skills and education of workforce as a major or very severe obstacle to growth. Signifi cance level: * = 10 percent; ** = 5 percent; *** = 1 percent.

TABLE G.3 Determinants of managerial perceptions of skills of workforce: Robustness to additional variables

Variable (1) (2) (4) (5) (6) Technological innovation 0.034 0.018 0.030 0.038 0.029 0.011*** 0.014 0.014** 0.010*** 0.016* Open 0.026 0.043 0.032 0.020 0.030 0.011** 0.023* 0.014** 0.012* 0.016* ISO certifi cate 0.003 0.012 Research and development 0.050 0.015*** tech_R~g 0.0002 0.001 Use of e-mail and Internet 0.032 0.011*** Use of computers 0.054 0.016*** Industry fi xed eff ects? Yes Yes Yes Yes Yes Country fi xed eff ects? Yes Yes Yes Yes Yes Country-industry fi xed eff ects? No No No No No Observations 7,266 4,608 4,704 6,363 3,324

Source: Almeida 2009b, based on World Bank IC Surveys database. Note: Dependent variable is a dummy variable equal to 1 where fi rms rate skills and education of workforce as a major or very severe obstacle to growth. ISO = International Organization for Standardization. Signifi cance level: * = 10 percent; ** = 5 percent; *** = 1 percent. 196 APPENDIX G

TABLE G.4 Determinants of time to fi ll skilled vacancies: Robustness to additional variables

Variable (1) (2) (3) (4) (5) Technological innovation 0.093 0.016 0.033 0.077 0.175 0.027*** 0.028 0.027 0.029*** 0.052*** Open 0.079 0.079 0.081 0.015 0.213 0.028*** 0.030*** 0.030*** 0.034 0.051*** ISO certifi cate 1.076 0.782 Research and development (R&D) 0.099 0.031*** Percentage of R&D/sales −0.0004 0.001 Internet use 0.177 0.033*** Computer use 0.180 0.059*** Basic fi rm-level controls included? Yes Yes Yes Yes Yes Industry fi xed eff ects? Yes Yes Yes Yes Yes Country fi xed eff ects? Yes Yes Yes Yes Yes Observations 3,970 2,893 2,893 3,455 1,434 R-squared 0.112 0.041 0.039 0.119 0.134

Source: Almeida 2009b, based on World Bank IC Surveys database. Note: Dependent variable is a dummy variable that assumes the value 1 if the fi rm fi lls skilled vacancies. Table reports the marginal eff ects (at mean values) on the fi rm’s propensity to fi ll skilled vacancies from probit regressions. Microfi rms (with fewer than 10 employees) are the omitted size group. Signifi cance level: *** = 1 percent.

TABLE G.5 Determinants of managerial perceptions of skills of workforce: Robustness to diff erent geographical areas

Cambodia China Indonesia Lao PDR Malaysia Mongolia Philippines Korea, Rep. Thailand Vietnam Variable (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Panel a. Excluding one country at a time Technological innovation 0.037 0.041 0.032 0.036 0.033 0.036 0.032 0.034 0.040 0.031 0.011*** 0.010*** 0.010*** 0.010*** 0.010*** 0.010*** 0.011*** 0.011*** 0.011*** 0.012*** Open 0.032 0.028 0.030 0.032 0.034 0.031 0.027 0.029 0.027 0.033 0.011*** 0.011** 0.011*** 0.011*** 0.011*** 0.010*** 0.011** 0.011*** 0.011** 0.012*** Industry fi xed eff ects? Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Country fi xed eff ects? Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Observations 7,393 6,434 7,182 7,621 7,141 7,672 7,247 7,342 6,480 6,24 4 Panel b. Restricting sample to only one country Technological innovation −0.001 0.003 0.066 0.013 0.063 −0.044 0.062 0.025 0.017 0.046 0.029 0.029 0.033** 0.033 0.036* 0.077 0.026** 0.024 0.028 0.020** Open −0.017 0.042 0.026 −0.014 −0.005 −0.008 0.089 0.045 0.048 0.026 0.037 0.028 0.038 0.045 0.035 0.090 0.033*** 0.033 0.029* 0.022 Industry fi xed eff ects? Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Country fi xed eff ects? Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Observations 407 1,421 678 240 723 188 616 470 1,384 1,577

Source: Almeida 2009b, based on World Bank IC Surveys database. Note: Dependent variable is a dummy variable equal to 1 where fi rms rate skills and education of workforce as a major or very severe obstacle to growth. Signifi cance level: * = 10 percent; ** = 5 percent; *** = 1 percent. 197 198 APPENDIX G

TABLE G.6 Determinants of managerial perceptions of skills of workforce: Robustness to diff erent geographic areas, by income levels

Low income Low-middle income Middle income Excluding capital city Variable (2) (3) (4) (5) Technology innovation 0.032 0.031 0.035 0.040 0.014** 0.016** 0.014** 0.011*** Openness 0.012 0.044 0.037 0.032 0.017 0.015*** 0.014*** 0.011*** Industry fi xed eff ects? Yes Yes Yes Yes Country fi xed eff ects? Yes Yes Yes Yes Observations 2,505 4,102 4,825 6,436 Source: Almeida 2009b, based on World Bank IC Surveys database. Note: Dependent variable is a dummy variable equal to 1 where fi rms rate skills and education of workforce as a major or very severe obstacle to growth. Signifi cance level: ** = 5 percent; *** = 1 percent.

TABLE G.7 Determinants of time to fi ll skilled vacancies: Robustness to diff erent geographical areas

China Indonesia Malaysia Korea, Rep. Thailand Vietnam Variable (1) (2) (3) (4) (5) (6) Panel a: Excluding one country at a time Technology innovation 0.098 0.029 0.094 0.098 0.088 0.086 0.029*** 0.026 0.028*** 0.028*** 0.031*** 0.028*** Openness 0.058 0.051 0.111 0.099 0.089 0.084 0.033* 0.028* 0.029*** 0.030*** 0.034*** 0.030*** Basic fi rm-level controls included? Yes Yes Yes Yes Yes Yes Industry fi xed eff ects? Yes Yes Yes Yes Yes Yes Country fi xed eff ects? Yes Yes Yes Yes Yes Yes Observations 3,460 3,534 3,565 3,755 2,916 3,840 R-squared 0.115 0.031 0.117 0.107 0.117 0.105

Panel b. Restricting sample to only one country Technology innovation −0.006 0.457 0.018 −0.040 0.081 0.050 0.054 0.113*** 0.068 0.057 0.048* 0.049 Openness 0.121 0.284 −0.082 −0.096 0.065 0.091 0.052** 0.129** 0.076 0.082 0.049 0.066 Basic fi rm-level controls included? Yes Yes Yes Yes Yes Yes Industry fi xed eff ects? Yes Yes Yes Yes Yes Yes Country fi xed eff ects? Yes Yes Yes Yes Yes Yes Observations 754 680 649 459 1,298 374 R-squared 0.026 0.068 0.040 0.081 0.019 0.088

TABLE G.8 Determinants of time to fi ll skilled vacancies: Robustness to diff erent geographical areas, by income levels

Low-middle Excluding High income Low income Income Middle income capital city Variable (1) (2) (3) (4) (5) Technology innovation −0.040 0.050 0.131 0.104 0.086 0.057 0.049 0.035*** 0.031*** 0.029*** Openness −0.096 0.091 0.144 0.105 0.066 0.082 0.066 0.035*** 0.032*** 0.032** Basic fi rm-level controls included? Yes Yes Yes Yes Yes Industry fi xed eff ects? Yes Yes Yes Yes Yes Country fi xed eff ects? Yes Yes Yes Yes Yes Observations 459 374 2,732 3,381 3,509 R-squared 0.081 0.088 0.125 0.109 0.109

Source: Almeida 2009 b, based on World Bank IC Surveys database. Note: Dependent variable is a dummy variable that assumes the value 1 if the fi rm fi lls skilled vacancies. Table reports the marginal eff ects (at mean values) on the fi rm’s propensity to fi ll skilled vacancies from probit regressions. Microfi rms (with fewer than 10 employees) is the omitted size group. Signifi cance level: ** = 5 percent; *** = 1 percent. 200 APPENDIX H

Appendix H Reasons for Skill Shortages in East Asia

FIGURE H.1 Reasons for skill shortages, FIGURE H.2 Reasons for skill shortages, by sector, Philippines by sector, Indonesia

emigration of too few locally skilled workers trained students

starting pay too low low quality local training

staﬀ turnover job turnover quality of education low starting wages quantity of education

emigration of skilled workers other

0510 15 20 25 30 35 40 0510 15 20 25 30 35 40 share of firms (%) share of ﬁrms (%) manufacturing services manufacturing services

Source: di Gropello, Tan, and Tandon 2010, based on 2008 Philippines Skills Source: di Gropello, Kruse, and Tandon 2011, based on 2008 Indonesia Survey. Skills Survey. REASONS FOR SKILL SHORTAGES IN EAST ASIA 201

FIGURE H.3 Three main causes of vacancies in FIGURE H.4 Three main causes of vacancies in manufacturing in Thailand identifi ed by fi rms manufacturing in Malaysia identifi ed by fi rms

applicants demand applicants demand high wage high wages

lack of suﬃcient universities don't produce graduates suﬃcient graduates

lack of required core skills applicants do not have required core skills lack of required technical skills applicants do not have required technical skills no applicants for unskilled positions

other high turnover of new staﬀ 0 102030405060 share of firms (%) other

0 5 10 15 20 25 30 share of firms (%)

Source: World Bank IC Surveys database: Thailand 2006. Source: World Bank IC Surveys database: Malaysia 2007.

FIGURE H.5 Main obstacles for recruiting, FIGURE H.6 Most important reasons for Vietnam vacancies in manufacturing in Mongolia

applicants demand lack of local high wages graduates

lack of suﬃcient graduates

workers prefer other jobs lack of required core skills

lack of low wage technical skills levels high staﬀ turnover

other other

0 0.5 1.01 1.5 2.02 2.5 3.03 3.5 4.04 4.5 0 51015 20 25 30 35 40 index share of firms (%) skilledskilled worker workersecondarysecondary technical technical collegecollege

Source: World Bank 2008, based on 2003 MOLISA-ADB survey on labor Source: World Bank IC Surveys database: Mongolia 2010. market. Note: Ranking of reasons for diffi culties. 202 APPENDIX I

Appendix I Comparative Demand- and Supply-Side Indicators

TABLE I.1 Comparative demand- and supply-side indicators, by country

Indicator Vietnam Cambodia Mongolia China Trends in tertiary education premium Increase Sharp increase Increase Increase (increase vis-à-vis secondary and below) Trends in tertiary-educated workforce Sharp increase Sharp increase Decrease Slight increase Trends in tertiary premium across No impact Increases No impact — generations

Tertiary education premium (increase vis-à-vis secondary and below), most recent available year (%) 55 90 70 60 Tertiary-educated workforce (%) 13 3.40 33 24 Weeks to ﬁ ll professional vacancies 2.5 4 4.5 (2003); About 5 (year) (2005) (2007) 6 (2009) (2003)

Trends in unemployment rate of Slight increase Decrease Slight decrease — tertiary graduates Unemployment of tertiary graduates, most recent available year (%) 1.22 2.11 6.70 6.86 Percentage employed as professionals or managers, most r ecent available year (%) 84 52 69 —

(continued next page) COMPARATIVE DEMAND- AND SUPPLY-SIDE INDICATORS 203

TABLE I.1 (continued)

Indicator Indonesia Philippines Malaysia Thailand Trends in tertiary education premium Stable/slightly Stable/slightly —Stable/slightly (increase vis-à-vis secondary and decreasing increasing increasing below) Trends in tertiary-educated workforce Sharp increase Increase — Increase Trends in tertiary premium across No impact for No impact for —Decrease generations youth youth Increase for 35–49 Increase for 35–49 Tertiary education premium (increase vis-à-vis secondary and below), most recent available year (%) 84 70 — 120 Tertiary-educated workforce (%) 21 27 — 21 Weeks to ﬁ ll professional vacancies About 2 (ICS 2003) About 3 (ICS 2003) Almost 6 (2002) About 6 (2004) (year) About 3 (SS 2008) About 5 (SS 2008) Almost 5 (2007) 7.5 (2006) Trends in unemployment rate of Slight decrease Slight increase — Slight decrease tertiary graduates Unemployment of tertiary graduates, most recent available year (%) 8.55 11.55 — 2.11 Percentage employed as professionals or managers, most r ecent available year (%) 45 59 — 68

Sources: Author elaboration on the basis of WDI database; Sakellariou 2010a, 2010b; di Gropello and Sakellariou 2010; World Bank IC Surveys database; di Gropello, Kruse, and Tandon 2011; di Gropello, Tan, and Tandon 2010. Note: — = not available; ICS = Investment Climate Survey; SS = Skill Survey. 204 APPENDIX J

Appendix J Skill Gaps in East Asia

FIGURE J.1 Skill gaps identifi ed by employers

a. Philippines: Three weakest generic b. Philippines: Three weakest job-speciﬁc skills skills for managers/professionals (and sources of job-speciﬁc skills) for managers/professionals

time management foreign degree initiative technical qualiﬁcation negotiation voc/tech qualiﬁcation leadership

problem solving experience in different field creativity experience in same field teamwork practical skills independent work

communication general experience

math theory

computer skills grades writing secondary diploma literacy

language local degree

0 246 8 10 12 14 0 5 10 15 20 25 percent percent

c. Indonesia: Share of respondents identifying a d. Indonesia: Share of respondents identifying a gap in generic skills, managers/professionals gap in behavioral skills, managers/professionals

basic skills (math, general behavioral skills literacy) ability to work independently

thinking skills risk taking and initiative

communication behavioral negotiation

computer skills team orientation

organization

English leadership

0 5 10 15 20 25 30 35 40 45 50 0510 15 20 25 30 35 40 45 percent percent

(continued next page) SKILL GAPS IN EAST ASIA 205

FIGURE J.1 (continued)

e. Indonesia: Three weakest job-specific skills (and f. Vietnam: Characteristics possessed by sources of job-specific skills) for managers/professionals recruits of different education levels

diploma job-related technical skills foreign university practical knowledge general work of the job experience grades and theoretical transcripts knowledge of the job local university good writing/ verbal skills practical knowledge work experience in foreign language/ different field communication work experience in same field punctuality primary education independence SMK

SMU experience theoretical knowledge 0 5 10 15 20 25 30 percent 0 2 4 6 8 10 12 14 16 18 20 percent technical school professional school college

g. Malaysia: Top three skills professionals and skilled h. Thailand: Top three skills professionals and skilled workers lack the most in doing their job (manufacturing) workers lack the most in doing their job (manufacturing)

English English information information technology skills technology skills technical skills technical skills communication communication skills skills numerical skills numerical skills problem solving problem solving skills skills leadership leadership

creativity creativity time time management management team work team work

social skills social skills

adaptability adaptability 024 6 8 10 12 14 0 10 20 30 40 50 60 70 percent percent

(continued next page) 206 APPENDIX J

FIGURE J.1 (continued)

i. Mongolia: Three most important skills workers lack in doing j. Cambodia: Skills and attributes that are their jobs (professionals and skilled workers) most diﬃcult to ﬁnd (professionals)

English analytical skills information technology

decision-making skills technical/professional

creativity experience

time management foreign language proﬁciency communication

work attitude leadership

social skills communication

numerical skills technical skills

adaptability education problem solving

teamwork computer skills

0 1020304050607080 0 10 20 30 40 50 60 70 percent percent management professional

Sources: di Gropello, Tan, and Tandon 2010, based on 2008 Philippines Skills Survey; di Gropello, Kruse, and Tandon 2011, based on 2008 Indonesia Skills Survey; World Bank 2008, based on 2003 Vietnam MOLISA-ADB survey on labor market; World Bank IC Surveys database: Malaysia 2007; World Bank IC Surveys database: Thailand 2004; World Bank IC Surveys database: Mongolia 2010; HRINC 2010, based on Cambodian Federation of Employers and Business Associations Youth and Employment Study (BDLINK Cambodia Co. 2008). Note: SMK = vocational secondary schools (Indonesia); SMU = general secondary schools (Indonesia). SKILL GAPS IN EAST ASIA 207

FIGURE J.2 Skill gaps according to employees

a. Indonesia: Skills most lacking among employees b. Malaysia: Skills most lacking among employees (proportion of respondents indicating particular skill (proportion of respondents indicating particular skill “one of the three most lacking”) for professionals and skilled workers “one of the three most lacking”) for professionals and skilled workers

writing English

English professional communication skills communication social skills team skills teamwork leadership leadership skills time management time management skills independence adaptability adaptability creativity/innovation skills creativity

numerical skills numerical skills

problem solving problem solving information technology computer skills technical/professional technical skills 024 6 8 10 12 14 16 18 0 10 20 30 40 50 60 70 80 percent percent

c. Thailand: Skills most lacking among employees d. Mongolia: Skills most lacking among employees (proportion of respondents indicating particular skill as (proportion of respondents indicating particular skill as “one of the three most lacking”) for professionals and skilled workers “one of the three most lacking”) for professionals and skilled workers

English English professional professional communication communication skills skills social skills social skills

teamwork teamwork

leadership skills leadership skills

time management skills time management skills

adaptability adaptability

creativity/innovation skills creativity/innovation skills

numerical skills numerical skills

problem solving problem solving information technology information technology skills skills technical/professional technical/professional skills skills 0 10 20 30 40 50 60 70 80 90100 0 10 20 30 40 50 60 percent percent

Sources: di Gropello, Kruse, and Tandon 2010, based on 2008 Indonesia Skills Survey; World Bank IC Surveys database: Malaysia 2007; World Bank IC Surveys database: Thailand 2004; World Bank IC Surveys database: Mongolia 2010. 208 APPENDIX K

Appendix K Skill Gaps across Professionals and Skilled Workers in Indonesia and the Philippines

FIGURE K.1 Key generic skill gaps (according to employers), Philippines

a. Managers b. Production/sales

time management problem solving initiative initiative negotiation leadership leadership time management problem solving negotiation creativity communication teamwork independent work independent work creativity communication language math computer computer writing writing teamwork literacy math language literacy

0 5 10 15 0 5 10 15 percent percent

Source: di Gropello, Tan, and Tandon 2010, based on 2008 Philippines Skills Survey.

FIGURE K.2 Key job-specifi c skill gaps (according to employers), Philippines

a. Managers b. Production/sales

foreign degree foreign degree technical qualification experience in different field vocational/technical qualification technical qualification experience in different field local degree experience in same field experience in same field practical grades general experience practical theory theory grades vocational/technical qualification secondary diploma general experience local degree secondary diploma

0 5 10 15 20 0 5 10 15 20 percent percent

Source: di Gropello, Tan, and Tandon 2010, based on 2008 Philippines Skills Survey. SKILL GAPS ACROSS PROFESSIONALS AND SKILLED WORKERS 209

FIGURE K.3 Share of respondents identifying a gap in generic skills, Indonesia

50 47 48 44 41 38 40 36 33 30 30

percent 20 13 14 10

0 manager/professional skilled worker basic skills (math and literacy) thinking skills behavioral skills computer skills English skills

Source: di Gropello, Kruse, and Tandon 2011, based on 2008 Indonesia Skills Survey.

FIGURE K.4 Share of respondents identifying a gap in behavioral skills, Indonesia

50 49 44 41 39 40 40 38 38 37 35 36 33 32 30 30 28 27 26

percent 20

0 manager/professional skilled worker behavioral skills (general) ability to work independently risk taking and initiative communication negotiation team orientation organization leadership

Source: di Gropello, Kruse, and Tandon 2011, based on 2008 Indonesia Skills Survey. 210 APPENDIX L

Appendix L Doctoral Degrees Earned in Science and Engineering

TABLE L.1 Science and engineering doctoral degrees earned in selected regions and locations, by fi eld (2000 or most recent year)

Math/ All doctoral Natural computer Agricultural Social/behavioral Region and location degrees All S&E sciencesa sciences sciences sciences Engineering Non-S&E All regions 207,383 114,337 46,715 7,389 7,761 20,054 32,418 93,046 Asiab 47,489 24,409 8,658 373 3,085 1,467 10,826 23,080 China (2001) 13,001 8,153 2,655 — 536 621 4,341 4,848 India (1997) 10,408 4,764 3,498 — 968 — 298 5,644 Japanc (2001) 16,078 7,401 1,586 — 1,241 610 3,964 8,677 Kyrgyz Republic 396 256 161 19 8 20 48 140 Korea, Rep. 6,143 2,865 614 247 242 108 1,654 3,278 Taiwan, China (2001) 1,463 970 144 107 90 108 521 493 Middle Eastb 5,759 2,902 1,307 241 265 495 594 2,857 Sub-Saharan Africab 2,064 679 253 0 142 143 141 1,385 Europeb 97,840 53,119 23,567 4,412 2,577 8,927 13,636 44,721 Americab 50,544 31,198 12,015 2,188 1,512 8,738 6,745 19,346 North America 46,475 28,590 10,824 2,095 1,039 8,421 6,211 17,885 South America 4,069 2,608 1,191 93 473 317 534 1,461

Sources: Chapman 2010; National Science Board, Science & Engineering Indicators 2004, http://www.nsf.gov/statistics/seind04/c2/c2s5.htm. Note: S&E = science and engineering; — = not available. Data for doctoral degrees use the International Standard Classifi cation of Education (1997) 6 level. S&E data do not include health fi elds. a. Includes physical, biological, earth, atmospheric, and ocean sciences. b. Includes only those locations for which relatively recent data are available. c. Includes thesis doctorates, called ronbun hakase, earned by employees in industry. SIMULATIONS FOR FINANCING HIGHER EDUCATION 211

Appendix M Simulations for Financing Higher Education

The simulation model presented in chapter 4 has been built so that it computes the projected intake, enrollment, and fl ow rates for an individual country’s higher education system (fi gure M.1). These estimates are based on population data and are adapted to enrollment and quality-related policy options (such as increasing the number of classrooms, construct- ing new laboratories, recruiting new instructors, providing faculty fellowships, providing bursaries to students, and so on). The ensuing number of enrollments by level and type of education, combined with the current and future modalities of resource utilization (teaching staff, equipment, infrastructure, and so on), provide estimates of future requirements of teachers, nonteaching staff, instructional materials, and educational facilities. These projected requirements, together with cost-related data and parameters, can give an estimate of fi nancial requirements and possible fi nancing gaps.

FIGURE M.1 Architecture of the simulation model

flow direction

graduates from a cycle or level of education

intake rate, macro- different + teachers American economic modalities of and other enrollment personnel; and educational American school resource ratios, and + classrooms budgetary expenditures enrollments at utilization flow rates and other indicators (recurrent different rooms; and education levels + educational investment) materials and equipment

Source: Adapted from UNESCO 2005. 212 APPENDIX M

Designed in this way, the model can provide simulations of both capital and recurrent expenditures over time. Total costs are given by the following equation t =+t t CRCId d d where t = year d = level of education t Cd = total costs t RCd = total recurrent costs, and t Id = investment.

Recurrent costs are derived by the following equations

t =+t t ++t t RCd CTd CMd CAd COd n k t = t CTd ∑∑ Tdij wdij i=1 j+1 t =×t CMd Ed CMMPSd t =×t CAd Ed CAPSd t =×t COd Ed COPSd

where t CTd = teachers’ costs t CMd = costs of materials t CAd = administrative costs t COd = other costs t Tdij = teachers by category i and level j t wdij = salaries by category and level CMPSd = per pupil cost for materials CAPSd = per pupil administrative cost, and COPSd = per pupil other costs.

And capital costs are derived by the following equation t =×⎡ t −−×t−1 ⎤ ICBPSd ⎣EaEd ()1 d ⎦

where

CBPSd = per pupil building cost, t Ed = enrollment, and a = replacement rate of buildings. References

Acworth, E. B. 2008. “University-Industry Altbach, P. G., and J. Salmi, eds. Forthcoming. Engagement: The Formation of the Knowledge The Road to Academic Excellence: The Mak- Integration Community (KIC) Model at the ing of World-Class Research Universities. Cambridge-MIT Institute.” Research Policy Washington, DC: World Bank. 37 (8): 1241–54. AmCham-China (American Chamber of Com- Aghion, P., M. Dewatripont, C. Hoxby, A. Mas- merce in the People’s Republic of China). Colell, and A. Sapir. 2007. “Why Reform 2010. “2010 Business Climate Survey Report.” Europe’s Universities?” Bruegel Policy Brief Beijing: AmCham-China. http://www.amcham 2007/04, Bruegel, Brussels. china.org/businessclimate2010. ———. 2008. Higher Aspirations: An Agenda Asia Business Council. 2009. Asia Business Coun- for Reforming European Universities. Bruegel cil Annual Survey 2009. Hong Kong: Asia Blueprint 5. Brussels: Bruegel. Business Council. ———. 2009. “The Governance and Perfor- Askling, B., M. Bauer, and S. Marton. 1999. mance of Research Universities: Evidence from “Swedish Universities towards Self-Regula- Europe and the U.S.” NBER Working Paper tion: A New Look at Institutional Autonomy.” Series 14851, National Bureau of Economic Tertiary Education and Management 5 (2): Research, Cambridge, MA. 173–93. Allen, M. 1996. “Research Productivity and Audretsch, D. B. 2008. The Entrepreneurial Soci- Positive Teaching Evaluations: Examining the ety. New York: Oxford University Press. Relationship Using Meta-Analysis.” Journal of Autor, D. H., F. Levy, and R. J. Murnane. 2001. the Association for Communication Adminis- “The Skill Content of Recent Technological tration 2: 77–96. Change: An Empirical Exploration.” NBER Almeida, R. 2009a. “Does the Workforce in East Working Paper Series 8337, National Bureau Asia Have the Right Skills? Evidence from of Economic Research, Cambridge, MA. Firm-Level Surveys.” Background paper pre- Barr, N. 2008. “Financing Higher Education.” In pared for Regional Study on Skills in East Asia, Annual World Bank Conference on Develop- World Bank, Washington, DC. ment Economics—Regional, 2008: Higher ———. 2009b. “Innovation and Openness in Education and Development, ed. Justin Yifu East Asia: Are They Increasing the Demand Lin and Boris Pleskovic, 143–74. Washington, for Educated Workers?” Background paper DC: World Bank. prepared for Regional Study on Skills in East Barro, R. J., and X. Sala-i-Martin. 1995. Eco- Asia, World Bank, Washington, DC. nomic Growth. New York: McGraw-Hill.

213 214 REFERENCES

Bastiaens, J. 2009. International Assistance and ———. 1998, 2007, 2009. Indonesia National State-University Relations (Studies in Higher Socioeconomic Survey (SUSENAS). Jakarta, Education). New York: Routledge. Indonesia. Baum, S., and K. Payea. 2005. Education Pays Bramwell, A., and D. A. Wolfe. 2008. “Univer- 2004: The Benefi ts of Higher Education for sities and Regional Economic Development: Individuals and Society. New York: College The Entrepreneurial University of Waterloo.” Board Publications. Research Policy 37 (8): 1175–87. BDLINK Cambodia Co. 2008. Youth and Brimble, P., and R. F. Doner. 2007. “University- Employment: Bridging the Gap. Youth Industry Links and Economic Development: Employment and Social Dialogue Project. The Case of Thailand.” World Development Cambodia Federation of Employers and Busi- 35 (6): 1021–36. ness Associations, Phnom Penh. Buchori, M., and A. Malik. 2004. “Higher Berdahl, R. O. 1971. Statewide Coordination of Education in Indonesia.” In Asian Universities: Higher Education. Washington, DC: Ameri- Historical Perspectives and Contemporary can Council on Education. Challenges, ed. P. Altbach and T. Umakoshi, Berman, E., J. Bound, and S. Machin. 1998. 249–78. Baltimore: Johns Hopkins University “Implications of Skill-Biased Technological Press. Change: International Evidence.” Quarterly Byun, K. 2008. “New Public Management in Journal of Economics 113 (4): 1245–80. Korean Higher Education: Is It Reality or Berry, C. R., and E. L. Glaeser. 2005. “The Diver- Another Fad?” Asia Pacifi c Education Review gence of Human Capital Levels across Cities.” 9 (2): 190–205. NBER Working Paper Series 11617, National Ca, T. N. 2006. “Universities as Drivers of the Bureau of Economic Research, Cambridge, Urban Economies in Asia: The Case of Viet- MA. nam.” Policy Research Working Paper Series Bertschy, K., M. A. Cattaneo, and S. C. Wolter. 3949, World Bank, Washington, DC. 2009. “PISA and the Transition into the Labor Cai, Y., and F. Yan. 2009. “The Responses of Market.” LABOUR: Review of Labour Eco- Private Higher Education Institutions to nomics and Industrial Relations 23 (Special Market-Oriented Environments in China—an Issue): 111–37. Institutional Approach.” Paper presented at the Bjarnason, S., K. M. Cheng, J. Fielden, M. J. 22nd CHER (Consortium of Higher Educa- Lemaitre, and D. Levy. 2009. A New Dynamic: tion Researchers) conference, Porto, Portugal, Private Higher Education. Paris: United September 10–12. Nations Educational, Scientifi c and Cultural Cao, Y. 2007. “Chinese Private Colleges and Organization (UNESCO). the Labor Market.” PhD diss., Educational Bladh, A. 2007. “Institutional Autonomy with Administration and Policy Studies, University Increasing Dependency on Outside Actors.” at Albany, State University of New York. Higher Education Policy 20: 243–59. Carney, R. W., and L. Y. Zheng. 2009. “Institu- Bloom, D., M. Hartley, and H. Rosovsky. 2004. tional Incentives to Innovate: An Explanation “Beyond Private Gain: The Public Benefi ts of for Singapore’s Innovation Gap.” Journal of Higher Education.” In International Hand- East Asian Studies (May–August). book of Higher Education, vol. 18, I, ed. CASS-IPS (Institute for Population Studies at the J.J.F. Forest and P. Altbach, 293–308. Oxford: Chinese Academy of Social Sciences). 1999, Oxford University Press. 2005. China Urban Labor Survey (CULS). Borghans, L., A. L. Duckworth, J. J. Heckman, CASS-IPS, Beijing, China. and B. ter Weel. 2008. “The Economics and Castellacci, F., and D. Archibugi. 2008. “The Psychology of Personality Traits.” Journal of Technology Clubs: The Distribution of Knowl- Human Resources 43 (4): 972–1059. edge across Nations.” Research Policy 37 (10): Boudard, E. 2001. Literacy Profi ciency, Earnings 1659–73. and Recurrent Training: A Ten Country Com- Chapman, B., and P. Drysdale. 2008. “Financing parative Study. Stockholm: Institute of Interna- Higher Education in East Asia.” East Asian tional Education, Stockholm University. Bureau of Economic Research Newsletter BPS (Badan Pusat Statistik). 1994, 1996, 1999, (August). http://www.eaber.org/intranet/pub- 2001, 2003, 2005, 2006. Indonesia National lish/Public/newsletters.php. Labor Force Survey (SAKERNAS). Jakarta, Chapman, D. W. 2010. “Higher Education Fac- Indonesia. ulty in East Asia.” Background paper prepared REFERENCES 215

for World Bank 2011, University of Minnesota. Dowrick, S. 2003. “Ideas and Education: Level or http://siteresources.worldbank.org/INTEAST Growth Effects?” NBER Working Paper Series ASIAPACIFIC/Resources/EastAsia-Higher 9709, National Bureau of Economic Research, EducationFaculty.pdf. Cambridge, MA. Chen, K., and M. Kenney. 2007. “Universities/ EdStats (Education Statistics) (database). World Research Institutes and Regional Innovation dataBank, World Bank, Washington, DC. Systems: The Cases of Beijing and Shenzhen.” http://databank.worldbank.org. World Development 35 (6): 1056–74. Estermann, T., and T. Nokkala. 2009. University Cloete, N. 2002. “South African Higher Educa- Autonomy in Europe I: Exploratory Study. tion and Social Transformation.” Higher Edu- Brussels: European University Association. cation Digest Supplement (Autumn). The Expert Group on New Skills for New Jobs. Cohn, E., S.L.W. Rhine, and M. C. Santos. 1989. 2010. “New Skills for New Jobs: Action Now.” “Institutions of Higher Education as Multi- A report prepared for the European Commis- Product Firms: Economies of Scale and Scope.” sion, European Union, Brussels. Review of Economics and Statistics 71 (2): Fajnzylber, P., and A. M. Fernandes. 2004. “Inter- 284–90. national Economic Activities and the Demand Cornford, F. M. 1908. Microcosmographia Aca- for Skilled Labor: Evidence from Brazil and demica: Being a Guide for the Young Academic China.” Policy Research Working Paper Series Politician. Cambridge: Bowes & Bowes. 3426, World Bank, Washington, DC. Darby, M. R., L. G. Zucker, and A. Wang. 2004. Farrell, D., and A. J. Grant. 2005. “China’s Loom- “Universities, Joint Ventures and Success in the ing Skilled Labor Shortage,” McKinsey Quar- Advanced Technology Program.” Contempo- terly (November). http://mkqpreview1.qdweb rary Economic Policy 22 (2): 145–61. .net/article_page.aspx?ar=1685. Deng, P. 1997. Private Education in Modern Feenstra, R. C., and G. H. Hanson. 1997. “For- China. Westport, CT: Praeger. eign Direct Investment and Relative Wages: Evi- di Gropello, E. 2006. Meeting the Challenges of dence from Mexico’s Maquiladoras.” Journal Secondary Education in Latin America and of International Economics 42 (3–4): 371–94. East Asia: Improving Efﬁ ciency and Resource Feldman, K. 1987. “Research Productivity and Mobilization. Directions in Development Scholarly Accomplishment of College Teach- Series. Washington, DC: World Bank. ers as Related to Their Instructional Effective- di Gropello, E., A. Kruse, and P. Tandon. 2011. ness: A Review and Exploration.” Research in Skills for the Labor Market in Indonesia: Higher Education 26 (3): 227–98. Trends in Skills Demand, Gaps, and Supply. Fielden, J. 2008. “Global Trends in University Directions in Development Series. Washington, Governance.” World Bank Education Paper DC: World Bank. Series, World Bank, Washington, DC. http:// di Gropello, E., and C. Sakellariou. 2010. “Indus- siteresources.worldbank.org/EDUCATION/ try and Skill Wage Premiums in East Asia.” Resources/278200-1099079877269/547664- Policy Research Working Paper Series 5379, 1099079956815/Global_Trends_University World Bank, Washington, DC. _Governance_webversion.pdf. di Gropello, E., H. Tan, and P. Tandon. 2010. Fielden, J., and N. LaRocque. 2008. “The Evolv- Skills for the Labor Market in the Philippines. ing Regulatory Context for Private Education Directions in Development Series. Washington, in Emerging Economies.” Education Working DC: World Bank. Paper Series 14, International Finance Corpo- Directorate General for Higher Education, Min- ration and World Bank, Washington, DC. istry of National Education. 2009. Directorate Fry, R. 2004. “Latino Youth Finishing College: General for Higher Education Annual Report The Role of Selective Pathways.” Pew Hispanic 2009. Indonesia. Center Report, Washington, DC. Doner, R. F., P. Intarakumnerd, and B. K. Ritchie. Geiger, R., and C. Sa. 2008 Tapping the Riches 2010. “Higher Education and Thailand’s of Science: Universities and the Promise of National Innovation System.” Background Economic Growth. Cambridge, MA: Harvard paper prepared for World Bank 2011, World University Press. Bank, Washington, DC. http://siteresources Gittleman, M., and E. N. Wolff. 1993. “Inter- .worldbank.org/INTEASTASIAPACIFIC/ national Comparisons of Inter-industry Wage Resources/Thailand-HENationalInnovation Differentials.” Review of Income and Wealth System.pdf 39 (3): 295–312. 216 REFERENCES

Glaeser, E. L. 2007. “Entrepreneurship and the Hattie, J., and H. W. Marsh. 1996. “The Relation- City.” NBER Working Papers Series 13551, ship between Research and Teaching: A Meta- National Bureau of Economic Research, Cam- Analysis. Review of Educational Research bridge, MA. 66 (4): 507–42. Goldberg, P. K., and N. Pavcnik. 2005. “Trade, Heckman, J. J., J. Stixrud, and S. Urzua. 2006. Wages, and the Political Economy of Trade “The Effects of Cognitive and Noncognitive Protection: Evidence from the Colombian Abilities on Labor Market Outcomes and Trade Reforms.” Journal of International Eco- Social Behavior.” Journal of Labor Economics nomics 66 (1): 75–105. 24 (3): 411–82. Gonzales, P., J. C. Guzman, L. Partelow, Hicks, M., and George, R. 1998. “Approaches E. Pahlke, L. Jocelyn, D. Kastberg, and T. Wil- to Improving Student Learning at the Univer- liams. 2004. Highlights from the Trends in sity of South Australia.” Presentation at the International Mathematics and Science Study HERDSA annual conference “Transformation (TIMSS) 2003. Washington, DC: National in Higher Education,” Auckland, New Zea- Center for Education Statistics. land, July 7–10. Gonzales, P., T. Williams, L. Jocelyn, S. Roey, Hill, H., and P. Tandon. 2010. “Innovation and D. Kastberg, and S. Brenwald. 2008. High- Technological Capability in Indonesia.” Back- lights from TIMSS 2007: Mathematics and ground paper prepared for World Bank 2011, Science Achievement of U.S. Fourth- and World Bank, Washington, DC. Eighth-Grade Students in an International Hoachlander, G., A. C. Sikora, and L. Horn. Context. Washington, DC: National Center 2003. Community College Students: Goals, for Education Statistics. Academic Preparation, and Outcomes. NCES Government of the Philippines. 2007. Comments 2003-164, National Center for Education Sta- of the National Economic and Development tistics. Washington, DC: U.S. Department of Authority on the Joint Resolution Creating Education. a Congressional Commission to Review and HRINC. 2010. “Higher Education and Skills for Assess the State of Competitiveness of Science the Labor Market in Cambodia.” Background and Technology. Manila: Government of the paper prepared for World Bank 2011, HRINC, Philippines. Phnom Penh, Cambodia. GSO (General Statistics Offi ce) [Vietnam]. 1992, Huisman, J., and J. Currie. 2004. “Accountabil- 1998, 2004. Vietnam Living Standards Survey ity in Higher Education: Bridge over Troubled (VLSS). Hanoi, Vietnam: General Statistics Water?” Higher Education 48 (4): 529–51. Offi ce IMF (International Monetary Fund). 2006. ———. 2006. Result of the Viet Nam House- World Economic Outlook 2006: Financial hold Living Standards Survey [VHLSS] 2006. Systems and Economic Cycles. Washington, Hanoi, Vietnam: General Statistics Offi ce. DC: IMF. ———. 2008. Result of the Survey on Household Information Office of the State Council of the Living Standards [VHLSS] 2008. Hanoi, Viet- People’s Republic of China. 2005. “Regional nam: General Statistics Offi ce. Autonomy for Ethnic Minorities in China.” Guan, J. C., R. C. Yam, and C. K. Mok. 2005. White Paper on Ethnic Minorities, Informa- “Collaboration between Industry and Research tion Offi ce of the State Council of the People’s Institutes/Universities on Industrial Innovation Republic of China, Beijing. http://www.china in Beijing.” Technology Analysis and Strategic .org.cn/e-white/20050301/index.htm. Management 17 (3): 339–53. Ingles, S. J., J. A. Owings, P. Kaufman, M. N. Hansen, M. T., H. W. Chesbrough, N. Nohria, Alt, and X. Chen. 2002. Coming of Age in the and D. N. Sull. 2000. “Networked Incubators: 1990s: The Eighth-Grade Class of 1988 12 Hothouses of the New Economy.” Harvard Years Later. Washington, DC: U.S. Department Business Review 78 (5): 74–84. of Education, Institute of Education Sciences, Hanushek, E. A., and L. Wößmann. 2007. Edu- National Center for Education Statistics. cation Quality and Economic Growth. Wash- Jakubowski, M., H. A. Patrinos, E. Porta, and J. ington, DC: World Bank. Wisniewski. 2010. “The Impact of the 1999 Hashimoto, K., and E. Cohn. 1997. “Economies Education Reform in Poland.” Policy Research of Scale and Scope in Japanese Private Univer- Working Paper Series 5263, World Bank, sities.” Education Economics 5 (2): 107–15. Washington, DC. REFERENCES 217

Jenkins, A., T. Vlackman, R. Lindsay, and Western Cape Province, South Africa.” World R. Paton-Salzburg. 1998. “Teaching and Bank, Washington, DC. Research: Student Perspectives and Policy Lai, M., and L.N.K. Lo. 2007. “The Changing Implications.” Studies in Higher Education Work Lives of Academics: The Experience of a 23 (2): 127–41. Regional University in the Chinese Mainland.” Johanson, R. K., and A. V. Adams. 2004. Skills Higher Education Policy 20 (20): 145–67. Development in Sub-Saharan Africa. Wash- Lane, K., and F. Pollner. 2008. “How to Address ington, DC: World Bank. China’s Growing Talent Shortage,” McKinsey Johnson, G. 1994 University Politics: F.M. Corn- Quarterly (July). http://www.mckinseyquar- ford’s Cambridge and His Advice to the Young terly.com/How_to_address_Chinas_growing_ Academic Politician. Cambridge: Cambridge talent_shortage_2156. University Press. LaRocque, N. 2002. Private Education in the Johnstone, B. D. 2004. “The Economics and Philippines: A Market and Regulatory Survey. Politics of Cost Sharing in Higher Education: Manila: Asian Development Bank. Comparative Perspectives.” Economics of Laursen, K., and A. Salter. 2004. “Searching Education Review 23 (4): 403–10. High and Low: What Types of Firms Use Uni- KAM (Knowledge Assessment Methodology) versities as a Source of Innovation?” Research (data tool). World Bank, Washington, DC. Policy 33: 1201–15. https://www.worldbank.org/kam. Lederman, D., and W. F. Maloney. 2003. Kapur, D., and M. Crowley. 2008. “Beyond the “R&D and Development.” Policy Research ABCs: Higher Education and Developing Working Paper 3024, World Bank, Washing- Countries.” Working Paper Series 139, Center ton, DC. for Global Development, Washington, DC. Lee, M. 1999. “Corporatization, Privatization, Kennedy, K. J., and Z. Lee. 2008. The Changing and Internationalization of Higher Education Role of Schools in Asian Societies: Schools for in Malaysia.” In Private Prometheus: Private the Knowledge Society. London: Routledge. Higher Education and Development in the Kim, A., and Y. Lee. 2008. Student Loan Schemes 21st Century, ed. P. Altbach, 137–61. West- in the Republic of Korea: Review and Recom- port, CT: Greenwood Press. mendations. Bangkok, Thailand: UNESCO. ———. 2004. “Malaysian Universities: Toward Kim, S., and J.-H. Lee. 2006. “Changing Facets Equality, Accessibility, and Quality.” In Asian of Korean Higher Education: Market Compe- Universities: Historical Perspectives and Con- tition and the Role of the State.” Higher Edu- temporary Challenges, ed. P. Altbach and cation 52 (3): 557–87. T. Umakoshi, 221–46. Baltimore, MD: Johns Kirsch, I. S., A. Jungeblut, L. Jenkins, and A. Kol- Hopkins University Press. stad. 1993. Adult Literacy in America: A First Lee, S. H. 1998. “Korean Private Higher Educa- Look at the Results of the National Adult Lit- tion Faces Economic Crisis.” International eracy Survey. Princeton, NJ: Educational Test- Higher Education 13: 9–20. ing Service. Levy, D. 2008a. “Exploring the Viability of a Kodama, F., and J. Suzuki. 2007. “How Japanese Semi-Elite Category.” Paper presented at the Companies Have Used Scientific Advances 33rd Annual Conference of the Association to Restructure Their Business: The Receiver- for the Study of Higher Education (ASHE), Active National System of Innovation.” World International Division, Jacksonville, Florida, Development 35 (6): 976–90. November 5–8. Kodama, T. 2008. “The Role of Intermedia- ———. 20 08b. “Private H igher E ducation’s tion and Absorptive Capacity in Facilitating Global Surge: Emulating U.S. Patterns?” University-Industry Linkages—an Empirical Paper presented at Privatization in Higher Study of TAMA in Japan.” Research Policy Education Conference, Samuel Nieman 37 (8): 1224–40. Institute for Advanced Studies in Science Koshal, R. K., and M. Koshal. 1999. “Econo- and Technology, Technion, Haifa, Israel, mies of Scale and Scope in Higher Education: January 6–8. A Case of Comprehensive Universities.” Eco- ———. 2010. “East Asian Private Higher Educa- nomics of Education Review 18 (2): 269–77. tion: Reality and Policy.” Background paper Kruss, G., and J. Lorentzen. 2007. “University- prepared for World Bank 2011, University of Industry Links for Development: The Case of Albany, State University of New York. 218 REFERENCES

Li, C., ed. 2005. Bridging Minds across the Michaelowa, K. 2007. “The Impact of Primary Paciﬁ c: U.S.-China Educational Exchanges, and Secondary Education on Higher Educa- 1978–2003. New York: Lexington Books. tion Quality.” Quality Assurance in Educa- Lin, J. 1999. Social Transformation and Private tion 15 (2): 215–36. Education in China. Westport, CT: Praeger. Miner, A. S., D. T. Eesley, M. Devaughn, and Lin, T.-C. 2004. “The Role of Higher Education T. Rura-Polley. 2001. “The Magic Beanstalk in Economic Development: An Empirical Study Vision: Commercializing University Inven- of Taiwan Case.” Journal of Asian Economics tions and Research.” In The Entrepreneurship 15 (2): 355–71. Dynamic: Origins of Entrepreneurship and the Lin, Y., Y. Xia, L. Feixuai, L. Jing, H. Weilie, and Evolution of Industries, ed. C.B. Schoonhoven W. Junyong. 2005. “The Beida Reform Trig- and E. Romanelli, 109–46. Palo Alto, CA: gers Chain Reactions.” Chinese Education Stanford University Press. and Society 38 (1): 49–61. Mingat, A., B. Ledoux, and R. Rakotomalala. Linh, V. H., L. V. Thuy, and G. T. Long. 2010. 2010. Developing Post-Primary Education in “Equity and Access to Tertiary Education: The Sub-Saharan Africa: Assessing the Financial Case of Vietnam.” Background paper prepared Sustainability of Alternative Pathways. Africa for World Bank 2011, Indochina Research & Human Development Series. World Bank: Consulting (IRC) and National Economics Washington, DC. University, Hanoi. Ministry of Education, Singapore. 2005. “Edu- Loyalka, P. K. 2009. “Is Aid Reaching Poor Stu- cation in Singapore: Corporate Brochure.” dents? The Distribution of Student Financial Singapore. Aid across Chinese Higher Education System.” Ministry of Education and Human Resources Paper presented at the 2009 International Development, Republic of Korea. 2010. “Brain Academic Symposium on Higher Education Korea 21, a Project for Nurturing Highly Finance, Beijing, November 7. Qualiﬁ ed Human Resources for the 21st Cen- Luan, Z. 2007. “Exploration of the Problem of tury Knowledge-Based Society.” Ministry Equal Access to Higher Education in Shang- of Education, Seoul. http://unpan1.un.org/ dong Province.” Unpublished master’s thesis, intradoc/groups/public/documents/apcity/ Shangdong Normal University, China. unpan015416.pdf. Lundvall, B.-Å. 2007. “National Innovation Sys- Mitchell, J. E., and D. S. Rebne. 1995. “Nonlinear tems: Analytical Concept and Development Effects of Teaching and Consulting on Aca- Tool.” Industry and Innovation 14 (1): 95–119. demic Research Productivity.” Socio-Economic Ma, W. 2010. “Equity and Access to Tertiary Planning Sciences 29 (1): 47–57. Education: Case Study—China.” Background Moeliodihardjo, B. Y. 2010. “Equity and Access paper prepared for World Bank 2011. http://site in Higher Education: The Case of Indonesia.” resources.worldbank.org/EASTASIAPACIFIC Background paper prepared for World Bank EXT/Resources/226300-1279680449418/ 2011, Fakultas Ilmu Komputer Universitas HigherEd_ChinaEquityStudy.pdf. Indonesia, Indonesia. Macerinskiene, I., and Vaiksnoraite, B. 2006. MOET (Ministry of Education and Training), “The Role of Higher Education to Economic Vietnam. 2005, 2010. University Survey. Development.” Vadyba (Management) 2 (11): Hanoi, Vietnam. 82–90. Mok, K. H. 2009. “The Quest for Regional Hub Martinez, M., and S. Klopott. 2005. The Link of Education: Searching for New Governance between High School Reform and College and Regulatory Regimes in Singapore, Hong Access and Success for Low-Income and Kong and Malaysia.” Paper prepared for the Minority Youth. Washington, DC: American East-West Senior Seminar on Quality Issues in Youth Policy Forum and Pathways to College the Emerging Knowledge Society. Kuala Lum- Network. pur, Malaysia, October. Marton, S. 2000. The Mind of the State: The ———. 2010. “Innovation and Higher Education: Politics of University Autonomy in Sweden, A Comparative Study of Five Asian Societies.” 1968–1998. Göteborg: BAS Publishers. Background paper prepared for World Bank Mei, T. A. 2002. Malaysian Private Higher Edu- 2011, Hong Kong Institute of Education, Hong cation: Globalization, Privatization, Transfor- Kong. mation and Marketplaces. London: ASEAN Moon, M., and K. Kim. 2001. “A Case of Korean Academic Press. Higher Education Reform: The Brain Korea 21 REFERENCES 219

Project.” Asia Pacifi c Education Review 2 (2): Internationalisation of Higher Education.” 96–105. OECD Policy Brief (August), OECD, Paris. Motohashi, K. 2005. “University-Industry Col- ———. 2005. Learning a Living: First Results laborations in Japan: The Role of New Tech- of the Adult Literacy and Life Skills Survey. nology-Based Firms in Transforming the Paris: OECD. National Innovation System.” Research Policy ———. 20 07a. OECD Education at a Glance 34 (5): 583–94. 2007. Paris: OECD. Mukherjee, H. 2010. “Access to and Equity in ———. 20 07 b. Program for International Student Higher Education: Malaysia.” Background Assessment 2006. Paris: OECD. paper prepared for World Bank 2011, World ———. 20 08a. The Internationalisation of Busi- Bank, Washington, DC. ness R&D: Evidence, Impact and Implica- Mullis, I., M. Martin, E. Gonzalez, K. Gregory, tions. Paris: OECD. R. Garden, K. O’Connor, S. Chrostowski, and ———. 20 08b. Tertiary Education for the Knowl- T. Smith. 2000. TIMSS 1999 International edge Economy. Paris: OECD. Mathematics Report: Findings from IEA’s ———. 20 09a. OECD Reviews of Tertiary Edu- Repeat of the Third International Mathemat- cation: China 2009. Paris: OECD. ics and Science Study at the Eighth Grade. ———. 20 09b. OECD Reviews of Tertiary Edu- Chestnut Hill, MA: Boston College. cation: Korea 2009. Paris: OECD. National Science Board. 2008. Science and ———. 20 09c. OECD Reviews of Tertiary Edu- Engineering Indicators 2008. Arlington, VA: cation: Japan. Paris: OECD. National Science Foundation. ———. 2010a. PISA 2009 Results: Overcom- National Statistical Office of Mongolia. 1998, ing Social Background: Equity in Learning 2002, 2007, 2010. Living Standards Measure- Opportunities and Outcomes. Volume II. ment Survey (LSMS). Ulaanbaatar, Mongolia: Paris: OECD. National Statistical Office. http://www.nso ———. 2010b. PISA 2009 Results: What Students .mn. Know and Can Do: Student Performance in Neave, G., and F. A. van Vught, eds. 1994. Gov- Reading, Mathematics and Science. Volume I. ernment and Higher Education Relation- Paris: OECD. ships across Three Continents: The Winds of Okada, A. 2005. “A History of the Japanese Change. Oxford: Pergamon. University.” In The “Big Bang” in Japanese Nelson, R. 1987. Understanding Technical Higher Education: The 2004 Reforms and the Change as an Evolutionary Process. Amster- Dynamics of Change, ed. J. G. Eades, R. Good- dam: Elsevier Science. man, and Y. Hada, 32–51. Melbourne: Trans Newman, J. H. 1976 [1st ed. 1852]. The Idea of a Pacifi c Press. University. Oxford: Clarendon Press. Orbeta, A. C., Jr. 2008. “Higher Education in the NIS (National Institute of Statistics). 1997, 1999, Philippines.” World Bank, Washington, DC. 2003, 2007, 2009. Cambodia Socio-Economic Background paper prepared for di Gropello, Survey (CSES). National Institute of Statistics. Tan, and Tandon. 2010, World Bank, Wash- Phnom Penh, Cambodia. ington, DC. NSO Philippines (National Statistics Office, Patton, D., L. Warren, and D. Bream. 2009. “Ele- Republic of the Philippines). 2000a, 2006a. ments That Underpin High-Tech Business Family Income and Expenditure Survey Incubation Processes.” Journal of Technology (FIES). Manila, Philippines: National Statis- Transfer 34 (6): 621–36. tics Offi ce. Permani, R. 2009. “The Role of Education in ———. 1988, 1991, 1994, 1997, 2000b, 2001, Economic Growth in East Asia: A Survey.” 2004, 2006b. Labor Force Survey (LFS). Asian-Pacific Economic Literature 23 (1): Manila, Philippines: National Statistics 1–20. Offi ce. Ping Chung, Y. 2007. The Student Loans Scheme NSO Thailand (National Statistical Offi ce Thai- in Hong Kong. Bangkok, Thailand: UNESCO land). 1990, 1992, 1994, 1996, 1998, 2000, Bangkok; Paris: International Institute of 2002, 2004, 2006. Thailand Socio-Economic Educational Planning. http://ddp-ext.world Survey (SES). Bangkok, Thailand: National bank.org/EdStats/CHNpub03.pdf. Statistical Offi ce. Postiglione, G. 2006. “Higher Education in China: OECD (Organisation for Economic Co-op- Perils and Promises for a New Century.” Har- eration and Development). 2004. “The vard China Review 5 (2): 138–43. 220 REFERENCES

Postiglione, G., and G. Mak, eds. 1997. Asian paper prepared for World Bank 2011, World Higher Education: An International Hand- Bank, Washington, DC. book and Reference Guide. Westport, CT: Salmi, J. 2007. “Autonomy from the State vs. Greenwood Press. Responsiveness to Markets.” Higher Educa- Potter, J. 2008. “Entrepreneurship and Higher tion Policy 20 (3): 223–42. Education: Future Policy Directions.” In ———. 2009. The Challenge of Establishing Entrepreneurship and Higher Education, ed. World-Class Universities. Washington, DC: J. Potter, chapter 14. Paris: OECD Local Eco- World Bank. nomic and Employment Programme. Salmi, J., and A. M. Hauptman. 2006. “Inno- Praphamontripong, P. 2010. “Intra-Sectoral vations in Tertiary Education Financing: A Diversity: A Political Economy of Thai Private Comparative Evaluation of Allocation Mecha- Higher Education.” PhD diss., Educational nisms.” Education Working Paper Series 4, Administration and Policy Studies, University World Bank, Washington, DC. at Albany, State University of New York. Salmi, J., and A. Saroyan. 2007. “League Tables PROPHE (Program for Research on Private as Policy Instruments: Uses and Misuses.” Higher Education) International Databases. Higher Education Management and Policy 19 Private and Public Higher Education Shares. (2): 24–62. University of Albany, State University of Schiller, D., and P. Brimble. 2009. “Capacity New York. http://www.albany.edu/dept/eaps/ Building for University-Industry Linkages in prophe/data/international.html. Accessed Developing Countries: The Case of the Thai October 2010. Higher Education Development Project.” Sci- Przirembel, C. 2005. “Transplanting Industries: ence, Technology and Society 14 (1): 59–92. The ICAR Project: Embedding the Automotive Schott, P. K. 2006. “The Relative Sophistica- Industry in South Carolina.” Paper presented tion of Chinese Exports.” NBER Working at MIT’s “1st International Conference on Paper 12173, National Bureau of Economic Local Innovation Systems,” Cambridge, MA, Research, Cambridge, MA. December 13. Scotchmer, S. 2004. Innovation and Incentives. Raza, R. 2010. “Higher Education Governance Cambridge, MA: MIT Press. in East Asia.” Background paper prepared Senker, J. 1995. “Tacit Knowledge and Models of for World Bank 2011, World Bank, Washing- Innovation.” Industrial and Corporate Change ton, DC. http://siteresources.worldbank.org/ 4 (2): 425–47. INTEASTASIAPACIFIC/Resources/Higher- Sianesi, B., and J. van Reenen. 2003. “The Returns EducationGovernance.pdf. to Education: Macroeconomics.” Journal of Rosenzweig, M. 2009. “Higher Education and Economic Surveys 17: 157–200. International Migration in Asia: Brain Circu- SJTU (Shanghai Jiao Tong University) Academic lation.” In Annual World Bank Conference Ranking of World Universities (ARWU). http:// on Development Economics 2008, Regional: www.arwu.org/. Higher Education and Development, ed. J. Y. Smeaton, B., and M. Hughes. 2003. “A Basis Lin and B. Pleskovic, 59–84. Washington, DC: for Skills: Investigating Intermediate Skills.” World Bank. Learning Skills and Development Agency, Round, A. 2003. “Higher Education and ‘Skills’ London. in ICT Disciplines: A Partial Review of the Lit- Smeby, J.-C. 1998. “Knowledge Production and erature.” Produced for the Council of Profes- Knowledge Transmission: The Interaction sors and Heads of Computing (CPHC), Swin- between Research and Teaching at Universi- don, United Kingdom. ties.” Teaching in Higher Education 3 (1): Rumbley, L., I. Pacheco, and P. Altbach. 2008. 7–20. International Comparison of Academic Sala- Sohn, D.-W., and M. Kenney. 2007. “Universities, ries. Boston: Boston College, Center for Inter- Clusters, and Innovation Systems: The Case national Higher Education. of Seoul, Korea.” World Development 35 (6): Sakellariou, C. 2010a. “Access to and Equity of 991–1004. Higher Education in East Asia.” Background Sorensen, C. W. 1994. “Success and Education paper prepared for World Bank 2011, World in South Korea.” Comparative Education Bank, Washington, DC. 38: 10–35. ———. 2010b. “Labor Market Outcomes of Steiner-Khamasi, G., and I. Stolpe. 2004. “Decen- Higher Education in East Asia.” Background tralization and Recentralization Reform in REFERENCES 221

Mongolia: Tracing the Swing of the Pendu- Quality of Cross-Border Education. Bangkok: lum.” Comparative Education 40: 29–53. UNESCO Asia and Pacific Regional Bureau Sukamoto. 2002. “Private Higher Education for Education. in Indonesia.” In The Report of the Second UIS (UNESCO Institute for Statistics) Data Regional Seminar on Private Higher Education: Centre. Montreal, Canada. http://www.uis Its Role in Human Resource Development in a .unesco.org/Pages/default.aspx. Globalised Knowledge Society, ed. UNESCO UIS. 2009. Global Education Digest 2009: Com- PROAP and SEAMEO RIHED, 39–46. Bang- paring Education Statistics across the World. kok, Thailand: UNESCO-RIHED. Montreal, Canada: UIS. Sutton Trust. 2004. “Select Government Fund Usher, A. 2005. A Little Knowledge Is a Dan- Matching Programmes: An Examination of gerous Thing: How Perceptions of Costs and Characteristics and Effectiveness.” CASE Benefi ts Affect Access to Education. Toronto, Europe, London. ON: Educational Policy Institute. Tan, E. A. 2010. “The State of the Philippines’ Valisno, M. D. 2002. “Private Higher Education National Innovation System.” Background in Philippines.” In The Report of the Second paper prepared for World Bank 2011, Uni- Regional Seminar on Private Higher Education: versity of the Philippines, Manila. http://site Its Role in Human Resource Development in a resources.worldbank.org/INTEASTASIA Globalised Knowledge Society, ed. UNESCO PACIFIC/Resources/philippine-national- PROAP and SEAMEO RIHED, 75–118. Bang- innovation-system.pdf. kok, Thailand: UNESCO-RIHED. Tangkitvanich, S., and A. Manasboonphempool. van der Sluis, J., M. van Praag, and W. Vijverberg. 2010. “Evaluating the Student Loan Fund of 2008. “Education and Entrepreneurship Selec- Thailand.” Economics of Education Review tion and Performance: A Review of the Empiri- 29: 710–21. cal Literature.” Journal of Economic Surveys Tansinsin, L. 2007. “Filipino Experience in Fos- 22 (5): 795–841. tering University-Industry Partnership.” World Vandenbussche, J., P. Aghion, and C. Meghir. Intellectual Property Organization, Geneva, 2006. “Growth, Distance to Frontier and Switzerland. Composition of Human Capital.” Journal of Tertiary Education Commission of New Zealand. Economic Growth 11 (2): 97–127. 2009. Tertiary Education Commission State- VCCI (Vietnam Chamber of Commerce and ment of Intent 2009–2012. New Zealand: Industry). 2009. Vietnam Business Annual Tertiary Education Commission. Report 2008. Hanoi, Vietnam: National Polit- THE (Times Higher Education) World University ical Publishing House. Rankings. http://www.timeshighereducation Wadhwa, V., G. Gereffi , B. Rising, and R. Ong. .co.uk/. 2007. “Where the Engineers Are.” Issues in TIMSS (Trends in International Mathematics Science and Technology Spring: 73–84. and Science Study) (database). U.S. Depar- WIPO (World Intellecutal Property Organiza- ment of Education, Institute of Education tion). 2009. World Intellectual Property Indi- Sciences (IES), National Center for Education cators 2009. Geneva: WIPO. Statistics, Washington, DC. http://nces.ed World Bank. 1997. China: Higher Education .gov/timss/. Reform. World Bank: Washington, DC. Tybout, J. 2000. “Manufacturing Firms in Devel- ———. 2003. World Development Report 2004: oping Countries: How Well Do They Do and Making Services Work for Poor People. Wash- Why?” Journal of Economic Literature 38: ington, DC: World Bank and Oxford Univer- 11–44. sity Press. UNESCO (United Nations Educational, Scientifi c ———. 2005. “Project Appraisal Document: and Cultural Organization). 2005. Education Managing Higher Education for Relevance and Policy and Strategy Simulation Model EPSSim Effi ciency Project.” Report Series 31644-ID. Version 2.1. Paris: UNESCO. Washington, DC: World Bank. ———. 2006. Higher Education in South-East ———. 2007. Malaysia and the Knowledge Econ- Asia. Bangkok: UNESCO Asia and Pacific omy: Building a World-Class Higher Educa- Regional Bureau for Education. tion System. Washington, DC: World Bank. UNESCO (United Nations Educational, Scien- ———. 2008. Vietnam: Higher Education and tifi c and Cultural Organization) Asia Pacifi c Skills for Growth. Washington, DC: World Quality Network. 2006. Regulating the Bank. 222 REFERENCES

———. 20 09a. Thailand Social Monitor: Towards Institutional Perspective.” European Review a Competitive Higher Education System in a 18 (supplement 1): S117–39. Global Economy. Bangkok, Thailand: World Yilmaz, Y. 2009. “Skills Development in East Bank. Asia: A Comparative Study of Supply of Educa- ———. 20 09b. Vietnam Knowledge Report tion and Training Systems.” Background paper Revised. Washington, DC: World Bank. prepared for Regional Study on Skills in East ———. 2010a. “Assessing the Quality of Educa- Asia, World Bank, Washington, DC. tion in Shanghai, China, Using PISA 2009.” ———. 2010. “Higher Education Institutions in World Bank, Washington, DC. Thailand and Malaysia: Can They Deliver?” ———. 2010b. “Emerging Stronger from the Cri- Background paper prepared for World Bank sis.” East Asia and Pacifi c Economic Update, 2011, World Bank, Washington, DC. Volume 1. Washington, DC: World Bank. Yonezawa, A., and M. Baba. 1998. “The Market ———. 2010c. “Indonesia: Higher Education Structure for Private Universities in Japan.” Financing.” World Bank, Washington, DC. Tertiary Education and Management 4 (2): ———. 2010d. “Mongolia Policy Note. Tertiary 145–52. Education in Mongolia: Meeting the Chal- Yusuf, S. 2003. Innovative East Asia: the Future lenges of the Global Economy.” Report No. of Growth. Washington, DC: World Bank. 52925–MN. World Bank, Washington, DC. Yusuf, S., and K. Nabeshima. 2009. Tiger Econo- ———. 2010e. “Program Appraisal Document mies under Threat: Comparative Analysis of for a Higher Education Quality and Capacity Malaysia’s Industrial Prospects and Policy Improvement Project in Cambodia.” Washing- Options. Washington, DC: World Bank. ton, DC: World Bank. ———. 2010. “From Technological Mastery to ———. 2010f. “Program Document for a Higher Innovation.” Background paper prepared for Education Development Policy Credit to Viet- World Bank 2011, World Bank, Washington, nam.” Washington, DC: World Bank. DC. ———. 2010g. “Vietnam High Quality Education Zaman, M. Q. 2004. Review of the Academic for All by 2020.” World Bank, Washington, Evidence on the Relationship between Teach- DC. ing and Research in Higher Education. Lon- ———. 2011. Putting Higher Education to Work: don: Department for Education and Skills Skills and Research for Growth in East Asia. Publications. Washington, DC: World Bank. Zhou, Y. 2008. The Inside Story of China’s World Bank IC (Investment Climate) Surveys High-Tech Industry: Making Silicon Valley (database). World Bank, Washington, DC. in Beijing. New York: Rowman & Littlefi eld https://www.enterprisesurveys.org. Publishers. World Development Indicators (WDI) database. Ziderman, A. 2003. Student Loans in Thailand: World dataBank, World Bank, Washington, Are They Effective, Equitable, Sustainable? DC. http://databank.worldbank.org. Policy Research and Dialogue Student Loans World Integrated Trade Solution (WITS) (data Schemes in Asia Series, vol. 1, no. 1. Bangkok: tool). World Bank, Washington, DC. https:// United Nations Educational, Scientific and wits.worldbank.org Cultural Organization and International Insti- Wright, M., B. Clarysse, A. Lockett, and M. tute for Educational Planning. Knockaert. 2008. “Mid-Range Universities’ ———. 2004. Policy Options for Student Loan Links with Industry: Knowledge Types and Schemes: Lessons from Five Asian Case Stud- the Role of Intermediaries.” Research Policy ies. Policy Research and Dialogue Student 37 (8): 1205–23. Loans Schemes in Asia Series, vol. 1, no. 6. Wu, W. 2010. “Higher Education Innovation in Bangkok: United Nations Educational, Scien- China.” Background paper prepared for World tifi c and Cultural Organization and Interna- Bank 2011, Virginia Commonwealth Univer- tional Institute for Educational Planning. sity, Richmond, VA. ———. 2006. “Student Loans in Thailand: From Yan, F. 2010. “The Academic Profession in Social Targeting to Cost Sharing.” Interna- China in the Context of Social Transition: An tional Higher Education 42: 6–8.

ECO-AUDIT Environmental Benefits Statement

The World Bank is committed to preserving Saved: endangered forests and natural resources. The • 15 trees Office of the Publisher has chosen to print • 10 million BTUs of Putting Higher Education to Work: Skills total energy and Research for Growth in East Asia on • 1,576 pounds of net recycled paper with 50 percent postconsumer greenhouse gases fiber in accordance with the recommended • 5,245 gallons of standards for paper usage set by the Green waste water Press Initiative, a nonprofit program support- • 695 pounds of ing publishers in using fiber that is not sourced solid waste from endangered forests. For more information, visit www.greenpressinitiative.org.

mpressive gains have been made in expanding access to higher education in East Asia over the past few decades, and the signiﬁ cance of higher Ieducation is expected to increase as developing economies face the challenge of sustaining growth in a competitive environment. Still, much work needs to be done if higher education is to realize its full potential to produce the skills and research needed for innovation and growth.

“Disconnects” between higher education institutions and the skills and research users and providers they interact with—fi rms, research institutes, earlier education institutions, and other skill providers—have undermined the functioning of the subsector. In order for higher education institutions to be more responsive to the labor market, they should be better aligned with what employers and employees need and better connected among themselves and other skills providers. They also require stronger linkages with fi rms and other research providers to deliver research that can enhance innovation and productivity. Enacting policy reforms in the areas of fi nancing, public sector management, and stewardship of the entire system will help the region achieve better skills and research outcomes and, ultimately, growth.

Putting Higher Education to Work: Skills and Research for Growth in East Asia takes an in-depth and comprehensive look at higher education in East Asia—how it has changed, how it will continue to evolve, and how it can be improved. Using innovative ﬁ rm surveys and the latest available evidence from the region, the authors identify functional skills needed for competitiveness and productivity, and look at how higher education systems can produce the type of skills and kind of commercially applicable research that will drive growth. Though focused on the developing countries in East Asia, the book’s methodologies, messages, and careful guidance will be of interest and value to researchers and policy makers in all regions where higher education and training are important issues.

This book is the ﬁ rst title in the East Asia and Paciﬁ c Regional Report Series, a series that presents analyses of issues relevant to the region, drawing on the global knowledge and experience of the World Bank and its partners. This series aims to inform public discussion, policy formulation, and development practitioners’ actions, and thereby turn challenges into opportunities.

ISBN 978-0-8213-8490-9

SKU 18490