Professional Documents
Culture Documents
Education to Work
Skills and Research for Growth in East Asia
Putting Higher
Education to Work
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
reflect the views of the Executive Directors of The World Bank or the governments they represent.
The World Bank does not guarantee the accuracy of the data included in this work. The boundaries,
colors, denominations, and other information shown on any map in this work do not imply any judge-
ment on the part of The World Bank concerning the legal status of any territory or the endorsement or
acceptance of such boundaries.
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
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-Specific 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 benefits 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 diversification 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 efficient 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 qualifications frameworks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
6.1 Expanding private higher education in the Republic of Korea . . . . . . . . . . . . . . . . . . 155
6.2 Resource diversification in China and Malaysia . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
6.3 Technology licensing offices 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 proficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 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 firms. . . 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 scientific 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 qualifications . . . . . . . . . . . . . . . . . . . 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-specific skill gaps in Vietnam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.14 Key job-specific skill gaps in Malaysia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.15 Average monthly salaries of Thai employees reporting a particular skill
as a top-three deficiency compared with employees not reporting skills
as a top-three deficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 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 Scientific 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 field 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 scientific and technical journal articles and number of researchers
in R&D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
CONTENTS ix
6.7 Asian students studying abroad in tertiary schools in the top-five 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-specific 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 identified
by fi rms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
H.4 Three main causes of vacancies in manufacturing in Malaysia identified
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 identified 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-specific 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 coefficients 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 coefficients 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
office, 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 field, 2006 or most recent year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.2 Student-to-faculty ratios in tertiary education, 2007 . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.3 Academic qualifications 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
E
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
agenda for policy makers across the the economy as a whole. Using innovative
globe. For countries seeking to spur growth, firm 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 firms’ 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 confident that this increasingly competitive global environment.
volume, the first in the East Asia and Pacific
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 Pacific Region
The measures proposed in this volume should World Bank
Acknowledgments
T
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-
guidance of Emmanuel Jimenez and sion), Annie Koh (Dean, Office 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 benefited 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.
xv
Abbreviations
xvii
xviii ABBREVIATIONS
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)
Fiscal year
January 1–December 31
Summary
1
2 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 bet ween 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 firms. 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 diversified 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 firms.
and policy failures. Public policy has the
potential to address them (although not all
Five disconnects the policies are strictly related to higher edu-
cation, 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
East Asia because 70 percent of all students
More efficient spending and financing
are enrolled in them. Yet decision-making
of higher education
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 inefficiencies 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 accountabilit y, 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
A
fundamental question facing East Asia’s drive toward greater productivity,
Asia, especially its low- and middle- growth, and technological development.
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 certification 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 benefits 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 institu-
tions need to have a strong role in research
provision and have strong links with fi rms
East Asia’s economic landscape
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 five 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 GROW TH THROUGH SKILLS AND RESEARCH 7
Chapters
Goals
1 Productivity Growth
System disconnects
3
Higher education institutions (HEIs) Causes of disconnects Disconnected actors
employers
poor information
weak incentives
low capacity
research institutions
earlier education
Finance
4
Stewardship
6
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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 (figure 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 significantly (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 significant 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 (figure 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.
financed 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
Taiwan, China
Singapore
Hong Kong SAR, China
technology clusters
upper China
middle
Malaysia
middle
Thailand
Philippines
lower
Indonesia
Vietnam
low
Cambodia
Mongolia
Lao PDR
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.
b. Hong Kong SAR, China, average for 2000–07.
els of productivity. From that perspective,
and because they cannot yet be classified 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 reflected
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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
FIGURE 1.3 Sectoral value added as a share of GDP, 1997 and 2007
a. 1997 b. 2007
high-income
high-income
economies
Hong Kong SAR, Hong Kong SAR,
China China
Singapore Singapore
Japan Japan
Mongolia Mongolia
middle-income
middle-income
Philippines Philippines
economies
economies
Indonesia Indonesia
China China
Thailand Thailand
Malaysia Malaysia
low-income
economies
low-income
Cambodia Cambodia
economies
Lao PDR Lao PDR
Vietnam Vietnam
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
China. These last four have some way to go TABLE 1.6 Manufacturing value added, lower-income economies
before they achieve per capita income par-
1990–99 2000–09
ity with Japan. Korea most closely approxi- Economy (average % of GDP) (average % of GDP) Change (%)
mates Japan in manufacturing breadth, while
Vietnam 15.2 20.3 34
Hong Kong SAR, China, and Singapore are
Cambodia 11.1a 18.0 62
the furthest removed. All have one thing in
Lao PDR 14.2 14.6b 3
common: they are at the technological fron-
tiers 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 elec-
tronics, petrochemical, metallurgical, auto- Malaysia, Thailand, the Philippines,
motive, and other fields, 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 purification, logistics, and use cessing activities are of greater significance
of information and communication technol- to Indonesia. All these countries have relied
ogy. All five 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
nificant 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 diversified 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 efficiency 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 fields 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
are in the middle cluster) began taking a seri-
Role and impact of higher
ous interest in innovation as a new growth
education
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 scientific, 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 flexibly, 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 significance 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 GROW TH THROUGH SKILLS AND RESEARCH 13
As a proponent of liberal education, the Oxford Higher education also generates economic ben-
scholar Cardinal Newman advocated for higher efits 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 benefits, higher educa-
Newman’s view is an important mission of higher tion provides several social benefits. Higher educa-
education, the benefits 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 benefits. tion tend to have higher standards of living and bet-
There is much evidence of the economic benefits 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 benefits 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 benefits
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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
100
Korea, Rep.
Finland
United States
80
New Zealand Denmark
Australia
Iceland
Spain OECD Sweden
tertiary school enrollment (% gross)
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
Philippines
Middle East and North Africa
Middle-income countries
China
20 Indonesia
Lao PDR
Cambodia
Sub-Saharan Africa
Low-income countries
0
0 10 20 30 40 50 60 70
GDP per capita (current US$, thousands)
FIGURE 1.5 Tertiary enrollment and labor productivity, 2005 a “respectable” higher education thresh-
old. And figure 1.5 reports a positive rela-
100 48 tion between labor productivity and higher
education.
tertiary school enrollment (% gross)
80
The relation between higher education
output per worker (constant
sia
ia
n
ep
ne
di
in
pa
an
ys
ne
Ch
bo
,R
ala
ail
Ja
do
ilip
a
m
Th
re
M
Ca
In
Ph
Ko
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 (figure 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
FIGURE 1.6 Educational attainment and skills proficiency 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
4
university-affiliated science parks, as well
350
as pursued policies leading to high enroll-
scale scores
300 3
ment in higher education (particularly in
250 2
science and engineering) since the 1970s. 25
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
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 pursued an agenda to develop itself as a hub
Switzerland Norway Canada United Bermuda Italy
States
for higher education, pouring resources into
levels improving quality, developing staff, fostering
500 b. Mean problem-solving scores
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
scale scores
300
2 driver of its economic growth. 27 Long-term
250
trends in R&D undertaken by higher edu-
200
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
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 vation outcomes in upper-income East Asia
Norway Switzerland 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 (figure 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 qualified staff, and located in steadily since the 1970s in most economies
more R&D-intensive industries.24 (figure 1.13), the GER still remains below par
The top technology cluster in East Asia can in developing East Asia (figure 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 GROW TH 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 firms
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
35 AUT
preneurs, managers, and skilled workers to CHE
(per million people)
30
perform well, thus supporting technological FIN
25 DNK SWE
mastery, productivity, and competitiveness.
These systems also help develop countries’ 20
DEU
Korea, Rep.
AUS GBR
technological capability by undertaking 15
BEL NZL FRA
research, supporting technology transfer, 10
NLD
ITA
and providing workers with (and upgrading) 5 NOR Philippines
BRA ARG Indonesia ESP
the skills for innovation. 0 RUS
0 5 10 15 20 CHL 25MEX 30 35 40
Skills and research support longer-run 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 Office data; WDI database.
nomic structures. Cross-country regressions
have shown that a college degree is associ-
ated 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
80
60
50
40
30
20
10
0
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Sources: UNESCO Institute for Statistics (UIS) Data Centre; WDI database.
80
patents granted by USPTO (per million people)
70
60
50
40
30
20
10
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.
1,000
scientific and technical journal articles (per million people)
800
600
400
200
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
350
300
per million people (BoP, current US$ millions)
royalty and license fees, receipts
250
200
150
100
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
a. Low-income economies
60
40
20
0
70
75
80
85
90
95
00
05
10
19
19
19
19
19
19
20
20
20
Lao PDR Vietnam Ghana Bangladesh Cambodia Nepal
b. Middle-income economies
60
gross enrollment ratio (%)
40
20
0
70
72
74
76
78
80
82
84
86
88
90
92
94
96
98
00
02
04
06
08
10
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
Lao PDR Vietnam Ghana Bangladesh Cambodia Nepal
Chile Mongolia Thailand Malaysia Brazil Philippines
Mexico China Indonesia India Pakistan
c. High-income economies
100
80
gross enrollment ratio (%)
60
40
20
0
70
72
74
76
78
80
82
84
86
88
90
92
94
96
98
00
02
04
06
08
10
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
20
HIGHER EDUCATION FOR GROW TH 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 eco-
nomic one. OECD
Countries will not fi nd a one-size-fits-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
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 specific 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-
specific 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 reflect 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 qualifications the number of workers with a tertiary educa-
are increasingly taken as indicators of a par- tion, the slightly increasing or flat 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 specific 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, flat 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
skill premium and proportion of
60 30
20
40 10
0
20
–10
0 –20
1997 2003–04 2007 1992 1998 2004 2006
year year
c. Mongolia d. Chinaa
90
70
80
skill premium and proportion of
0 0
1998 2002 2007 1999 2005
year year
e. Philippines f. Indonesia
80 100
skill premium and proportion of
70
80
skilled workforce (%)
60
50 60
40
30 40
20
20
10
0 0
1988 1991 1994 1997 2001 2004 2006 1994 1999 2001 2003 2005 2007
year year
g. Thailand
140
120
skill premium and proportion of
100
skilled workers
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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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’ significant 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
figure 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 flexibility 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 firms 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, firm 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 firm 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 (figure 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 GROW TH 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 firms
future growth and development patterns.47
a. Foreign/nonforeign firms
In a context in which low-income countries 60
need to break into manufacturing and engen-
Indonesia
der a technological capability, and middle- 50
income countries need to move further up the
TABLE 1.7 Regression coefficients 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.
Significance level: * = 10 percent; ** = 5 percent; *** = 1 percent.
26 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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-specific skills and several thinking and be more important in services. (Further dis-
behavioral skills, reflecting a change in skill aggregation of skill demand49 in other studies
demand. Job-specific 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 significant 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
significant fluctuations in importance across creativity (at least for a few workers in the
TABLE 1.8 Importance of technical, thinking, and behavioral skills for professionals
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 (figure 1.17) and the
Philippines, as well as negotiation and leader- a. Skills by sector
80
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 30
experience in the field. 20
As a further illustration of the importance 10
of practical orientation, labor force surveys 0
nonexporter exporter nonexporter exporter
suggest that fi rms often prefer workers with
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 behavioral skills
ates are, for instance, employed as profes-
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 figure 1.18 also show that a
and professionals with a university educa- significant share of tertiary education gradu-
tion (figure 1.18). In Mongolia skilled work- ates, particularly those educated at TVET
ers with TVET are also paid much more than and college level, find 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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-specific 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, nificantly, 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 suf-
ficient. 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 firms 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, finance, 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 GROW TH 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
Mongolia
Mongolia
clerical/service/sales
clerical
skilled workers
service/sales
unskilled
managers managers
professionals
professionals
Indonesia
Indonesia
clerical/service/sales
clerical
skilled workers
service/sales
unskilled
managers
managers
professionals
professionals
Thailand
clerical/services/sales
Thailand
clerical
skilled workers
unskilled service/sales
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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
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 innova- tion 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-specific 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 identifies 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 for-
mal education to equip them for their working life, Source: The Expert Group on New Skills for New Jobs 2010.
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 firm and because the Japan
returns accrue quickly. Once process innova-
Singapore
tion, which is generally incremental, gath-
Hong Kong SAR, China
ers momentum and its utility is widely per-
Korea, Rep.
ceived, R&D gains stronger adherence (most
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 firms 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 inter-
Cambodia
est in innovation. Such encouragement can
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 financing. In particu-
lar, in lower- and middle-income countries Source: WDI database (latest year, 2002–07).
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 justification 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 find 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
There are certainly several arguments that in the United States and Japan provide fur-
research facilitates teaching, with benefits ther evidence of product-specific 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 fields. They fi nds product-specific 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 findings. 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
efits 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 repu-
tation, and thus attracts high-quality stu-
dents and faculty. Teaching and research
Conclusion
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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
FIGURE 1.20 Correlation between higher education R&D and depend on their position within the tech-
scientific and technical journals nology cluster. For instance, developing the
technological and engineering capacity of
1,400
workers and building some limited research
scientific and technical journal articles
1,000 Singapore
and the Philippines at their stage of techno-
800
logical development but would not be suf-
600 ficient for China. The bottom line is that all
400
Japan
countries need to start moving up within
Mongolia China
Korea, Rep their income and cluster group and beyond,
200
Philippines Malaysia and higher education can help them to do
Vietnam 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-
patents granted by USPTO 2009
35
AUT middle-income economy according to the
CHE
gross national income per capita Atlas method
(per million people)
30
25
FIN applied to 2009 data. Hence, it is considered
JPN
CAN a middle-middle-income economy in this
20
KOR book. This classification is also justifiable
AUS
15 GBR from the fact that all available higher educa-
ISR
10 BEL
FRA tion indicators are for 2009 or earlier.
ITA 4. The World Development Report uses the
5
MYS MEX
RUS ARG ZAF CHL
ESP BRA gross national income per capita Atlas method
IDN IND THA
0 applied to 2009 data (latest available update).
0.0 PHL 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 Office 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
classification is also justifiable 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 classifica-
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 GROW TH THROUGH SKILLS AND RESEARCH 35
clusters are more closely aligned to economy nontechnologically innovative firms within
performance. countries that are relevant. By the same
8. China’s technological prowess in bio- token, the firm sample of Indonesia distorts
technology and nanotechnology might enable somewhat the results by including more
it to cross this threshold in the next decade. firms close to the technology frontier.
9. Because of this technical advancement, export- 23. Yusuf 2003.
ing firms 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 firms (Yilmaz 2009). cation systems in these countries contribute to
10. Schott 2006. innovation at the firm level. The firms, 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, financial 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 firms 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” (definition adopted in the World Bank’s on-the-job exposure and previous experience
Investment Climate Surveys). This definition, 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 firm. 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
firms than some other samples. It is the in Southeast Asia (Tan 2010).
comparisons between technologically and 35. Senker 1995.
36 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
36. One needs to interpret the trends carefully in 41. The main data set used here is a large firm-
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 firms according to their
38. The regression framework for estimating edu- industry and location within each country.
cation (and industry) wage premiums follows The final sample has 9,776 firms distrib-
the approach used by Goldberg and Pavcnik uted across a wide range of sectors (manu-
(2005) and other researchers. Specifically, 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) reflecting worker i’s industry ery, nonmetallic and plastic materials, paper,
affiliation: textiles, and wood and furniture.
42. The relationship with technological innova-
ln(wijt) = H′ijtbH + Sijt · Iijt spjt + Iijt wpjt + eijt
tion was shown in figure 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 firms Machin (1998); Fajnzylber and Fernandes
confirm 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 firms), 44. This limitation is despite the effort to control
and 95 percent of firms (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 finance 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, finance 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 GROW TH 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 fications 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-specific), 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 firms (or firms
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.
definitions (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 findings. 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
T
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
productivity 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 fulfilling 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 fies. 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 fields.
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
pool, and the share of graduates in science,
Higher education and skills for
technology, engineering, and mathematics
growth: The main issues
(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 significance regulatory economic and policy uncertainty,
39
40 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 (figure 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 confirms 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 fill 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 firm size
ficulties 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 finance, and several indicators
more than four weeks in China and Mongolia of technological intensity and firm perfor-
(figure 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-
(figure 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. reflected 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 (figure 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 firms’ degree of openness and upper-middle-income and upper-income
FIGURE 2.1 Business climate obstacles and skill bottlenecks FIGURE 2.2 Time to fill professional vacancies
weeks
5
index
1.0 4
3
2
1
0.5 0
sia
p.
lia
ia
nd
il
ne
az
di
in
di
ys
Re
na
go
ila
ne
Ch
bo
In
Br
pi
ala
et
a,
on
a
do
ilip
Th
Vi
re
M
M
Ca
In
Ph
Ko
0
East Asia other
a
sia
DR
ia
lia
p.
m
ne
di
in
an
ys
Re
na
go
ne
oP
Ch
bo
pi
ala
ail
et
a,
on
do
ilip
m
La
Th
Vi
re
M
M
Ca
In
Ph
Ko
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
weeks
40 4
percent
30
3
2
20
1
10
0
0
lia
sia
nd
ia
ne
ys
go
ila
ne
pi
.
DR
ia
sia
lia
ala
ep
ne
di
in
an
on
a
ys
do
na
ilip
go
ne
Th
oP
,R
Ch
bo
pi
M
ala
ail
M
et
In
on
do
a
ilip
Ph
m
La
Th
Vi
re
M
Ca
In
Ph
Ko
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).
41
42 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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-
firms engaging in technological
25 Indonesia
Vietnam
20
15 Philippines Higher education for the labor market
Lao PDR
10 Korea, Rep. Even though tertiary gross enrollment ratios
Cambodia
5 (GERs) are below par in several countries
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 per-
formance 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
help explain why employers in most coun-
firms using the Internet (%)
60
Indonesia Thailand
50 tries in the region fi nd the current quantity
Malaysia of university graduates, or quantity of edu-
Vietnam
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
10
of tertiary graduates may not be fully aligned
when the GER has been subject to recent
0 changes (such as recent growth). The propor-
0 10 20 30 40 50 60
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
exporting firms (%)
50
Philippines
Vietnam (figure 2.4). In contrast, the gaps between
40 Lao PDR Malaysia GERs and the ratio of university graduates in
30 Cambodia
Korea, Rep. Mongolia and Thailand suggest that enroll-
20
ment growth has been more recent.
10
Third, “below par” does not mean that
0
0 10 20 30 40 50 60 70 the lack of supply of tertiary graduates is an
nonexporting firms (%) immediate cause of skill gaps in all countries.
Source: Almeida 2009a, based on World Bank ICSs (various years).
While demand for higher education is ris-
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 MEE TING ITS PROMISES? 43
TABLE 2.1 Regression coefficients 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 economies economiesb
a
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 fill skilled vacancies.
Significance 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 qualifications
when assessing how much the number of ter-
tiary graduates may be a constraint, which 20
follows here.
Several critical indicators are reviewed for
15
each country. The main demand-side (labor
market) indicators include the level and trend
percent
lia
sia
nd
ne
na
go
ila
ne
bo
pi
et
on
a
do
ilip
m
Th
Vi
In
Ph
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
side and supply-side indicators suggest that 100 trends in education
enrollment ratio
premium
the overall quantity of tertiary graduates is
8
still insufficient to respond to the needs of
3
the economy and a supply push is thus war- 20
ranted to increase the number of graduates. tertiary-educated days to fill
professional vacancies
While not all indicators are consistent (some workforce
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
tertiary education premium
a result, for all three countries, the right side
60
of the graphs tends to be more developed
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 sec-
tor combined with a very low and stagnant 24 25
tertiary-educated days to fill
tertiary-educated workforce (appendix C). professional vacancies
workforce
This finding suggests the need for Cambodia
to widen its very low curriculum diversifica-
tion (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 manufactur-
ing 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
days to fill
have seen stagnating returns and increasing tertiary-educated
workforce professional vacancies
rates. The recent postcrisis recovery should,
however, have put things back on track.
In the case of Vietnam, the evidence con- 84
professionalization rate
fi rms the perception of employers about the
lack of sufficient tertiary graduates as a key
Source: Appendix I.
reason for skill gaps (appendix H).
46 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 findings highlight a need to focus on
the number of tertiary-educated workers in
tertiary gross trends in education
enrollment ratio premium
these two countries rather than on a supply
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
workforce professional vacancies
the transition period; in Mongolia much of
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 ter-
tiary education quantity gaps limited to criti-
tertiary education premium
cal sectors such as services (the Philippines)
100
and manufacturing (Indonesia), but no imme-
diate overall quantity gap.
tertiary gross trends in education The Philippines seems to experience an
enrollment ratio premium
insufficient supply of tertiary graduates in ser-
40
17 vices. This is because of higher and increas-
ing 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 flat 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 figures 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- sufficient. 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, sufficient 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 confirmed
enough graduates (appendix H). The gap in by supply-side indicators for the Philippines
IS HIGHER EDUCATION MEE TING 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
for skills, as also illustrated by a longer-term trends in education
enrollment ratio premium
wage compression on tertiary workers. 8 16 –5
Broader policies to enhance the demand for
tertiary workers may also be needed (assum- 10
21
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 fig-
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-significant but decreasing tertiary education premium
length of time to fi ll professional vacancies, 70
combined with very low (and decreasing) sig-
nificance of the shortage of new university tertiary gross trends in education
graduates as a reason for vacancies, but still enrollment ratio premium
30
relatively low tertiary GERs, points to pos- 5
sible demand constraints in relation to new
15
graduates. However, without data on the 27
days to fill
current stock of tertiary-educated workers tertiary-educated
professional vacancies
workforce
(such as returns, ratio, and employment per-
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 coefficients 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 figure 2.8).
matters. Regional tertiary completion rates The preceding analysis suggests that
for the International Standard Classification 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 (figure 2.8). Rather than increas-
ing enrollment rates (and another reason not
Higher education for innovation
to do it), this evidence confirms 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 (figure 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 there-
fore be made. First, this evidence suggests
60
that while there may be a minimum coverage
50
threshold to spur innovation, the volume of
gross tertiary completion (%)
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
50
half of students earn science and engineer-
ing degrees in China, Japan, Singapore, and
40
Thailand. Between half and one-third of stu-
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 stu-
dents earn their degrees in science and 0
engineering in most of East Asia’s low- and
DR
sia
ia
nd
lia
ne
di
in
ys
na
go
ila
ne
oP
Ch
bo
pi
ala
middle-income countries. While the East
et
on
a
do
ilip
m
La
Th
Vi
M
Ca
In
Ph
Asian average is slightly higher than the G7
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, Scientific 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 constrain-
ing 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
Singapore 59 2004
extent, the Philippines (figure 2.11). Inas-
Japan 63 2005
much as these disparities are correlated with
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 definition 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 definition 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
FIGURE 2.10 Enrollment shares in science and engineering, latest nearly fivefold 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
percent
DR
lia
s ia
SA sia
Ko ina
Sin p.
re
ne
di
Re
po
na
go
ne
y
oP
Ch
bo
ala
et
ga
a,
on
do
ilip
m
La
R,
Vi
re
M
M
Ca
In
Ph
80 100
80
60
60
40
40
20 20
0 0
a
m
d
lia
a
sia
lia
sia
a
in
s
ne
an
di
in
di
ne
an
na
na
Ch
go
go
ne
ne
Ch
bo
bo
pi
ail
pi
ail
et
et
on
on
do
do
ilip
m
ilip
m
Th
Vi
Th
Vi
M
M
Ca
Ca
In
In
Ph
Ph
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
a
sia
lia
d
di
an
na
go
ne
bo
ail
et
on
do
m
Th
Vi
M
Ca
In
low-income middle-income
countries countries
enrollment completion
Source: Sakellariou 2010a, based on labor force and household surveys (latest year).
51
52 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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
Regression model (1) (2)
that supporting disadvantaged groups can prevent
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 significant indi- Source: World Bank 2010a.
cator of performance. a Similarly, results of regres- Note: Standard errors are noted in brackets.
Significance level: *** = 1 percent.
sion analyses looking at the determinants of grade
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 significance 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 insignificant 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-
a. World Bank 2010a.
ity of their schools.d 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 MEE TING ITS PROMISES? 53
FIGURE 2.12 Tertiary unemployment rates and time to fill professional vacancies
14
12
10
0
a
lia
sia
ia
ne
di
in
an
ys
na
go
ne
Ch
bo
pi
ala
ail
et
on
do
ilip
m
Th
Vi
M
M
Ca
In
Ph
unemployment rate of tertiary graduates (percent)
weeks to fill professional vacancies
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,
but gaps in technical skills in the Philippines
Gaps in thinking, technical, and
become more important.
behavioral skills
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 deficiencies, 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
Mongolia
Philippines
Indonesia
Thailand
Malaysia
more so as the push to innovation intensifies 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 finding 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 Vietnam, 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, reflecting 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 MEE TING ITS PROMISES? 55
Gaps in technical skills are also evident, FIGURE 2.13 Key job-specific skill gaps in Vietnam
particularly in practical job knowledge.
While gaps in generic skills are on the rise,
gaps in job-specific technical skills (practical
job-related technical skills
and theoretical knowledge of the job) remain
pervasive, all the more in relation to current
needs. This is particularly the case in Vietnam, practical knowledge
of the job
where only a small share of college graduates
are said to possess good technical knowl- theoretical knowledge
edge of their job (figure 2.13), and Indonesia, of the job
where technical gaps are significant 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 significant number of employees still
foreign language/
feel that they do not have the needed techni- communication
cal or professional skills (figure 2.14), and the
problem grows in relation to the importance
punctuality
given to these skills.
In the three countries for which the the-
ory versus practice differentiation exists independence
(Vietnam, Indonesia, and the Philippines, as
illustrated in appendix J), practical knowl-
edge 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 specific skills
Source: World Bank 2008, based on 2003 MOLISA-ADB survey on labor market.
(including practical knowledge) acquired Note: Characteristics possessed by recruits of different 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 firms found that
(China), find that poor technical skills are 44 percent of executives cited insufficient 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 firms
technical skills. (The same report notes that and employees at all employment levels.29
the quality of university graduates is a seri-
ous issue in Vietnam, where firms 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-five 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
FIGURE 2.14 Key job-specific skill gaps in Malaysia international integration continues, particu-
larly 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 (figure 2.15) and Malay-
time management skills sia (figure 2.16) further suggest that employ-
ers 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 pre-
miums in relation to professionals with very
technical/professional skills poor levels of these skills.30
0 10 20 30 40 50 60 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 deficiencies 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 (figure 2.17). Similarly, in Vietnam
expected exception of the Philippines and about 30 percent of those surveyed reported
Malaysia, both of Anglo-Saxon tradition, significant 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 MEE TING 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 deficiency compared with
education cycle, and others such as practi- employees not reporting skills as a top-three deficiency
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.
information technology skills
Higher education has, however, a critical
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
technology providers. Other countries’ TABLE 2.6 World Intellectual Property Organization patent filings by
receipts are insignificant (figure 2.19). origin and office, 2007
In sum, Korea could emerge as a net Patent filings
provider of technology in the near future. Patent filings by office by origin
Singapore is actively engaged in the trade of Economy Total Resident Nonresident Total
technologies in both directions through for-
Japan 396,291 333,498 62,793 501,270
eign affiliates. China is currently the leading
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
Vietnam 0 0 0 13
The quality and research capacity of univer-
sities play a key role in determining techno- Source: WIPO 2009.
Note: Data for Indonesia and the Philippines are for 2006.
logical capability and innovation, though
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
U.S. dollars (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 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 confirm
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
U.S. dollars (000)
2,000
its top 500 institutions from 16 to 30 in a
1,500 five-year period.
Published scientific output provides
1,000
additional evidence that most East Asian
500 economies lag behind OECD countries.
Unsurprisingly, members of the lead-
0 ing technology cluster score the highest
1995 1997 1999 2001 2003 2005 2007 2009
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 edu-
cation 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
International rank
Regional 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 MEE TING ITS PROMISES? 61
as noted, receive more patents than other FIGURE 2.20 Scientific 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 scientific and techni-
cal papers when data are scaled to the total a. East Asia
populations. With the notable exception of
high-income
Singapore
countries
Mongolia
and the United States (figure 2.20). Philippines
China is trying hard to move up to the top Indonesia
cluster by giving the highest priority to tech- China
lower-income
Vietnam
Lao PDR
tertiary education and R&D, and over these
Cambodia
past 15 years it is closing the gap, as measured
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
high-income
countries
Ghana
results. They are, in fact, much further behind, Nepal
trailing also several other middle-income Bangladesh
countries outside the region. From 2000 to 0 200 400 600 800 1,000
scientific and technical articles per
2005, Korea produced an annual average of million inhabitants
more than 20,000 scholarly publications,
over 10 times more than the number of pub-
Source: WDI database.
lications produced in Thailand and Malaysia.
62 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
FIGURE 2.21 Leading ways of acquiring technological innovation in firms, Malaysia, Mongolia,
and Thailand
2%
Malaysia, 2007
2%
Thailand, 2006
1%
Mongolia, 2004
0 20 40 60 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
declared using research centers or uni- FIGURE 2.22 Sources of product innovation at firm 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 tech-
nology 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 universi-
ties as a high and medium source)43 and in
Singapore by more than 10 percent.44 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 findings are somewhat conditioned by
the ICS samples, which mainly cover manu-
the skills required by the labor market and
facturing firms and a generally significant pro-
to support innovation, while also starting
portion of large and foreign-owned firms. 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 firm
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
firms, skills become less of an obstacle.
but skill-related gaps are persistent. Gaps
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 firms according to their industry
is not performing as it should, which is dis- and location within each country. The final
cussed in chapter 3 through the perspective sample consists of 10,215 firms in a wide
of disconnects. range of sectors (manufacturing 78 percent,
64 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
25. For Indonesia this is confirmed 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 scientific 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 figure 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 confidence 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 specific data on sources of tech-
greater use of nontechnical graduates in roles nological innovation at firm 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.
filled by workplace trainees (Round 2003). 44. Mok 2010.
Disconnects in Higher
Education 3
T
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
shortcomings 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 significant 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-
ers and users (see figure 3.1): between higher
The first disconnect: Between
education institutions (HEIs) and skill users;
higher education and employers
between universities and firms (now seen
(skill users)
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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
employers
poor information
weak incentives
low capacity
research institutions
earlier 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 qualified (and less productive) workers or
their subject based on the advice and wishes—and, to scale back their growth ambitions. It can also stifle 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 five 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 field of study and then signal overall graduate unemployment levels
because of market need. (Among university and voca- and supply-demand mismatches in particular fields.
tional student respondents, 59 percent chose their Additionally, universities could review their course
field 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
DISCONNEC TS 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 fields 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 fields
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 (figure 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 (figure 3.2). Far fewer Low enrollment in science, technology, engi-
students are in other fields. neering, and math (STEM) fields 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
60
50
40
percent
30
20
10
0
Cambodia Lao PDR Vietnam Mongolia Indonesia Thailand Malaysia Korea, Rep. Singapore Japan
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 fields is visible (figure 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 offer-
ings, such as Cambodia, many students can-
Misalignment of institutional mix
not obtain jobs because of a lack of oppor-
tunity to study certain fields 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 nontra-
ditional delivery, such as distance higher
120 education. (Appendix A illustrates the mix
100 in each country.)
Evidence suggests that the institution-
degree mix is misaligned with labor market
percent
80
needs in several countries, with implica-
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
0 Vietnam,6 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-
60 facturing and services).7 Most countries also
have still relatively low shares of postgradu-
50
ate students.
percent
40
TVET/non-TVET
30
Often difficult employment prospects for
20 graduates and relatively low tertiary com-
pletion 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
linear (tertiary education wage premium
graduates alongside unfilled positions in
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
DISCONNEC TS 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 stu-
dents are enrolled in TVET (figure 3.4). In
countries
income
Malaysia nearly 60 percent of tertiary students
high-
Korea, Rep.
are in TVET institutions. Nearly a third of
Thailand’s upper-secondary students pursue Malaysia
TVET qualifications, and almost 20 percent
of tertiary students do. Indonesia has recently Thailand
considered having 70 percent of its upper-
middle-income
secondary students in TVET programs. China
countries
Tertiary TVET has several advantages.
It shapes the skill set of the next generation Indonesia
of workers, provides students with read-
ily employable skills, and can help make up Philippines
student undersupply in some fields. A recent
World Bank study on the Philippines, for Mongolia
example, found that postsecondary TVET
institutions have higher labor market rele-
low-income
Vietnam
countries
FIGURE 3.5 Employers’ perceptions on general and TVET tertiary education, Indonesia and the Philippines, 2008
a. Indonesia b. Philippines
14
share of university institutions
100
by weakness
by strength
60 8
50
40 6
30 4
20
10 2
0 0
h
qu ache g
ls
h c ities
dy
ha th in ds
LM ry
st
ty
on
ty
dy
ce
y
ink
ed
die
ilit
cit
ilit
gt
in
kil
co
st
ali
ali
an
e
stu
stu
ati
yl
ch
len
ne
ne
pa
ab
tab
du
rs
tu
cil
qu
qu
ev
fic
r
tea
of
of
t
ca
ap
fs
a
us
ap
cle
rel
ing
ies
res of f
ali
ho
oL
s
lds
rch
ind
of
ad
d
qu
cy
ing
ilit
LM
ch
te
i
ty
t
fie
fie
gt
arc
ty
ea
LM
fac
ce
er
tea
ali
ng
len
ks
ali
res
of
e
an
ch
lin
qu
ty
tea
ev
oc
rie
rel
yt
va
ilit
tab
ap
ad
c. Indonesia d. Philippines
16
100
share of postsecondary TVET
14
90
institutions by weakness
institutions by strength
80 12
70 10
60
8
50
40 6
30 4
20
10 2
0 0
h
ing
ills
ies
ity
dy
s
try
st
ty
on
ty
dy
ce
y
k
ed
ed
die
cit
ilit
lin
gt
co
ali
ali
an
stu
stu
bil
ati
sk
ilit
us
ch
len
ne
ne
pa
tab
tu
try
qu
qu
ev
ind
pa
fic
fac
er
tea
of
of
ca
fs
LM
LM
us
ap
cle
rel
ch
ing
ies
ali
ca
ho
lds
lds
of
th
rch
ind
of
ad
tea
qu
cy
to
ing
ilit
LM
rch
ch
wi
ty
fie
fie
gt
ty
ea
LM
fac
ce
er
tea
ali
ng
ea
len
ks
ali
res
of
an
ch
qu
res
lin
ha
qu
ty
tea
ev
oc
rie
rel
yt
va
ilit
tab
ap
ad
Sources: (a), (c) di Gropello, Kruse, and Tandon 2011; (b), (d) di Gropello, Tan, and Tandon 2010.
TABLE 3.1 Earned science and engineering doctoral degrees, selected region or economy and selected field,
2006 or most recent year
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,
country
income
high-
Vietnam 2.6
ple, has a fairly low ratio of students enrolled
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 par-
ticular in middle- and lower-income coun-
tries. 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 (figure 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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
60
years—not just for research but also for skill
development and training as in the Philip-
50 pines and Vietnam—they have remained
mostly informal and piecemeal.
40 Overall, information on graduate employ-
ment, labor markets, and skills is still weak
percent
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,
Cambodia 23:1
secondary, and tertiary teachers who in turn
China 19:1
shape the quality and relevance of the entire
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,
35
which grew far faster than faculty numbers,
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
ratio
20
the Organisation for Economic Co-operation
and Development (OECD) average of 15 to 1 15
(table 3.2),19 having climbed or stayed rela- 10
tively constant in most countries (except for
the Philippines; figure 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
diversification (particularly the possibility of Malaysia Korea, Rep. Japan
offering college education and STEM fields,
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 figure
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 qualifications, and of faculty with at least a master’s degree.
76 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
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 diversified institutional setting. Lower-income
countries face a strong double quantity-quality
80 challenge (particularly Vietnam).
60 Weak incentives
percent
lia
sia
n
ep
ne
di
in
pa
an
na
go
ne
Ch
bo
,R
pi
Ja
et
on
do
lip
ea
m
Th
,
Vi
an
M
Ca
r
In
i
Ph
Ko
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
DISCONNEC TS 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 disincen-
Singapore
tive to tailor programs to the needs of the
high-income
economies
local community. Korea, Rep.
Japan
Hong Kong SAR, China
The second disconnect: Between Thailand
higher education and companies
middle-income
(research users) Philippines
economies
Mongolia
Engagement is weak. The limited contribu-
Malaysia
tion of universities to technology adaptation
Indonesia
and upgrading in companies is suggestive of
a disconnect between companies and uni- China
economies
versities in research and technology mostly
income
Vietnam
low-
applicable to lower- and middle-income Cambodia
countries but also to some upper-income 0 1 2 3 4 5 6 7
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-
high-income
United States
economies
Mexico
are fairly developed. By contrast, were one to
India
consider only research and technology, they
Chile
would be higher in Japan.
Even so, the picture that emerges is one Brazil
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
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 findings 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 firm 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 firms identified 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 firms 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 flexibil- 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 identified 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 firms in Thailand.36
Elsewhere, an electronics firm in the
Philippines identifi ed the lack of qualifi ed
Reasons for the disconnect
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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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), difficulty reaching
country) generally and university offerings mutually agreeable profit-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 scientific 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
1,200
to bring new technologies to market.41
scientific and technical journal articles
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 teach- ing, 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
specifically 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 fierce 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 educa-
tion’s share in both national R&D expen-
diture and personnel has in fact trended
The fourth disconnect: Among
downward in recent years, accounting for
higher education institutions
8.5 percent and 14.6 percent, respectively, in
themselves and between
2007. Moreover, interaction between public
these institutions and training
research institutes and universities largely
providers (horizontal disconnect
focuses on recruiting university graduates.52
across skill providers)
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
DISCONNEC TS 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 certification.
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 efficiency 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 flexible 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
industry in curriculum development, and
Reasons for the disconnect
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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
Philippines
Tertiary and earlier academic
Hong Kong SAR, China performance
Academic performance at the tertiary level
Mongolia
is closely related to earlier academic perfor-
mance. 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
DISCONNEC TS 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
percent
percent
25 25
20 20
15 15
10 10
5 5
0 0
1997 2003–04 2007 1992 1998 2004 2006
c. Philippines d. Indonesia
70 60
60 50
50
40
40
percent
percent
30
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)
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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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-
middle-income
Mongolia
in reading and math subjects 67 was partly
reversed between 2006 and 2009, under-
scoring its significant gaps in relation to the
Indonesia OECD average in 2009 (figure 3.16). The
country also needs to reverse its stagnating
China
TIMSS results.
Second, countries with a higher TIMSS
score in math and science have higher STEM
low-income
countries
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 significant inequities in Economy 1999 2003 2007
access to and survival in prior education lev-
Taiwan, China 585 585 598
els across all dimensions. Indonesia seems to
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 reflect 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
others 2000.
reasons). The gender inequities in favor of 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
(figure 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
United States 515 527 520
benefits of promoting girls’ education.
International average 488 473 500
Results of PISA 2006 show the benefits of
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 significant 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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
Shanghai, China interventions focus on fee exemptions).72
While they start early, inequities by eth-
Singapore
nic group are most often further exacerbated
in the transition between secondary and ter-
tiary education, raising additional issues for
Hong Kong SAR, China addressing this disadvantage.73 Even the few
benefits 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 (figure 3.22).74
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 SGP
600 HKG
students at universities are inherently linked
550
to issues of access and equity because suc-
TIMSS math
cessful performance in these examinations 500 MYS
is the result of prior academic experience. 450
Evidence for this fi nding is seen notably in 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
enroll in high-quality tertiary education insti- KOR
SGP
tutions.77 While the Ministry of Education 550 HKG
and Training sets the cutoff marks needed to
TIMSS science
500 MYS
enroll in tertiary institutions for mathemat-
ics, physics, and chemistry, higher-quality 450 IDN
universities require better entrance examina- 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 officials 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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-
700
a. Math ing enough effort at that level.
650
Curricula may not be sufficiently well
KOR SGP aligned between the secondary and tertiary
600 JPN
cycles, the case in Mongolia and Thailand.86
TIMSS math
550
This misalignment may be partly responsible
500 MYS for some of the dropouts between secondary
450 THA and tertiary education. Yet countries such as
400 IDN
PHL
Korea and Singapore have been successful at
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 depends on inclusiveness early on, requiring
650
600
a bold and comprehensive approach to tackle
KOR JPN disadvantage. There is, for instance, still too
550 SGP little use of distance education at the second-
TIMSS science
HKG
500 MYS ary level, and scholarships to cover transport
450 costs to urban centers, to address the urban-
IDN
400 rural gap, are limited.88 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 specific measures, implement-
ing comprehensive approaches to tackle qual-
ity 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 difficulty 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
inf luence outcomes later, requiring at nonmonetary barrier to access, particularly
minimum a coordinated effort to improve for disadvantaged groups whose parents
DISCONNEC TS IN HIGHER EDUCATION 91
FIGURE 3.19 Estimated survival rates for children ages 13–19, Cambodia, Indonesia, and Thailand
percentage completed
percentage completed
75 75
50 50
25 25
0 0
0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12
grade grade
male female male female
percentage completed
75
75
50 50
25 25
0 0
0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12
grade grade
urban rural urban rural
75 75
50 50
25 25
0 0
0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12
grade grade
poorest quintile richest quintile poorest quintile richest quintile
percentage completed
75 75
50 50
25 25
0 0
0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12
grade grade
urban rural poorest quintile richest quintile
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
50 70
60
40 50
percent
40
30 30
20
20 10
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 con-
strained 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 arrange-
ments of the government in relation to the
60 education sector, also influences the way
50 interrelationships are understood and built
completion rate (%)
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
80
Conclusion and moving forward
This chapter has argued that failure to 70
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).
Note: Tertiary education is proxied by university education in Mongolia and the Philippines;
systems will respond the same way to higher 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
tremendous role. Figure 3.25 shows a simpli- conduct research, hire high-quality faculty,
fied 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 five 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 figures are simpli- Finally, funding for scholarships can help
fied 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 influence all disconnects with connects. Greater and more efficient 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 efficient public spending
scholarships. Low enrollment in STEM fields 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
fields. 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
DISCONNEC TS IN HIGHER EDUCATION 95
The overall policy direction, oversight, regulation, addressing scale economies needed to finance 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 benefits difficult 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 stifling 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
higher
capacity
Financing higher education
stronger and employers
incentives
higher education
and companies
(as research users)
HEIs between
themselves and with
training providers
better
information
higher education and
Management
employers
stronger
incentives
higher education
and companies
(as research users)
HEIs between
themselves and with
training providers
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
and by putting in place the right legal and 6. World Bank 2008.
financial incentives (addressing incentive 7. Smeaton and Hughes (2003) define 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
required to operate effectively in jobs with a
In light of these considerations and the
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: financ- 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
that this poor mix may be contributing to the university 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.
Scientific 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 field 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 find
17. Only Thailand and Vietnam provide a dis- that faculty in second-tier European univer-
aggregation of tertiary education by field 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 scientific 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 firms’ risk aversion has generated
positive correlation between teaching and a conflicting 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 firm Alpha Catalyst found that
combination of several factors, including the Malaysian firms, 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 financial 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.
DISCONNEC TS IN HIGHER EDUCATION 99
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 Office 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 scientificize Ministry of Education (Fielden 2008).
the whole people, performed in conjunction 91. Some policies involve actions to support firms
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 fiscal, financial,
science soon followed at the tertiary level. and liberalization policies to encourage inno-
86. Several reports (World Bank 2009b, 2010d) vation in firms. 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 ulti-
mately decides the success of higher educa-
tion reform.
92. Rumbley, Pacheco, and Altbach 2008.
Financing Higher Education
4
S
olid financing 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
but 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 financing goes to institutions
regardless of whether they are addressing
Financing needs
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 efficiently 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
be very costly without greater efficiency 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.
The fi nancing gaps that these projections
Simulation analysis
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, Scientific 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 qualifications, 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 financ-
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 qualifications, 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 qualification 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 fig-
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
2011 61.7 151.8 90.1
While these estimates are several orders
2013 61.7 186.9 125.2
of mag nit ude above what I ndonesia,
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
Source: Authors’ calculations based on model presented in appendix M.
be put in context. The estimates are broadly
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
Gap between
US$7,000 in Chile.1 The United States
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
average is US$11,512. 2 2013 11.6 359.6 125.2
2015 11.6 430.7 174.6
2017 11.6 536.4 259.5
Funding priorities
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 finance. quality trade-offs. It may also be difficult 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 (figure 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
fications 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 qualifications). 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 edu-
cation 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
fields, 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 fields, and allowing for greater curric- tackle inclusiveness.
ulum and degree diversification 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 benefits 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 five countries,
but only around a tenth in the Philippines .70
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-benefit 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 uni-
versity 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.
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 significantly declined, East Asian economies
particularly in comparison to its middle-in-
come neighbors (figure 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 pri-
20
mary and secondary education, providing a
larger potential pool for tertiary education.6
0
But tertiary education policies have comple-
percent
mented these efforts by supporting the higher –20
transition of secondary graduates to tertiary
education. –40
Since 1998 Vietnam has instituted policies
to encourage fee deductions and exemptions, –60
benefiting poor and ethnic minority students.
In 2006 about 22 percent of disadvantaged –80
university students were benefiting from sig-
nificant fee deductions of up to 50 percent of –100
Cambodia Vietnam Indonesia Mongolia Thailand
tuition.7 low-income economies middle-income economies
Vietnam has also had student aid schemes enrollment completion
since the mid-1990s. Since 2006 stu-
dents 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 con-
firms that the program is well targeted, with aid instruments to performance (or at least
benefits 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 significant 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
five in a rural area, the loans are forgiven. after graduation.
Universities have the autonomy to admin-
C hi n a . Tu it ion fe e s i n 20 0 4 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 first- 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$806 in these univer- covers only a small proportion of students.
sities, and for languages and medicine, A short repayment period is also taxing for
Net payment
Average Living cost Scholarship Work study Net payment (% of gross
tuition (US$) (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.
40
Other lower- and middle-income coun-
% of household income
35 32
tries in the region have had a more piecemeal
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
10
Indonesia. Private spending, primarily 5
for tuition fees and levies, constitutes the 0
bulk of financing for higher education.14 public education private education all education
In 2009 the average spending per student
per year was about US$2,200 in public
Source: World Bank 2010c.
institutions (tuition fees, student fees, and
living expenses) and about US$1,200 in pri-
vate institutions. Supporting one tertiary
education student can cost up to a third of Cambodia. Scholarships are the primary
yearly income (figure 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 (figure 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 significantly less than the cost of study
110 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 (figure 4.5). These results could be improved with better
Sending a child to school thus represents a targeting. There is also evidence suggest-
significant 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 figures 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 fields 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
3,000 70
2,500 60
50
% of total expenditure
2,000
40
baht
1,500
30
1,000
20
500 10
0 0
poorest second third fourth richest
income quintile
monthly expenditure on education (left axis)
educational expenditure (right axis)
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 tar-
geted 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 ne ed-based s chola rsh ips , b e c ause
whereas upper-income students dominated merit-based ones do not promote inclusive-
fields 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 specific 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-
sities (reaching 60 percent of overall enroll-
How to fund priority activities
ment 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 significant 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
costs can be high, the social benefits are even spending significantly 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 finance 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 (figure 4.7).
ing student loans. Within a coherent financ- 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 financing 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 diversification (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 (figure 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
Japan 0.61
the region can suggest some room for real-
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
needs in other parts of the education system24 Indonesia 1.2
and the fairly low overall education spending Philippines 0.34
as a share of GDP in several countries (figure 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
share of GDP (%)
ing in most countries since the beginning
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
efficiency 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 first imperative will be to ensure
higher public spending shares for research,
Increasing efficiency 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 Efficiency 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
FIGURE 4.7 Public expenditure per tertiary pupil as a share of GDP per capita, 2006–07
60
50
40
30
20
10
0
Cambodia Lao PDR Vietnam
40 40
20 20
0 0
Malaysia Mongolia Philippines Thailand Brazil Chile India Mexico
80 80
60 60
40 40
20 20
0 0
Hong Kong SAR, Japan Korea, Rep. Finland United Kingdom United States
China
funds into journal articles than Japan and within teaching on non-STEM fields. Malay-
Korea (figures 4.12 and 4.13). sia’s high spending indicators in relation to
Some of these differences may simply be the outcomes suggest some inefficiency.
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 inefficient 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
teaching), which makes selectivity impor- a. Low-income economies
150
tant. Harvard University had an endowment
of US$37 billion, and annual spending of
120
US$3.2 billion, and per student spending of
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.
FIGURE 4.9 Public tertiary education expenditure as a share of total public education expenditure, 2006–07
education expenditure (%)
share of public
30 30
20 20
10 10
0 0
Cambodia Lao PDR Vietnam Bangladesh Ghana Nepal
40
30
share of public
30
20
20
10 10
0 0
Malaysia Mongolia Philippines Thailand Brazil Chile India Mexico South Africa
30 30
20 20
10 10
0 0
Hong Kong Japan Korea, Rep. Singapore Finland United Kingdom United States
SAR, China
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
0 1 2 3 4 5 6
share of GDP (%)
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 identified, 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 competi-
Korea, Rep.
tive governance framework for all universi-
ties (discussed in the next chapters). Also Malaysia
useful is a competitive (or other performance-
Thailand
driven) process to target the resources at the
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 effective-
ness of both teaching and research (even in Vietnam
fields that have less scope to be financed 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
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 fields identified 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 cur-
rent king, Rama IX. Several scholarships are given Source: World Bank 2009a.
118 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 more selective resource allocation. Higher
40 Korea, Rep.
education institutions submit their fund-
% of GDP per capita
35
ing proposals along with their development
30
Philippines
Vietnam
plans, with key performance indicators. 30
25
Cambodia The funds are then used to fund equipment,
20
facility improvement, and staff upgrading and
15
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 10 20 30 40 50 60 70 sustain financial viability—and they encour-
STEM share (%) age faculty to devote time to teaching and
research, improving quality across the system
Sources: EdStats database; UIS Data Centre.
(box 4.4). Student fees can then be used by
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 govern-
ment could simply remove or reduce fund-
journals per million people
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 difficult. One way around this is
200 to have diverse financing mechanisms beyond
100 Malaysia
the base fund, to incentivize institutions’ per-
Thailand
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 direc-
tion 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 i n resou rce allocations across dents that graduate, sometimes at higher
FINANCING HIGHER EDUCATION 119
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: = noneⲐnot applicable; = partial.
a. Formula funding incorporating performance indicators piloted in five autonomous universities (and some nonautonomous universities) in Indonesia.
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 fields of study and equity, though they may also use them to
or with specific 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, efficiency, grants for autonomous universities (and
120 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
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 benefits over a flat 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 efficiency 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 loan 41
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 insti-
tution (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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 profits 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 floor—a guaranteed minimum that Source: Sutton Trust 2004.
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 ified period. Without the income-contingent part of
the loan, he or she may have to declare bankruptcy, the loan, a borrower may have difficulty 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 efficiency (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 inflation.
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 efficiency 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) inefficiencies 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
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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 fiscal 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 fidu-
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 qualified 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
schemes, helping ensure sustainability.
Summary of policy priorities
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 financing 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 fields, 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 financing to finance research, STEM fields,
1. OECD 2008b.
and inclusiveness. (Private funding would 2. OECD 2008b.
complement public funding in financing some 3. As visible from the snapshot presented in
activities and focus on increasing coverage table 4.5 and further evidence from Salmi
and diversification.) 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 efficiently, 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
rigorous econometric analysis (World Bank
Mobilizing and prioritizing public funding.
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 figures 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.
fields, and scholarships. 13. The Philippines has done very little overall in
terms of equity (Orbeta 2008).
Increasing the efficiency of public fund-
14. World Bank 2010c.
ing. Because few countries have much flex- 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 efficiency 17. World Bank 2010e.
requires being more selective and perfor- 18. These figures are not directly comparable
mance based in the way public funds for with the amounts in figure 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.
21. Singh 1973 cited in Mukherjee 2010.
Leveraging private funds and correcting
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 financing simu-
helping to recover costs. lations for Sub-Saharan Africa under five
126 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 flow 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 finance 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 qualification, 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
F
inancial capacity forms an essential public sector can help tackle the disconnect
pillar of effective higher education. between higher education institutions and
What 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 efficiency 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 sufficient 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 difficult 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’ fulfilling 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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
fi rst of three accountability relationships—a
Global moves to autonomy “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 justified 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.
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 tech-
nological frontier.
Enhanced autonomy can also support a
better match between output from tertiary politicians policy makers
T he mai n a rg u ment for h ig her i nstit utions’ 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 University a 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
nificant 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 (fifth disconnect). accountability.
The evidence on the effects of autonomy in
East Asia is less widespread than for OECD
Effects of autonomy at
countries, as real autonomy is still rare or,
institutional level
in some countries, very new.9 Still, the par-
tial 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
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-profit
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. E lsewhere , re c ent at tempt s at refor m i n
(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 man-
age 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
flexibility.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 financial 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. Significant 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
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
governments directly determine salaries.a 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-
ing a decentralized approach; countries such as Greece and Portugal follow a
hired by the institution, which also sets salaries 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 flexibility in
The other area where higher education terms of hiring and firing 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 staffing autonomy.
committed to introduce fi xed-term contracts.
Korea, by contrast, continues to limit staffing Procedural autonomy
autonomy in these institutions. Even more than for substantive autonomy,
Middle-income countries’ institutions seem procedural—especially financial—autonomy
to have partial autonomy on staffing, though differs widely between high-income econo-
it is quite restrictive. Institutions in Indonesia mies and other categories.
MANAGING PUBLIC HIGHER EDUCATION 135
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
fit. 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 finan- 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 financial
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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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
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.
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
university councils is second to that of the
Incomplete accountability
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 deficiencies 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 qualifications 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 figure 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
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 fit between supply and
financing, 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 difficult. strategic vision for higher education, coun-
tries 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
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 officials
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. Differs 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
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 five-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
TABLE 5.4 External quality assurance, East Asia
Type of body Type of system Body funding source Requirementa Disclosure
Semi- Government-
Independent autonomous represented
Economy body body body Accreditation Audit Assessment Government Institutions Others Mandatory Voluntary Complete Limited None
High-income
Japan X X X X X X X X X
Singapore X X X X X X X
Hong Kong SAR, Chinab X X X X X X X X
Korea, Rep. X X X X X X X X
Middle-income
Malaysia X X X X X X
Thailand X X X X X X
China X X X X X X
Indonesia X X X X X
Philippines X X X X X
Low-income
Vietnam X X X X X
Lao PDR X X X X X
Cambodia X X X X X
Sources: Raza 2010 based on Expert Survey, except Singapore, for which information taken from Ministry of Education, Singapore 2005.
a. This requirement is for public and private institutions, unless otherwise stated.
b. For Hong Kong SAR, China, quality assurance is mandatory for public universities and voluntary for private universities.
141
142 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN 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-qualified or more faculty. Indeed, if
well developed and accompanied by adequate fac-
ulty 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 hig her 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 difficult 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
financing is still in its infancy. be more generous with substantive autonomy
than procedural autonomy, though this is
now changing. In East Asia, too, governments
Moving forward 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 staffing policies, hir-
Accountability lines to nongovernment ing and firing 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 fiduciary role. tant, because it is a critical mechanism of
The compact relationship (see figure 5.1) external efficiency. 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 efficient outcomes. This will require
two main types of measures, both of which
Autonomy
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 staffing. 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 qualification 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 financing. 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 insufficient 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 financing, 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 manage-
ment and protecting the capacity of institu-
Creating an enabling environment
tions to fix 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,
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 qualifications 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 five years, achieving full implementation by year
or even in the same education level) and should 2000.
therefore be supported, their design and implemen-
tation are complex. Challenges include developing Sources: Australian Qualifications Framework Web site, www.aqf.edu.au; Johanson
standards-based on-the-job analysis, preparing new and Adams 2004.
MANAGING PUBLIC HIGHER EDUCATION 145
lack of comprehensive reform in legislation, represents between a fifth 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 efficiency, 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 first 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
E
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-
many 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 financing; 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 fulfill 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 firms 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 offices (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 con-
sider related issues and challenges, progress
will be difficult. Without a bold, systematic,
Providing effective coordination and comprehensive vision—and processes for
among government bodies 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 five 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 STE WARDSHIP FOR HIGHER EDUCATION 149
120
100
80
percent
60
40
20
0
p.
na
sia
ia
DR
lia
m
ne
in
di
in
pa
an
ys
Re
na
go
hi
ne
oP
Ch
Ch
bo
pi
ala
ail
Ja
,C
et
a,
on
do
ilip
La
Th
R,
Vi
re
M
an
M
Ca
In
SA
Ph
Ko
iw
ng
Ta
Ko
ng
Ho
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 official) 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 figures are more broadly inclusive. There are also differences in how enrollment and other variables are
measured. Data years are the most recently available by country (generally 2006, 2007, or 2008).
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 significant 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 fields more expensive
(including Japan and Thailand) show higher to offer, such as engineering and information
152 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 (figure
and Thailand).10 However, the main demand- 6.3). Other sources confirm this evidence.
absorbing field 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 (figure 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 confirms 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
% of graduates
60
50
40
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 STE WARDSHIP FOR HIGHER EDUCATION 153
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 deliv-
ery to its full potential. This failure has impli-
gross enrollment ratio (%)
enrollment (millions)
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
In countries such as China and Vietnam,
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
71
73
75
77
79
81
83
85
87
89
91
93
95
97
99
01
03
05
07
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
20
20
20
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 (figure 6.5).
education and the government’s insuffi- Private delivery also has the potential to
cient supply of public universities provided lead to system diversification 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 diversification 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-
officially recognize current private institu- tions 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 STE WARDSHIP FOR HIGHER EDUCATION 155
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
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, engag-
ing 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
percent
25
and reputation (“garage”22) and those show- 20
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
1994 1996 1998 2000 2002 2004 2006 2008 2009
fields), often through inferior resources, a
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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 figure 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
became coupled with an approach to cut
Main issues for private delivery
the brain drain of domestic students going
in East Asia
abroad and to attract thousands of foreign
Two main higher education issues face low- students to Malaysia, specifically 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 insufficient 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 STE WARDSHIP 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 figure 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
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 specific 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 figure 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 findings 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 Japan 36 about 70 percent. 37 Overall, East
1996 private higher education act), though, Asia’s tuition share accounts for more than
PROVIDING STE WARDSHIP FOR HIGHER EDUCATION 159
China Malaysia
• Funding comes from overseas Chinese and • Foreign funding is permitted for PHEIs.
domestic businesses, with much initiative by • Business firms open their own institutions or buy
professors and administrators from public institutions set up by individual proprietors.
universities. • Various sorts of foreign financial 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 affiliated 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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 efficient 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 sufficient 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” fields, 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 sufficiently 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 benefits 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 benefits of such a policy
good benefits but did little research. in coverage.
On the basis of the preceding review, the Competitive research funding could also
following assessment of the financing 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 STE WARDSHIP FOR HIGHER EDUCATION 161
with firms;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 fields, 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 firms 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 firms 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
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 STE WARDSHIP FOR HIGHER EDUCATION 165
Technology Transfer Initiative (TTI) in Ire- by providing matching funds to firms 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 firms 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 financial 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 firms 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 firms 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 specific
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.
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 con-
tact between expertise in the participating universi- Source: Technology Transfer Initiative; see http://www.biotechnologyireland
ties and local industry. .com/pooled/profiles/BF_COMP/view.asp?Q=BF_COMP_9249.
166 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
Knowledge-integrating communityb
A knowledge-integrating community supported a. T. Kodama 2008.
by the Cambridge–MIT Institute (a collaboration b. Acworth 2008.
20
in Germany. This list of receiving countries is
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
0
Asian students.77 United United Australia Japan Germany
Prog ra m mobi l it y (wh ich i nvolve s 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 uni-
versities 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH IN EAST ASIA
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 difficult 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 par-
ing and research skills. These opportunities ticularly active in cross-border education dur-
can help address the first 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 STE WARDSHIP 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 profitable, 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 coun-
tries are less likely to return than foreign- ***
trained students from high-income countries. The previous parts of this study docu-
mented some of the main constraints that
Moving forward East Asia’s higher education systems are fac-
ing. 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
identified 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 inefficiencies 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 efficiency 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 PUT TING HIGHER EDUCATION TO WORK : SKILLS AND RESEARCH FOR GROW TH 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.
21. Levy 2010.
Two steps forward, one step back. East
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 fiscal framework for PHE (PHE promo-
middle-income economies are struggling to tion laws and decrees, PHE growth targets,
replicate the standard bearers’ success. These fiscal 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,
by lack of systematic and fully comparable
Notes 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 officially 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 STE WARDSHIP 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 five universities in the
50. These are typically business, finance, lan- United States derive significant income from
guages, and various social sciences. licensing of research findings 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, firm, 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 firms, 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
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
Appendix B
Economic Indicators
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
percent
percent
10 20
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
percent
30
percent
4 20
2 10
0 0
1988 1994 2001 2006 1988 1994 2001 2006
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
percent
percent
15
15
10 10
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
percent
15
percent
10
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
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).
TRENDS IN RETURNS TO SKILL AND SHARE OF SKILLED WORKERS, BY SECTOR 181
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
percent
25
percent
4 20
15
2 10
5
0
1999 2005 0
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
percent
percent
6
30
4
20
2 10
0 0
1998 2002 2007 1998 2002 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).
182 APPENDIX D
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
% change
–1
–2
–3
–4
–5
–6
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
industry code
change in industry premium change in skill premium
FIGURE D.2 Change in unstandardized weighted industry and skill premiums by sector, Philippines,
1988–2006
40
30
20
10
% change
–10
–20
–30
–40
–50
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
industry code
change in industry premium change in skill premium
FIGURE D.3 Change in unstandardized weighted industry and skill premiums by sector, Indonesia,
1990–2004
15
10
0
% change
–5
–10
–15
–20
–25
1 2 3 4 5 6 7 8
industry code
change in industry premium change in skill premium
FIGURE D.4 Change in unstandardized weighted industry and skill premiums by sector, Vietnam,
1992–2006
10
6
% change
–2
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
industry code
change in industry premium change in skill premium
FIGURE D.5 Change in unstandardized weighted industry and skill premiums by sector, Cambodia,
1997–2007
15
10
5
% change
–5
–10
1 2 3 4 5 6 7 8 9
industry code
change in industry premium change in skill premium
FIGURE D.6 Change in unstandardized weighted industry and skill premiums by sector, China,
1995–2005
10
0
% change
–5
–10
–15
–20
1 2 3 4 5 6 7 8
industry code
change in industry premium change in skill premium
FIGURE D.7 Change in unstandardized weighted industry and skill premiums by sector, Mongolia,
1998–2007
15
10
5
% change
–5
–10
1 2 3 4 5 6 7 8
industry code
change in industry premium change in skill premium
Appendix E
Openness, Technology, and Demand for Tertiary
Graduates, Regression Tables
TABLE E.1 Openness, technological innovation, and the demand for skills
TABLE E.2 Openness, technological innovation, and the demand for skills: Robustness to technological
variables
TABLE E.3 Openness, technological innovation, and the demand for skills: Robustness to alternative samples
TABLE E.4 Openness, technological innovation, and the demand for skills: Robustness to alternative
samples
Appendix F
Demand for Job-Specific and Generic Skills in East Asia
FIGURE F.1 Key job-specific 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
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
0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80
percent percent
managers skilled workers managers skilled workers
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
experience
loyalty
technical skills
ethnic quotas
interpersonal
skills
language
experience proficiency
othera
0 5 10 15 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
0 5 10 15 20 25 30 0 10 20 30 40 50 60 70 80 90
percent percent
analytical skills
decision-making skills
experience
work attitude
communication
technical skills
education
computer skills
0 10 20 30 40 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)
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
time management 4
adaptability 2
caring for others 80 creativity/innovation 5
numerical skills 11
problem solving 6
technical/professional skills 11
0 20 40 60 80 100 0 5 10 15 20 25 30
percent percent
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
0 20 40 60 80 100 0 2 4 6 8 10 12 14 16 18 20
percent percent
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.2 Determinants of managerial perceptions of skills and education of East Asian workforce
TABLE G.4 Determinants of time to fill skilled vacancies: Robustness to additional variables
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 fixed effects? Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Country fixed effects? 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 6,244
TABLE G.6 Determinants of managerial perceptions of skills of workforce: Robustness to different geographic areas,
by income levels
TABLE G.7 Determinants of time to fill skilled vacancies: Robustness to different geographical areas
TABLE G.8 Determinants of time to fill skilled vacancies: Robustness to different 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 firm-level controls included? Yes Yes Yes Yes Yes
Industry fixed effects? Yes Yes Yes Yes Yes
Country fixed effects? 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 firm fills skilled vacancies. Table reports the marginal effects (at mean values) on the firm’s propensity to
fill skilled vacancies from probit regressions. Microfirms (with fewer than 10 employees) is the omitted size group.
Significance 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
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
0 5 10 15 20 25 30 35 40 0 5 10 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 identified by firms manufacturing in Malaysia identified by firms
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 10 20 30 40 50 60
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 1.0
1 1.5 2.02 2.5 3.0 3 3.5 4.0 4 4.5 0 5 10 15 20 25 30 35 40
index share of firms (%)
skilled worker
skilled worker secondary
secondarytechnical college
technical college
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 difficulties.
202 APPENDIX I
Appendix I
Comparative Demand- and Supply-Side Indicators
Appendix J
Skill Gaps in East Asia
math theory
computer skills
grades
writing
secondary diploma
literacy
0 2 4 6 8 10 12 14 0 5 10 15 20 25
percent percent
thinking skills
risk taking and initiative
communication
behavioral
negotiation
organization
English
leadership
0 5 10 15 20 25 30 35 40 45 50 0 5 10 15 20 25 30 35 40 45
percent percent
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
adaptability adaptability
0 2 4 6 8 10 12 14 0 10 20 30 40 50 60 70
percent percent
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
numerical skills
technical skills
adaptability
education
problem solving
0 10 20 30 40 50 60 70 80 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
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
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
adaptability adaptability
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
a. Managers b. Production/sales
0 5 10 15 0 5 10 15
percent percent
Source: di Gropello, Tan, and Tandon 2010, based on 2008 Philippines Skills Survey.
a. Managers b. Production/sales
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
50 47 48
44
41
40 38
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.
50 49
44
41 40
40 38 38 39 37
35 36
33 32
30
30 28 27
percent
26
20
10
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 field
(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 Classification of Education (1997) 6 level. S&E data do not include health fields.
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 pro-
jected intake, enrollment, and flow rates for an individual country’s higher education system
(figure 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 (teach-
ing staff, equipment, infrastructure, and so on), provide estimates of future requirements
of teachers, nonteaching staff, instructional materials, and educational facilities. These pro-
jected requirements, together with cost-related data and parameters, can give an estimate of
fi nancial requirements and possible fi nancing gaps.
flow direction
Designed in this way, the model can provide simulations of both capital and recurrent expen-
ditures over time. Total costs are given by the following equation
Cdt = RCdt + Idt
where
t = year
d = level of education
Cdt = total costs
RCdt = total recurrent costs, and
Idt = investment.
CM = Edt × CMPSd
t
d
where
CTdt = teachers’ costs
CMdt = costs of materials
CAdt = administrative costs
COdt = other costs
Tdijt = 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.
where
CBPSd = per pupil building cost,
Edt = enrollment, and
a = replacement rate of buildings.
References
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.
BDLI N K 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 Pacific 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, Scientific 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 Benefits 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 Proficiency, 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 Efficiency 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 / IN TEASTASIAPACIFIC/ 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 Office) [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.
Office 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 Office. 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 Office. 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 Office 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. L ederman, D., and W. F. Maloney. 20 03.
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- ———. 2008b. “Private Higher Education’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 E ducation 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. Qualified 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
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 Benefits 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 / I N T EASTASI A 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, Scientific ———. 2005. “Project Appraisal Document:
and Cultural Organization). 2005. Education Managing Higher Education for Relevance and
Policy and Strategy Simulation Model EPSSim Efficiency 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
tific and Cultural Organization) Asia Pacific Skills for Growth. Washington, DC: World
Quality Network. 2006. Regulating the Bank.
222 REFERENCES
———. 2009a. 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-
———. 2009b. 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 Pacific 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 & Littlefield
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 tific 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 informa-
tion, visit www.greenpressinitiative.org.
I
mpressive gains have been made in expanding access to higher education
in East Asia over the past few decades, and the significance of higher
education 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.
This book is the first title in the East Asia and Pacific 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