Report No: ACS3321




People's Republic of China
Developing Skills for Economic Transformation and
Social Harmony
Yunnan Province



May 30, 2013




EASHE


EAST ASIA AND PACIFIC
Standard Disclaimer:


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


Copyright Statement:


The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of
applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will
normally grant permission to reproduce portions of the work promptly.

For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center,
Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/.

All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H
Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org.
iii



                                                              Contents

Acknowledgments ............................................................................................................. viii
Background to This Study .................................................................................................... x
Executive Summary ..............................................................................................................1
Chapter 1         Introduction ...................................................................................................13
                  Context of Yunnan ........................................................................................14
                  Definition and Measurement of Skills ..........................................................16
                  Conceptual Framework .................................................................................17
Chapter 2         Skills Challenges: Demand, Gaps, and Mismatches .....................................20
                  Labor Market Trends in China and Yunnan .................................................20
                  Skills Composition of the Current Workforce ..............................................26
                  Trends in Returns to Education.....................................................................32
                  Wage Returns to Skills..................................................................................37
                  Demand for Skills: Evidence from STEP Employer Survey ........................38
                  Skills Gaps and Mismatches .........................................................................42
                  Drivers of Future Demand for Skills ............................................................46
                  Summary .......................................................................................................51
Chapter 3         Overview of Yunnan’s Education and Training System ..............................55
                  Formal Education System .............................................................................55
                  Non-formal Training .....................................................................................58
Chapter 4         Strengthening Technical and Vocational Education and Training ...............61
                  Legal and Policy Framework ........................................................................61
                  Governance and Management.......................................................................62
                  School-Industry Links ...................................................................................64
                  Quality Assurance .........................................................................................67
                  Curriculum Reform .......................................................................................74
                  Employment and Pathways ...........................................................................77
                  Financing Formal TVET ...............................................................................79
                  Summary .......................................................................................................92
Chapter 5         Non-formal Training .....................................................................................96
                  Rural Work Force and Surplus Labor Transfer ............................................96
                  Rural-to-Urban Migration and Employment: Policy Evolution ...................97
iv


                   Demand for and Supply of Rural Worker Training ......................................98
                   Managing and Financing Rural Worker Training Programs ......................100
                   Content and Delivery of Rural Labor Training...........................................101
                   Impact of Rural Training: Evidence from Honghe Prefecture....................102
                   Summary and Policy Implications ..............................................................105
Chapter 6          Work-Based Training in Urban Yunnan .....................................................108
                   Demand for and Supply of Work-Based Training in Urban Yunnan .........108
                   Policy Framework .......................................................................................109
                   Participation in Work-Based Training ........................................................109
                   Financial Input in Work-Based Training ....................................................113
                   Training Implementation in SMEs..............................................................115
                   Summary and Policy Implications ..............................................................116
Chapter 7          Developing Cognitive Skills in Schools: Lessons from PISA Shanghai 2009121
                   PISA and Skills ...........................................................................................121
                   PISA Shanghai 2009 ...................................................................................122
                   Comparing School Characteristics ..............................................................127
                   School Characteristics and Student Performance .......................................127
                   Lessons from Shanghai ...............................................................................133
Chapter 8          Conclusions and Policy Implications ..........................................................136
                   Implications for Cross-Subsector Issues .....................................................138
                   Implications for Subsector-Specific Issues .................................................140
Appendix A. Returns to Education and Skills ................................................................143
Appendix B. Education and Training System in Yunnan ................................................154
Appendix C. Honghe Rural Labor Training Impact Evaluation....................................158
Appendix D. PISA Shanghai ...........................................................................................161


Figures

Figure 1.1 Contribution of Tertiary, Secondary, and Primary Sector to GDP Growth in Yunnan,
1957–2010..................................................................................................................................... 15
Figure 1.2. Conceptual Framework for Study of Skills Demand and Supply in Yunnan............ 18
Figure 2.1. Composition of Employed Persons at Year-End by Sector in Yunnan and China,
1980–2010..................................................................................................................................... 20
Figure 2.2. Quarterly Number of Job Vacancies by Sector, Yunnan, 2010–11 .......................... 21
Figure 2.3. Workers in Urban Areas by Ownership, Yunnan ....................................................... 23
Figure 2.4. Trends in real wages (Index 1994=100) of China and Yunnan ................................ 23
Figure 2.5. Real wages (Index 1994=100) by types of ownership ............................................... 24
v


Figure 2.6. Labor productivity of selected countries ................................................................... 25
Figure 2.7. Labor productivity of Yunnan compared to selected provinces ................................ 25
Figure 2.8. Education Composition of the Employed, Yunnan and China, 2001–10.................. 26
Figure 2.9. Proportion of Total Employed with Upper Secondary or Tertiary Education in China
by Province, 2010 ......................................................................................................................... 27
Figure 2.10. Education Composition of Male and Female Workers in Yunnan, 2010................ 27
Figure 2.11. Educational Attainment by Human Resources Type, 2008..................................... 29
Figure 2.12. Education Composition in Non-public Enterprises by Industry, 2008.................... 30
Figure 2.13. Share of Certified Workers by Type and Level, 2008 ............................................. 30
Figure 2.14. Share of Certified Workers at Medium and Senior Level in Non-public Enterprises,
2008............................................................................................................................................... 31
Figure 2.15. Percent Change in Wages Relative to Middle School Graduates, Yunnan and China,
1998–2012..................................................................................................................................... 35
Figure 2.16. Changes in Marginal Returns to Education in the 1990s in Latin America ............ 36
Figure 2.17. Number of Workers Twelve Months Ago, Currently, and Expected in Twelve
Months .......................................................................................................................................... 39
Figure 2.18. Percent of Firms Reporting Various Job-Related Skills as Important to Employee
Retention ....................................................................................................................................... 40
Figure 2.19. Percent of Firms Reporting Various Personality Traits as Important to Employee
Retention ....................................................................................................................................... 41
Figure 2.20. Percent of Firms Reporting Various Factors as Important to Employee Retention 42
Figure 2.21. Firms Facing Factors Constraining Growth in Yunnan and China, 2005 ............... 43
Figure 2.22. Rankings of Labor Factors Impeding Business Operation and Growth .................. 43
Figure 2.23. Percent of Employees Considered Qualified for their Jobs..................................... 44
Figure 2.24. Quarterly Job Vacancies and Applications in Yunnan, 2010–11 ............................. 44
Figure 2.25. Percent of Firms Encountering Problems Hiring Employees at Various Levels .... 45
Figure 2.26. Real GDP and Number of People Employed in Yunnan, Real and Projected, 1957 –
2015............................................................................................................................................... 48
Figure 2.27. Real GDP and Employment by Sector, 1957–2008 ................................................ 48
Figure 3.1. Structure of the Formal Education System, Yunnan ................................................. 56
Figure 4.1. Horizontal Arrangements of TVET, Yunnan ............................................................ 62
Figure 4.2. Vertical Arrangements of TVET, Yunnan ................................................................ 63
Figure 4.3. Pathways Taken by Secondary TVET Graduates in Yunnan, 2008 .......................... 77
Figure 4.4. Public Spending on Education as a Share of Total Government Spending, 2010 .... 81
Figure 4.5. International Comparison of Yunnan’s Public Expenditure on Education, 2009 ..... 81
Figure 4.6. Budgetary Recurrent Expenditure per Student by Level of Education in Yunnan,
2010............................................................................................................................................... 83
Figure 4.7. Budgetary Recurrent Expenditure per Student on Secondary TVET, 2009 .............. 83
Figure 4.8. Budgetary Recurrent Expenditure per Student in Yunnan, 2009 .............................. 84
Figure 4.9. Total Budgetary and Nonbudgetary Expenditure by Level of Schooling in Yunnan,
2009............................................................................................................................................... 85
Figure 4.10. Household Income and Educational Expenditure in Yunnan, 2010........................ 86
Figure 4.11. Perceptions of Financial Burden Posed by Education Expenditures in Yunnan, 2010
....................................................................................................................................................... 87
Figure 4.12. Gross Enrollment Ratio in Yunnan at Upper Secondary Level by City and
Prefecture, 2010 ............................................................................................................................ 88
vi


Figure 4.13. Student-Teacher Ratio Compared to Recurrent Expenditure per Student in Yunnan,
2010............................................................................................................................................... 90
Figure 4.14. Operating Conditions of Secondary TVET Institutions in Yunnan, 2010 .............. 91
Figure 5.1. Number of Workers in Rural Areas and Share of Total Employment in Yunnan,
1978–2010..................................................................................................................................... 97
Figure 5.2. Demand for and Supply of Rural Labor Training ..................................................... 99
Figure 6.1. Demand for and Supply of Work-Based Training in Urban Yunnan ...................... 108
Figure 6.2. Percentage of Employees Receiving Training, China and Various Cities .............. 110
Figure 6.3. Provision and Duration of Training on the Premises of the Workplace, Kunming 112
Figure 6.4. Percent of Workers Receiving Employer-Organized Formal Training Outside the
Workplace ................................................................................................................................... 113
Figure 6.5. Percent of Wage Bill and Compensation Fund Spent by Firms on Outside Training
Providers ..................................................................................................................................... 114
Figure 6.6. Average Amount Spent per Employee on External Training Providers ................. 115
Figure 7.1. University Participation at Age 21 by PISA Reading Proficiency at Age 15 in
Canada, 2009............................................................................................................................... 122
Figure 7.2. PISA Reading Performance of Shanghai and OECD Countries, 2009 ................... 123
Figure 7.3. Proportion of Resilient Achievers and Disadvantaged Low Achievers .................. 125
Figure 7.4. Per Student Expenditure at the Pre-primary Level, Yunnan ................................... 134
Figure 7.5. Enrollment Rates in Pre-primary Schools, Yunnan ................................................ 134


Tables

Table 2.1 Total Number of Staff and Workers at Year-End in Yunnan by Industry, 2007 and
2010............................................................................................................................................... 22
Table 2.2. Number of People Employed in Non-public Enterprises, by Industry and Occupation,
2008............................................................................................................................................... 28
Table 2.3. Rates of Return to Schooling, Yunnan and China, 1988–2012 .................................. 34
Table 2.4. Percent of Firms Encountering Hiring Problems, Type A Occupations .................... 46
Table 2.5. Percent of Firms Encountering Hiring Problems, Type B Occupations ..................... 46
Table 2.6. Selected Goals Specified in Yunnan’s 12th Five-Year Plan ...................................... 47
Table 2.7. Projected Demand for Skills in Yunnan, 2010–20 ..................................................... 51
Table 3.1. Impressions of the TVET and General Education Systems, Yunnan ......................... 58
Table 4.1. Extent of School-Enterprise Collaboration, Kunming, Yunnan ................................. 65
Table 4.2. Standards for the Setup of Secondary and Tertiary TVET Institutions, Yunnan ....... 68
Table 4.3. Selected Teacher Qualifications and Standards for Secondary and Tertiary TVET
Institutions..................................................................................................................................... 69
Table 4.4. Provinces Meeting the Establishment Standards for Secondary TVET by Region,
2009............................................................................................................................................... 70
Table 4.5. Provinces Meeting the Establishment Standards for Tertiary TVET by Region, 2009
....................................................................................................................................................... 71
Table 4.6. Comparison of ISCO-08 and China’s Occupational Classifications by Major Groups
....................................................................................................................................................... 72
Table 4.7. Year-End Employment Rates in Yunnan, 2010.......................................................... 78
Table 4.8. Private Provision of Education in Yunnan, 2010 ....................................................... 85
vii


Table 4.9. Comparison of Enrollment Rates and Advancement Rates in Yunnan and China, 2010
....................................................................................................................................................... 87
Table 4.10. Selected Standards for Secondary TVET Institutions in Yunnan, 2010................... 88
Table 4.11. Selected Standards for Tertiary TVET Institutions in Yunnan, 2010....................... 89
Table 5.1. Major Training Programs for Rural Workers in Yunnan ......................................... 100
Table 5.2. Content and Provision of Rural Worker Training Programs in Yunnan .................. 101
Table 5.3. Training in Rural Honghe ......................................................................................... 103
Table 5.4. Estimated Effects of Training on Net Income per Person, 2007 and 2008 .............. 105
Table 7.1. Mean PISA Scores, Shanghai and OECD, 2009 ...................................................... 123
Table 7.2. PISA Scores of Schools in Shanghai and OECD Average ....................................... 126
Table 7.3. Characteristics and Performance of Middle Schools ................................................ 128
Table 7.4. Characteristics and Performance of General High Schools ...................................... 130
Table 7.5. Characteristics and Performance of Vocational Schools .......................................... 131


Boxes

Box 2.1. Yunnan Non-public Enterprise Survey, 2008 ............................................................... 28
Box 2.2. CHIP 1988, 1995, 2002 and RUMiC 2007 ................................................................... 33
Box 2.3. STEP Employer Survey 2012 ....................................................................................... 38
Box 3.1. Government Subsidies for Non-formal Training in Yunnan ........................................ 59
Box 4.1. Overcoming Fragmented Governance of TVET: Xinjiang and Shanghai .................... 63
Box 4.2. APINDO, the Employer Association of Indonesia ....................................................... 64
Box 4.3. Ningbo's Regulation for Promoting TVET School-Industry Collaboration ................. 66
Box 4.4. Vocational Education Groups in Yunnan...................................................................... 67
Box 4.5. Turkey: Involving Industry and Training Providers in the Development of the National
Qualification Framework (NQF) .................................................................................................. 73
Box 4.6. Hong Kong SAR, China’s Qualification Framework ................................................... 74
Box 4.7. Main Results and Outputs of the Implementation of TVET Curriculum Reform in
China ............................................................................................................................................. 75
Box 5.1. Yunnan’s Policies on Rural Worker Training, 2004–08 ............................................... 98
Box 5.2. The Honghe Rural Household Survey and Analytic Sample ...................................... 103
Box 5.3. Consolidating Funds for Non-formal Training ........................................................... 106
Box 6.1. The Korean Employment Insurance System ............................................................... 117
Box 6.2. Singapore’s Skills Development Fund (SDF) ............................................................. 119
Box 6.3. Providing Institutional and Technical Assistance to SMEs: Korea’s SME Training
Consortium.................................................................................................................................. 119
Box 7.1. Definitions of the PISA Domains................................................................................ 121
Box 7.2. PISA Shanghai 2009 ................................................................................................... 126
viii



                                Acknowledgments

The design and preparation of the study was led by Xiaoyan Liang (Task Team Leader)
under the guidance of Xiaoqing Yu (EAP HD Director) and Luis Benveniste (EAP Sector
Manager). The report was co-authored by Xiaoyan Liang and Shuang Chen (Consultant),
with significant contributions from a few researchers who partnered with us to carry out
background studies:

Prof. Xingxu Li of the Yunnan University of Finance and Economics conducted the
initial analysis and background study on the Honghe Rural Training Impact Evaluation.
He has also given us access to Yunnan’s 2008 Survey of Nonstate Enterprises. Under the
leadership of Prof. Li, the Yunnan University of Finance and Economics was contracted
to implement the STEP Household Survey and the STEP Employer Survey. Their
excellent work in the field has provided this report with the most up-to-date and first-
hand data.

The Shanghai Academy of Educational Services under the leadership of Director
Shuchao Ma, Vice-Director Yang Guo, and Mr. Chen Zhang, conducted a review of the
implementation and outcomes of the TVET curriculum reforms in Guangdong, Liaoning,
Shandong, Yunnan, and Shanghai. The Academy also provided us with data on the
financing, standards and status of TVET.

Prof. Linlin Pu, Prof. Xing Dong, and Dr. Fang Lu from the Yunnan National University
were the principal investigators for the SABER-WfD benchmarking study. Their input in
the data collection instrument has provided critical information for analyzing Yunnan’s
TVET system.

Dr. Zhenyi Guo conducted a background study on work-based training and rural labor
training and allowed us to use her related research as an input for the work-based training
chapter.

Mr. Bo Yun, Mr. Haijun Liu, and Ms. Yaqin Li, from the Yunnan Academy of
Educational Sciences, prepared a background report on the status of TVET in Yunnan.

Further, the report benefitted from upstream review by Prof. Simpson Poon from the
Technological and Higher Education Institute of Hong Kong and the Hong Kong
Vocational Training Council, and Prof. Sung Joon Paik from the Korea Development
Institute School of Public Policy and Management. Within the Bank, peer reviewers,
including Jee-peng Tan (Advisor), Mamta Murti (Advisor), and Andreas Blom (Lead
Education Specialist, Africa Region), have provided valuable comments. Former EAP
Sector Manager Eduardo Bustillo Velez has also provided constructive feedback.
Emanuela Di Gropello (Sector Leader, AFTHD) advised the team on the initial
conceptual framework. Dewen Wang (Social Protection Economist EASHS) offered
valuable input on non-formal training and training funds. Harry Patrinos (Lead
Economist, HDNED) gave comments on the initial analysis of the PISA data.
ix


The study benefitted from a multidonor trust fund for the data collection and analysis of
STEP Household and Employer Surveys. The HDNSP and HDNED teams, led by Maria
Laura Sanchez-Puerta and Alexandria Valerio, provided excellent technical assistance
and other support both to the Bank team and the local implementers.

Finally, the study would not have been possible without the support and collaboration
from the officials in the Yunnan Provincial Department of Education who had
demonstrated ample commitment and professionalism. Officials and colleagues in the
Yunnan Department of Human Resources and Social Security and Poverty Alleviation
Office, and Provincial Research Institute also provided valuable inputs.
x


                           Background to This Study

The World Bank has a long history of investing in China’s technical and vocational
education and training (TVET), dating back to the 1963 Higher Education Project, which
benefited selected tertiary TVET programs in project schools. The initial World Bank
projects for TVET in China were designed mostly on the national level, supporting
selected institutions in multiple provinces. Examples of such projects include the
Vocational Education Project I, the Vocational Education Project II, the Labor Market
Development Project, and the Migrant Training Project.

Starting in 2006, based on strong analytical work and increased demand from various
provinces, the World Bank began a direct policy dialogue with interested provinces, and
in turn started supporting provincial level projects in TVET. These include the
Guangdong Technical and Vocational Education Project and the Liaoning and Shandong
Technical and Vocational Education Project, both of which are currently under
implementation. Progress to date indicates that sub-national lending tends to result in
more focused and efficient policy dialogue and project implementation, especially when a
good national policy framework is in place and provincial “good practices� are sought. In
particular, the World Bank is able to fill an important gap by introducing international
experience and innovation in school-based reforms, spreading such reforms within and
outside the project provinces, supporting learning activities to capture lessons from
reforms, and informing government policies and investments in the sector.

In 2010, through the Ministry of Finance and the National Development and Reform
Commission, the World Bank received a project proposal from the Yunnan Department
of Education for a US$50 million loan to support the development of TVET in the
province. Both the provincial counterparts and the World Bank team felt the need to first
develop a common knowledge base to help design the lending operation. As a result, the
Yunnan Department of Education requested that the World Bank conduct a background
study for the Yunnan Technical and Vocational Education and Training Project. As a
result of consultation with their counterparts and within the World Bank, the scope of the
study was broadened beyond a review of the formal TVET system to include skills-
demand analysis and a review of the work-based and non-formal training systems.

The study is aimed at facilitating policy development leading to a demand-driven, high-
quality, and equitable education and training system conducive to lifelong learning. The
experience of Yunnan also sheds light on skills development in China, as the analysis in
this report situates Yunnan in the broader national context. The intended audience
includes policymakers in the Ministries of Education, Human Resources and Social
Security, Agriculture, and Finance, and the Poverty Alleviation Office and National
Development and Reform Commission. The report would also be of interest to
researchers and development workers interested in understanding skills development in
China.
xi


Abbreviations

      ATC       Average treatment effect on the controls

      ATE       Average treatment effect

      ATT       Average treatment effect on the treated

      CHIP      China Household Income Project

      GDP       Gross Domestic Product

      GER       Gross Enrolment Ratio

      HEI       Higher Education Institution

      HRSS      Human Resources and Social Security

      ICA       Investment Climate Assessment

      NQF       National Qualification Framework

      OECD      Organisation for Economic Co-operation and
                Development

      PISA      Programme for International Student Assessment

      R&D       Research and Development

      RUMiC     Longitudinal Survey on Rural Urban Migration in China

      SAR       Special Administrative Region

      SDF       Skill Development Fund

      SMEs      Small and medium-sized enterprises

      SOEs      State-owned enterprise

      STEP      Skills Toward Employment and Productivity

      TVET      Technical and Vocational Education and Training

      UNDP      United Nations Development Programme
1



                                Executive Summary

China has achieved impressive growth over the last three decades and has now become
the second-largest economy in the world. To sustain its growth, China is transitioning
from an investment-led, high-carbon growth model to a consumption-led, green growth
model, less reliant on low-cost manufacturing and more on technology and innovation.
Skills development has been a key factor enabling China’s unprecedented growth, and
will continue to play a vital role in sustaining its ongoing economic transformation and
pursuit of a harmonious society.

Located on the southwestern border of China, Yunnan is a medium-sized Chinese
province with abundant natural resources and high levels of ethnic diversity. Although
Yunnan is still one of the poorest provinces in China, it has experienced rapid economic
growth rates over the last decade and is expected to maintain an annual growth rate of 10
percent or higher. The recent national Bridgehead Strategy has further positioned Yunnan
as a strategic gateway in the Southwest region, providing tremendous new opportunities
for its development. Is Yunnan ready to meet the skills challenges posed by rapid
economic growth? What critical skills are needed for workers to help achieve Yunnan’s
economic and social transformation? Is Yunnan’s current education and training system
capable of supplying sufficient and relevant skills demanded by the economic and social
transformation?

This report consists of three parts. The first part, Skills Challenges: Demand, Gaps, and
Mismatch, examines the sources of the mismatch of supply and demand for skilled labor
in Yunnan. The second part, Education and Training, sets out the challenges facing
Yunnan as it seeks to strengthen the Technical and Vocational Training and Education
(TVET) system, improve access to education and the quality of educational outputs,
encourage more robust work-based and rural training, and invest additional resources in
general education. Finally, in Conclusions and Policy Implications, the report lays out
specific policy proposals that cross both sector lines and address sector-specific issues.

Skills Challenges: Demand, Gaps and Mismatches
Yunnan is primarily an agricultural economy and rural society. Currently over 60 percent
of workers are still engaged in the agricultural sector, compared to the national average of
37 percent. Over the last decade, however, the share of employment in the secondary and
tertiary sectors has been on the rise. The tertiary sector has seen the fastest growth in
employment. Within the secondary sector, employment in mining and construction has
also maintained strong growth. Further, after three decades of opening up and attendant
economic development, the non-public sector now generates almost half of total urban
employment, as the share of employment in state-owned enterprises and collectives has
dropped significantly.

The labor force in Yunnan generally possesses low levels of education and skills. The
majority of the workers have completed only nine years of compulsory education or less.
2


Skills certification rates are also low among professional and technical workers and
operative frontline workers. In non-public enterprises, only about 24 percent of the
professional and technical workers and 12 percent of operative frontline workers have
obtained skills certification or professional ranking. Consequently, Yunnan workers have
on average lower wages and productivity than the national average.

Even with rapid large-scale education expansion and reform, average rates of return to
education have risen from 3 percent in 1988 to 7 percent in 2002. The most dramatic
increase has occurred at the tertiary level. Bachelor’s degree holders earned 79 percent
more than middle school graduates in 2002, and the gap is even larger among workers in
urban Kunming as of 2012. Gaps in returns have been widening between tertiary and
secondary education and between Bachelor’s degrees and associate degrees. The trend is
similar to patterns found across China, indicating strong and growing demand for
workers with tertiary education. Although returns to secondary TVET were historically
higher than those to general high school, at the upper secondary level, our data suggest
that the gaps are narrowing in recent years, implying that the secondary high school
degree and TVET degree are less distinguished in Yunnan’s labor market.

Selected cognitive, technical, and non-cognitive skills are associated with higher wages
even after controlling for educational attainment. In particular, workers who perform at
higher levels of reading and writing tend to have higher earnings. Their jobs tend to be
more cognitively challenging and autonomous. Computer and foreign language skills are
associated with higher wages across and within occupations and sectors. In addition,
workers who are more creative and open to new challenges also earn significantly more
even after controlling for their technical and cognitive skills and educational attainment.

From the employers’ perspective, high-skilled professionals, technicians and operative
workers are in high demand. In deciding which employees to retain after a probation
period, employers consider non-cognitive skills such as “communication skills� and
“leadership skills� along with job-specific technical skills to be the most important.

The latest policy developments, as laid out in Yunnan’s 12th Five-Year Plan, are
expected to further drive the demand for skills. Projected economic growth, together with
the structural adjustment within and across sectors, will create more demand for workers
in the secondary and tertiary sectors. More importantly, the successful implementation of
the Bridgehead Strategy, which would help Yunnan further accelerate its growth and
open up to neighboring countries and provinces, relies on an abundant supply of highly-
skilled professionals and technical workers.

Faced with an increasing demand for skills, especially for professional and technical
workers and operative frontline workers, the stock of skills among Yunnan’s current
labor force appears insufficient. Skills gaps have become more apparent. Throughout
2010 and 2011, only 35 to 45 percent of the job openings in Yunnan were filled, and only
50 to 55 percent of job applicants were hired. The vast majority of firms surveyed in
Kunming considered “difficulties with finding experienced workers� and the “technical
and vocational education and training of workers� to be the most problematic labor
factors inhibiting business growth. The lack of required skills has been the most
3


commonly cited problem in hiring for both highly-skilled professionals and technical
occupations, as well as operative frontline positions. In addition, the mismatch of wage
expectations poses a major issue for firms hiring high-skilled managers, professionals,
and technicians. For construction workers and operative frontline jobs, lack of applicants
is an issue.

Education and Training
Yunnan has a more than sufficient supply of labor, and yet the large population does not
possess a high stock of skills. As the shortage of high-skilled workers becomes more
apparent, Yunnan also faces a surplus of low-skilled laborers, especially among the 65
percent of its population still residing in rural areas. An imbalance in the demand and
supply of skills exists in almost all country contexts. It is imperative to develop an
education and training system that is responsive to the ever-changing demand from the
economy and society at large.

This report identifies challenges associated with the need to improve the four major areas
of the education and training system: (i) the formal TVET system; (ii) non-formal
training; (iii) work-based training; and (iv) general school education.

Challenges to Strengthening the Formal TVET System
Governance and Management: The legal framework for the TVET sector in Yunnan
adheres closely to Chinese national laws and regulations. A relatively complete set of
legal instruments ensures education and training rights for all citizens and provides
direction for governance and management, quality assurance arrangements, financing of
TVET, and collaborating with enterprises.

The Ministries of Education and Human Resources and Social Security are the two
departments substantially in charge of the management of all TVET, with the former
responsible for all secondary and tertiary educational institutions and the latter
responsible for vocational training institutions. The Ministry of Education considers itself
the overall leader for coordinating system reform in TVET. Interdepartmental TVET
Coordination Committees have been established at the national level and in a few
provinces in an effort to overcome fragmentation. In reality, these committees are
advisory with little executive power and membership is restricted to government officials.
Industry and social groups have little formal representation in the management of the
TVET system or in the running of public TVET schools.

School-industry links. The involvement of industry in the governance and management of
TVET institutions would provide school leaders with crucial support in developing
programs to provide the skills enterprises need. With industry participation, students
could gain relevant workplace skills before graduation. Currently, at the school level,
collaboration with industry is often ad hoc and dependent on personal connections.
Collaboration tends to be limited to recruiting graduates or providing internships and
training. Efforts to establish formal school-industry links through school-industry
collaboration advisory committees and sector-specific advisory committees have recently
started to emerge with support from the national model TVET school project and
4


development partners such as the World Bank. Very few enterprises, however, are
involved in the development of training programs.

Curriculum Reform: In recent years, schools have become more aware of the need to
develop programs that are aligned with the demands of the labor market and society and
have started to update their curriculums accordingly. A lack of up-to-date and industry-
led competency-based standards at the system level, however, hampers curriculum
reform at the school level. The development by individual schools with their own set of
standards for teaching may be a practical solution in the short term. However, in the long
run, Yunnan would be better served by a systematic province-wide approach to
developing competency-based standards with inputs from industry that schools can use as
the basis for curriculum development. Lack of resources at the school level has further
impeded the successful implementation of curriculum reform. Schools are constrained by
teacher capacity, high student-teacher ratios, and outdated facilities and equipment.

Quality Assurance: There is no independent accreditation agency for training providers.
Administrative departments of the Ministries of Education and Human Resources and
Social Security have recently updated the standards for establishing TVET institutions on
both the secondary and tertiary levels to reflect the government’s most recent talent
development plans. Although the standards are comprehensive and specific to each type
of institution, they reflect educational “inputs,� including physical space, equipment per
student, student-teacher ratios, and the proportion of instructors who obtain the double
qualification of teacher qualifications and technical skills certificates, rather than the
“outputs� of specific skills training and certifications. In reality, some institutions,
particularly secondary TVET institutions located in rural areas, continue to have
difficulty meeting the standards.

Skills Certification: An increasing number of students are required to obtain skills
certifications in addition to their academic diplomas. The occupational classification and
competency standards for each occupation, however, have not keep up with the evolving
demands of the market. Evidence shows that existing skills certificates do not carry much
weight among employers; indeed, the most recent quarterly job vacancy data show that
only about 20 percent of job openings in Yunnan require a skills certification. Only a
minority of existing workers in Yunnan have obtained certification.

Financing: Yunnan’s level of total public spending on education is high compared to
other provinces and even to some developed countries. Yunnan allocates over 6 percent
of its gross domestic product and over 19 percent of total government spending to
education as compared to less than 4 percent and about 16 percent, respectively,
nationwide. Yet, owing to the large number of students in its educational system and low
level of economic development, expenditure per student in Yunnan is lower than the
national average at every level of education. Overall resources appear insufficient.

There are huge disparities in education financing between rural and urban Yunnan. As
education finance is decentralized to local governments, including TVET at the
secondary level, schools in poor rural areas struggle to cover both personnel and
operating costs. Particularly hard hit are rural vocational high schools in Yunnan.
5


Province-wide, rural vocational high schools average a per pupil expenditure of Y2333,
which is less than two-thirds of the provincial average of Y3960. Special funds for TVET
from provincial and national governments are awarded to provide incentives and
motivation to improve quality and innovation. However, these funds tend to be allocated
based on the quality of the proposals submitted, which results in resource-rich schools
benefiting disproportionately more than resource-poor schools. Although competitively
awarded funds can be effective in promoting quality improvement in the long term, it is
important to consider equity-based financing in the short term to bring all schools to a
minimum level of inputs and standards. The rural vocational high schools across China
and Yunnan will need renewed policy direction and favorable financing from the
provincial and central governments in order to close the resource gap, and enable all to
participate on a level playing field with the rest of the country.

Private provision of TVET in Yunnan is emerging with 12 percent of secondary TVET
students and 25 percent of tertiary TVET students enrolling in private institutions.
Yunnan can explore expanding the private provision of training, particularly at the
tertiary level, that is, beyond its current level to mobilize additional resources and to have
a healthy dose of “competition� in the system. Across both public and private institutions,
the level of private spending on TVET in Yunnan remains lower than the national
average. In this context, private support could be more strongly mobilized.

Challenges to Improving Access and Quality of Non-formal Training
Non-formal education and training constitutes a significant safety net and provides
second chances for those who have left the formal TVET system. In Yunnan’s context, a
significant number of unskilled rural surplus laborers, accounting for a third of its rural
population, need to be trained and transferred to the secondary or tertiary sectors. Those
who remain in the farming sector must upgrade their skills to boost agricultural
productivity.

Both the central and provincial governments have introduced a number of training and
employment schemes targeted at rural surplus labor. Our quantitative evaluation using
propensity score matching has illustrated the positive and statistically significant effect of
those training programs: Rural households who participated in training in 2006 earned 26
percent more income per person in 2007 and 11 percent more income per person in 2008
than if they did not participate. On average, having household members attend training
increases the net income per person in 2007 and 2008 by over 20 percent.

Our analysis has shown the economic returns of rural labor training—in particular
highlighting that although training is given to individual household members, the effect
of training seems to spill over to the net income of the entire family. However,
implementation of rural labor training faces a number of challenges:

First, financial input in rural labor training appears insufficient, directly affecting the
quality of training delivery. For example, due to the relatively short duration of most
training programs, the skills acquired through training are rather limited. Guidance
training, the main form of rural worker training, covers only basic work- and living-
6


related orientation for rural workers before migration. Only a quarter of all the training
programs actually equip participants with practical skills.

Second, participation rates in rural training are low. Many rural laborers do not have the
minimum education level (usually completion of middle school) required by certain
training programs. Some are unaware of the value of training, and some are concerned
that the training might take too much of their time away from their paid agricultural
activities. Households in ethnic minority villages are also less likely to participate.

Third, given the low educational attainment of the rural training participants, training
materials should be tailored to their specific needs, and bridge the gaps between their lack
of preliminary knowledge and necessary skills.

Finally, because multiple agencies are involved in the management, financing, and
delivery of non-formal training, there is a lack of coordinated policy and implementation.
Regular monitoring and evaluation of the programs is also missing.

Challenges to Incentivizing and Supporting Firms to Provide Work-Based
Training
Enterprises play an important role in training and upgrading the skills of their employees.
Globally available evidence demonstrates that investment in work-based training is
correlated with higher firm productivity. Before the economic reform, work-based
training was provided by the government in state-owned enterprises. The government’s
role now, however, is more that of a regulator than a provider. Most recently, the Chinese
government stipulated that enterprises devote 1.5 to 2.5 percent of their total wage bill to
employee training. However, compliance with the law and the details of its
implementation are left very much in the hands of the enterprises themselves. As a result,
how and the extent to which work-based training is provided varies greatly.

Although survey results indicate that more than 90 percent of the firms in Yunnan and
China overall have provided training, the majority of enterprises in Yunnan have
provided training to fewer than 10 percent of their employees. Most training is conducted
in-house by firm managers, technical personnel or peers, or through on-the-job learning.
Very few firms have paid external trainers to conduct training.

Small and medium-sized enterprises find it particularly difficult to comply with the
requirement that 1.5 to 2.5 percent of payroll be spent on training. Financial constraints
aside, the capacity and willingness to provide employee training are major issues. It is
thus important for the government to assess the particular needs of these enterprises and
employ different incentive and subsidizing mechanisms, instead of relying solely on
compulsory requirements.

The relevance and quality of training are also called into question. Strategic planning for
upgrading skills, particularly among small and medium-sized enterprises, is lacking. Thus,
technical assistance, in addition to financial assistance, is needed to enable firms to
conduct training-needs assessments and develop training plans. A monitoring and
evaluation system also needs to be in place for quality assurance.
7


Challenges to Investing in General Education for the Development of Cognitive
Skills
Cognitive skills form the core part of skill sets required by the economy and society as a
whole. Drawing on 2009 data from the Shanghai Programme for International Student
Assessment (PISA), we investigated what school-level characteristics correlate with
higher cognitive skills among 15-year-olds, including mathematical, reading, and science
skills, as measured by PISA. We then examined Yunnan’s school systems in light of
lessons learned from Shanghai.

Shanghai’s experience shows that school resources matter. School resources include
instructional resources and facilities, as well as the extracurricular activities available to
students. They also include abundant teacher resources, high rates of good teacher
qualifications, and supportive student-teacher relations. Shanghai’s experience also
suggests that teacher involvement in decision making about a school’s resource allocation
and curriculum might be beneficial for both general high schools and vocational schools.

Lessons from Shanghai also show the benefits of pre-primary education on students’
subsequent cognitive skill development. Children who have attended pre-primary
schools for at least one year perform significantly better across all subjects on PISA.
Currently, however, participation in pre-primary education is limited in Yunnan, and
varies greatly across cities and prefectures. Investing in and popularizing pre-primary
education in Yunnan might be a cost-effective way of developing cognitive skills in early
youth.

Conclusions and Policy Implications
The Chinese TVET system is vast and complicated. Every reform effort, such as
curriculum development and teacher training, has enormous resource implications.
Accordingly, developing and implementing a set of centralized policies poses enormous
challenges, especially for a resource-constrained province like Yunnan. The analyses in
the report reveal certain policy gaps and challenges that have hampered Yunnan’s
education and training system from operating at its full potential. In light of global trends,
the report offers a broad set of policy recommendations.

A number of policy areas such as governance and coordination, industry participation,
financing, quality assurance, and monitoring and evaluation cut across sectors and pertain
not just to the formal TVET system, but also to non-formal education, and even to work-
based training. Specific recommendations for non-formal training, work-based training,
especially for small and medium enterprises, and pre-primary education follow.

Implications for Cross-Subsector Issues

1. Governance and Coordination
The effective functioning of all TVET subsectors could be improved by greater
coordination among the relevant stakeholders in government, education, and industry.
8


In the short- and medium-term, it is feasible to develop and strengthen the role of existing
coordination mechanisms, such as the creation of Inter-departmental TVET Coordination
Committees, in areas such as policy development, planning, and service delivery at both
the provincial and local levels. The operations of Yunnan’s committee could be
regularized and the committee’s functioning improved.

In the long-term, Yunnan might consider consolidating policy making, planning,
financing, and service delivery of TVET into a central government agency, such as a new
Skills Development Authority, or alternatively, into one of the existing ministries. This
would eliminate the distinction between schools governed by the Ministry of Education
and those governed by the Ministry of Human Resources and Social Security, as
Shanghai has already done for its secondary TVET institutions.

Consolidated governance and management may be even more pertinent for non-formal
training, which currently appears even more fragmented and less structured. The strongly
vested interests of various stakeholders in this sector may mean that consolidation will be
a long-term effort.

2. Industry Participation
Industry and employer involvement can play a key role in ensuring that the educational
and training system, especially at the TVET level, is responsive to the demand side of the
skills equation. Emerging industry participation has yet to be systematically incorporated
into the governance structure of any of the education and training sectors. Industry
involvement should be institutionalized and expanded to include a diverse range of issues
related to skills development, including setting skills development priorities, developing
competency standards for skills certification, allocating resources, and making policies
regarding work-based training. In many countries, employer or industry-based bodies
participate at different levels, providing advice, taking part in setting development
strategies, voicing the needs and demands of industry, and even exerting decision-making
power.

At the school level, Yunnan could introduce a legal framework and implementation
guidelines to promote school-industry collaboration. The framework and regulations
should establish the rights and obligations of employers, schools, teachers, students, and
all relevant stakeholders in the school-industry partnership.

Similar principles are applicable to non-formal education and work-based training. For
example, industry and employers could be further involved in managing work-based
training funds, designing training programs, and assessing training participants.

3. Quality Assurance
Under an effective qualifications framework, skills certifications are demanded by
employers and sought by students and employees. Certifications are accepted measures of
competency and substantive learning. A well-developed framework could address both
accreditation of providers and certification of learners.
9


The existing qualifications framework needs to be updated with standards and
competencies that reflect the demands of the labor market. Competency standards for
occupations and skills certification should be based on input from industry, employers,
education and training providers, and other stakeholders. An effective qualifications
framework can also facilitate recognition of non-formal learning, integrating different
kinds of non-formal education and training, and improving their quality and efficiency.
The qualifications framework should respect all forms of learning, that is, assessing
learners based on competencies rather than credentials, and should not disadvantage one
type of learning against another.

At the operational level, there are various important parameters for Yunnan to consider as
it develops its qualifications framework. For example, to what extent will Yunnan rely
on the existing national framework in developing one more specific to its provincial
context? How will the coordinating entity be chosen, and how will industry input be
systematically incorporated into the development of the framework and each of the
underlying standards? To what extent should the framework take into consideration
regional and even international standards? It might be most feasible to develop a
framework first in key industries with mature, organized private sectors before extending
it to all sectors and occupations. Good examples can be found in Turkey, Australia, New
Zealand, Hong Kong SAR, China, Singapore, and the European Union. Ultimately what
is most important is the development of a consensus specific to Yunnan regarding the
scope and depth of the framework and operational details.

4. Financing TVET
School resources appear insufficient in the current system, especially for secondary
TVET— even though public expenditures are high when measured as a percentage of
Yunnan’s provincial gross domestic product and total government spending. Yunnan
could explore ways to increase private expenditures in TVET and, at the same time, use
its existing public resources more efficiently.

Returns to private financing of upper secondary and tertiary education are high, however,
and continue to increase. Yunnan could mobilize more private resources to increase the
level of investment in education and training leading to even greater returns. Yunnan
should be able to increase the relatively low level of private financing and provision in
the current system, letting high-income households share the cost of education and
training, while further diversifying and freeing up other sources of revenue for education
and training. The Yunnan government’s recent announcement of regulations and
establishment standards for private TVET institutions is an important first step.

At the same time, more public resources could be directed toward alleviating the huge
disparities that exist in secondary TVET between rural and urban areas and between
schools. Special funds for TVET tend to disproportionally benefit already stronger
schools. Public finance should be targeted at helping TVET schools in disadvantaged
localities, low-income families, and poor-performing schools with more resource
constraints.
10


Finally, Yunnan could consider introducing demand-side financing mechanisms, such as
vouchers to improve the efficiency of public spending and quality of training.

5. Monitoring and Evaluation
Monitoring and evaluation provide feedback on the extent to which policies have been
implemented across the system, what outcomes have been achieved, and what
interventions have been effective or ineffective in achieving these outcomes. The very
fact that the monitoring system exists and works creates an incentive for local authorities
and schools to implement required policies and adhere to standards. Evaluation findings
can be further used to promote good practices within the system and, in turn, replicate
them on a wider scale.

Although this analysis identifies certain policy gaps in Yunnan’s skills development
system and proposes broad policy suggestions, to a large extent, many of the challenges
and issues arise not from a lack of central policies but from a lack of consistent and full
implementation of those policies, particularly in rural and disadvantaged areas. This
suggests the need to invest further in monitoring and evaluation of existing policies and
programs. Commissioned special studies and ad hoc impact evaluations currently exist,
but conducting routine monitoring and evaluation in a more systematic manner and with
more sophisticated methodologies would be helpful.

In addition, the use of pilot and experimental programs, and an analysis of their results, is
critical to ensuring the long-term health of the education and training system. China has
been learning from other TVET systems around the world, but its unique social, cultural,
and economic context make the adoption of any imported good practice a challenge.
Even within China, practices differ among provinces. Policies and practices that are
effective in Shanghai will not easily be transplanted to Yunnan. Developing a unique
Yunnan-style TVET system is the best and only policy choice. In this context, the
willingness to experiment and evaluate is key.

Implications for Subsector-Specific Issues
1. Non-formal Training
China’s industrial sector growth and economic development strategies have led to the
need to retrain large numbers of rural workers outside of the formal educational system
for transfer to the secondary and tertiary sectors. In Yunnan, the rural surplus labor
amounts to a third of the total rural labor force. Yunnan’s rural working-age population
averages only seven years of education, and more than 85 percent of rural workers have
received no training. Those who remain in the agricultural sector also need training in
both higher-value-added agricultural activities and in technology to improve their
productivity. Non-formal training is a critical means of implementing the necessary rural
training.

Yunnan faces challenges in improving both the delivery and content of, and access to,
non-formal training. The design and implementation of non-formal training need to be
better informed by a needs assessment and demand analysis. Forms of delivery and the
11


content of training should be tailored to the specific needs and circumstances of training
participants, such as rural laborers with relatively low levels of education and literacy,
and the employers in the industrial sectors in which they may ultimately work. An
improved qualifications framework would provide a mechanism that would enable non-
formal training programs to lead to skills certification. Participation in non-formal
training is hampered by barriers such as the requirement of a minimum education level, a
lack of awareness of the value of such training, and the quality and relevance of the
training itself. Encouraging both enterprises and learners to value skills certifications, and
adding flexibility that accounts for prior work and life experience can help overcome
these barriers.

Public, private, and industry sources of funding for non-formal training could be
consolidated into a single fund. A set of transparent funding criteria for the disbursement
of training funds would be necessary. To assure quality of training, the funding criteria
should be outcome-based rather than input-based. Training vouchers would be another
option, creating a competitive market for non-formal training, while giving participants
more choices.

2. Incentives and Technical Assistance for Work-Based Training
China’s ambitious economic goals require a well-educated and highly-skilled labor force.
Work-based training will be the means by which enterprises improve the skills and
performance of their existing employees.

Employers are currently required to dedicate 1.5 to 2.5 percent of their total wage bill to
training, and Yunnan must ensure the effective implementation of this requirement. Other
mechanisms that provide more incentives and quality assurance for work-based training
should also be considered. Korea’s Employment Insurance Scheme is a good example
that might be emulated in Yunnan.

In addition to financial assistance, the government should provide technical and
institutional assistance to facilitate the provision of work-based training, especially by
small and medium-sized enterprises. Technical assistance is particularly required in the
areas of needs assessment, training design and implementation, and monitoring and
evaluation. Institutional support might include organizing groups of small and medium-
sized enterprises to reduce the cost of training design and delivery. Government can also
systematically help firms build partnerships with training providers.

3. Pre-primary Education
Yunnan could invest more heavily in and popularize pre-primary education. Ample
research demonstrates that early childhood development programs, such as nutrition and
health services, parenting interventions, and center-based programs can improve
children’s physical and cognitive development, as well as social and emotional wellbeing.
Data from the Shanghai PISA and other cross-national evidence show the benefits of
early childhood education for children’s later cognitive development. Currently,
government investment and enrollment rates in pre-primary schools and early childhood
12


development programs in Yunnan are low. Investing in pre-primary education might be a
relatively cost-effective way to improve the skill levels of a young generation.
13


Chapter 1      Introduction


China is one of the world’s fastest growing economies with an average annual growth
rate of 9 percent over the last three decades. By 2010, China had surpassed Japan to
become the second largest economy in the world by gross domestic product (GDP), as
measured in purchasing power parity (PPP) terms. As of 2008, fewer than 30 percent of
the population lived on less than US$2 a day (the official definition of poverty), an
almost 40 percent reduction in the poverty rate within a decade. However, as of 2010,
China— even as the world’s second largest economy—ranks significantly in GDP per
capita, at about US$4,393.

According to the 12th Chinese National Economic and Social Development Five-Year
Plan (2011–15, the 12th Five-Year Plan), China’s annual GDP growth will slow, but still
remain above a rate of 7 percent. The Chinese government is keenly aware that it is
critical to sustain economic growth and social development. Indeed, the government is
aware that it will also be necessary to successfully shift economic growth from low to
high value-added, and from investment-led to consumption-led industries. This would
ideally be accomplished within a framework of inclusive growth and environmental
sustainability. Improving the quality and productivity of its labor force is explicitly put
forth as an enabling factor to the success of the 12th Five-Year Plan.

The education and skills of the workforce have played a key role in enabling China’s
unprecedented economic and social development over the last three decades. China has
made phenomenal improvements in the quantity and quality of its human resources since
the 1980s. Within just a few decades, China has universalized a minimum of nine-years
of basic education, developed general and vocational secondary education, and rapidly
expanded higher education. The gross enrollment ratio has increased at the lower
secondary level from 52 to 98 percent, at the upper secondary level from 29 to 78 percent,
and at the tertiary level from under 3 percent to a positive 23 percent. At the same time,
the overall quality of China’s labor force has improved significantly. Out of a population
of 1.3 billion, China’s economically active population totaled 0.78 billion in 2010. Of the
761 million total employed, 114 million are considered high-skilled “talents.� The
average years of schooling of the labor force has now reached 9 years, while new labor
market entrants have on average 12 years of schooling.

However, when compared to the more developed countries, China remains a human
resource-poor country. At the tertiary level, the gross enrollment ratio of 26 percent
remains below the world average of 30 percent. China has only about 40 research and
development personnel per 10,000 laborers, as compared to 110 in the United Kingdom,
115 in the Russian Federation, and 137 in Japan. According to the Academic Ranking of
World Universities 2010, only 6.8 percent of the world’s top 500 universities are in China,
as compared to 30 percent in the United States. None of the Chinese universities are in
the top 100. For example, Peking University and Tsinghua University ranked 151st and
200th, respectively.i
14


The majority of China’s labor force is unskilled or low skilled. Nearly half the workers
have attained only the nine-year compulsory education, 24 percent have completed only
primary education, and more than 3 percent remain illiterate. Among technical workers,
only a quarter qualified as high-skilled workers with skills certifications.ii Furthermore,
among China’s 253 million rural workers, including 159 million migrant workers, only
about 23 percent have attained a high school education or above, and as many as 69
percent of the rural workers have not received any type of training (National Bureau of
Statistics of China 2008).

Context of Yunnan

Located on the southwestern border of China, Yunnan is a medium-sized province with
abundant natural resources, including minerals, forests, animals, and plants. Yunnan has a
population of 46 million, 30 million of whom reside in rural and 16 million in urban areas.
Nearly 84 percent of Yunnan’s land mass is rural and mountainous. Yunnan further
distinguishes itself with its ethnic diversity: one-third of its population, or 15 million
people, belong to 26 different non-Han ethnic groups.

The total GDP of Yunnan has increased from Y8.4 billion in 1980 to Y722 billion in
2010, registering an annual growth rate of over 10 percent. However, compared to the
rest of China, particularly the coastal provinces, Yunnan’s economy is still
underdeveloped. Its GDP per capita in 2010 was Y15,752, only slightly more than half
the national average of Y29,992. The annual net income per capita of Yunnan’s rural
population was only Y3,952 (US$623) in 2010. About 22 percent of its population, or 10
million people, still live under the national poverty line of Y2,300, making Yunnan the
third poorest province in China after Gansu and Guizhou. At the end of 2011, Yunnan
had 73 national-level poverty counties, the highest number among all provinces.

Like the rest of China, the structure of Yunnan’s economy has been undergoing a
fundamental transformation. In 1980, 43 percent of Yunnan’s GDP came from the
primary sector. By 2010, the contribution of the primary sector was only 15 percent; with
the other 85 percent coming from the secondary and tertiary sectors (figure 1.1).
15

Figure 1.1 Contribution of Tertiary, Secondary, and Primary Sector to GDP Growth in Yunnan, 1957 –
2010

 100%
  90%
  80%
  70%
  60%
                                                                                 Tertiary Sector
  50%
                                                                                 Secondary Sector
  40%
                                                                                 Primary Sector
  30%
  20%
  10%
     0%
          1987
          1988
          1989
          1990
          1991
          1992
          1957
          1962
          1965
          1966
          1967
          1968
          1969
          1970
          1971
          1972
          1973
          1974
          1975
          1976
          1977
          1978
          1979
          1980
          1981
          1982
          1983
          1984
          1985
          1986



          1993
          1994
          1995
          1996
          1997
          1998
          1999
          2000
          2001
          2002
          2003
          2004
          2005
          2006
          2007
          2008
          2009
          2010
Source: Yunnan Statistical Yearbook 2011.

As urbanization continues and development patterns shift, the demand for skills will also
change, and the resulting skill bottlenecks are likely to constrain growth. Although more
than 60 percent of Yunnan’s labor force is still engaged in the primary sector, the demand
for skilled labor in the secondary and tertiary sectors, especially Yunnan’s pillar sectors
such as mining, energy, biotechnology, and tourism, continues to increase. Skills gaps are
emerging. At the same time, employers have been increasingly faced with problems such
as insufficient numbers of applicants, irrelevant skills, and unmatched expectations,
suggesting skills mismatches.

The educational attainment of Yunnan’s labor force has been rising over the past decade.
The proportion of workers who are illiterate has dropped from 20 percent in 2000 to only
11 percent in 2010, while the share of workers with high school degrees and above,
increased from more than 6 percent to more than 9 percent. Yet the average educational
attainment of workers in Yunnan has consistently lagged behind the national average. By
the end of 2010, Yunnan still had a higher share of illiterate workers (7 percent versus 3
percent nationwide), a higher share of workers with only a primary education (47 percent
versus 24 percent nationwide), and a lower share of workers with high school education
or above (14 percent versus 24 percent nationwide).

Looking ahead, fueled by global forces and government economic development policies,
China’s future growth will involve radical transformations: an economy that will be
increasingly led by consumption rather than investment, a move from high-energy to
low-carbon energy sources, and greater reliance on innovation over low-cost
manufacturing. Aligned with those principles, Yunnan has also laid out its economic
development strategy as articulated in its 12th Five-Year Plan. It includes the following
targets:
16


         ï‚·   A GDP growth rate of 10 percent or higher with a goal of total GDP of
             Y1,163 billion by 2015;
         ï‚·   Percentage composition of the primary, secondary, and tertiary sectors of
             15:43:42;
         ï‚·   Increase in expenditures on research and development to 1.4 percent of GDP;
         ï‚·   Average level of education of 9 years;
         ï‚·   Energy consumption reduced by 12 percent;
         ï‚·   Forestation increases to 5.45 percent;
         ï‚·   Urbanization rate increases to 44 percent by 2015 and to 50 percent by 2020;
         ï‚·   Disposable income increases to Y31,722 per urban resident and Y8,081 per
             rural resident.

Furthermore, the national Bridgehead Strategy has positioned Yunnan as a strategic
gateway in the southwest of the country. The ambitious development goals and the
resulting government investment not only present tremendous opportunities, but also
pose major skills challenges for Yunnan.

The objectives of this study are to provide an in-depth analysis of the current and future
demand for skills, identify skills gaps and sources of mismatches, and review the formal
and non-formal education and training systems. For example, what levels and
composition of skills are in demand? Are the current skills of the labor force adequate to
meet the skills demand? What roles can education and training play in bridging the skills
gaps and addressing the skills mismatches? We attempt to address these questions in this
report.

Definition and Measurement of Skills

Definitions of the term “skill� vary greatly in the current literature. For the purposes of
this report, skills are categorized and defined as the following:

     ï‚·   Generic skills

a. Cognitive skills include subject-specific knowledge, such as literacy and numeracy, as
well as non-subject-specific skills, such as critical and creative thinking, and problem
solving.

b. Non-cognitive skills include characteristics not included under cognitive skills, for
example, behavioral skills such as perseverance, self-discipline, teamwork, and the ability
to negotiate conflict and manage risks.

     ï‚·   Technical and vocational skills

A mix of knowledge and skills used to perform specific jobs that rely on clearly defined
tasks. For the purposes of this report, we define them as job-specific skills as opposed to
generic skills that are transferable across jobs.

The measurement of skills used in the various datasets and in this report includes:
17


     ï‚·   Educational attainment

Educational attainment comprises widely-available measures, which could potentially
indicate a broad range of skills—generic and technical, cognitive and non-cognitive—at
different levels.

     ï‚·   Skills certification

Rates of skills certification among graduates and employees, as well as the levels of
certification, are important measures of technical and vocational skills.

     ï‚·   Skills Toward Employment and Productivity (STEP) Survey Measures

The household and employer surveys of STEP provide direct measures of cognitive, non-
cognitive and technical skills through tests. The survey also links individual skills with
labor market outcomes (Puerta et al. 2012).

     ï‚·   Programme for International Student Assessment measures

The Programme for International Student Assessment provides measures of generic, non-
subject based, cognitive skills acquired by 15-year-olds in the domains of reading,
mathematics and science literacy (OECD 2009).

Conceptual Framework

This study employs a demand-supply framework, as illustrated in figure 1.2. It starts with
a demand-side analysis in chapter 2, examining historical trends in demand for skills,
revealing the types of skills in demand, and projecting future demand for skills driven by
economic growth and policy development. Chapter 2 also highlights the emerging skills
shortages and mismatches in Yunnan. The rest of the report focuses on the access, quality,
and relevance of Yunnan’s education and training system and how effective it is in
supplying the skills in demand.
18

Figure 1.2. Conceptual Framework for Study of Skills Demand and Supply in Yunnan




Note: TVET = Technical and Vocational Education and Training.

Skills development takes place in formal programs that lead to recognized diplomas and
qualifications. It also takes place through non-formal training, which tends to be less
structured, of shorter duration, and often does not lead to a formal certification. It is
important to note that the line between formal and non-formal training is increasingly
blurred in the Chinese context: Formal educational institutions are increasingly being
contracted to provide non-formal training, while students and employees have increasing
opportunities to qualify for formal certifications based on a series of short-term, non-
formal training courses.

An overview of Yunnan’s formal and non-formal education and training system is
presented in chapter 3. Chapter 4 focuses on the formal Technical and Vocational
Education and Training (TVET) system, examining its governance, industry participation,
curriculum reforms, quality assurance, and finances. Analysis of the formal education and
training system focuses mainly on secondary and tertiary Technical and Vocational
Education and Training. Due to the limited scope of the report, we are unable to conduct
an in-depth analysis of Yunnan’s basic education system or the academic tracks of higher
education.

Chapters 5 and 6 address two major training programs outside the formal education
system: non-formal training for rural workers and work-based training for urban workers,
both of strategic importance. Finally, Chapter 7 draws on lessons from the Shanghai
Programme for International Student Assessment to demonstrate the role of schools in
developing the cognitive skills of 15-year-olds. The report concludes with a summary of
19


findings and a set of policy recommendations for meeting the skills challenges and
improving the education and training system.

Admittedly, education and training are only part of the equation for a balanced skills
development system. The recent World Bank initiative, Stepping Up the Skills for More
Jobs and Higher Productivity (World Bank 2010), broadly defines a skills development
system to include five steps: (1) getting children off to the right start; (2) ensuring that all
students learn; (3) building job-relevant skills; (4) encouraging entrepreneurship and
innovation; and (5) facilitating labor mobility and job matching.

In view of the broad scope of a skills development system, this report focuses primarily
on steps (2) and (3). The report does not venture into labor market policies such as hukou
(household) registration, minimum wages and pensions, which are also integral parts of a
skills supply and demand system.

References
OECD (Organisation for Economic Co-operation and Development). 2009. PISA 2009
Assessment Framework. Paris: OECD.

National        Bureau          of       Statistics    of                      China.          2008.,
http://www.stats.gov.cn/tjfx/dfxx/t20080220_402463674.htm.

Sanchez-Puerta, Maria Laura, Alexandria Valerio, Gaelle Pierre, and Sergio Urzua. 2012.
“STEP Skills Measurement Surveys Methodology Note Draft,� World Bank, Washington,
DC.

World Bank. 2010. Stepping Up the Skills for More Jobs and Higher Productivity.
Washington DC: World Bank.

Yunnan Bureau of Statistics. 2011. Yunnan Statistical Yearbook. Beijing: China


Notes

i
   Rankings vary by the methodologies used. For example, the Times Higher Education World University
Rankings (2011–12) ranked Peking University and Tsinghua University 49th and 71st, respectively. The
U.S. News & World Report ranked the two universities 44th and 48th, respectively, for the same year.
ii
    China’s vocational certifications are granted at five levels: Level 1 (senior technicians), Level 2
(technicians), Level 3 (senior-skilled workers), Level 4 (medium-skilled workers) and Level 5 (junior-
skilled workers).
20


Chapter 2       Skills Challenges: Demand, Gaps, and Mismatches


This chapter reviews the skills challenges in Yunnan using a demand-supply framework.
Data for both the provincial and national levels allow us to compare Yunnan against
national trends and situate the results within a broader context. The chapter begins with
an overview of the employment trends and skills composition of the current labor force.
Increasing demand for skilled labor is implied by the rising educational attainment of the
labor force, along with the increasing wage returns to education in the job market. Data
from household and employer surveys shed further light on the specific cognitive, non-
cognitive and technical skills in demand from both employee and employer perspectives.
Our analysis also indicates emerging skills shortages and identifies skills mismatches by
examining problems encountered by firms in hiring new employees. Recent
developments in national and provincial policies are expected to pose further skills
challenges, as Yunnan is presented with immense opportunities for economic growth,
structural change, and greater openness through the Bridgehead Strategy.

Labor Market Trends in China and Yunnan

By the end of 2011, employment in Yunnan totaled over 27 million. The agricultural
sector still dominates Yunnan’s total employment with nearly 17 million workers,
although its share has been declining steadily since the early 1990s. Compared to the
national averages, Yunnan has historically had a higher share of employment in the
primary sector, and a lower share in the secondary and tertiary sectors (figure 2.1).
Across China, the share of primary sector employment has dropped by 32 percent
percentage points from 1980 (69 percent) to 2010 (37 percent), while the share of tertiary
sector employment has more than doubled from 13 percent to 34 percent within the last
three decades. The composition of the primary, secondary, and tertiary sectors in 2010
was 37 percent, 29 percent, and 35 percent nationwide, respectively, in contrast to
Yunnan’s composition of 60 percent, 13 percent, and 27 percent, respectively.

Figure 2.1. Composition of Employed Persons at Year-End by Sector in Yunnan and China, 1980 –2010
a.      Yunnan

                                        100%
                  % of total employed




                                        80%

                                        60%

                                        40%

                                        20%

                                         0%
                                               1980
                                               1981
                                               1982
                                               1983
                                               1984
                                               1985
                                               1986
                                               1987
                                               1988
                                               1989
                                               1990
                                               1991
                                               1992
                                               1993
                                               1994
                                               1995
                                               1996
                                               1997
                                               1998
                                               1999
                                               2000
                                               2001
                                               2002
                                               2003
                                               2004
                                               2005
                                               2006
                                               2007
                                               2008
                                               2009
                                               2010




                                                                     Year
                                               Primary   Secondary    Tertiary
21


b.       China

                                       100%
                                        90%



                 % of total employed
                                        80%
                                        70%
                                        60%
                                        50%
                                        40%
                                        30%
                                        20%
                                        10%
                                         0%
                                              1980
                                              1981
                                              1982
                                              1983
                                              1984
                                              1985
                                              1986
                                              1987
                                              1988
                                              1989
                                              1990
                                              1991
                                              1992
                                              1993
                                              1994
                                              1995
                                              1996
                                              1997
                                              1998
                                              1999
                                              2000
                                              2001
                                              2002
                                              2003
                                              2004
                                              2005
                                              2006
                                              2007
                                              2008
                                              2009
                                              2010
                                                                           Year
                                                   Primary   Secondary      Tertiary


Source: Yunnan Statistical Yearbook 2011.

Growth in employment in Yunnan in recent years has been driven mostly by demand from
the tertiary sector. Quarterly job vacancy data show that the tertiary sector had the largest
number of job openings throughout 2010 and 2011, followed by the secondary sector
(figure 2.2). In 2011 alone, Yunnan’s tertiary sector generated 453,298 job openings.

Figure 2.2. Quarterly Number of Job Vacancies by Sector, Yunnan, 2010–11

 140000

 120000

 100000

     80000                                                                                                   Primary

     60000                                                                                                   Secondary
                                                                                                             Tertiary
     40000

     20000

         0
             2010/1                       2010/2    2010/3   2010/4      2011/1   2011/2   2011/3   2011/4

Source: HRSS 2010a, 2011.

Within the tertiary sector, the largest increases in employment occurred in relatively
smaller-sized sectors (table 2.1): for example, between 2007 and 2010, employment in
leasing/trade/business services increased by 49 percent, followed by services to
households/other services (36 percent), and real estate (32 percent). In terms of the
absolute number of workers, education and public administration/social organization
together employed over 0.99 million workers in 2010. The secondary sector, with its
22


large work force, has also seen continued growth in employment. For example,
employment in mining and construction expanded by 21 percent and 14.5 percent,
respectively, between 2007 and 2010.

Table 2.1 Total Number of Staff and Workers at Year-End in Yunnan by Industry, 2007 and 2010


                                                                       2007     2010
                                                                                          Total growth
     Sector                         Industry
                                                                         (10,000              (%)
                                                                         persons)
 Primary
 Farming, forestry, animal husbandry and fishery                      15.1     13.1     -13.1

 Secondary
 Mining                                                               11.9     14.4     20.9
 Manufacturing                                                        53.0     55.1     3.9
 Production and supply of electricity, gas and water                  7.5      8.2      9.4
 Construction                                                         29.9     34.2     14.5

 Tertiary
 Transportation, storage and post                                     12.0     12.6     4.8
 Information transmission, computers service and software             2.9      2.9      0.0
 service
 Wholesale   and retail trades                                        13.3     14.9     11.5
 Hotels and catering services                                         5.6      5.6      0.0
 Finance                                                              7.0      8.1      16.7
 Real estate                                                          2.9      3.8      31.6
 Leasing trade and business service                                   4.7      6.9      49.0
 Scientific research, technology service and geological               5.4      6.0      9.4
 prospecting
 Water    conservancy, administration of environment and public       4.2      4.5      7.1
 facilities
 Services to households and other services                            0.6      0.8      36.4
 Education                                                            47.9     50.5     5.4
 Health care, social security and social welfare                      13.1     14.7     12.5
 Culture, sports and entertainment                                    3.4      3.4      2.4
 Public administration and social organization                        40.4     44.0     8.8
Source: Yunnan Statistical Yearbook 2010, 2011.

Measured by ownership of employer, Yunnan’s employment composition has shifted
away from the dominance of state-owned enterprises (SOEs) as the non-public sector
increasingly generates employment (figure 2.3). At the beginning of China’s economic
reforms in 1978, SOEs dominated employment in Yunnan with an 88 percent share.
Three decades later, the share of SOEs has dropped to 30 percent, while the non-public
sector (including self-employment) constitutes 46 percent of the total employment in
urban Yunnan (with nearly 4 million workers).
23

Figure 2.3. Workers in Urban Areas by Ownership, Yunnan


     Number of workers (10,000)   700
                                                                        Others
                                  600
                                  500
                                                                        Urban Private Enterprises and Self-
                                  400
                                                                        employed Individuals
                                  300
                                                                        Other ownership
                                  200
                                  100
                                                                        Collectives
                                   0
                                        1978
                                        1980
                                        1985
                                        1990
                                        1995
                                        1998
                                        1999
                                        2000
                                        2001
                                        2002
                                        2003
                                        2004
                                        2005
                                        2006
                                        2007
                                        2008
                                        2009
                                        2010
                                                                        SOE
                                         Year

Source: Yunnan Statistical Yearbook 2011.

Along with GDP and employment, real wages in China and Yunnan have seen
tremendous growth since the early 1990s (Figure 2.4). Across China, real wages have
increased fivefold. Real wages of workers in state-owned enterprises surpass those in
other types of enterprises, and the disparities have been growing.

The growth in real wages has been uneven across ownership types and regions. In
Yunnan, real wages have increased fourfold between 1994 and 2010, but the gaps
between China and Yunnan have been widening over the same period (figure 2.4). In
particular, the growth seems to be driven solely by SOEs and urban collectives in Yunnan,
as real wages in other types of enterprises have been almost unchanged between 1994 and
2010. In 2010, real wages of workers in SOEs and urban collectives are more than four
times those in other types of enterprises in Yunnan (figure 2.4).

Figure 2.4. Trends in real wages (Index 1994=100) of China and Yunnan

 600

 500

 400

 300                                                                    China
                                                                        Yunnan
 200

 100

          0
                                   2000
                                   1994
                                   1995
                                   1996
                                   1997
                                   1998
                                   1999

                                   2001
                                   2002
                                   2003
                                   2004
                                   2005
                                   2006
                                   2007
                                   2008
                                   2009
                                   2010




Source: China Statistical Yearbook 2011; Yunnan Statistical Yearbook 2011
24

Figure 2.5. Real wages (Index 1994=100) by types of ownership

a. China:

 600

 500

 400

                                                                  SOEs
 300
                                                                  Collectives
 200                                                              Other

 100

     0
         1998
         1994
         1995
         1996
         1997

         1999
         2000
         2001
         2002
         2003
         2004
         2005
         2006
         2007
         2008
         2009
         2010
b. Yunnan:

 600

 500

 400

                                                                  SOEs
 300
                                                                  Collectives
 200                                                              Other

 100

     0
         1998
         1994
         1995
         1996
         1997

         1999
         2000
         2001
         2002
         2003
         2004
         2005
         2006
         2007
         2008
         2009
         2010




Source: China Statistical Yearbook 2011; Yunnan Statistical Yearbook 2011

During the same period, China’s labor productivity, defined by GDP per person
employed, has more than quadrupled. Much of the gains in labor productivity were
driven by manufacturing and expansion of competition (World Bank 2012). However,
despite the huge gains over the last two decades, the level of labor productivity of China
remains lower compared to other middle-income and high-income countries in East Asia,
such as Japan, Korea, Thailand and Malaysia (figure 2.6). The gap is even larger when
compared to the average of OECD member countries. Significant regional disparities in
25


productivity also exist within China (figure 2.7). For example, the productivity of the
Eastern region is more than twice that of the Western region. The productivity of workers
in Yunnan increased by 76 percent between 2004 and 2009, but remains at one of the
lowest levels in China. Even within the Western provinces, the productivity of Yunnan’s
labor force in 2009 was the second lowest, after the Guizhou province.

Figure 2.6. Labor productivity of selected countries

                                               60000
          employed (constant



                                               50000
           GDP per person




                                                                   China
             1990 PPP $)



                                                                   OECD members
                                               40000
                                                                   Japan
                                               30000               Korea, Rep.
                                                                   Philippines
                                               20000
                                                                   Malaysia
                                               10000               Thailand
                                                                   Indonesia
                                                   0
                                                       1991
                                                       1992
                                                       1993
                                                       1994
                                                       1995
                                                       1996
                                                       1997
                                                       1998
                                                       1999
                                                       2000
                                                       2001
                                                       2002
                                                       2003
                                                       2004
                                                       2005
                                                       2006
                                                       2007
                                                       2008
                                                       2009
                                                       2010




Source: World Development Indicator

Figure 2.7. Labor productivity of Yunnan compared to selected provinces

                                           7
     GDP per pseron employed (CNY10,000)




                                           6

                                           5

                                           4                                   2004
                                           3                                   2005

                                           2                                   2007
                                                                               2009
                                           1

                                           0




Source: World Bank 2012
26


Skills Composition of the Current Workforce

The educational attainment of the labor force in Yunnan has been increasing over the last
decade. As shown in figure 2.8, the proportion of workers who are illiterate has dropped
from 20 percent in 2000 to 11 percent in 2010. Over the same period, the share of
workers with a high school degree or above has increased from 6.5 percent to 9.3 percent,
and the share of workers with tertiary education has increased from 1.4 percent to 3.2
percent.

Figure 2.8. Education Composition of the Employed, Yunnan and China, 2001 –10

a.                           Yunnan                                                                                            b.           China

                           100%                                                                                                100%




                                                                                                         % of total employed
     % of total employed




                           80%                                                                                                  80%

                           60%                                                                                                  60%

                           40%                                                                                                  40%

                           20%                                                                                                  20%

                            0%                                                                                                      0%




                                                                                                                                                                                   2007
                                                                                                                                         2001
                                                                                                                                                2002
                                                                                                                                                       2003
                                                                                                                                                              2004
                                                                                                                                                                     2005
                                                                                                                                                                            2006


                                                                                                                                                                                          2008
                                                                                                                                                                                                 2009
                                                                                                                                                                                                        2010
                                  2001
                                         2002
                                                2003
                                                       2004
                                                               2005
                                                                      2006
                                                                             2007
                                                                                    2008
                                                                                           2009
                                                                                                  2010




                                                                 Year                                                                                                 Year
                 Illiterate                                   Primary school
                                                                                                                     Illiterate                                  Primary school
                 Junior secondary                             Senior secondary
                 college and above                                                                                   Junior secondary                            Senior secondary


Source: China Statistical Yearbook, Yunnan Statistical Yearbook 2002–11.

Despite the increase in educational attainment over the past decade, the vast majority of
Yunnan’s current workforce is still composed of workers with only a basic education or
less. As of 2010, 33 percent of workers in Yunnan had completed only compulsory
education, and 47 percent had only a primary school education. Yunnan has a much
higher share of illiterate workers (7 percent) than the national average (3 percent).
Moreover, skilled workers with a high school education and above constitute only about
14 percent of Yunnan’s labor force, the third lowest in China, after Tibet and Guizhou
(figure 2.9).
27

Figure 2.9. Proportion of Total Employed with Upper Secondary or Tertiary Education in China by
Province, 2010

                           70.0
                           60.0
     % of total employed



                           50.0
                           40.0
                           30.0
                           20.0
                           10.0
                            0.0

                                       Chongqing
                                             Anhui




                                          Zhejiang
                                            Jiangxi
                                             Hebei




                                            Tianjin
                                             Gansu




                                             Hubei




                                            Beijing
                                            Hainan



                                            Shanxi




                                         Shanghai
                                           Qinghai




                                           Shaanxi
                                             Fujian



                                           Ningxia
                                          Xinjiang
                                               Jilin

                                          Liaoning
                                              Tibet




                                      Guangdong
                                          Guangxi
                                            Henan




                                           Jiangsu
                                          Guizhou
                                           Yunnan




                                          National
                                           Sichuan




                                        Shandong
                                            Hunan



                                     Heilongjiang




                                  Inner Mongolia
                                      upper secondary   tertiary

Source: China Labor Statistical Yearbook 2011.

On average, female workers are less educated than male workers in Yunnan. Female
workers account for 46 percent of the total employed in Yunnan. As shown in figure 2.10,
a larger proportion of female workers are illiterate (10 percent versus 4 percent of male
workers) or have only a primary education (50 percent versus 44 percent of male
workers). Among high-skilled workers, however, the gender gaps are narrower: 12
percent of female workers have an upper secondary education or above, compared to 15
percent of male workers.

Figure 2.10. Education Composition of Male and Female Workers in Yunnan, 2010

                           60.0
     % of total emloyed




                           50.0
                           40.0
                           30.0
                           20.0
                           10.0
                            0.0                                         Male
                                                                        Female




                                      Education attainment

Source: China Labor Statistical Yearbook 2011.

To further investigate the skills composition specific to each industry and occupation, we
draw on data from a human resources survey of 3,000 non-public enterprises in Yunnan
conducted in 2008 (box 2.1). In 2010, non-public enterprises contributed to more than 40
28


percent of Yunnan’s economy and generated almost half of the urban employment. The
majority of non-public enterprises in Yunnan are concentrated in the tertiary sector (61
percent), but non-public enterprises in the secondary sector employ the highest proportion
of workers (52 percent), followed by the tertiary sector (36 percent), and the primary
sector (13 percent). The manufacturing, wholesale/retail trades, construction, and mining
industries employ the largest number of workers (figure 2.6).

                         Box 2.1. Yunnan Non-public Enterprise Survey, 2008

The Yunnan Non-public Enterprise Survey used in this report was conducted in 2008 by the
Yunnan Statistical Bureau. The survey selected 46,208 enterprises across Yunnan using stratified
random sampling by industry and size. Workers are categorized by occupation into: (1)
managerial personnel (��管�人员); (2) professional and technical personnel (专业技术人员);
and (3) frontline/operative skilled workers (技能人员).

Within each occupation, workers are further divided into senior-, medium- and junior-level
workers. For workers who have obtained vocational certification, senior-level workers correspond
to those certified at Levels 1 through 3, medium-level workers correspond to those certified at
Level 4, and junior-level workers correspond to those certified at Level 5.

Human resources in Yunnan’s non-public enterprises are composed mostly of
frontline/operative skilled workers (29 percent), followed by professional and technical
personnel (18 percent), and managerial personnel (17 percent). Across sectors (table 2.2),
the tertiary sector has the highest proportion of professional and technical personnel, led
by health care/social security/social welfare, scientific research/technology
service/geological prospecting, and education. The secondary sector, including
construction, manufacturing, and mining, employing the largest number of workers, also
has one of the highest proportions of frontline/operative skilled workers.

Table 2.2. Number of People Employed in Non-public Enterprises, by Industry and Occupation, 2008

Industry                                           Managerial         Professional  Frontline
                                                   Personnel          and technical skilled
                                                                      personnel     workers
Farming, forestry, animal husbandry and                  8,667             5,303       15,462
fishery
Mining                                                  27,786             19,232            53,023
Manufacturing                                           84,295             84,117           206,503
Production and supply of electricity, gas and            3,108              4,660             3,572
water
Construction                                            43,352             89,575           112,762
Transportation, storage and post                         8,665              6,678            19,705
Information transmission, computer services              3,700              3,633             2,469
and software services
Wholesale and retail trades                             63,448             43,193           44,353
Hotels and catering services                            12,011              6,430           17,841
Finance                                                  1,436              1,625            1,157
Real estate                                             14,694             14,436           11,295
29

Leasing trade and business services                   12,716               16,443             11,713
Scientific research, technology services and           3,325                7,777              1,863
geological prospecting
Water conservancy, administration of                   1,617                1,970             1,903
environment and public facilities
Services to households and other services              3,924                4,034             5,145
Education                                              4,644               10,954             5,439
Health care, social security and social welfare        1,530                8,185             1,373
Culture, sports, and entertainment                     2,517                1,589             2,442
Source: Yunnan Non-public Enterprise Survey 2008.

Almost half (forty-nine percent) of workers in non-public industries in Yunnan have
completed only middle school or less, and only 18 percent of them hold a tertiary degree.
Within the non-public enterprises, managerial personnel have the highest education
levels, with over 40 percent holding tertiary degrees. By contrast, over half of
frontline/operative skilled workers have only a middle school degree or less, and fewer
than 10 percent of them hold tertiary degrees (figure 2.11).

Figure 2.11. Educational Attainment by Human Resources Type, 2008

 60%
                                                                    Master's
 50%
 40%
 30%                                                                Bachelor's
 20%
 10%                                                                Associate
  0%

                                                                    High School (including
                                                                    vocational)
                                                                    Middle school and below


Source: Yunnan Non-public Enterprise Survey 2008.

Figure 2.12 illustrates the education composition by industry, ranked by proportion of
workers with tertiary degrees. The finance industry has the highest share of educated
workers (72 percent with tertiary degrees and 97 percent with high school degrees or
above), followed by scientific research/technology service/geological prospecting (63
percent with tertiary degrees and 90 percent with high school degrees or above). By
contrast, industries in the primary and secondary sectors, such as mining,
farming/forestry/animal husbandry/fishing, manufacturing, and construction, have the
highest proportion of workers with only a middle school education or below.
30

Figure 2.12. Education Composition in Non-public Enterprises by Industry, 2008


                                               Finance
         Scientific Research, Technology Service and…
   Information Transmission, Computers Service and…
                                            Education
                  Leasing Trade and Business Service
                                           Real Estate
       Health Care, Social Security and Social Welfare
                         Wholesale and Retail Trades
               Water Conservancy, Administration of…
 Production and Supply of Electricity, Gas and Water
          Services to Households and Other Services
                   Culture, Sports and Entertainment
                                         Construction
                     Transportation, Storage and Post
                                       Manufacturing
                         Hotels and Catering Services
    Farming, Forestry, Animal Husbandry and Fishery
                                                Mining
                                                     0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

        Master's   Bachelor's    Associate   High school (including Vocational)   Middle school and below

Source: Yunnan Non-public Enterprise Survey 2008.

In addition to education composition, certification rates are low among professional and
technical personnel and frontline/operative skilled workers (figure 2.13). Overall, roughly
24 percent of professional and technical personnel and 12 percent of frontline/operative
skilled workers have obtained professional or vocational certification. Among those
certified, the majority are junior-level workers. Only about 2 percent of professional and
technical personnel and 2 percent of frontline/operative skilled workers at the senior level
are certified.

Figure 2.13. Share of Certified Workers by Type and Level, 2008


                             Senior     Medium      Junior




                   12.16%



                                                             5.98%
                   8.89%
                                                             3.93%
                   2.48%                                     1.96%
         Professional and technical                Frontline skilled worker

Source: Yunnan Non-public Enterprise Survey 2008.
 31


 Figure 2.14 depicts the share of medium- and senior-level professional and technical and
 frontline/operative workers who have skills certification or professional rankings. Among
 the five industries that employs the highest proportion of professional and technical
 personnel, education and scientific research/technology service/geological prospecting
 also have a relatively high proportion of medium- and senior-level technicians and
 professionals who are certified. In comparison, production and supply of electricity, gas
 and water has one of the highest proportions of professional and technical workers, but
 fewer than five percent of them are certified.

 Certification rates are even lower among frontline/operative skilled laborers. In particular,
 although the agricultural sector employs one of the highest proportions of
 frontline/operative skilled workers, fewer than 0.3 percent of its medium- and senior-
 level frontline/operative workers are certified. Likewise, within the secondary sector,
 mining consists of 28 percent frontline/operative skilled workers, but fewer than 0.3
 percent of its frontline/operative skilled workers at the medium or senior level are
 certified. In both manufacturing and construction, which employ a large number of
 people and a high proportion of frontline/operative skilled laborers, only 6 to 7 percent of
 the medium- or senior-level employees are certified.

 Figure 2.14. Share of Certified Workers at Medium and Senior Level in Non-public Enterprises, 2008

 a.       Professional and technical personnel

                  Real Estate
                    Education
         Scientific Research,…
                Construction
              Manufacturing
                      Finance
Transportation, Storage and…
 Leasing Trade and Business…                                            Certified senior-level
  Information Transmission,…                                            professional/technical personnel
 Health Care, Social Security…
Wholesale and Retail Trades                                             Certified medium-level
   Farming, Forestry, Animal…                                           professional/technical personnel
        Water Conservancy,…
Hotels and Catering Services
 Services to Households and…
   Production and Supply of…
                       Mining
         Culture, Sports and…
                             0%   5%     10%    15%    20%    25%
 32


 b.      Frontline/operative skilled workers

   Scientific Research, Technology…
                         Education
          Production and Supply of…
                        Real Estate
                    Manufacturing
                      Construction
       Services to Households and…
      Hotels and Catering Services                                   Certified senior-level skilled
      Wholesale and Retail Trades                                    workers
 Transportation, Storage and Post
  Health Care, Social Security and…                                  Certified medium-level skilled
       Leasing Trade and Business…                                   workers
               Water Conservancy,…
        Information Transmission,…
Culture, Sports and Entertainment
                           Finance
         Farming, Forestry, Animal…
                            Mining
                                  0%   5%   10%   15%   20%   25%


 Source: Yunnan Non-public Enterprise Survey 2008.

 Low certification rates reflect, in part, the relatively low skill levels of workers. However,
 they also call into question the effectiveness and relevance of the skills certification
 system in Yunnan. Employers do not seem to demand that their employees be certified.
 The low recognition of skills certification among employers not only calls into question
 the validity and relevance of skills certificates, but also reduces the incentives for
 employees to acquire certified skills or for employers to provide skills training.

 Trends in Returns to Education

 At the same time that the educational attainment of the population and labor force has
 been improving rapidly, demand for education and skills continues to rise as evidenced
 by dramatic increases in the past decade in the returns to education, especially at the high
 school level and beyond. Using household survey data from 1988 to 2012, we estimate
 the private returns to schooling in Yunnan and China and examine how they have
 changed over time from the late 1980s through the 2000s.

 Data for 1988, 1995, and 2002 were drawn from the Chinese Household Income Project
 (CHIP) (box 2.2). For more recent years, we use data from the Longitudinal Survey on
 Rural Urban Migration in China (RUMiC) 2007 to examine national trends. Yunnan was
 not included in the RUMiC 2007 sample. To complete our analysis, we draw from the
 most recent STEP Household Survey 2012, which sampled 2,000 individuals from 2,000
 households in urban Kunming, the capital city of Yunnan (Sanchez-Puerta et al. 2012). A
 description of the analytic sample from the STEP Household Survey is available in
 Appendix A. It is important to note that the weighted sample only represents the urban
 population in Kunming. Although not representative of the entire province, workers in
33


the urban Kunming sample have higher education levels and are concentrated in the
service sector.

                         Box 2.2. CHIP 1988, 1995, 2002 and RUMiC 2007

The three CHIP surveys for 1988, 1995, and 2002 were conducted by researchers from the
Chinese Academy of Social Sciences together with associated Chinese and international scholars,
with the assistance of the National Bureau of Statistics (Keith and Zhao 1993; Riskin, Renwei,
and Shi 2001; Li, Luo, Wei, and Yue 2008).

RUMiC 2007 was initiated by a group of researchers at the Australian National University, the
University of Queensland, and the Beijing Normal University and was supported by the Institute
for the Study of Labor, which provides the Scientific Use Files. Financial support for RUMiC
was provided by the Australian Research Council, the Australian Agency for International
Development, the Ford Foundation, the Institute for the Study of Labor, and the Chinese
Foundation of Social Sciences.

This study draws on the urban household component of each wave of the survey and restricts the
samples to workers with positive yearly earnings. Owners of private or individual enterprises are
not included. Yunnan was sampled in the three waves of CHIP, but not in RUMiC 2007.

We employ the following Mincer equation (Mincer 1974) for our calculation:

                                         (2.1) where is the yearly wage, is the years
of schooling, and is the years of labor market experience. Here, we use total years of
schooling reported by individualsi and the estimated years of labor market experience
based on age and years of schooling.ii We also estimate the returns to each level of
schooling. In addition, we control for gender, ethnicity,iii type of company ownership,
city, and industry. Table 2.3 presents the estimated results from the OLS regressions.

Rates of return to an additional year of education have been increasing throughout the
1990s in both Yunnan and China. Liu, Park, and Zhao (2010) attributed the rising returns
to education throughout the 1990s to skill-biased technical progress, specifically, changes
in available technologies and institutional changes associated with the economic
transition and maturation of China’s urban labor market. Another important contributing
factor is the changes in industrial wage rents, reflecting the growing specialization of
more-educated workers in higher rent industries relative to less-educated workers.

Although national data suggest dramatic increases in returns to education during two
periods of time, 1992 to 1994, and 1994 to 1999 (Zhang, Zhao, and Park 2005),
significant changes in returns to education in Yunnan do not seem to emerge until the
latter half of the 1990s. Based on our data in Yunnan, the change is most pronounced
during the period between 1995 and 2002, when the rate of return to an additional year of
education increased by 3.18 percentage points. During the 2000s, nationwide data suggest
a slight decrease in the average rates of returns to an additional year of education, most
likely due to the decreased returns to education below high school. Data from urban
Kunming suggest that the rates of returns to an additional year of education remain at
around 8 percent in 2012.
34

Table 2.3. Rates of Return to Schooling, Yunnan and China, 1988 –2012

                                              1988        1995       2002     2007      2012a
  Per additional year (%)        Yunnan     3.37        3.99       7.17     —          8.02
                                 China      3.24        4.32       7.05     6.10       —
  Percentage increase in
  wages relative to middle
  school (%)
    High school                  Yunnan     4.73        9.16       19.01    —          19.84
                                 China      7.33        10.37      12.30    11.74      —
     Secondary Technical         Yunnan     14.68       18.65      25.11    —          24.86
     and Vocational
     Education and Training
     (TVET)
                                 China      12.41       23.49      29.69    32.31      —
     Tertiary, associate         Yunnan     22.38       23.74      53.73    —          57.93
     degree
                               China     18.77    32.71     52.50           57.62      —
     Bachelor’s and above      Yunnan 30.34       37.71     79.32           —          112.55
                               China     30.08    43.48     77.54           96.60      —
    N                          Yunnan 893         1082      833             —          1026
                               China     6788     11137     9147            5695       —
Source: Authors’ calculation using CHIP 1988, 1995, 2002, RUMiC 2007        and STEP   Household
Survey 2012. Full models presented in appendix A.
Note: — = Not available.

a. STEP Household Survey 2012 was sampled only in Kunming.

Table 2.3 also lists the returns by education level expressed as a percentage change in
wages relative to those of middle school graduates. There are positive and significant
returns to every level of education beyond middle school, and the rates of return increase
with the level of education. Returns to secondary TVET appear to be higher than returns
to general high school both in Yunnan and nationwide.

In figure 2.15, we graphed our results to better illustrate the changes in returns to
different levels of schooling over time. Between 1988 and 2002, although returns to
schooling have been increasing rapidly at every level of education beyond the
compulsory level, the most dramatic increase is seen on the tertiary level. The increase in
returns to tertiary-level education in Yunnan has been even more dramatic than the
increase in the national average. In 1995, individuals in Yunnan with tertiary associate
degrees earned 24 percent more than middle school graduates. Less than a decade later, in
2002, the difference had become as large as 54 percent. Even more, for workers with
Bachelor’s degrees and above, the wage difference above middle school graduates
increased from 38 percent in 1995 to 79 percent in 2002. In 2010, earnings of Bachelor’s
degree holders and above in urban Kunming were more than double those of middle
school graduates.
35


As the returns to tertiary degrees dramatically increase, the gap between returns to
tertiary education and secondary education have been widening. Since 2002, returns to
tertiary education, especially Bachelor’s degrees, continued to increase, while returns to
general high school graduates seem to have stabilized. At the same time, the gap between
returns to tertiary associate degrees and Bachelor’s degrees has also been increasing.
National data suggest that between 2002 and 2007, the gap continued to widen: returns to
Bachelor’s degrees increased by over 19 percent, while returns to secondary TVET and
tertiary associate degrees exhibited only slight increases as compared to the previous
decade. These changes reflect the increasingly stronger demand for workers with tertiary
education relative to those with secondary education; within the tertiary level, Bachelor’s
degrees are rewarded in the job market at increasingly higher rates over associate degrees.

Figure 2.15. Percent Change in Wages Relative to Middle School Graduates, Yunnan and China, 1998–
2012
a.       Yunnan

 120.00%

 100.00%

     80.00%
                                                           High school

     60.00%                                                Secondary TVET
                                                           Tertiary, associate
     40.00%
                                                           Bachelor's and above

     20.00%

      0.00%
              1988     1995      2002      2012


b.       China

 120.00%
 100.00%
     80.00%                                                High school

     60.00%                                                Secondary TVET

     40.00%                                                Tertiary, assssociate
                                                           Bachelor's and above
     20.00%
      0.00%
              1988     1995      2002      2007

Source: Authors’ calculations using CHIP 1988, 1995, 2002, RUMiC 2007 and STEP Household
Survey 2012.
36


The increasing returns to tertiary education are not a pattern unique to Yunnan: similar
trends have been found in other medium- and low-income countries during the 1990s (see,
for example, Blom, Holm-Nielsen, and Verner 2001; Fiszbein, Patrinos, and Giovagnoli
2007; Behrman, Birdsall, and Szekely 2003; Riboud, Savchenko, and Tan 2006).

Figure 2.16 shows the patterns in Latin America in the 1990s (Behrman, Birdsall, and
Szekely 2003). Colclough, Kingdon, and Patrinos (2010) suggest that the causes of such
patterns lie mainly on the supply side: educational expansion at the basic level leads to
rising qualifications required for particular jobs, making it increasingly difficult to find
jobs without an education beyond the basic level. On the demand side, increased
openness to trade and foreign investment might also have driven up demand for high-
skilled workers.

Figure 2.16. Changes in Marginal Returns to Education in the 1990s in Latin America




Source: Behrman, Birdsall and Szekely 2003.

On the one hand, these trends send a strong signal that government should continue to
expand the upper secondary and tertiary levels of education. On the other hand, the high
private returns to higher levels of education also imply that, for relatively poor provinces
such as Yunnan, public financing of upper secondary and tertiary education should focus
more on equity and supporting low-income students, while letting high-income families
share the private costs.

At the upper secondary level, both in Yunnan and nationwide, the returns to secondary
TVET appear to be higher than the returns to a general high school education.
Nationwide, the gap between returns to general a high school education and secondary
TVET has been widening over time, with returns to high school slightly increasing
between 1995 and 2002, and eventually decreasing between 2002 and 2007. In contrast
to the national trend, in Yunnan, returns to a general high school education exhibit a
37


steeper increase between 1995 and 2002 than returns to secondary TVET, narrowing the
gap between the two. Our data from Kunming suggest that in 2012, holders of secondary
TVET diplomas only earn about five percent more than holders of high school diplomas.
The narrowing gap between secondary TVET and general high school in Kunming seems
to suggest that secondary high school degrees and TVET degrees are becoming less
differentiated in terms of their returns in Yunnan’s labor market.

We also tested whether rates of return to an additional year of education vary by gender,
ethnicity, or enterprise ownership. The results are presented in appendix A. Nationwide,
women have higher returns to education, and the difference is statistically significant.
The difference was most pronounced in 2002, at 3 percent, and dropped slightly to 1
percent in 2007. Workers of ethnic minority groups do not seem to have returns to
education that are much different from those of Han ethnicity. In 2007, we are able to
categorize firm ownership, and we find that employees of SOEs, government agencies,
and public institutions all have lower returns to education than employees of entities that
are not publicly owned, and the difference is statistically significant. Employees of SOEs
seem to have as much as 4 percent lower returns to education compared to non-public
enterprises. Although we also conducted the same analysis using Yunnan’s sub-sample,
most of these differences are not statistically significant. In 1995, women seemed to have
2 percent higher returns to education. In urban Kunming, working in government
administration also generates 3 percent higher returns for each additional year of
education than working for other types of employers.

Wage Returns to Skills

Trends in rates of returns imply an increasing demand for workers with upper secondary
education and above. In this section, we estimate returns to cognitive, non-cognitive and
technical skills beyond education as measured by the STEP household survey.
Definitions of all measures used in the analysis and full estimation results are presented
in appendix A.

Among cognitive skills, reading and writing skills are associated with higher wages.
Workers who read at work earn on average 8 to 12 percent more than those who do not
read. Furthermore, the difference remains positive and significant after we control for
education levels, occupations and sectors. Workers whose work involved writing more
than 25 pages earn about 19 to 23 percent more than those who do not write at work. In
comparison, more advanced levels of numeracy skills are not associated with higher
wages except that people who need to perform simple tasks such as measuring and
calculating prices earn 11 percent more than those who do not do any math at work,
holding education levels constant. However, the association is no longer significant once
we control for occupations, suggesting that the demand for basic numeracy skills might
be specific to occupation. In addition, higher-wage jobs appear to be more cognitively
challenging as workers who have to spend at least 30 minutes thinking at work every day
earn 20 to 25 percent more than those whose jobs barely involve any thinking, and the
difference is significant even after controlling for education levels, occupations and
sectors.
38


A number of technical skills appear to be associated with higher wages across and within
industries and occupations. For example, jobs that involve computer use pay 9 to 17
percent more than those which do not— even after controlling for education levels,
sectors and occupations. Higher paying jobs also appear to offer workers more freedom
and autonomy. Fluency in a foreign language is associated with about 11 percent higher
wages. The difference is no longer significant once we control for non-cognitive skills
Interestingly, jobs that involve more contact with people outside of co-workers, such as
firm clients, seem to pay less.

We also estimated the returns to non-cognitive skills, and the results are presented in
Appendix A. Most of the non-cognitive skills do not seem to be significantly associated
with wages, except for openness, which measures creativity and how open people are to
new experiences. Openness seems to be correlated with higher education levels, but even
after controlling for education levels, sector, industry, and cognitive and technical skills,
higher measures of openness remain significantly associated with higher wages.

Demand for Skills: Evidence from STEP Employer Survey

Analysis of the STEP Household survey data from Kunming suggests that jobs that
involve more reading, writing and thinking pay better. It also suggests that technical
skills such as computer use and foreign languages are demanded across occupations and
sectors, in addition to level of education. Moreover, people rated to be more open also
earn higher wages in our analysis. In this section, we examine the demand for skills from
the employers’ perspective. Which occupations are in high demand? What cognitive,
non-cognitive and technical skills are most sought after?

According to the STEP Employer Survey (box 2.3), the demand for workers in every type
of occupation has been growing, but the fastest growth in demand is for technicians,
professionals, and managers.

                               Box 2.3. STEP Employer Survey 2012

The STEP Employer Survey was conducted in Kunming, Yunnan, in 2012. The survey selected
300 non-public enterprises in Kunming using stratified random sampling based on industry and
firm size. The survey contains such information as the firms’ background information, hiring and
recruiting practices, employers’ perception of vocational and general education, and
compensation.

The survey helps identify the cognitive, non-cognitive and technical skills most used and sought
after by employers in each occupational category. Ten occupational categories are grouped into
type A, which demands more skills (managers, professionals, technicians/associate
professionals), and type B (clerical support workers, service workers, sales workers, skilled
agricultural/forestry/fishery workers, construction/craft and related trades workers, plant and
machine operators/assemblers, and elementary occupations).

In figure 2.17, the starting point of each arrow represents the average number of workers
in each category 12 months before, while the endpoint represents the average number of
workers expected within 12 months after the survey. The purple dots mark the average
39


number of workers during the time of the survey, and the blue bars represent the expected
growth rate. Among all occupations, the number of technicians/associate professionals is
expected to double in 12 months’ time, the highest growth rate. The number of managers
is also expected to grow by 40 percent. Kunming currently has the highest number of
plant and machine operators/assemblers/drivers, and the number is expected to continue
to grow by over 20 percent. In addition, the number of construction/craft and related
trades workers is also expected to grow by over 30 percent in 12 months’ time.

Figure 2.17. Number of Workers Twelve Months Ago, Currently, and Expected in Twelve Months


                                                Growth rate expected next year   How many current are there in the firm

                                         100%                                                                             50
     Growth rate expected in 12 months




                                          90%                                                                             45
                                          80%                                                                             40




                                                                                                                               Number of workers
                                          70%                                                                             35
                                          60%                                                                             30
                                          50%                                                                             25
                                          40%                                                                             20
                                          30%                                                                             15
                                          20%                                                                             10
                                          10%                                                                             5
                                           0%                                                                             0




Source: STEP Employer Survey 2012.
Note: The starting point of each arrow represents the average number of workers in each category
twelve months before the survey; the endpoint represents the average number of workers
expected within twelve months after the survey. The purple dots mark the average number of
workers during the time of the survey, and the blue bars represent the expected growth rate.

The STEP Employer Survey also shows the specific skills in demand for different types
of occupations. Figure 2.18 illustrates the job-related skills considered most important by
firms deciding which new employees to retain after a probation period. For type A
occupations, 34 percent of the firms consider “leadership skills� as the most important,
followed by “job-specific technical skills,� and “communication skills.� For type B
occupations, “communication skills�, “job-specific technical skills,� and “numeracy
skills� are most commonly cited as the most important job-related skills.
40

Figure 2.18. Percent of Firms Reporting Various Job-Related Skills as Important to Employee Retention

a.        Type A occupations
                                           Leadership skills                                        33.56%
                                Job-specific technical skills                              27.16%
                                     Communication skills                         16.84%
            Ability to read and write in a foreign language             5.82%
                                                      Literacy         5.02%
                                                    Numeracy          4.24%
                              Creative and critical thinking        2.40%
                                     Problem solving skills        1.90%
                                           Teamwork skills        1.51%
                            Ability to work independently        0.94%
                                  Time management skills         0.60%


b.        Type B occupations

                          Communication skills                                                42.41%
                    Job-specific technical skills                                23.12%
                                     Numeracy                           13.23%
                               Teamwork skills                  7.26%
                       Time management skills                5.22%
                Ability to work independently             2.59%
                          Problem solving skills         1.91%
                                        Literacy        1.41%
                               Leadership skills        1.22%
                  Creative and critical thinking       0.79%
 Ability to read and write in a foreign language       0.56%

Source: STEP Employer Survey 2012.
Note: Type A occupations include managers, professionals, technicians/associate professionals;
type B occupations include clerical support workers, service workers, sales workers, skilled
agricultural/forestry/fishery workers, construction/craft and related trades workers, plant and
machine operators/assemblers, and elementary occupations.

Figure 2.19 reveals the personality traits most highly valued by firms when deciding
which new employees to retain after a probation period. For both types of occupations,
over 70 percent of the firms consider conscientiousness (“Does a thorough job, is hard
working, does things efficiently�) as the most important personality trait.
41

Figure 2.19. Percent of Firms Reporting Various Personality Traits as Important to Employee Retention

a.      Type A occupations


       Conscientiousness                                                               73.51%

 Openness to experience                      20.56%

          Agreeableness       2.62%

            Extraversion      2.09%

      Emotional stability    1.22%




b.      Type B occupations


       Conscientiousness                                                               72.87%

            Extraversion               13.44%

          Agreeableness            8.41%

 Openness to experience       2.78%

      Emotional stability     2.23%



Source: STEP Employer Survey 2012.
Note: Type A occupations include managers, professionals, technicians/associate professionals;
type B occupations include clerical support workers, service workers, sales workers, skilled
agricultural/forestry/fishery workers, construction/craft and related trades workers, plant and
machine operators/assemblers, and elementary occupations.

Finally, firms were asked to rank among job-related skills, personal characteristics, and
personality traits in deciding which new hires to retain (figure 2.20). Not surprisingly, for
both types of occupations, the vast majority of the firms consider job-related skills to be
the top factor affecting their decision to retain new hires after a probation period.
Personality traits are considered more important than personal characteristics such as
appearance, age, and gender.
42

Figure 2.20. Percent of Firms Reporting Various Factors as Important to Employee Retention


                             Job-related skills
                                        67%
                                         59%




                                                                                 Type A

                              7%                                                 Type B
                           17%                25%
                                             24%
        Personal
                                                      Personality traits
      characteristics




Source: STEP Employer Survey 2012.
Note: Type A occupations include managers, professionals, technicians/associate professionals;
type B occupations include clerical support workers, service workers, sales workers, skilled
agricultural/forestry/fishery workers, construction/craft and related trades workers, plant and
machine operators/assemblers, and elementary occupations.

Skills Gaps and Mismatches

Compared to other parts of China, a skills shortage was not yet a major factor impeding
Yunnan’s growth in the mid-2000s. Thirteen percent of the enterprises surveyed by the
World Bank’s Investment Climate Assessment (2005) in Yunnan indicated worker skills
and education level as the most or second most severe factors constraining growth,
placing it 7th among the 14 factors listed on the survey. As shown in figure 2.21, the top
constraints faced by the firms are electricity, access to finance, and transport. In
comparison, when we take into account all 120 cities across China, as many as 28 percent
of the firms surveyed list worker skills and education level as the most severe constraint,
the second most common of all factors.

Yunnan undertook economic reforms at a relatively later date than the rest of China,
especially in the coastal areas. Thus, the survey likely reveals that the challenges faced
by Yunnan are different from those faced by major cities in other parts of China:
infrastructure and access to finance were the most pressing concerns for firms’ growth in
Yunnan in 2005.
43

Figure 2.21. Firms Facing Factors Constraining Growth in Yunnan and China, 2005


                                                    China     Yunnan

                                          Electricity
                                  Access to finance
                                          Transport
  anti-competition behaviors by other enterprises
                                      Financing cost
   Unstable economic and administration policies
                worker's skills and education level
                                Local protectionism
                                 Tax Administration
                                     water services
 Access to infomormation on laws and regulations
                                        Crime, theft
                                    Communication
                                           Customs
                                                        0   0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5

Source: World Bank 2005.
Note: The figure shows the percentage of firms that consider the factor as the most or second-
most severe factor constraining growth.

The most recent employer survey conducted in Kunming in 2012 suggests that the
situation has changed: labor factors are now considered the most pressing in
constraining operations and growth. Among various labor factors, “finding workers with
previous experience� and “[TVET] of workers� are the two most severe factors impeding
the operation and growth of business, as indicated by figure 2.22. In comparison, the
general education of workers and availability of labor are considered less problematic.

Figure 2.22. Rankings of Labor Factors Impeding Business Operation and Growth


              Can you please indicate how problematic each of the following
              labor factors is for the operation and growth of your business?
                             (1=no problem, 5=severe problem)
 FINDING WORKERS WITH PREVIOUS EXPERIENCE                                                             3.54
                                                                                               3.18
            PAYROLL TAXES AND SOCIAL SECURITY…                                          2.79
                                                                                       2.71
              GENERAL EDUCATION OF WORKERS                                            2.69
                                                                                     2.60
                                 MINIMUM WAGE                                      2.42
                                                                                  2.35
        EMPLOYMENT PROTECTION LEGISLATION/…                                    2.11

Source: STEP Employer Survey 2012.
44


Moreover, on average, firms in Kunming considered only around 82 to 84 percent of their
employees qualified for the job. Illustrated by the histogram in figure 2.23, fewer than 45
percent of the firms considered over 90 percent of their employees qualified for their jobs.
Nearly 10 percent of the firms considered 30 percent or fewer of their managers,
professionals, and technicians qualified.

Figure 2.23. Percent of Employees Considered Qualified for their Jobs

                  50%
                  40%
     % of firms




                  30%
                  20%
                                                                                       Type A
                  10%
                                                                                       Type B
                  0%




                              % of employees qualified for the job

Source: STEP Employer Survey 2012.
Note: Type A occupations include managers, professionals, technicians/associate professionals;
type B occupations include clerical support workers, service workers, sales workers, skilled
agricultural/forestry/fishery workers, construction/craft and related trades workers, Plant and
machine operators/assemblers, and elementary occupations.

Potential skills gaps and mismatches are further indicated by the job vacancy data.
Throughout 2010 and 2011, although the total number of job openings in Yunnan
exceeded the pool of applicants, only around 50 to 55 percent of the applicants were
hired, filling 35 to 45 percent of the total openings (figure 2.24).

Figure 2.24. Quarterly Job Vacancies and Applications in Yunnan, 2010–11

 250000

 200000

 150000
                                                                  # openings
 100000
                                                                  # applicants
  50000                                                           # applicants hired
                  0




Source: HRSS 2010a, 2011.
45


To further characterize the potential skills gaps and mismatches, we draw on insights into
firm hiring of new employees as revealed by the STEP Employer Survey. According to
the survey, for firms in Kunming, it takes on average 34 days to fill a
manager/professional/technician position, that is, much longer than the 16 days needed to
fill a less-skilled type B position. On average, each job opening receives 14 to 16
applications. Moreover, more than 9 percent of the firms have used contractors for skill
shortages in type A positions, and more than 17 percent have used contractors for type B
positions.

More firms have encountered problems hiring professional and technical workers than
hiring workers in lower-skilled occupations (figure 2.25). Among firms that have tried to
hire in the past year, 89 percent have encountered problems hiring sales workers, 74
percent hiring professionals, and 72 percent hiring technicians/associate professionals. In
comparison, only 1.4 percent and 30 percent of the firms have encountered problems in
trying to hire skilled agricultural/forestry/fishery workers and elementary occupation
workers, respectively.

Figure 2.25. Percent of Firms Encountering Problems Hiring Employees at Various Levels

                                                                                                                                                     89%
                                                                                                                     72%             74%
                                                                   58%        63%                64%
                                     37%              46%
                     30%
        1%
     forestry and…




                                   related trades…



                                                     operators,…


                                                                   Managers




                                                                                                 Service workers


                                                                                                                   Technicians and
                                                                              Clerical support




                                                                                                                                                     Sales workers
                     occupations




                                                                                                                                     Professionals
     agricultural,


                     Elementary




                                                     Plant and
                                    Constuction,




                                                      machine




                                                                                                                    professionals
                                      Craft and
        Skilled




                                                                                                                      associate
                                                                                  workers




Source: STEP Employer Survey 2012.

Table 2.4 and table 2.5 show the specific problems encountered in hiring. Across type A
occupations, the lack of skills and mismatches of expected wages are the two most
common problems encountered. Lack of skills is identified as an issue by 73-81 percent
of the firms hiring professionals and technicians/associate professionals. For type B
occupations, “applicants lacked required skills� is again one of the most commonly cited
problems encountered in hiring. In particular, over 91 percent of the firms reported the
problem when hiring plant and machine operators/assemblers/drivers. Further, in hiring
construction/craft and related trade workers, over 70 percent of the firms encounter the
problem of “no or few applicants.� Unmatched wage expectation is also a common
problem in the hiring of clerical support workers, service workers, skilled
agricultural/forestry/fishery workers, and elementary occupations. Working conditions
have posed challenges in hiring service workers and elementary occupation workers.
 46

 Table 2.4. Percent of Firms Encountering Hiring Problems, Type A Occupations

                                                                                   Technicians
                                                                                   and associate
                                                   Managers          Professionals professionals
 There are no or few applicants                    40                38            34
 Applicants lacked required skills                 58                81            73
 Applicants expected wages higher than we can
 offer                                             77                56              63
 Applicants did not like working conditions        22                32              51
Source: STEP Employer Survey 2012.
Note: The two most common problems are highlighted in gray.

 Table 2.5. Percent of Firms Encountering Hiring Problems, Type B Occupations

                     Clerical    Service     Sales      Skilled     Constru     Plant      Element
                     support     workers     workers    agricultu   ction,      and        ary
                     workers                            ral,        craft and   machine    occupati
                                                        forestry    related     operator   ons
                                                        and         trades      s, and
                                                        fishery     workers     assembl
                                                        workers                 ers,
                                                                                drivers
 There are no or
 few applicants    20       25         35          34               71          43         9
 Applicants
 lacked required
 skills            60       61         66          32               51          91         77
 Applicants
 expected wages
 higher than we
 can offer         65       56         76          38               61          27         74
 Applicants did
 not like working
 conditions        49       73         66          34               24          23         88
 Source: STEP Employer Survey 2012.
 Note: The two most common problems are highlighted in gray.


 Drivers of Future Demand for Skills

 As China’s economic reforms deepen and spread to the interior provinces, the enormous
 opportunities for development in Yunnan will also bring new skills challenges. The
 central government has projected three major transformations as the guiding principles
 for China’s economic growth in the new era moving from: (1) an investment-led to
 consumption-led growth; (2) a high-energy to a low-carbon based economy; and (3)
 relying on low-cost manufacturing to relying on science, technology, and innovation.
 Guided by these principles, Yunnan has been pursuing its own development strategy of
 building a strong province with a green economy, diverse culture, and a bridgehead for
47


opening China to South Asia and Southeast Asia. Economic growth and structural
changes, as well as the Bridgehead Strategy, as specified in Yunnan’s 12th Five-Year
Economic Development Plan (12th Five-Year Plan), are expected to further drive the
demand for skills.

Driver 1: Economic Growth and Structural Changes
Yunnan’s real gross domestic product (GDP) has been growing at an annual rate of over
10 percent in recent years. The 12th Five-Year Plan sets goals for Yunnan of maintaining
its annual growth rate and reaching a GDP of more than Y1,163 billion in 2015. Table
2.6 lists some of the other goals set out in the plan pertaining to economic growth and
structural changes.

Table 2.6. Selected Goals Specified in Yunnan’s 12th Five-Year Plan

               Indicator                       2010              2015        Annual growth        Total
                                                                                                growth
 Yunnan GDP (2010Y)                  722 billion            > 1163 billion   > 10%           n.a.
                       a
 Industrial composition (%)          15:45:40               12:46:42         n.a.            n.a.
 Light:heavy sector (%)              46.2:53.8              48.2:51.8        n.a.            n.a.
 Value added by tertiary sector (%)  40                     > 42             n.a.            >2
 Share of employment in tertiary 28.8                       35.7             n.a.            6.9
 sector (%)
 Urbanization rate (%)               36                     45               2               9
 New urban employment (persons)      24 million             22 million       n.a.            n.a.
Source: Yunnan 12th Five-Year Plan (2011–2015).

a. Consisting of primary, secondary, and tertiary sectors.

How much of the GDP growth can be translated into employment and jobs? Although the
number of people employed has been increasing with real GDP, employment growth
rates are much lower than GDP growth rates. Figure 2.26 illustrates the trends in GDP
growth and employment in Yunnan over the past decades.
48

Figure 2.26. Real GDP and Number of People Employed in Yunnan, Real and Projected, 1957 –2015

 16000                                                                      3000
 14000
                                                                            2500
 12000
                                                                            2000
 10000
  8000                                                                      1500
  6000
                                                                            1000
  4000
                                                                            500
  2000
      0                                                                     0
          1974
          1975




          2010
          2011
          1957
          1962
          1965
          1966
          1967
          1968
          1969
          1970
          1971
          1972
          1973
          1976
          1977
          1978
          1979
          1980
          1981
          1982
          1983
          1984
          1985
          1986
          1987
          1988
          1989
          1990
          1991
          1992
          1993
          1994
          1995
          1996
          1997
          1998
          1999
          2000
          2001
          2002
          2003
          2004
          2005
          2006
          2007
          2008
          2009
          2012
          2013
          2014
          2015
              Real GDP(1952=100)                Projected GDP
              Number of people employed

Source: Yunnan Statistical Yearbook 2011; Yunnan 12th Five-Year Plan (2011–2015).

The relationship between real GDP growth and job growth rates also varies by sector, as
shown in figure 2.23. The primary sector, while employing the largest number of laborers,
has seen the slowest growth in real GDP. Further, total employment has been decreasing
in recent years. The secondary sector in Yunnan has seen rapid growth in real GDP, but
employment has remained flat with slight increases in recent years. In comparison, while
the tertiary sector hasn’t grown as quickly as the secondary sector in terms of real GDP, it
has the fastest growth in demand for labor.

Figure 2.27. Real GDP and Employment by Sector, 1957–2008

a. Primary sector
          40000                     1800
          35000                     1600
          30000                     1400
          25000                     1200
                                    1000
          20000
                                    800
          15000                     600
          10000                     400
           5000                     200
              0                     0
                   1957
                   1968
                   1973
                   1978
                   1983
                   1988
                   1993
                   1998
                   2003
                   2008




                  Real GDP   Employment
49


b. Secondary sector

 40000                       1800
 35000                       1600
 30000                       1400
 25000                       1200
                             1000
 20000
                             800
 15000                       600
 10000                       400
  5000                       200
     0                       0
          1957
          1968
          1973
          1978
          1983
          1988
          1993
          1998
          2003
          2008



         Real GDP     Employment


c. Tertiary sector

 40000                      1800
 35000                      1600
 30000                      1400
 25000                      1200
                            1000
 20000
                            800
 15000                      600
 10000                      400
  5000                      200
     0                      0
          1957
          1968
          1973
          1978
          1983
          1988
          1993
          1998
          2003
          2008




         Real GDP     Employment

Source: Yunnan Statistical Yearbook 2011.

Over the next five years, the secondary and tertiary sectors are expected to continue to
grow, with the tertiary sector expected to generate increasing employment. The structure
of the economy is to be further optimized and diversified, leveraging Yunnan’s unique
resources and advantages. The composition of Yunnan’s economy has been gradually
moving away from one dominated by agriculture. In 2010, the secondary sector
contributed the most to Yunnan’s economy (45 percent), followed by the tertiary sector
(40 percent) and the primary sector (15 percent). The shares of the secondary and tertiary
sectors in the economy are expected to continue to grow by 2 percent each over the next
five years. Within the secondary sector, the government particularly encourages the
development of the labor-intensive light sector. Furthermore, the tertiary sector is
expected to account for 36 percent of total employment by 2015, a 7 percent increase
from 2010.
50


Driver 2: Bridgehead Strategy
Recognizing its strategic location, the Chinese government has formulated a Bridgehead
Strategy (桥头堡) to position Yunnan as: (i) a strategic gateway to the Southwest; (ii) an
experimental zone for opening up China’s southwestern borders; (iii) a base for export-
oriented, distinct, and competitive industries in the western region; (iv) a treasure house
of biodiversity and a safety net for the ecosystem in the Southwest; and (v) a national-
level model zone for ethnic harmony and border stability.

The national strategy is expected to further accelerate Yunnan’s provincial development
and strengthen its openness and exchange with neighboring countries and provinces.
Yunnan has been upgrading its transportation infrastructure connecting to Southeast Asia
and building a pipeline for oil and gas supplies. More cooperation and exchange is
expected, both internationally with the Greater Mekong Sub-region and domestically
with other economic zones within China.

Yunnan will adopt both the strategy of “coming in,� by allowing more foreign investment
and attracting financial and human capital from abroad, and the strategy of “going out,�
by leveraging the resources and demand in the foreign market and increasing the scale
and effectiveness of investments abroad. The Bridgehead Strategy will not only create
more employment in the secondary sector, but more importantly, drive the demand for
high-skilled professionals and technicians in transportation, logistics, international trade,
and finance.

In order to sustain economic growth and reach its development goals, Yunnan will need
to increase the number of skilled laborers, improve the quality of human resources, and
optimize the skills composition of its labor force. Specific targets have been laid out in
Yunnan’s 12th Five-Year Plan, as well as in its Medium- and Long-Term Plan for Talent
Development.iv Table 2.7 lists Yunnan’s most important targets, a list that can also be
read as the projected demand for skills. Consistent with our previous analysis, Yunnan
has aimed to raise its overall level of human resources, with a focus on developing
professional and technical skills. The plans also list 10 key sectors of strategic importance,
as well as urgent needs for skills concentrated in the secondary and tertiary sectors.
51

Table 2.7. Projected Demand for Skills in Yunnan, 2010 –20

                                                                      2010          2015            2020
 Overall goals
 Average years of education received by new labor                                              n.a.
 force (years)                                                    8          12.5
 Total amount of talent resources (millions of people)                       n.a.
                                                                  2.51                         5
 Talent share of total human resources (%)                        n.a.       n.a.              13
 Proportion of GDP spent on human resources (%)                   n.a.       n.a.              10
 Talent contribution rate (%)                                     n.a.       n.a.              20
   GERa at upper secondary level (%)                              65         85                n.a.
 R&Db spending as percent of GDP (%)                              0.6        1.5               n.a.
   Develop high-level innovative talent
 Proportion of labor force with tertiary education (%)            n.a.       n.a.              15
 R&Db employees per 10,000 laborers                               n.a.       n.a.              15
 Selected types of talent
 Party and government officials holding college degree            n.a.       n.a.              87
 and above (%)
 Enterprise management talent                                     n.a.       575,400           729,500
 Professional and technical talent (millions)                     n.a.       1.36              1.58c
 High-skilled workersd
 Total number of high-skilled workers                 420,000 613,900           811,900
 Number certified at Levels 1 and 2                   45,000 70,000             100,000
 High skilled workers’ share of total skilled labor (%)
                                                      21         25             26
 Rural practical talent (millions of people)          n.a.       1.03           1.57
 Ten industries urgently needing talent               tobacco, energy, nonferrous metals,
                                                      black metals, petroleum and
                                                      chemicals, biotechnology, tourism
                                                      and culture, commercial and trade
                                                      circulation, equipment
                                                      manufacturing, optic-electronics
Source: Yunnan 12th Five-Year Plan; Yunnan Talent Development Plan, Yunnan High-Skilled
Talent Cultivation Medium- and Long-Term Plan (2010–20).
Note: n.a. = not applicable.
a. GER = gross enrollment ratio.
b. R&D = research and development
c. Target distribution is senior (10 percent), medium (40 percent), junior (50 percent).
d. High-skilled workers are roughly defined as technicians who are certified as senior (Level 3), technicians
(Level 2) and senior technicians (Level 1).

Summary

In this chapter, we have reviewed the education and skills composition of the current
workforce in Yunnan, which reveals an increasing demand for skills. Labor force and
employer surveys shed light on the specific skills in demand and reveal emerging skills
gaps and mismatches in Yunnan.

Employment in Yunnan has been historically dominated by the agriculture sector, but the
shares of secondary and tertiary employment have been on the rise over the past decade.
52


In recent years, employment has been driven mostly by the tertiary and secondary sectors.
At the same time, employment in SOEs and collectives has been declining. Today the
non-public sector absorbs nearly half of the total employment in Yunnan.

Although the education levels of the labor force have been increasing, the skills levels of
the labor force in Yunnan remain low compared to the national average, as well as to
other parts of China. A majority of workers have completed only compulsory education
or less. Further, the primary and secondary sectors, employing the largest number of
workers, are composed mostly of frontline/operative skilled workers. Skill certification
rates are also low among technical, professional, and frontline/operative skilled workers.

Nevertheless, a rising demand for skills in Yunnan is indicated by the rising wage returns
to education. Rates of return for an additional year of education in Yunnan have
increased from more than 3 percent in 1988 to more than 7 percent in 2002. During the
same period, returns to every level of education beyond the junior secondary level
increased in Yunnan, with the most dramatic increase occurring at the tertiary level. Gaps
in returns between tertiary and secondary education and between Bachelor’s degrees and
associate degrees have been widening, indicating a particularly strong demand for higher-
skilled workers with tertiary education, especially those with Bachelor’s degrees.

A number of cognitive, non-cognitive and technical skills measured by the STEP
Household survey are associated with higher wages— even after controlling for level of
education. According to our analysis of data from Kunming, higher-wage jobs require
more intense reading, writing and thinking skills. In addition to higher levels of education,
computer and foreign language skills are technical skills in demand across occupations
and sectors. Among non-cognitive skills measures, openness appears consistently
associated with higher wages.

Demand for higher skills is further revealed from employers’ perspectives. Firms in
Yunnan increasingly seek to hire high-skilled professionals and technicians, and the
demand for frontline/operative skilled workers remains high. Apart from job-specific
technical skills, firms also consider “leadership skills� and “communication skills� to be
among the most important job-related skills. Conscientiousness has been cited by over 70
percent of the firms as the most important non-cognitive skill.

As the demand for skills has grown, skills shortages and mismatches have begun to
emerge. Among various labor factors, “finding workers with previous experience� and
“[TVET] of workers� appear to be the factors most severely impeding the operation and
growth of businesses. Skills shortages and mismatches have caused problems for firms
trying to hire new employees and prevented them from filling job vacancies. Particularly,
in hiring for managerial, professional, and technical positions, the “lack of required skill�
and “mismatch of wage expectation� are the two most common problems. For
frontline/operative skilled workers, “lack of required skill� and “lack of applicants� are
the two major problems.

The latest policy developments specified in Yunnan’s 12th Five-Year Plan are expected
to intensify the demand for skills in the future. To sustain economic growth and carry out
53


the Bridgehead Strategy, Yunnan will need to increase the overall quality of human
resources and optimize the composition of skills within its labor force. A key to bridging
the gaps and addressing the skills challenges faced by firms in Yunnan lies in
strengthening its education and training system. How can the quality and goals of the
formal education and training system be better aligned with the skills demands of
employers? How can Yunnan promote the access to, relevance, and quality of non-formal
training to improve the skills of employees? These issues will be addressed in detail in
the following chapters.

References
Behrman, J., N. Birdsall, and M. Szekely. 2003. “Economic Policy and Wage
Differentials in Latin America.� Center for Global Development Working Paper 29,
Washington, DC.

Blom, A., L. Holm-Nielsen, and D. Verner. 2001. “Education, Earnings and Inequality in
Brazil, 1982–1998: Implications for Education Policy.� Peabody Journal of Education
76 (3–4): 180–221.

Chinese Household Income Project. 1988.

Chinese Household Income Project. 1995.

Chinese Household Income Project. 2002.

China Labor Statistical Yearbook. 2011. Beijing: China Statistics Press.

China Statistical Yearbook. 2011. Beijing: China Statistics Press.

Colclough, Christopher, Geeta Kingdon, and Harry Patrinos. 2010. “The Changing
Pattern of Wage Returns to Education and its Implications.� Development Policy Review
28 (6): 733–47.

Fiszbein, A., H.A. Patrinos, and P.I. Giovagnoli. 2007. “Estimating the Returns to
Education in Argentina Using Quantile Regression Analysis: 1992–2002.� Economica 53
(1–2): 53–72.

Keith, Griffin, and Renwei Zhao. 1993. The Distribution of Income in China. London:
Macmillan.

Li, Shi, C. Luo, Z. Wei, and X. Yue. 2008. “Appendix: The 1995 and 2002 Household
Surveys: Sampling Methods and Data Description.� In Inequality and Public Policy in
China, ed. B. Gustafsson, S. Li and T. Sicular, 337–54. Cambridge: Cambridge
University Press.

Longitudinal Survey on Rural Urban Migration in China (RUMiC). 2007.

Mincer, J. 1974. Schooling, Experience, and Earnings. New York: NBER Press.
54


Sanchez-Puerta, Maria Laura, Alexandria Valerio, Gaelle Pierre, and Sergio Urzua. 2012.
“STEP Skills Measurement Surveys Methodology Note Draft,� World Bank, Washington,
DC.

Riboud, M., Y. Savchenko, and H. Tan. 2007. The Knowledge Economy and Education
and Training in South Asia. Washington, DC: World Bank.

Riskin, Carl, Zhao Renwei, and Li Shi, 2001. China's Retreat from Equlity: Income
Distribution and Economic Transition. Armonk, N.Y.: M.E.Sharpe.

Skills Toward Employment and Productivity Employer Survey. 2012.

Skills Toward Employment and Productivity Household Survey. 2012.

World Bank. 2005. Investment Climate Assessment. Washington DC: World Bank.

———. 2012. China 2030 Conference Edition. Annex Table 10 (pg. 214).

Yunnan Bureau of Statistics. 2008. Yunnan Non-public Enterprise Survey. Beijing:
China Statistics Press.

———. 2011. Yunnan Statistical Yearbook. Beijing: China Statistics Press.

HRSS (Yunnan Department of Human Resources and Social Security). 2010a. Quarterly
Vacancy Data.

———. 2010b. Yunnan Medium- and Long-Term Plan for Talent Development,
http://yn.yunnan.cn/html/2010-12/13/content_1438313.htm.

———. 2011. Quarterly Vacancy Data.

———. Yunnan High-Skilled Talent Cultivation Medium- and Long-Term Plan (2010–
20).


Notes

i
   Except for 1988, where we estimated the number of years according to the highest level of education
reported by individuals.
ii
    Years of experience = age minus years of schooling minus six.
iii
    Due to the limitations of the data, minority is not controlled for in estimating returns for Yunnan
(Kunming) 2012.
55


Chapter 3      Overview of Yunnan’s Education and Training System


Yunnan’s education and training system is the key to improving the skills of its population and
ensuring a sufficient supply of skills to meet future challenges. This chapter offers an overview
of the formal and non-formal education and training systems in Yunnan.

According to the provincial Department of Human Resources and Social Security, Yunnan’s
population totaled 46 million in 2010 and grew at a rate greater than 6 percent. In 2010, 72
percent of Yunnan’s population was of working age. The total workforce amounted to 36 million
in 2010, of whom 0.67 million were new entrants into the labor force. Demographically, Yunnan
has a high birth rate of 13 percent as of 2010, which is more than one percent higher than the
national average. Youth aged birth to 14 years account for 21 percent of Yunnan’s total
population.

Although the total supply of labor seems to be sufficient, the large population does not seem to
ensure an attendant high stock of skills. Less than 15 percent of the total population has attained
education beyond the compulsory level, compared to 23 percent nationwide. The illiteracy rate
remains at 6 percent, almost 2 percent higher than the national average. At the same time that
skills gaps are starting to emerge, Yunnan is also faced with a labor surplus, especially among
the 65 percent of its population still residing in rural areas.

In China, education and training take place in both formal and non-formal settings. In this report,
all diploma-oriented education programs are treated as part of the formal education and training
system. Also included are adult learning and distance learning programs. Non-formal education
and training include various non-diploma-oriented training programs.

Formal Education System

Figure 3.1 illustrates the structure of the formal education system in China. Highlighted in gray
is the vocational track at different levels. The system starts with kindergarten and preschool for
early childhood development at the pre-primary level. Compulsory education starts at age six,
with a duration of nine years through the end of middle school at the lower secondary level.
There are a small number of vocational schools at the lower secondary level.

At the upper secondary level, Technical and Vocational Education and Training (TVET) is
provided at vocational high schools, secondary specialized schools, and secondary skilled-
worker schools. The main governance bodies are the Ministry of Education, which supervises
vocational high schools and secondary specialized schools, and the Ministry of Human
Resources and Social Security, which supervises secondary skilled-worker schools. At the
tertiary level, TVET is provided mostly in TVET colleges, advanced specialized schools, and
advanced skilled-worker schools. Some universities also offer vocational programs. Most of the
TVET programs grant associate degrees. Although the majority of secondary TVET graduates
enter the job market, some of them advance into tertiary-level programs, mostly associate degree
programs or other TVET programs.
56


The share of Yunnan’s students in basic-level education (currently enrolling over 6 million
students), although gradually decreasing, is more than four times that of students in upper
secondary-level education (appendix table B.1). The shares of students in the secondary and
tertiary education systems are expanding, as more students are admitted than the number of
graduates. At the secondary level, the size of the general education and the vocational tracks are
about the same, with each enrolling about 0.6 million students. At the tertiary level, Bachelor’s
degree programs currently enroll more students (0.26 million) than associate degree programs
(0.18 million), and are expanding more quickly.

Figure 3.1. Structure of the Formal Education System, Yunnan




Despite the large number of students enrolled, enrollment rates and advancement rates are low
and vary greatly across cities and prefectures within Yunnan (figure 3.2). For example, although
the net enrollment rates at the primary level have exceeded 99 percent, only 96 percent of the
primary school graduates advance to the lower secondary level. In Baoshan city, the
advancement rate is only 85 percent. Further, only 73 percent of middle school graduates
advance to the upper secondary level, and in 10 out of the 16 cities and prefectures, advancement
rates to the senior secondary level are even lower than the provincial average. The highest
57


advancement rate (94 percent in Yuxi) is more than double the lowest rate of 46 percent in
Nujiang.

Figure 3.1. Advancement Rates Across Cities and Prefectures

a. Advancement to lower secondary school
         105

         100

           95

           90

           85

           80

           75
                      Yuxi




                     Pu'er
                   Qujing




                   Banna
                  Honghe
                 Baoshan




                  Lincang
                   Lijiang



                Chuxiong




                       Dali
                Wenshan
                  Yunnan
                 Kunming




                Zhaotong




                  Nujiang
                     Diqin
                  Dehong



b. Advancement to upper secondary school

 100
  90
  80
  70
  60
  50
  40
  30
  20
  10
   0
             Yuxi
          Qujing




            Pu'er



         Honghe
        Baoshan




         Lincang
          Lijiang



       Chuxiong



          Banna
              Dali
       Wenshan
       Zhaotong




         Dehong
         Yunnan
        Kunming




         Nujiang
            Diqin




Source: Yun, Liu, and Li 2011.
Note: Blue and red bars represent advancement rates above and below the provincial average,
respectively.

The relevance and quality of the formal education system are also called into question.
According to the Skills Toward Employment and Productivity (STEP) Survey, among employers
in Kunming (table 3.1), only 56 percent of the firms agree that the current TVET system meets
the skill needs of employers adequately. More specifically, as many as 65 percent of the firms
agree that the current TVET system does not produce enough workers with up-to-date
knowledge of methods, materials, and technology. Over half of the firms also agree that the
58


current TVET system does not produce enough people with a “good attitude and self-discipline�
or “practical skills.�

Likewise, fewer than half of firms consider the general education system to be adequately
meeting the skills needs of employers. Sixty-seven percent of firms think that the general
education system fails to produce enough people with up-to-date knowledge of methods,
materials, and technology. Over half of the firms think that the current general education system
does not produce enough people with either the level or the kinds of skills needed. Aside from
job-related skills, 55 percent of firms think that the general education system does not produce
people with a “good attitude and self-discipline.� Employers’ evaluations shed light on the
weaknesses of the current formal education and training system. Specifically, the current system
fails to keep up with the ever-evolving skills needs of employers, and both the TVET and general
systems are unable to produce enough students with important non-cognitive skills, such as good
attitude and self-discipline.

Table 3.1. Impressions of the TVET and General Education Systems, Yunnan

                                                                                  General education
                                                                           TVET       system
  Meets the skill needs of employers adequately                             56           47
  Does not produce enough people with the level of skills needed by
  employers                                                                 40           56
  Does not produce enough people with the kinds of skills needed by
  employers                                                                 47           50
  Does not produce enough people with up-to-date knowledge of
  methods, materials, and technology                                        66           67
  Does not produce enough people with practical skills                      58           49
  Does not produce enough people with good attitude and self-
  discipline                                                                59           55
 Source: STEP Employer Survey 2012.

Non-formal Training

According to the 2011 quarterly job vacancy data, the largest group of job applicants in Yunnan
is the unemployed (including the newly unemployed, employment-to-unemployed, and others),
followed by rural laborers, new college graduates, and laid-off workers. Efforts to improve the
skills and retrain the current labor force rely on various training programs outside of the formal
education system.

In 2010, the Yunnan Department of Human Resources and Social Security, the Development and
Reform Commission, and the Department of Finance jointly issued an Announcement on
Implementing Special Vocational Training Plan (HRSS 2011). The Announcement identifies six
major types of training programs to be subsidized by public funds: (1) in-service and transfer
training for employees; (2) migrant employee training; (3) rural labor transfer training and labor
preparatory training; (4) reemployment training for the unemployed; (5) entrepreneurship
training; and (6) urgently-needed talent training.
59


The Announcement states that the government should devote at least 15 percent of its
Employment Categorical Fund to subsidizing providers of vocational training. Training
providers, including skilled worker schools, public employment training centers, and private
institutions wishing to participate in subsidized training programs (such as migrant worker
training, reemployment training, and rural labor transfer training) have to go through the initial
application and review process, followed by regular inspection. Training providers participating
in public vocational training programs also get government support for improving infrastructure
and operating conditions.

For work-based training directed at employees, such as in-service, transfer, and migrant
employee training, the government mandates that enterprises retain 1.5 to 2.5 percent of the total
payroll for an Employee Education and Training Fund. In addition, at least 60 percent of the
employee training fund should be spent on technical workers, especially frontline/operative
workers. The government also provides additional subsidies to help employers suffering
financial difficulties, and as incentives for enterprises and individuals (box 3.1).

                     Box 3.1. Government Subsidies for Non-formal Training in Yunnan

ï‚· In-service and transfer training for employees
Enterprises suffering financial difficulties that are participating in the unemployment insurance program
can draw funds from unemployment insurance to finance training and subsidize up to 50 percent of
training fees. Unemployed personnel receive Y700 upon completion of training programs, Y800 upon
certification, Y1300 upon completion of entrepreneurship training, with an additional Y100 for securing
new employment.

ï‚· Migrant employee training
Employers can draw from the Employee Education Fund to provide training for rural migrant workers
within 6 months of their employment. For each migrant worker who gets certified, employers receive a
subsidy of Y400, equal to 50 percent of training fees.

ï‚· Rural worker training
Rural students graduating from middle school or high school who are unable to continue their studies are
encouraged to participate in labor preparatory training, and receive Y800 in subsidies upon certification.

ï‚· Reemployment training for the unemployed
Subsidies of Y688–800 upon certification or reemployment. Without reemployment, payment of the 60
percent of the subsidy for training and certification.

ï‚· Entrepreneurship training
Y300 for participating in training. Y1,000 for starting up a company hiring three to five persons. Y2,000
for starting up a company hiring more than five persons. An additional Y1,000 for first-time college
graduate entrepreneurs.

ï‚· Urgently needed talent training
Companies are encouraged to set up an employment category fund. Government subsidies are granted on
the principle of “whoever pays for training gets subsidies.�

Source: Announcement on Implementing Special Vocational Training Plan 2010.
60


There are three major types of external training providers: skilled-worker schools, employment
training centers, and non-public training agencies. According to government statistics, non-
public training agencies provided training to the largest number of people in 2009, followed by
public employment training centers, and then skilled-worker schools (appendix table B.2).
Employment training centers, training mostly laid-off, unemployed workers, and rural workers,
are characterized by high proportions of part-time teachers, high student-teacher ratios, and the
short duration of their training programs: seventy-seven percent of the trainees at public
employment training centers receive training for less than six months.

Participants may or may not be certified upon completing non-formal training. According to
2009 statistics (appendix table B.2), almost all trainees who attend skilled-worker schools are
certified upon finishing, and 15 percent are certified at Level 3 (senior), Level 2 (technicians), or
Level 1 (senior technicians). Non-public agencies, training the highest number of people, also
have a certification rate of 90 percent among their trainees. Moreover, the average employment
rates among graduates from employment training centers and non-public agencies are around 76
percent.

Studies on non-formal training in China are scarce, and there is no systematic review of how
well the formal and non-formal education and training systems perform. We fill this gap by
offering an in-depth analysis in the following chapters. Chapter 4 reviews the formal TVET
system of Yunnan, critically examining its legal framework, governance, curriculum, and finance.
Chapter 5 evaluates the impact of rural labor transfers and skills training. Chapter 6 offers
insights into employee training in urban enterprises in Yunnan. Finally, Chapter 7 draws on
Shanghai’s Program for International Student Assessment data to shed light on the role of
schools in developing student literacy, numeracy, and science skills.

References
Skills Toward Employment and Productivity (STEP) Employer Survey. 2012.

Yun, Bo, Haijun Liu, and Yaqin Li. 2011. “云���业技术教育�展研究报告.�

HRSS (Yunnan Department of Human Resources and Social Security). 2010. Announcement on
Implementing             Special             Vocational        Training          Plan,
http://www.ynhrss.gov.cn/readinfo.aspx?InfoId=9987.

———. 2011. Quarterly Vacancy Data.

———. 2011. Yunnan Department of Human Resources and Social Security. 年云��人力资
�       和   社      会      �      障     事      业    �     展     统    计     公     报 ,
http://www.ynhrss.gov.cn/readinfo.aspx?InfoId=82abbe9f37444b379de71346ac0b1262.
61


Chapter 4      Strengthening Technical and Vocational Education and
               Training


A central part of any skills development system is the formal technical and vocational
education and training (TVET) system. In 2010, there were more than 300 secondary and
51 tertiary TVET institutions in Yunnan, enrolling 663,400 and 183,300 tertiary students,
respectively (appendix table B.3). Between 10 and 12 percent of the total TVET
enrollment is in private institutions. In addition to enrolling full-time students, many of
the formal TVET institutions also offer non-formal training programs to adults. This
chapter provides an overview of the legal framework, governance and management,
school-industry links, quality assurance, curriculum reforms, and financing of the formal
TVET system in the national and provincial contexts.

Legal and Policy Framework

Since the founding of the People’s Republic of China in 1949, the government has
enforced a series of laws and regulations to guide and regulate the provision of TVET. A
relatively established legal and policy framework exists at both the national and
provincial levels. In 1986, the Education Law was promulgated, specifying that the
responsibility for the provision of post-secondary education lies with the provincial and
national governments, while the responsibility for basic education and secondary
education lies with local governments. Clause 19 of the Education Law states that
government will provide conditions that enable all citizens to receive technical and
vocational education and training, including adult education.

In 1995, the Labor Law specified that the state take various measures to promote
employment, develop vocational education, set labor standards, and gradually improve
the life of workers. The People’s Congress promulgated the Vocational Education Law in
1996, which established the guiding principles for TVET and broadly defined the TVET
system to include all forms of pre- and in-service training for all ages. The law also
established the right of all citizens to receive training and included provisions for
establishing nongovernmental TVET institutions, collaborating with enterprises, and
financing TVET.

In 2002, the State Council announced the Decision to Vigorously Promote the Reform
and Development of Vocational Education. The Decision increased support for and
leadership in vocational education and expanded the scale of vocational education. In
2005, the State Council published the Decision to Vigorously Develop Vocational
Education, which reiterates the strategic importance of developing vocational education
for social economic development and calls for deeper employment-oriented reforms and
a higher quality of TVET institutions. The latter Decision encourages the mobilization of
non-public resources and links between school and industry. It further pledges to
diversify sources of funding and strengthens leadership for TVET.
62


Most recently, in recognition of the strategic importance of education and training in
upgrading labor force skills and facilitating the shift of economic growth patterns, the
Chinese government renewed its commitment to education and training by launching two
important policy documents in June 2010: the Medium- and Long-Term Education
Development Plan (2010–20) and the Medium- and Long-Term Talent Development Plan
(2010–20). Both plans emphasize the role of industries and enterprises in TVET
development and call attention to the need to develop vocational education and training
for rural areas and rural workers. In addition, China is now in the process of revising its
Vocational Education Law and developing a Lifelong Learning Law. Following national
law and policy, local governments develop specific plans and targets in accordance with
the national framework.

Governance and Management

The chief administrative authority in China is the State Council. Provision of education
and training is largely public, through authorities in the Ministry of Education and the
Ministry of Human Resources and Social Security (HRSS) and their local departments.
For certain economic sectors such as transportation, health, and agriculture, line
ministries share responsibility in delivering sector-specific training. There is no single
government ministry or agency that is in charge of TVET affairs in China, though
Ministry of Education has been charged with a leading role in implementing the TVET
system reform under the direction of the Medium- and Long-Term Education
Development Plan, with input from a broad spectrum of stakeholders including the public
and private sectors, donors, and training providers (figure 4.1).

Figure 4.1. Horizontal Arrangements of TVET, Yunnan


                                        Horizontal Arrangements
                           Ministry of Education                             Ministry
                                                                  Ministry
             Leading Role in Planning, coordinating, macro-                     of        NDRC
                                                                  of HRSS
                                managing                                     Finance


Source: Adapted from Liu 2010.
Note: NDRC = National Development and Reform Commission.

Vertically, there are four levels of government structure: central government, provincial
governments, prefectures, and counties or cities (figure 4.2). The central government
develops a national policy framework, and local governments undertake the main
responsibility for developing and implementing TVET programs.
63

Figure 4.2. Vertical Arrangements of TVET, Yunnan


     Vertical Arrangements   Ministry of Education                                 Central Government:
                                                                                   Leading role

                                 Provincial Level
                                                                                   Local
                                 • 23 provinces, 5 autonomous regioins, 4
                                   municipalities                                  Governments—
                                                                                   TVET Departments
                                        Prefecture Level                           under local
                                        •333 prefectures                           educational
                                                                                   authorities--
                                               County Level                        undertake the
                                               •2860 Counties
                                                                                   main
                                                                                   responsibilities for
Source: Adapted from Liu 2010.                                                     TVET development

In an effort to improve TVET coordination, the National Inter-Ministry TVET
Coordinating Committee was set up at the national level, composed of members from the
Ministry of Education, HRSS, the Ministry of Finance, the National Development and
Reform Commission, and the Ministry of Agriculture, among others. At the local level, a
number of provinces have also established such coordinating committees. Examples can
be found in Xinjiang and Shanghai, which provide good models for overcoming
fragmented governance of TVET at the provincial level (box 4.1). Xinjiang replicates the
national coordination model at the provincial level; Shanghai goes further, however, to
consolidate functions such as admissions, curriculum and teaching, teacher management,
and student registration under the Education Bureau.

                             Box 4.1. Overcoming Fragmented Governance of TVET: Xinjiang and Shanghai

Xinjiang and Shanghai have taken two different approaches to addressing the fragmented
governance of TVET at the provincial and city levels.

In 2006, Xinjiang set up an overarching Provincial Vocational Education Leading Group in 2006.
The deputy secretary of Xinjiang's Party Committee chairs the Leading Group, and the vice
president of Xinjiang in charge of education serves as the co-chair and executive director. The 22
members in the Leading Group represent the Departments of Education, HRSS, Organization,
Propaganda, Finance, Economy and Trade, and Poverty Relief, the Agricultural Office, the
Development and Reform Commission, Labor Union, the Association for the Handicapped, the
Women’s Federation, and the Youth League.

The Vocational Education Leading Group is jointly operated with the Xinjiang Employment
Office. The Leading Group created a platform for coordinating policy development, but it does
not consolidate functions from various government authorities.

Shanghai has taken a more comprehensive approach. Starting in the fall semester of 2006, the
Shanghai government placed the admission, curriculum and teaching, teacher management, and
student registration of skilled-worker schools under the authority of the Shanghai Education
64

Bureau, without changing financing sources or human resources management. Educational
authorities on the district level are in direct charge of the establishment, abolishment, inspections,
student admissions and registration of local skilled-worker schools. City-level skilled-worker
schools are under the direct leadership of the Shanghai Education Bureau. Previously, the 41
skilled-worker schools in Shanghai, including 28 nongovernmental schools, were under the
management of the Labor and Social Security Bureau, while the secondary vocational schools
and specialized schools were under the management of the Education Bureau. Although there
have been no formal reviews, informal interviews with Education Bureau officials and school
leaders in Shanghai indicate that the consolidation was much appreciated and has resulted in a
number of positive developments, such as more streamlined admissions and registration and
equitable teacher management and career development opportunities.

School-Industry Links

Key elements of a demand-driven education and training system include a governance
and management structure that can systematically incorporate industry representation at
the system as well as the school level. In China, governance and management of TVET
remain a government affair. Although employers and industry associations are sometimes
consulted in the policymaking process, the consultation is often ad hoc, and the role of
industry and employers remains advisory. The establishment of coordination bodies for
TVET governance has been carried out without formal input from industry. Within Asia,
Indonesia’s establishment of APINDO, the Employers Association, provides a good
example of how to institutionalize the involvement of employers and industries in human
resource development on the apex level (box 4.2).

                       Box 4.2. APINDO, the Employer Association of Indonesia

APINDO (the Employer Association of Indonesia) was recognized in 1975 by decree of the
Minister of Manpower, and mandated by KADIN Indonesia (the Indonesian National Chamber of
Commerce and Industry), as the sole body representing employers in all the tripartite councils on
industrial relations and manpower affairs at the national, provincial, and district levels, and in the
international forum.

APINDO is an independent organization. It has approximately 5,000 members in all parts of
Indonesia with branches in each of the 27 provinces and in the very many districts or industrial
centers. The membership of APINDO includes private enterprises, state-owned enterprises, local
companies, joint ventures, and cooperatives. The mission of APINDO is to enhance harmonious
industrial relations, represent the Indonesian business community in manpower institutions, and
protect, defend, and empower all businesses, especially its member businesses.

At national level, APINDO is represented in tripartite bodies including the National Council for
Tripartite Cooperation, the National Productivity Council, and the National Committee for
Productivity Award, the National Wages Council, the National Council for Occupational Safety
and Health, the National Council for Work Training, and the Central Committee for Settlement of
Industrial Disputes. APINDO is also represented in several bodies at the provincial level, such as
the Provincial Tripartite Councils, the Provincial Councils on Manpower Affairs, and the
Provincial Committees for Settlement of Industrial Disputes.
65


Service to membership at the national or provincial and district levels includes consultation on
industrial relations and manpower development, legal assistance, representation at labor courts,
and training programs on manpower affairs.

Strong government policies, as well as a robust economy and labor market, have led
TVET institutions across China, including in Yunnan, to collaborate in various ways with
business enterprises. Forms of collaboration include contracting with enterprises for
graduate employment, teacher training, student internships during their last year of study,
equipment donation, and establishing school-run enterprises for training and production.

The capacity of different schools for the planning and management necessary to
strengthen links with industry varies greatly. National model secondary TVET schools
and tertiary TVET schools tend to have a greater capacity to develop in-depth
collaborations with enterprises. Schools constrained by limitations of capacity and
resources tend to engage in more rudimentary forms of collaboration with local
enterprises. From the employers’ point of view, industry has little incentive to provide
systematic and sustained inputs to the management of TVET schools. Not all industry
settings are conducive to student participation: simulating various production problems
for training in a real-time production setting or shutting down production equipment for
training purposes may be prohibitively costly. According to the STEP Employer Survey,
approximately 57 percent of workplaces in Kunming have regular contacts with
educational or training institutions. Among those, most school-enterprise collaborations
take the form of staff recruitment and employee training. From 56 to 65 percent of firms
also provide work experience for students in either professional or technical occupations
or relatively lower skilled and more frontline/operative type B occupations. However,
fewer than a quarter of firms are involved in curriculum development for training and
educational institutions; even fewer are involved in testing of students. Table 4.1 shows
the percentage of enterprises collaborating with schools in various ways for both type A
and type B occupations.

Table 4.1. Extent of School-Enterprise Collaboration, Kunming, Yunnan

                           Type of Collaboration                            Type A      Type B
 Recruitment of staff                                                      85          86
 Provides work experience for students (internships and apprenticeships)   56          65
                                                                           56
 Uses the institution for further training of your firm's existing employees           48
 Provides feedback to the institution for their curriculum development     19          25
 Participation in testing of students                                      4           4
 Other                                                                     10          17
Source: STEP Employer Survey 2012.
Note: Type A occupations include managers, professionals, technicians/associate professionals;
Type B occupations include clerical support workers, service workers, sales workers, skilled
agricultural/forestry/fishery workers, construction/craft and related trades workers, plant and
machine operators/assemblers, and elementary occupations.

Overall, most of these collaborations rely on personal connections and remain ad hoc,
idiosyncratic efforts, without system support. No institutionalized mechanism enables
66


schools to formally incorporate industry in their governance, management, and
educational affairs. Public TVET institutions do not yet have governance boards that
include industry representation. For those schools that have established some kind of
board or committee, the role of industry remains mostly advisory. In comparison, private
TVET institutions are legally required to have a school board with diverse representation,
and they tend to have stronger links with industry than do public institutions. One of the
few exceptions is the city of Ningbo in Zhejiang province, which has taken a lead in
establishing policy and legal assurances for systematic school-industry collaboration (see
box 4.3).

           Box 4.3. Ningbo's Regulation for Promoting TVET School-Industry Collaboration

Ningbo’s Regulation for Promoting TVET School-Industry Collaboration was promulgated in
2008 and enacted on March 1, 2009. Selected provisions from the regulation include:

ï‚·    Government should encourage, support, and promote school-industry collaboration and
     establish a system guided by government and coordinated by industry associations
ï‚·    TVET institutions should take initiative and collaborate with business and industry in the
     field of student internships, curriculum development, admissions and employment, instructor
     training, employee training, and continuing education
ï‚·    TVET instructors should receive at least two months of industry training every two years
ï‚·    Business and industry should accept students and instructors for paid internships and
     industry training
ï‚·    TVET institutions should provide vocational ethics and safety education, provide accidental
     injury insurance for student interns, and assign guidance instructors to supervise student
     internships.

In 2012, Ningbo further enacted implementation methods for the regulation. The implementation
methods protect the rights and welfare of student interns and propose to establish (1) categorical
grants directed at school-industry collaboration, (2) a public online service platform to facilitate
collaboration and communication, and (3) a joint conference system involving various
stakeholders including government officials, schools, and industries.

In response to the demand and pressure from industry and interested enterprises, some
schools, particularly those in areas with strong and active local private sector
involvement, started to experiment with management models within legal constraints.
One such model that has emerged in recent years is the vocational education group (�教
集团), under which a cluster of similar enterprises and a group of TVET institutions form
a joint management group that oversees vocational education. The platform involves
resource sharing, student internships in enterprises, industrial experts instructing in
schools, and collaboration in development of training materials. The vocational education
group provides an integrated approach to linking industry, research and development, and
education and training (box 4.4). The model promises to provide an intermediate and
practical mechanism for institutionalizing industry involvement in the running of TVET
schools, even before Yunnan develops a comprehensive legal and policy framework to
sustain the collaboration.
67


                         Box 4.4. Vocational Education Groups in Yunnan

As of 2010, Yunnan had 20 provincial-level vocational education groups, covering light industry,
tourism, mining, transportation, agriculture, economic management, art and crafts, construction,
forestry, and energy. Additionally, there are 10 prefecture- and city-level vocational education
groups in Kunming, Yuxi, Dali, Honghe, and Zhaotong.

For example, the Yunnan Forestry Vocational Education Group, initiated by the Southwestern
Forestry University, integrates 26 entities in forestry, including TVET colleges, skilled-worker
schools, research institutions, and businesses. Drawing on resources from each institution, the
Group provides students with industry-commissioned training and effectively integrates
vocational education with industry training. The Group also facilitates research and development
in forestry and, importantly, supports the use of research and development in products and
services. The Group also explores new ways to finance vocational education.

Quality Assurance

Reflecting the education and human resources dual-track provision of TVET, there are
also two parallel arrangements for quality assurance for education and training providers
in Yunnan. The Ministry of Education first promulgated standards for the establishment
of secondary and tertiary TVET institutions in 2000. The standards were revised and the
revisions formally adopted in 2010. In 2012, HRSS updated its previous skilled-worker
school standards and issued the latest standards for the establishment of skilled-worker
schools, advanced skilled-workers schools, and technicians’ academies. In addition,
Yunnan started implementing an updated version of its Non-public Education Institute
Management Measures in May 2012, including additional establishment standards for
private secondary TVET schools (in accordance with the national standards for secondary
TVET) and private vocational training providers offering non-diploma-oriented training.
68

Table 4.2. Standards for the Setup of Secondary and Tertiary TVET Institutions, Yunnan

                    Secondary     Tertiary   Skilled-         Advanced       Technician   Private
                    TVET          TVET       worker           skilled-       academy      vocational
                                             school           worker                      training
                                                              school                      provider
 Campus area        √             √          √                √              √            √
 Training           √             √          √                √              √            √
 facilities and
 equipment
 Minimum            √             √          √                √              √            √
 enrollment
 School leader      √             √          √                √              √            √
 and
 management
 Teacher            √             √          √                √              √            √
 number and
 qualification
 Stable source      √             √          √                √              √            √
 of financing
 Curriculum and     √             √          √                √              n.a.         √
 teaching
 learning plan
 Teaching           n.a.          n.a.       n.a.             √              √            n.a.
 research group;
 major leaders
 Minimum            n.a.          √          √                √              √            n.a.
 majors
 Industry-school    n.a.          n.a.  school-        at least 3 at least 5 n.a.
 link                                   industry       enterprises enterprises
                                        collaboration per major    per major
                                        committee
Source: Ministry of Education 2000; Ministry of Education 2010b; HRSS 2012a; HRSS 2012b;
HRSS, 2012c; Yunnan Department of Education 2012.

Table 4.2 lists areas of requirement according to existing standards. The standards are
input-based, requiring a minimum level of school infrastructure, training facilities and
equipment, number of majors, and qualifications of institutional head and teachers. They
also require schools to have a stable source of financing and to set up specific curriculum
and teaching plans. At the tertiary level, HRSS requires schools to have teaching research
groups and an academic leader for each major. For skilled-worker schools at both the
secondary and tertiary levels, HRSS also requires each school to have a system for
industry-school collaboration: skilled-worker schools are required to set up a school-
industry collaboration committee, and each major in advanced skilled-worker schools or
technician academies has to partner with a minimum number of enterprises.
   69

   Table 4.3. Selected Teacher Qualifications and Standards for Secondary and Tertiary TVET Institutions

                       Secondary        Tertiary        Skilled-       Advanced         Technician     Private
                       TVET             TVET            worker         skilled-         academy        vocational
                                                        school         worker                          training
                                                                       school                          provider
Number of full-time    > 60             > 70 with n.a.                 n.a.             n.a.           Provincial
teachers                                Bachelor's                                                     (20); city
                                        degree                                                         (15);
                                                                                                       county (10)
Maximum share of       n.a.             n.a.            33             33               33             75
part-time teachers
(%)
Teacher:student        1:20             n.a.            1:20           1:20             1:18           n.a.
ratio
Minimum share of       20               n.a.            20             20               25             n.a.
teachers with
practical experience
(%)
Minimum share of       n.a.             n.a.            70             70               70             n.a.
technical teachers
and trainers (%)
Qualifications         Senior           Associate       Technical           Technical
                                                                       Technical           Medium
                       professional     senior          teachers            teachers
                                                                       teachers            professional
                       ranking (20      professional    certified at        certified at ranking (at
                                                                       certified at
                       percent);        ranking (20     Level 5 or     LevelLevel  5    5 least     15
                       double           percent); at    above (100          (100
                                                                       (100percent         percent)
                       qualification    least    two    percent), at        percent), at
                                                                       ), at Level 4
                       sa         (30   teachers per    Level 4 (30    (40  Level 4 (60
                       percent);        major with      percent);           percent);
                                                                       percent);
                       at least two     double          trainers            trainers
                                                                       trainers
                       teachers per     qualificatio    certified at        certified at
                                                                       certified at
                       major with       nsa             Level      3   LevelLevel  3    3
                       medium                           (100           (100 (100
                       professional                     percent)            percent),
                                                                       percent), at
                       ranking                                              certified at
                                                                       Level 1 or 2
                                                                       (45  Level 1 or 2
                                                                            (50
                                                                       percent); 50
                                                                            percent); 60
                                                                       percent
                                                                            percent of
                                                                       technical
                                                                            technical
                                                                       teachers
                                                                            teachers
                                                                       and trainers
                                                                       are  and trainers
                                                                                 all-
                                                                               b
                                                                            are
                                                                       around         all-
                                                                            aroundb
   Source: Ministry of Education 2000; Ministry of Education, 2010b; HRSS 2012a; HRSS 2012b;
   HRSS 2012c; Yunnan Department of Education 2012.
70


a. Teachers with both teacher qualifications and technical skills certification.

b. All-around teachers can teach both theoretical and technical classes.

Minimum standards are also specified for the number and qualifications of teachers (table
4.3). Qualifications are often measured by teachers’ professional ranking or skills
certification. Faculty at TVET schools is composed of teachers teaching general courses
and teachers and trainers teaching technical courses and workshops. Rather than setting
the same minimum qualifications and experience level for all teachers, the standards are
set up to ensure the faculty is composed of a minimum percentage of teachers with skills
certifications or professional rankings at a certain level. For example, technical teachers
and trainers should constitute at least 70 percent of the faculty at skilled-worker schools,
and all of them are required to obtain skills certification at level 5. For secondary TVET,
at least 20 percent of the full-time teachers are required to have previous experience in
the industry. Part-time teachers, though often rich in industry experience, cannot exceed a
third of all faculty members in skilled-worker schools.

These standards serve as the basis for the initial licensing of institutions and for the
annual inspection conducted by the local government authorities. Schools not meeting the
standards are issued a warning and put on probation. Schools that do not improve during
the probation period may be prohibited from admitting any students. Local governments
are encouraged to help schools reach the standards through macrolevel planning, targeted
public investment, and integrating and reallocating resources.

In reality, however, these accreditation standards serve more as targets for existing
schools than as minimum standards. The current operating conditions of many TVET
schools across China, especially at the secondary level, are below the specified
accreditation standards. For example, none of the provinces meet at least five of the six
major indicators listed in table 4.4. Shanghai, Guangdong and Zhejiang, all in east China,
perform relatively better, reaching four of the standards. A number of provinces in west
China, including Yunnan, do not reach any of the six standards.

Table 4.4. Provinces Meeting the Establishment Standards for Secondary TVET by Region, 2009

                  Enrollment         Student:t    Full-time        Floor       Value of       Number of
                                     eacher       teachers with    area per    equipment/f    computers
                                     ratio        both degrees     student     acility per    per 100
                                                  and skills       (m2)        student (Y)    students
                                                  certification
                                                  (%)
  National        > 1200 students    20:1         30               20          2500           15
  standards
     East         All                Tianjin,     0                0           All except     All except
                                     Beijing,                                  Hunan,         Jiangsu,
                                     Liaoning                                  Fujian,        Fujian,
                                                                               Hebei          Shandong,
                                                                                              Hebei,
                                                                                              Hainan
   71

         Middle        All except        Jilin        0                0           0                Hunan
                       Heilongjiang,
                       Jilin, Shanxi,
                       Hunan
         West          All except        0            Ningxia,         0           Tibet            Xinjiang
                       Inner                          Xinjiang
                       Mongolia,
                       Gansu,
                       Yunnan,
                       Xinjiang
     Number of         8                 27           29               31          26               23
     provinces not
     meeting the
     standards
    Source: Zhang and Ma 2011.

   In comparison, some tertiary TVET institutions operate at levels higher than the national
   standards. Tertiary TVET institutions in all provinces have reached minimum enrollment
   and minimum value of equipment and facilities. Seven provinces meet all seven
   indicators listed in table 4.5. Yunnan is meets only three indicators including enrollment,
   percentage of teachers with post-graduate degrees, and value of equipment and facilities.

   Table 4.5. Provinces Meeting the Establishment Standards for Tertiary TVET by Region, 2009

                Enrollment   Student-teacher ratio            Full-time       Floor area per    Value of
                                                              teachers        student (m2)      equipment/
                                                              with post-                        facilities
                                                              graduate                          per student
                                                              degrees (%)                       (Y)
National        > 2000       18:1                             15              14                4000
standards       students
   East         All          All except Shanghai,             All             All except        All
                             Shandong, Hebei, Tianjin,                        Hainan
                             Jiangsu, Liaoning, Beijing
  Middle        All          All except Henan, Shanxi,        All             All except        All
                             Jilin, Heilongjiang, Jiangxi                     Anhui, Jilin,
                                                                              Shanxi
  West          All          All except Guangxi, Yunnan,      All except      All except        All
                             Chongqing, Xinjiang,             Qinghai         Guizhou,
                             Qinghai, Xizang                                  Gansu,
                                                                              Yunnan,
                                                                              Qinghai,
                                                                              Sichuan,
                                                                              Ningxia
Number of     0           18                                  1               10                0
provinces
not meeting
standards
    Source: Zhang and Ma 2011.
72


Quality assurance is also exercised in the Chinese TVET system by skills certification.
Students obtain an education diploma upon graduation, but, increasingly, they also sit for
skills certification tests administered by HRSS, usually at the end of their study, to
increase their employability. In Yunnan’s skilled-worker schools alone, 77 percent of
graduates in 2010 were certified at Level 4 (medium-skilled workers) and 18 percent
were certified at Level 3 (senior-skilled workers).

In 1999, under the leadership of HRSS, China started to implement its own occupational
classification. There are eight main categories of occupations and 1,838 occupations.
Although over the years, the government has made revisions to the definitions and
standards for some occupations, most were developed during China’s planned economy
period. Moreover, the Chinese classification of occupations differs significantly from the
International Standard Classification of Occupations (ISCO) last revised in 2008. There is
no direct correspondence between China’s and the international classifications, making
international comparisons a challenge (table 4.6).

Table 4.6. Comparison of ISCO-08 and China’s Occupational Classifications by Major Groups

                   ISCO-08                                              China
     ï‚·   managers                                  ï‚·   unit head (of government organization, party
     ï‚·   professionals                                 organization, enterprise, or public institution
        technicians         and      associate        事业��)
         professionals                             ï‚·   professional and technical workers
     ï‚·   clerical support workers                  ï‚·   clerical and related workers
     ï‚·   service and sales workers                 ï‚·   business and service personnel
     ï‚·   skilled agricultural, forestry, and       ï‚·   agricultural, forestry, fishery, and water
         fishery workers                               conservancy workers
     ï‚·   craft and related trades workers          ï‚·   production      and    transport   equipment
     ï‚·   plant and machine operators and               operators and related workers
         assemblers                                ï‚·   military
     ï‚·   elementary occupations                    ï‚·   other unclassified
     ï‚·   armed forces occupations

Similarly, China’s existing competency standards for occupations have not kept up with
the evolving demands of industry and do not reflect the competencies required. Not all
existing competency standards have been broken down into discrete modules, which
would allow flexibility in training. Both occupations and the corresponding competencies
required have been changing rapidly, and the classification and competency standards
need to be updated accordingly with further input from industry. Box 4.5 describes how
Turkey involved stakeholders in the development of its National Qualification
Framework.
73


     Box 4.5. Turkey: Involving Industry and Training Providers in the Development of the National
                                    Qualification Framework (NQF)

Although qualification frameworks are government-led initiatives in many countries, social
partners can be actively involved in the process.

The NQF in Turkey dates back to a technical-vocational education and training reform process,
begun in the 1990s and supported by the World Bank, through which occupational standards were
developed to link both formal and non-formal training to the labor market. In 2006, the
Vocational Qualifications Authority was established under the Ministry of Labor and Social
Security.

One of the strengths of the emerging Turkish NQF is the voluntary participation and involvement
of stakeholders. Social partners (state, employers, employees, education and training providers)
are represented on the general board of Vocational Qualification Authority, enabling them to
express their needs and priorities and to set strategies for the system accordingly. Stakeholders
have a say at every stage of the process, and the system is shaped through consensus. This
involvement enhances the sense of ownership and has a positive effect on the outcomes in the
system.

Specifically, industries are represented by various institutions and organizations in the General
Assembly of the Vocational Qualification Authority. By signing the protocols of cooperation with
the Vocational Qualifications Authority, members of these institutions and organizations are
authorized to take part in NQF-related activities (such as occupational standards development).

Source: Allais 2010

Because skills certification is a key mechanism linking labor market demand with
education and training, developing a skills certification system with updated and
industry-led competency standards and occupational classifications reflecting industry
demand and recognized by employers is vital for the development of a demand-driven
skills development system. The lack of regular reviews and involvement of industry in
the development and update of competency standards may have contributed to the
seeming lack of credibility of the current certification system. Evidence shows that
existing skills certificates do not carry much weight among employers. According to the
non-public enterprise survey in Yunnan, only 24 percent of the professional and technical
personnel and 12 percent of frontline/operative skilled workers have obtained
professional or vocational certification. The most recent quarterly job vacancy data also
show that only about 20 percent of job openings in Yunnan require a skills certification.

The experience of Hong Kong shows how a comprehensive qualification framework not
only involves industry in developing competency standards, but also includes
mechanisms for assuring quality of training programs, facilitating articulation between
academic and vocational studies and recognizing prior learning (box 4.6).
74


                     Box 4.6. Hong Kong SAR, China’s Qualification Framework

Hong Kong SAR, China endorsed the establishment of a seven-level cross-sector qualification
framework and its associated quality assurance mechanism in February 2004. Qualifications
recognized under the qualification framework are outcome-based and apply to both academic and
vocational sectors.

Industries play a pivotal role in developing competency standards in the vocational sector. The
Specification of Competency Standards (SCS) comprises both the competency standards at
various levels and the assessment guidelines. SCSs are formulated by Industry Training Advisory
Committees (ITACs), consisting of representatives from employers, employees, and relevant
professional bodies. So far, ITACs have been formed for 18 industries. To ensure continued
relevance of an SCS, the ITACs concerned must review and update the SCS regularly to keep
abreast of the latest manpower requirements. The SCSs developed by industries also support the
Applied Learning Courses under the new academic structure for senior secondary education.
Upon completion of industry-wide consultation on the draft SCSs, training providers develop
appropriate training programs. The programs and related qualifications will be recognized under
the qualification framework if they are quality-assured by the Hong Kong SAR, China Council
for Accreditation of Academic and Vocational Qualifications.

The competency standards are presented as "units of competency." In this way, trainees will be
free to acquire any units of competency specified in the SCS according to their own needs and to
accumulate credits for obtaining various qualifications. These may be qualifications at different
levels within the same functional area or qualifications at the same level across a number of
functional areas or whole qualifications (such as certificates and diplomas). A credit
accumulation and transfer system enables learners to systematically accumulate the credits of
learning and training gained from various courses with a view to converting the accumulated
credits into a recognized qualification. By referring to the SCSs, employers may also provide
tailor-made in-service training to individual employees or use the SCSs as yardsticks for
identifying personnel with suitable skills and knowledge for recruitment or promotional purposes.

Qualifications are not confined to academic and training attainment. A recognition of prior
learning (RPL) mechanism developed by individual industries enables employees of various
backgrounds to receive formal recognition of the knowledge, skills, and experience already
acquired. The RPL mechanism operates based on the SCSs formulated by the respective
industries to ensure its credibility. The RPL mechanism has been implemented in seven industries
and will be extended to more sectors by phases.

Source: Hong Kong SAR China Qualification Framework 2004.

Curriculum Reform

In 2002 the State Council published the Decision to Promote the Reform and
Development of Vocational Education ( 大力推进�业教育改 � 与 � 展 的 决 定 )
and set the overall direction for TVET reform in china. An important clause stated:

Vocational schools and training institutions should adapt to the structural adjustment in
economy, the technological progress and the changes in the labor market. This includes
adjusting the major programs, developing majors facing the newly developed industries
and modern service industries, increasing the relevance and unique features of the
75


programs. The training of technical workers, especially senior technicians, should be
strengthened. To promote curriculum reform, developing curriculum that reflects new
knowledge, technology and techniques and that is unique to vocational education. To
strengthen the link between school and enterprises, between school and industries,
building systems for tight connection between vocational school and the labor market. (§
3, cl. 8)

TVET institutions across China have become keenly aware of the importance of
delivering relevant training programs. Schools across the country are engaged to varying
degrees in the process of updating training programs. To examine the implementation and
impact of the TVET curriculum reform, we commissioned a survey sampling 40 TVET
institutions at both the secondary and tertiary levels from Liaoning, Guangdong,
Shandong, Yunnan, Shanghai and Chongqing. Box 4.7 lists the major findings of the
survey. Overall, the survey shows that TVET curriculum reform has been commonly
recognized and implemented across China and focused on school-industry collaboration
and practice-theory combination. The majority of schools surveyed have conducted
skills-demand analyses, set up a sector-specific advisory committee and a school-industry
collaboration committee, developed and implemented new curriculum standards,
conducted evaluations of teachers, and solicited feedback on the reforms.

     Box 4.7. Main Results and Outputs of the Implementation of TVET Curriculum Reform in China

Demand Analysis
ï‚·   Two-thirds of the majors surveyed have conducted skills-demand analyses; however, only
    42 percent of schools use this information to adjust and optimize curriculums and programs
ï‚·   Sources of information include enterprises (36 percent), industry associations (23 percent),
    Internet (20 percent), and HRSS authorities (10 percent)

Industry Participation
ï‚·     Sector-specific advisory committees have been established by 85 percent of schools and 73
      percent of majors
ï‚·     School-industry collaboration committees have been established by 76 percent of schools
      and 65 percent of majors
     Among those involved in developing competency-based training standards （人�培养方
      案）, industry experts account for only 25 percent. Others include teachers (25 percent),
      principals (21 percent), TVET experts (14 percent), and skills certification experts (11
      percent)

Curriculum Development and Implementation
ï‚·     Forty percent of majors comply with national vocational qualification standards, 33 percent
      use industry standards, and 23 percent have developed their own competency standards
ï‚·     Before implementing a new curriculum, 41 percent of schools need approval by a sector-
      specific advisory committee, 24 percent need approval by academic deans, 21 percent need
      approval by the school principal, 9 percent need approval by a higher authority such as the
      Department of Education, and about 4 percent need approval by an industry authority
ï‚·     Sixty-nine percent of majors have strictly followed their new curriculums, 28 percent have
      implemented most parts

School Resources
76


ï‚·    Only 37 percent of majors have a student-teacher ratio lower than 20 to 1
ï‚·    Only 90 percent of schools and majors have established teacher files and teacher training
     plans
ï‚·    Facilities and equipment meet the instructional requirements of only 32 percent of majors

Evaluation and Quality Assurance
ï‚·    Seventy-six percent of majors have developed teacher evaluation systems, 79 percent
     regular graduate tracer studies and employer surveys
ï‚·    Only 35 percent of schools have a third-party education quality evaluation report
ï‚·    Seventy-four percent of majors conduct feedback surveys among teachers, students, and
     administrators during the process of curriculum implementation
ï‚·    Ninety-five percent of graduates must take vocational qualification tests to get certified

Source: Background Study by Shanghai Academy of Education Science.

However, our study also reveals several challenges encountered in the process of reform.
First and foremost, the overall lack of up-to-date, industry-led, competency-based
occupational standards seems to have hampered the overall quality of training programs
at the school level. For 23 percent of the majors surveyed, schools had to develop their
own standards. Only 40 percent of the majors surveyed are able to use HRSS-developed
competency standards as basis for revising their training programs. For a third of the
majors, schools used standards developed by industry. Updating the existing standards
and developing standardized competencies are crucial for reducing the costs and
improving efficiencies of curriculum development on the school level.

Second, schools have reported a lack of participation and involvement by industry.
Existing efforts are often the result of personal connections between school leaders and
industry representatives. Without sufficient industry participation, schools do not have
enough expertise to develop new curriculums. Although a majority of schools have
started conducting demand analyses, only 42 percent are able to use the information from
the analysis to change the curriculum and programs. Industry involvement in the
development of standards and curriculum can be further strengthened. Industry experts
account for only one quarter of those involved in the development of new training
programs. Only 41 percent of schools require the new training programs to be approved
by a sector-specific advisory committee before implementation.

Last but not least, schools are ill suited to be the sole designers and implementers of
curriculum reform. For example, schools lack the resources and information, which could
be provided by the government and other authorities on the macrolevel, to conduct a
scientific and comprehensive analysis of a market. Moreover, the lack of school resources
has impeded the successful implementation of curriculum reform in TVET: the capacities
of teachers are limited, the student-teacher ratio is high, and facilities and equipment are
inadequate for instruction. For the most part, only 28 percent of the newly developed
training programs can be implemented.
77


Employment and Pathways

To further examine the relevance and quality of Yunnan’s TVET system, we look into
the employment rates among graduates of different levels of education (figure 4.3). On
the secondary level, the overall employment rate among secondary TVET graduates has
been around 95 percent in recent years. According to the statistics provided by the
Graduate Employment Service Center, in 2008, 73 percent of the graduates were
employed by enterprises of varied ownership, 12 percent became individual business
owners, and 15 percent went on to the next level of education. The majority of the
graduates found employment in the tertiary sector. In addition, approximately 35 percent
of the graduates found employment outside Yunnan, including a small number working
abroad. The same statistics show that culture/art/athletics, information technology,
medicine/health, manufacturing, and commerce/trading/tourism had the highest
employment rates.

Figure 4.3. Pathways Taken by Secondary TVET Graduates in Yunnan, 2008

a. pathway                              b. sector                        c. geography


            Pathway                            Sector                          Geography
       Enterprises                         Primary     Secondary                 Yunnan
                                           Tertiary                              Outside Yunnan
       Individual Business Owner
                                                                                 Abroad
                                                                                        2%
       Advance to higher level of                     14%
       schooling
           15%
                                                                                34%
      12%                                  55%              31%
                                                                                             64%
                     73%



Source: Yun, Liu and Li 2011.

At the tertiary level in 2011, the initial employment rate among graduates was 84 percent,
and the year-end employment rate was almost 96 percent. The 2011 year-end
employment rates were 96 percent for Bachelor’s degree holders and associate degree
holders. According to the statistics in 2010, however, although overall employment rates
were similar, as many as 44 percent of the graduates with associate degrees had flexible,
that is, part-time or temporary, employment (table 4.7).
 78

 Table 4.7. Year-End Employment Rates in Yunnan, 2010

                   Post-graduate            Bachelor's            Associate               Total
                 number     percent   number      percent   number     percent   number       percent
Graduates        6,212      100       47,757      100       48,113     100       102,082      100
Employed         5,854      94        45,408      95        45,803     95        97,065       95
   Government
and political    429        7         2,369       5         1007       2         3,805        4
parties
   State and
collective       2,661      43        6,524       14        4,254      9         13,439       13
institutions
   Non-public
                 143        2         644         1         657        1         1,444        1
institutions
   SOE           978        16        7,184       15        5,989      12        14,151       14
   Non-public
                 761        12        9,737       20        7,471      16        17,969       18
enterprises
   Flexible
                 413        7         10,370      22        21,081     44        31,864       31
employment
   Continued
                 311        5         3,207       7         2,152      4         5,670        6
education
Unemployed       358        6         2,349       5         2,310      5         5,017        5
  Source: Yun, Liu and Li 2011.

 To improve the attractiveness of TVET, there are multiple pathways within TVET and
 between TVET and general education. In recent years, policies expanding admissions and
 diversifying admissions criteria increasingly encourage secondary TVET students to
 advance their studies. For example, an increasing number of Bachelor’s degree programs
 in universities have opened admission to secondary TVET graduates. Some colleges and
 universities set admissions quotas specifically for secondary TVET graduates from
 counterpart schools or programs. Some secondary TVET schools have even established
 collaboration agreements with institutions overseas. Guangdong recently gave autonomy
 to over ten tertiary TVET institutions for the purposes of organizing entrance exams and
 admitting secondary TVET graduates. Unlike general track programs, admissions to
 tertiary TVET programs often take into account students’ specialized skills and skills
 certification in addition to their academic scores on the entrance examination.

 There are also opportunities for students to move from the vocational to the general
 education track (专�本). There are three types of entrance examinations for vocational
 students seeking to transfer to general academic programs. First, students completing
 their associate degrees or tertiary TVET programs can transfer for two more years of
 study to a Bachelor’s degree program through entrance examinations. Another type of
 degree program is called “self-study� for certain degree holders. A second entrance
 examination is organized every year specifically for “self-study� students holding tertiary
 TVET or associate degrees. Upon completing the program, they are granted a certified
 “self-study� Bachelor’s degree. In addition, an adult college entrance examination, held
79


every fall, provides the chance for any associate degree holder to enter an adult
Bachelor’s degree program.

Creating pathways alone, however, is not sufficient to enhance the status and
attractiveness of TVET programs in Yunnan. In 2010, the actual rate of advancement to
further education was only about 15 percent among secondary TVET graduates and 5
percent among tertiary TVET graduates (figure 4.3 and table 4.7). These rates are
relatively low compared to more developed economies, such as Shanghai, where in 2012,
as many as 33 percent of secondary TVET graduates advanced to tertiary institutions. In
the Republic of Korea, more than two-thirds of secondary TVET graduates continued to
tertiary education in 2010.

Financing Formal TVET

Financing of education in China comes from five main sources: government fiscal
allocations to public schools, income from undertakings, including tuition, and student
fees, investment by private school sponsors, external donations, and other income
(appendix table B.4). Government fiscal allocations constituted about 85 percent of total
financing for TVET in Yunnan in 2009.

Resource Allocation
Public TVET schools in China and Yunnan rely heavily on fiscal allocations from the
central government’s categorical funds and equity projects and local governments. The
government employs a combination of supply- and demand-side allocation mechanisms.
At the secondary TVET level, schools receive block grants from the relevant authority,
based on the number of enrolled students, at an average of Y2,700 recurrent expenditure
minimum per student in secondary TVET. The actual amount of allocation varies by area.
Provincial-level schools tend to receive the highest amounts, followed by city-level and
county- and district-level schools. At the tertiary level, there are 34 provincial-level
tertiary institutions in Yunnan. In 2011, the allocation was Y10,000 per student for
students in Bachelor’s degree programs and Y5,000 for students in associate degree or
TVET programs. In 2012, the allocation is expected to reach Y12,000 per student in
Bachelor’s degree programs and Y6,000 per student in associate degree or TVET
programs. Schools use this block funding, combined with student fee income and
revenues from other income generating activities, to carry out all educational functions,
including the payment of teaching staff.

In recent years, the central government, through the Ministry of Education, the Ministry
of Finance, and the NDRC, has introduced schemes to boost the funding for TVET. The
most prominent program is the national Model TVET School Project at both the
secondary and tertiary levels. Usually these are nationally competitive multiyear grants.
Proposals are evaluated by a panel of national experts; thus only schools with quality
proposals are awarded grants and the accompanying status. Typically the Model TVET
School grant requires a 1:1 counterpart contribution from the provincial government. A
competition-based approach provides incentives and motivation for improving quality
and innovation in TVET, but tends to benefit more resourceful schools, which are more
likely to produce high quality proposals and meet the criteria for grants, at the expense of
80


weaker schools. Similarly, Yunnan’s provincial government allocates on average Y20
million categorical grants each year to vocational education, mostly for buying new
equipment, building new facilities, employing experienced teachers, and training teachers
for secondary TVET institutions. TVET institutions in rural and ethnic minority areas are
further supported by funds for poverty alleviation, science and technology promotion,
ethnic minorities, and migrant settlement. Initially these grants were also allocated on a
competitive basis, which presents the same equity issues as those that arise under the
Model TVET School Project.

Other than block grants based on enrollment and special funds earmarked for
development, however, there is not yet a regular performance-based funding arrangement
for TVET schools.

Government Spending on Education
In 2010, China’s total education financing from all sources was Y1,956 billion, of which
Y1,467 billion were fiscal allocations for education (including budgetary allocations,
education surcharges at all levels of government, and enterprise- and school-generated
funds), which accounted for almost 4 percent of gross domestic product. Budgetary
allocations, together with education surcharges at all levels of government, totaled
Y1,416 billion, accounting for nearly 16 percent of total government spending.

Yunnan’s total spending on education amounted to Y44 billion in 2009. Fiscal allocation
accounted for a larger proportion of the total allocation to education in Yunnan (over 84
percent) compared to the national average (74 percent). Yunnan’s budgetary allocation,
mostly from the provincial and local governments, accounted for 82 percent of the total
funding on education. If we include both the budgetary allocation and the education
surcharges, public spending on education in 2010 was Y44 billion in Yunnan, accounting
for 19 percent of the total government spending, the highest share in the southwestern
region and among the top five provinces in the nation (figure 4.4). The total amount of
public spending on education ranked second in the southwestern region, after Sichuan
province. Even compared to developed countries, Yunnan’s public spending on education
is high as a percentage of its total government expenditure and gross domestic product
(figure 4.5).
81

Figure 4.4. Public Spending on Education as a Share of Total Government Spending, 2010



                                              24
     Share of total government spending (%)


                                              22                                  Fujian

                                              20                                                         Zhejiang
                                                                                          Yunnan Hebei      Henan
                                                                                      Guangxi                                          Guangdong
                                                           Hainan             Guizhou
                                                                                Xinjiang       Beijing              Shandong
                                              18                                                                           Jiangsu
                                                                         Gansu Shanxi
                                                                                          Anhui
                                                                                     Shaanxi
                                              16                           Chongqing
                                                                      Tianjin  Jiangxi     Hunan
                                                                                   Inner Mongolia         Sichuan
                                                                          Jilin      Hubei Liaoning
                                                         Ningxia
                                              14
                                                          Qinghai             HeilongjiangShanghai
                                              12
                                                        Tibet
                                              10
                                                   0            200             400             600             800             1000         1200
                                                           Budgetary allocation on education (including education surcharges), million

Source: Ministry of Education 2010a.

Figure 4.5. International Comparison of Yunnan’s Public Expenditure on Education, 2009

a. total government expenditure

                                              30
     % of total government expenditure




                                              25
                                              20
                                              15
                                              10
                                               5
                                               0
                                                          China, Hong Kong…




                                                            United States of…
                                                       China, Macao Special…




                                                         United Kingdom of…
                                                                       Latvia




                                                                     Iceland
                                                                       Serbia




                                                                  Lithuania
                                                                    Slovakia




                                                                    Sweden
                                                                      Poland
                                                                  Bermuda
                                                                        China



                                                             Czech Republic
                                                                         Italy




                                                                   Hungary




                                                                   Portugal
                                                                        Brazil

                                                                      France
                                                                        Spain




                                                                     Finland




                                                                  Denmark
                                                                   Monaco




                                                                  Mongolia




                                                                   Slovenia




                                                               Netherlands
                                                                     Yunnan



                                                                    Norway
82


b. gross domestic product

                10
                 8
     % of GDP

                 6
                 4
                 2
                 0




                              United…
                        China, Hong…
                      China, Macao…




                      United States…
                               Serbia




                               Latvia




                             Iceland
                          Lithuania




                            Sweden
                            Slovakia




                            Hungary




                            Portugal
                           Bermuda
                                China




                                 Italy



                              Poland
                     Czech Republic




                                Brazil

                              France



                             Finland




                           Denmark
                            Monaco




                                Spain




                          Mongolia




                            Slovenia




                       Netherlands
                             Yunnan



                             Norway
Source: OECD EdStats; Ministry of Education 2009.

Although the high share of government spending demonstrates Yunnan’s commitment to
education, the amount of public spending per student appears inadequate. Yunnan’s
budgetary recurrent expenditure per student is lower than the national average at every
level of education (appendix table B.5). The gaps are particularly pronounced at the
compulsory level: for 2010, the per student recurrent expenditure is approximately 18
percent lower than the national average. At the tertiary level, including Bachelor’s and
associate degree programs, the recurrent expenditure reached Y8,526 per student (figure
4.6). At the secondary level, per student recurrent expenditure is Y4,728 for secondary
TVET, slightly higher than that for general secondary schools.

Recurrent expenditure can be further divided into personnel expenditure (including wages,
allowances, bonuses, social security benefits, and financial aid) and public expenditure
(operating expenses, administrative and office expenses, purchasing of equipment and
facilities, and maintenance expenses). At the tertiary level, over half of the recurrent
expenditure is public, whereas at the lower levels of schooling the majority of the
expenditure is devoted to personnel expenses (figure 4.6). In particular, at the compulsory
level and in general high schools, approximately three-quarters of the total recurrent
expenditure is devoted to personnel. For secondary TVET, over 61 percent of recurrent
expenditure was spent on personnel.
83

Figure 4.6. Budgetary Recurrent Expenditure per Student by Level of Education in Yunnan, 2010

 10000
  8000
  6000                                                                      3908.8
  4000
                             3186.74         3288.83        2875.01
  2000       2483.68                                                       4606.43
             802.56          1162.33         1026.96        1853.42
     0
          Primary School General Middle General High Secondary TVET Regular Higher
                            School        School                      Education
                                                                     Institutions

                                    Public    Personnel

Source: China Education Finance Statistical Yearbook, 2010; appendix table B.5.

As education finance is largely decentralized in China, the level of public spending on
education varies greatly by city and prefecture within Yunnan. The variance is most
pronounced for secondary TVET schools and ranges from Y10,788 recurrent expenditure
per student in Pu’er to Y2,600 per student in Chuxiong (figure 4.7). More than half of the
cities and prefectures spend less than the provincial average of Y4,265.

Figure 4.7. Budgetary Recurrent Expenditure per Student on Secondary TVET, 2009

          12000
          10000
           8000
           6000
           4000
           2000
              0




         Source: Yun, Liu, and Li 2011.

Inequalities in spending across cities and prefectures are correlated with the urban-rural
gap. Per student expenditure in rural schools is lower than average at all levels of
schooling with the exception of primary schools in Yunnan (figure 4.8). The gap is most
pronounced in vocational high schools; per student expenditure at rural vocational high
schools is 41 percent lower than the provincial average. Further, public expenditure and
personnel expenditure in rural vocational high schools are as much as 64 percent and 28
percent, respectively, lower than the total averages.
84

Figure 4.8. Budgetary Recurrent Expenditure per Student in Yunnan, 2009

                                     4500
                                     4000


           expenditure per student
             Budgetary recurrent
                                     3500
                                     3000
                                     2500
                                     2000                                                        Total
                                     1500                                                        Rural
                                     1000
                                      500
                                        0
                                            Vocational HS General HS     General       Primary
                                                                       Middle School   School

        Source: China Education Finance Statistical Yearbook 2010.

Private Spending and Provision of Education
Since the early 1980s, the financing of education in China has been shifting rapidly
toward a decentralized system with a diversified revenue base (Tsang 1996). In addition
to budgetary expenditures coming from fiscal allocation and categorical funds,
nonbudgetary sources include local education surcharges and levies, allocations by
enterprises, social contributions, and school fees. Specifically, it is stipulated in the
Vocational Education Law that at least 30 percent of the local education surcharges have
to be allocated to vocational education (20 percent for areas that have not yet popularized
compulsory education). In this subsection, budgetary expenditures are referred to as
“public� expenditures and nonbudgetary expenditures as “private.�

Private spending as a percentage of total per student expenditure is generally lower in
Yunnan than the national average (appendix table B.7). Within Yunnan, the share of
private funding is the highest for tertiary associate degree programs (45 percent),
followed by tertiary Bachelor’s degree programs and pre-primary school (figure 4.9). For
secondary TVET, nonbudgetary expenditure accounts for about 39 percent of the total
expenditure per student, lower than the national average of 43 percent.
85

Figure 4.9. Total Budgetary and Nonbudgetary Expenditure by Level of Schooling in Yunnan, 2009

                                         18000

         Total expenditure per student
                                         16000
                                         14000
                                         12000
                                         10000
                                          8000
                                          6000
                                          4000
                                          2000                                                     Non-budgetary
                                             0                                                     Budgetary




     Source: China Education Finance Statistical Yearbook 2010.

Table 4.8 further illustrates the proportion of students enrolled in private schools at
different levels in Yunnan. Most private education in Yunnan takes place at the pre-
primary and tertiary levels. For TVET, although the proportion of non-public spending is
relatively low, 26 percent of students are enrolled at the 10 non-public tertiary institutions,
and 12 percent are enrolled at the 54 secondary TVET institutions, levels similar to the
national average. In comparison, Yunnan has much lower levels of private provision at
the primary and lower secondary levels (only about 2 percent) and for general upper
secondary education (only about 6 percent).

Table 4.8. Private Provision of Education in Yunnan, 2010

                                                                                                   Total          Total
                                                                      Number                    enrollment,    enrollment,
   Level                                          School type        of schools   Enrollment   Yunnan (%)       China (%)
Tertiary                                    Regular HEIs             17           107,101      24              n.a.
                                            Independent Academies    7            59,905       23              n.a.
                                            TVET/associate degrees   10           47,196       26              n.a.
Upper                                       TVET                     54           85,795       12              14
secondary                                   General                  50           36,347       6               10
Lower
                                          79          48,524         2                                         8
secondary     General
Primary       General                     114         100,531        2                                         5
Pre-primary n.a.                          2,906       415,481        42                                        47
Source: Yun, Liu, and Li 2011. National averages are authors’ calculations based on                            Ministry of
Education 2010a statistics.
Note: HEI = Higher Education Institution.
86


Private spending on education also includes funds provided by students’ families. To
investigate household expenditures on TVET, we draw on a survey conducted in 2010
among 900 students from 10 schools in Kunming, Qujing, Dali, and Tengchong. Students
at secondary TVET are subsidized Y1,500 per year nationwide for living expenses. Most
recently, rural students, students from poor families, and students studying in agricultural
programs have been exempted from secondary TVET tuition fees by the State Council.
On average, personal education expenditures are Y5,739 per student per year, accounting
for about 27 percent of per capita family income. Tertiary TVET students spend the most
on education (Y7,353), accounting for 40 percent of their per capita family income. At
the secondary level, skilled-worker school students spend the most, accounting for 25
percent of their per capita family income. (See figure 4.10.)

Figure 4.10. Household Income and Educational Expenditure in Yunnan, 2010


  40000
  35000
  30000                                                                                                             Household income
                                                                                                                    per capita (yuan)
  25000
  20000
  15000                                                                                                             Personal
                                                                                                                    expenditure on
  10000
                                                                                                                    schooling (yuan)
     5000
       0
                Total      Tertiary TVET   Skilled worker school   Vocational High school   Secondary Specialized
                                                                                                  School

Source: Yun, Liu, and Li 2011.

The survey also asks students about their perception of the financial burden posed by
education. On average, only 7 percent of the students surveyed expressed “no difficulties,�
and nearly 53 percent reported “with some difficulties but bearable.� As many as 37
percent of the students think the cost of TVET education is “unbearable� and were forced
to borrow money, and 51 percent of tertiary TVET students agree. At the secondary level,
although the majority of students consider the financial burden to be bearable, 17 to 29
percent of the students think the cost is “unbearable� and also rely on borrowed money.
87

Figure 4.11. Perceptions of Financial Burden Posed by Education Expenditures in Yunnan, 2010

 100%                                                                       Unbearable, have to
                                                                            borrow money
  50%                                                                       Some hardships, but
                                                                            bearable
     0%                                                                     No hardships
           Total   Tertiary TVET Skilled worker Vocational    Secondary
                                      school    High school   Specialized
                                                                School


Source: Yun, Liu, and Li 2011.

Efficiency of Public Spending on TVET
Because we do not have data on TVET graduates’ labor market outcomes across Yunnan,
we use enrollment rates and advancement rates, as well as indicators of education
resources, such as student-teacher ratios, floor areas per student, and construction areas of
school buildings (total size of school building divided by number of students) to examine
the efficiency of Yunnan’s financing of formal TVET.

Although standardized testing scores are not available for benchmarking the performance
of Yunnan’s education system against other provinces, enrollment rates and advancement
rates beyond the primary level lag behind the national average (table 4.9). Only 73
percent of the middle school graduates continued on to the upper secondary level, 15
percent lower than the national average. Even at the compulsory level, only 96 percent of
primary school graduates continue on to the lower secondary level. Gross enrollment
rates in Yunnan at the senior secondary level are only 65 percent in 2010, as compared to
the gross enrollment rate at the senior secondary level in Shanghai of nearly 97 in 2009.
Among cities and prefectures in Yunnan, Qujing has achieved a gross enrollment ratio of
91 percent at the upper secondary level, in sharp contrast to the 34 percent in Zhaotong,
the lowest rate in the province (figure 4.12).

Table 4.9. Comparison of Enrollment Rates and Advancement Rates in Yunnan and China, 2010

                                                                                   Yunnan               China
   Enrollment rates of school-age children              100                                       100
   Advancement rates to lower secondary school          96                                        99
   Advancement rates to upper secondary school          73                                        88
  Source: China Statistical Yearbook 2011; Yun, Liu, and Li 2011.
88

Figure 4.12. Gross Enrollment Ratio in Yunnan at Upper Secondary Level by City and Prefecture, 2010


     Gross enrolment ratio at   100
     upper secondary level %
                                80
                                60
                                40
                                20
                                 0




Source: Yun, Liu, and Li 2011.
Note: Red bars indicate rates below, and blue bars indicate rates above, the provincial average.

Poor operating conditions suggest that the average investment in TVET is insufficient,
especially at the secondary level. Although TVET schools face difficulties meeting the
national standards across China, the issue is particularly compelling for Yunnan. For
example, table 4.10 compares the actual operating conditions of schools in Yunnan
against the national standards and the national average on major indicators. In secondary
TVET institutions, the student-teacher ratio in Yunnan (29:1) far exceeds the national
standard of 20:1, while the 17 percent of full-time teachers with both degree and skills
certifications is well below the national minimum of 30 percent and the national average
of 21 percent. In terms of the infrastructure and facilities, Yunnan is also lagging behind:
floor area per student is only 9 square meters, well below the national standard of 20
square meters and the national average of 12 square meters. The value of equipment and
facilities per student and the number of computers are also far from adequate to meet the
national standard and match the national average.

Table 4.10. Selected Standards for Secondary TVET Institutions in Yunnan, 2010

                                                          Full-time                                    Number
                                                          teachers with   Floor area   Value of
                                  Enrollment   Student-                                                of
                                                          both degrees    per          equipment/f
                                               teacher                                                 computers
                                                          and skills      student      acilities per
                                  (persons)    ratio                                                   per 100
                                                          certification   (m2)         student (Y)
                                                                                                       students
                                                          (%)
 National > 1,200         20:1         30                20         2,500                              15
 standard 1,672
 National                 27:1         21                12         2,331                              13
 s
 average
 Yunnan 1,414             29:1         17                9          1,337                              11
Source: Ministry of Education 2010b; China Education Statistical Yearbook 2010.
Note: Secondary TVET institutions in this table include secondary specialized                          schools and
vocational high schools, but do not include skilled-worker schools.

On the tertiary level (figure 4.11), Yunnan’s TVET institutions meet the national
standards and exceed the national average on major indicators. For example, Yunnan has
89


achieved an average student-teacher ratio of 18:1. The percentage of full-time teachers
with post-graduate degrees has reached nearly 33 percent, more than double the national
minimum standard of 15 percent. The value of equipment and facilities and floor area per
student, although slightly lower than the national average, meet the national standard.

Table 4.11. Selected Standards for Tertiary TVET Institutions in Yunnan, 2010

                                                     Full-time
                                                     teachers      Floor area   Value of
                                      Student-
                      Enrollment                     with post-    per          equipment/fac
                                      teacher
                      (persons)                      graduate      student      ilities per
                                      ratio
                                                     degrees       (m2)         student (Y)
                                                     (%)
 National
                     > 2,000         18:1          15          14     4,000
 standards
 National average 7,754              17:1          32          15     6,115
 Yunnan              5,236           18:1          33          14     4,615
Source: Ministry of Education 2000; Ministry of Education 2004; China Education Statistical
Yearbook 2010.
Note: Secondary TVET institutions in this table include secondary specialized schools and
vocational high school, but do not include skilled-worker schools.

In vocational high schools, Yunnan’s recurrent expenditure per student and average
student-teacher ratio are close to the national average. With similar levels of per student
expenditure, Yunnan has a lower student-teacher ratio than Guizhou, Sichuan, and
Shaanxi in southwestern China. However, compared to Qinghai, also with a similar level
of per student expenditure, Yunnan has a much higher student-teacher ratio. For
secondary specialized schools, Yunnan spends slightly more than the national average
and yet has a higher student-teacher ratio. Compared to Gansu and Shaanxi, Yunnan
appears to have a higher level of expenditure per student but also a higher student-teacher
ratio. (See figure 4.13.)
90

Figure 4.13. Student-Teacher Ratio Compared to Recurrent Expenditure per Student in Yunnan, 2010

            a. vocational high schools

                             35
                                                    Guizhou
                             30                      Sichuan
     student teacher ratio




                                                           Shaanxi
                             25           Yunnan
                                                           National
                             20

                                                        Qinghai
                             15

                             10
                                  0      2000          4000        6000          8000       10000     12000    14000   16000
                                                                Recurrent expenditure per student


b. specialized secondary schools

                             50
                             45
     student teacher ratio




                             40
                             35
                                      Sichuan
                             30                                 Yunnan
                                            National
                                                              Shaanxi
                             25
                                            Gansu
                             20
                             15
                             10
                                  0    2000        4000        6000       8000      10000     12000    14000   16000
                                                          Recurent expenditure per student


Source: Expenditure data from Ministry of Education 2010a; student-teacher ratio from China
Statistical Yearbook 2011.

Efficiency gaps can be further illustrated by inequalities in educational resources
available to secondary TVET institutions within Yunnan. The national standards specify
a minimum of 33 square meters of floor area per student, 20 square meters of
construction area of school building per student and a maximum student-teacher ratio of
20:1. However, as indicated by figure 4.14, basic operating conditions vary greatly across
the province, and only a few cities and prefectures meet the national standards. For
example, floor area per student ranges from 127 square meters in Nujiang to only 15
91


square meters in Lincang. Only five cities and prefectures have a student-teacher ratio of
lower than 20:1, while Qujing and Lincang have student-teacher ratios of 50:1.

Figure 4.14. Operating Conditions of Secondary TVET Institutions in Yunnan, 2010

a. Floor area per student

 140

 120

 100

  80

  60

  40

  20

      0




b. Construction area per student

 30

 25

 20

 15

 10

  5

  0
92


c. Student-teacher ratio

 60

 50

 40

 30

 20

 10

  0




Source: Yun, Liu, and Li 2011.
Note: Red bars indicate city and prefecture averages not meeting national standards.

Summary

This chapter provides a comprehensive review of the formal TVET system in Yunnan,
with a specific focus on governance and management, industry involvement, curriculum
reforms, quality assurance, and finance.

Although management of the formal TVET system in China has traditionally been
fragmented, involving multiple ministries and departments, in recent years
interdepartmental TVET coordination committees have been established at the national
level and in a few provinces. Governance and management of TVET in China are still the
responsibility of government, with limited industry involvement. Although employers
and industry associations are sometimes consulted in the policymaking process, their role
remains advisory and their involvement is often ad hoc.

With the wide implementation of curriculum reforms, schools have become more aware
of the need to align training programs and curricula with labor market demand. Efforts to
establish formal school-industry links through collaborative advisory and sector-specific
advisory committees have started to emerge, with support from the National Model
TVET School project and development partners such as the World Bank. However, the
links are often limited to recruiting graduates or providing internships and training. Very
few enterprises are involved in curriculum development or assessment of students.
Furthermore, the limited capacity of teachers, inadequate facilities and equipment, and
lack of school resources have impeded the implementation of the TVET curriculum
reform.

Quality assurance is carried out through enforcement of standards for training providers
and a nationwide skills certification system. Responding to the most recent national
93


policies, the Ministry of Education and HRSS have updated the standards for establishing
TVET institutions on both the secondary and tertiary levels. Although the standards are
comprehensive and specific to the types of institutions, many existing TVET institutions,
especially those at the secondary level, face difficulties in attaining them. The issue is
particularly pressing for resource-constrained provinces such as Yunnan. The standards
continue to serve as benchmarks for local governments as they strive to improve the
overall conditions of TVET institutions through better resource allocation and planning.

Quality assurance is also exercised at the student level through a certification system
developed by the HRSS. An increasing number of students are pursuing skills
certificates in addition to their academic diplomas upon graduation. However, the current
certificates suffer from lack of authority, and are not yet fully recognized by employers.
Only a minority of existing workers in Yunnan has obtained certification, and only 20
percent of the new job openings over the past year required a skills certification for new
hires. The existing occupational classification and competency standards for each
occupation have not been fully updated to reflect the changing demands of the market.

Yunnan’s education finance relies mostly on public sources of funding. Yunnan’s total
public spending on education amounted to 6 percent of its gross domestic product and 19
percent of total government spending— a high level compared to other provinces in
China and even to developed countries. Yet, owing to the large size of its education
system, Yunnan’s recurrent expenditure per student is lower than the national average at
every level of education. Overall, school resources appear insufficient. The gap is
particularly large at the compulsory level. Furthermore, the level of per student
expenditure varies greatly across cities and prefectures, as well as between rural and
urban areas. With education finance decentralized, poor rural areas have difficulty
allocating adequate funds. Particularly hard hit are the rural vocational high schools.

The poor performance of the education system in Yunnan is suggested by enrollment and
advancement rates that are lower than the national average. Within Yunnan, there exist
huge disparities in school resources and educational quality across cities and prefectures,
and between rural and urban areas. In particular, disparities among TVET institutions
appear even more pronounced than those among general institutions. These disparities
suggest that there is potential to further increase the efficiency of public spending on
education in Yunnan.

Private provision of TVET in Yunnan is emerging with 12 percent of secondary TVET
students and 25 percent of tertiary TVET students enrolling in private institutions. There
is potential for Yunnan to further explore expanding the private provision, in particular at
the tertiary level, beyond its current level to mobilize additional resources and to have a
healthy dose of “competition� in the system. Across both public and private institutions,
the level of private spending on TVET in Yunnan remains lower than the national
average. Indeed, private support could be more strongly mobilized.

The allocation of public funds in TVET is primarily driven by student enrollment and a
per student expenditure formula. In addition, the government provides special funding
opportunities which have been mainly competition based. Although competition-based
94


funding provides incentives and motivation for improving quality and innovation in
TVET, equity-based resource allocations for disadvantaged schools can enable these
schools to make critical improvements in educational quality. Demand-side financing is
limited to waiving fees for poor rural students and the RMB 1,500 across the board
subsidy for living expenses. There is no other demand-side financing such as training
vouchers.

References
Allais, Stephanie. 2010. The Implementation and Impact of National Qualifications
Frameworks: Report of a Study in 16 Countries. Geneva: ILO.

China Education Finance Statistical Yearbook. 2010. Beijing: China Statistics Press.

Hong Kong Qualification Framework. 2004. http://www.hkqf.gov.hk/.

HRSS (Department of Human Resources and Social Security). Yunnan Department of
Human Resources and Social Security. 2010. Quarterly Vacancy Data.

———. 2011. Quarterly Vacancy Data.

———2012a. Advanced Skilled Worker School Establishment Standard (Provisional),
http://www.gov.cn/gongbao/content/2012/content_2163427.htm.

———. 2012b. Skilled Worker School Establishment Standard (Provisional),
http://www.gov.cn/gongbao/content/2012/content_2163427.htm.

———. 2012c. Technicians Academy Establishment Standard                     (Provisional),
http://www.gov.cn/gongbao/content/2012/content_2163427.htm.

Liu, Peijun. 2010. “TVET System and its Innovation Strategy in China.� Presentation at
the South-South Study Visit on Skills and TVET, “TVET System & Its Innovation
Strategy in China—For a More Flexible, Responsive and Attractive System,� Guangdong,
China, November 1.

Ministry of Education. 2000. No. 14: Tertiary TVET Establishment Standard
(Provisional),
http://www.moe.edu.cn/publicfiles/business/htmlfiles/moe/moe_621/200409/3208.html

———. 2004. No. 2: Regular higher institutions basic operating condition indicators
(Provisional),
http://www.moe.edu.cn/publicfiles/business/htmlfiles/moe/moe_2/201006/88606.html.

———. 2009. Announcement on 2009 National Education Finance Implementation, 教
育 部 国 家 统 计 局 财 政 部 关 于 2009 年 全 国 教 育 � 费 执 行 情 况 统 计 公 告 ,
http://www.moe.edu.cn/publicfiles/business/htmlfiles/moe/s3040/201012/112378.html.
95


———. 2010a. 教育部国家统计局财政部关于 2010 年全国教育�费执行情况统计公
å‘Š                                                                                  ,
http://www.moe.edu.cn/publicfiles/business/htmlfiles/moe/s3040/201112/128871.html.

———. 2010b. No. 12: Secondary TVET School Establishment                    Standard,
http://www.gov.cn/gongbao/content/2010/content_1745852.htm.

OECD.Stat Extracts Database. Paris: OECD. http://stats.oecd.org/

State Council. 2002. Decision to Promote the Reform and Development of Vocational
Education ( 大 力 推 进 � 业 教 育 改 � 与 � 展 的 决 定 ),
http://www.gov.cn/gongbao/content/2002/content_61755.htm

Tsang, M.C. 1996. “Financial Reform of Basic Education in China.� Economics of
Education Review 15 (4): 423–44.

Yun, Bo, Haijun Liu, and Yaqin Li. “云���业技术教育�展研究报告.� 2011.

Yunnan Department of Education. 2012. Yunnan Non-public Education Institute
Management Measures, http://www.mbjyw.com/html/201208/20120829306.shtml

Zhang, Chen and Shuchao Ma. 2011. “我国�业学校办学�件评价和预警机制研究.�
中国高教研究�志, no. 8
96


Chapter 5       Non-formal Training


The previous chapter provided an in-depth review of the formal Technical and Vocational
Education and Training sector. In this chapter, we turn to the non-formal training sector.
Recognizing the importance of creating second chances and building a lifelong learning
framework for their citizens, many countries are taking steps to improve their non-formal
education and training systems through increasing both access and quality. China must also
aggressively face these challenges.

As China’s economic reforms and urbanization accelerate, a large population of rural laborers,
previously engaged in agricultural activities, has become idle. The training of surplus rural labor
is of paramount importance in facilitating the transfer of rural workers to the secondary and
tertiary sectors and preparing them for migration across regions.

This chapter presents an overview of the context of rural labor training, the evolution of the
policy framework, and the major training projects implemented in rural Yunnan. The analysis is
further enriched by a quantitative evaluation of the impact of rural training on household income
using data drawn from a survey of 3,000 rural households in Honghe Prefecture, Yunnan.

Rural Work Force and Surplus Labor Transfer

Training the largely unskilled rural working population is a key strategy of the Chinese
government’s poverty alleviation efforts. China has set target urbanization rates of 44 percent in
2015 and 50 percent in 2020, up from the current rate of 31 percent. In order to reach those
targets, a significant number of rural unskilled laborers must be transferred to the secondary or
tertiary sector. Those who remain in the agricultural sector need training in both higher-value-
added agricultural activities and technology to improve their productivity.

Yunnan’s structural adjustment in agriculture started in the 1990s, with policies such as
“shrinking grain production and expanding trading� and the later promotion of “characteristic,�
“highly efficient,� “large-scale,� and “export-oriented� agriculture. These policies are aimed at
improving agriculture efficiency and increasing rural income. Yet, because the rural population
is large and farmland is limited, rural poverty remains an issue. Governments at both the central
and local levels are aware that the solution lies in reducing the number of farmers, transferring
rural surplus labor to the secondary and tertiary sectors, turning producers of agricultural produce
into consumers, and reducing agricultural production while increasing the demand for it. These
strategies are expected not only to increase the income of the rural population, but also to
promote the development of “socialist new rural areas,� and a “well-off society� and bridge the
urban-rural gaps.

Yunnan has a more than sufficient stock of rural labor engaged in agricultural activities. Since
1978, the size of the rural workforce has almost doubled, although its share of the total labor
force has gone down by 5 percentage points (figure 5.1). As of the end of 2010, the rural labor
force totaled 22 million, constituting approximately 78 percent of Yunnan’s total labor force.
Two-thirds of the rural labor force is between the ages of 16 and 40. Among the rural workers,
97


78 percent were engaged in agricultural activities, while 9 percent and 13 percent were engaged
in the secondary and tertiary sectors, respectively.

Figure 5.1. Number of Workers in Rural Areas and Share of Total Employment in Yunnan, 1978 –2010

 2500                                                                86.0%

 2000                                                                84.0%
                                                                     82.0%
 1500
                                                                     80.0%
 1000
                                                                     78.0%
  500                                                                76.0%
     0                                                               74.0%
         1978
         1980
         1985
         1990
         1995
         1998
         1999
         2000
         2001
         2002
         2003
         2004
         2005
         2006
         2007
         2008
         2009
         2010
         Number of workers in rural areas     Share of the total employed

Source: Yunnan Statistical Yearbook 2011.

The provincial Department of Human Resources and Social Security (HRSS) estimated in 2011
that surplus labor was 7.6 million, one-third of Yunnan’s total rural labor force (HRSS 2011). By
the end of 2011, more than 5 million of the surplus rural laborers found employment in
secondary or tertiary sector, while 2.5 million remain to be transferred. In 2011 alone, 1.7 million
rural surplus laborers were transferred, with 1.3 million finding employment within Yunnan and
0.4 million outside Yunnan (HRSS 2012). The rate of transfer of rural labor in Yunnan is lower
than the national average.

Yunnan’s rural working age population averages only seven years of education. Nearly 9 percent
of the rural working age population is illiterate; most have attained only primary school (32
percent) or lower secondary education (39 percent). Only about 20 percent have had upper
secondary education or beyond. Furthermore, as many as 86 percent of the rural workers have
not received any training (National Bureau of Statistics of China 2008). The number of rural
professional and technical personnel amounted to less than 0.2 percent of the total human
resources in rural Yunnan.

The rural laborers who manage to transfer also lack skills and training. Transferred laborers
average a little more than eight years of education, slightly higher than the average for Yunnan’s
rural population, but still less than the level of compulsory education. The majority of transferred
workers have attained only lower secondary education. Only 2 percent of the current transferred
labor force received training before their transfers.

Rural-to-Urban Migration and Employment: Policy Evolution

Since 1949, China’s national policy on rural-to-urban migration and employment has undergone
three major phases:
98


Phase 1: Planned period (1949–77)

During this period in the broader context of China’s planned economy, labor mobility was
restricted between rural and urban areas, as well as across provinces. The residential registration
(hukou) policy strictly prohibited unplanned migration.

Phase 2: Restricted migration (1978–99)

With China’s economic reforms, the “family contract responsibility system� was introduced in
rural areas. Formerly collectively owned lands were contracted to households, releasing some
farmers from agricultural activities. During the same period, rural laborers were permitted, but
only on a restricted basis, to migrate to and get employment in urban areas. Gradually, the
government started regulating the recruitment process for rural migrants, including issuing urban
employment permits and temporary urban residential permits.

Phase 3: Guided and facilitated migration (2000 to present)

As the migration rate accelerated, the government has adapted its role from restricting migration
toward guiding it. For example, the national government initiated the Sunshine Project, the Rural
Labor Skills Training and Employment Project, and the Rain and Dew Project to prepare
migrants for the migration process and their ultimate employment in the cities. Yunnan has also
issued a series of provincial policies to implement training projects among rural workers (box
5.1).

                      Box 5.1. Yunnan’s Policies on Rural Worker Training, 2004–08

2004:
The policy entitled Suggestions on Accelerating Rural Labor Force Transfer by Yunnan Provincial
People’s Government proposed the Project of Training a Million Rural Workers

2005:
The Announcement on Distributing the 2005–2010 Training Plans for Rural Workers of Yunnan by the
General Office of Yunnan Provincial People’s Government set an annual target of providing training to
500,000 transferring workers. The ratio of guidance training to skills training was set at 4:1

2006:
The Announcement on Distributing the Implementation of Enhancing Employment and Re-employment
by Yunnan Provincial People’s Government pledged to provide policy assurance for subsidizing
Yunnan’s rural workers seeking employment in urban areas

2008:
The Announcement on Distributing the Implementation of the Employment Promotion Law by Yunnan
Provincial People’s Government specified that “unemployed, ex -servicemen and rural workers in urban
areas who participate in training will have their training fees subsidized�

Demand for and Supply of Rural Worker Training

Rural laborers need training in order to transfer to secondary and tertiary industries or across
regions. They need not only to acquire technical skills specific to the sector they are transferring
99


to, but they also need guidance training that prepares them for migration to urban areas.
Guidance training usually consists of an orientation covering topics such as basic rights at the
work place, laws and regulations, and advice on urban living and employment. Rural laborers
who remain in agricultural activities need training that would equip them with practical skills to
improve productivity.

Various training providers serve the rural population, including the Technical and Vocational
Education and Training schools, employment-training centers, and employers, all of which also
serve urban workers. In addition, the agricultural broadcasting schools and agricultural training
centers under the Department of Agriculture and the Poverty Alleviation Office deliver training,
especially agricultural skills training, to rural workers in local villages. Figure 5.2 illustrates the
training demand and supply among rural workers.

Figure 5.2. Demand for and Supply of Rural Labor Training

a. demand                                                   b. supply




                    • skill training                                            Department of Human
                      (specific to                                               Resource and Social
                      secondary                                                       Security
     Transferring     and tertiary                     Training provider:
        labor         sector)                      •Formal TVET schools            Department of
                    • Guidance                     •Employment training             Agriculture
                      training                      centers
                                                   •Employers
                                                   •Agricultural Broadcasting    Poverty alleviation
                                                    Schools                            office
                                                   •Agriculturual Training
                                                    Centers
                    • agricultural skill                                           Department of
      Farmers         training                                                       Education




Note: TVET = Technical and Vocational Education and Training

During the 11th Five-Year Plan period (2006–10), 6.5 million rural workers received training, of
whom nearly 5 million received guidance training and over 1.5 million received skills training. A
total of nearly 870,000 took part in the skills certification, over 750,000 of whom have
successfully obtained skills qualifications.
100


Managing and Financing Rural Worker Training Programs

Table 5.1 lists the major rural training programs carried out in Yunnan in recent years, organized
by Department of Agriculture, HRSS, and the Poverty Alleviation Office. In addition, Yunnan’s
Department of Education has been involved in training school-aged youth in rural Yunnan on
topics such as agriculture as part of the formal education curriculum. There are also training
programs initiated by the Trade Union, the All China Women’s Federation, the CCP Youth
League Committee, and the Department of Science and Technology.

The costs of training are shared between public finance (at the central, provincial, and local
levels) and participants. Governments at all levels have set up categorical funds for rural training,
which are disbursed to rural trainees in the form of cash, vouchers, or reduced training fees. In
addition, the All China Women’s Federation and the Youth League raise donations through
foundations devoted to conducting rural training for target groups.

Table 5.1. Major Training Programs for Rural Workers in Yunnan

  Provincial agency     Years            Training programs          Government spending        Number
                                                                                              receiving
                                                                                               training
 Department of         2004–     Sunshine Project (national rural   Y194.3 million           1 million
 Agriculture           09        labor transfer training project)   (Y104.3 million from
                                 Provincial-level rural labor       the national fund; Y90
                                 transfer training project          million from the
                                                                    provincial projects)
 HRSS                  Since     Rural Labor Skills and             Through 2009, Y314       804,000
                       2006      Employment Project                 million from special
                                                                    national government
                                                                    fund
 Poverty               2004–     Rain and Dew Project (labor        Y452 million from        1.5 million
 Alleviation Office    09        transfer training in poor areas,   Poverty Alleviation
                                 including skills and guidance      Fund
                                 training)
Source: Li 2012.

Several measures are in place to assure the quality of training. For example, Yunnan’s provincial
rural transfer labor training program requires that a panel of experts review the qualifications,
facilities, experience working with rural areas, and proximity to rural areas of participating
training providers. Potential training providers are not eligible to enter the bidding process if the
employment rate of their trainees is under 80 percent. Furthermore, to ensure the employability
of trainees, public training funds are allocated so as to give priority to targeted programs
contracted by employers.

The existing governance and finance of non-formal training suffer from three major challenges.
First, fragmented governance and financing reduce the efficiency and effectiveness of the system.
For example, the management of rural labor transfer training alone involves the Departments of
Finance at all governmental levels, along with HRSS, Education, Science and Technology, and
Construction. In poor areas the Poverty Alleviation Offices participate as well. Yunnan has set
up a provincial-level rural labor transfer leadership office (operating in the Agriculture
101


Department), but it is unknown whether or to what extent the coordinating efforts have been
effective.

Second, although competitive bidding is used for some training programs and training providers
are given autonomy in admitting students, conducting training, and establishing connections with
employers, there are no accountability measures to assure the quality of training. Allocation of
training funds is based mostly on the number of trainees and the unit cost of training, instead of
training outcomes such as certification and employment rates. In some cases, trainees and
training providers are reimbursed on condition that the trainees pursue skills certification. Given
the low credibility among employers of skills certification in general, certification rates do not
necessarily translate into higher employment rates or guarantee a high quality of training. In
addition, the requirements for auditing and monitoring are only roughly implemented, giving
training providers room to misuse public resources for their own interests.

Finally, overall training funds appear insufficient. The total government spending on rural
worker training in Yunnan was Y60 million in 2009 and Y65 million in 2010. The maximum
subsidy of Y600 per person is far below the cost of training a primary lathe worker or a welder,
which exceeds Y2,000, or the cost of training a chef, which exceeds Y1,300. As a result rural
training has been limited mostly to short-term, non-skill-oriented guidance training.

Content and Delivery of Rural Labor Training

Rural labor training programs in Yunnan are managed by three different government agencies:
HRSS, the Department of Agriculture, and the Poverty Alleviation Office. The programs offered
by each agency vary in duration, focus, and form of delivery. Table 5.2 shows the range of
program characteristics for each agency.

Table 5.2. Content and Provision of Rural Worker Training Programs in Yunnan

 Government        Types         Areas        Duration        Entry           Training        Final
   agency                                                 requirements      provided by   certification
Yunnan          Guidance      Law and        7–16         Middle           County        No
Department      training      policy         days         school           vocational
of                            concerning                  graduate,        high schools,
Agriculture:                  urban                       healthy          rural
Sunshine                      employment                                   broadcasting
Project                                                                    school
                Skills        More than
                training      20
                              agriculture-
                              related
                              fields
Yunnan          Guidance      Law and        At least No                   Accredited      Qualified
HRSS            training;     policy         one                           training        trainees obtain
                              concerning     month;                        providers, at   skills
                              urban          cooking                       home or local   certificates;
                              employment     approxim                      villages        those not
                              ;              ately 45                                      qualified
                Skills        Computer,      days                                          receive
102

               training     horticulture,                                             module
                            animal                                                    attendance
                            husbandry,                                                recognition
                            stitching and
                            sewing
Yunnan       Guidance       Law and       3 days      Middle         County           Usually no,
Poverty      training;      policy                    school         Vocational       but students
Alleviation                 concerning                graduate,      High School,     can take skills
Office: Rain                urban                     healthy        at home or       tests
and Dew                     employment                               local villages   administered
                            ;                                                         by HRSS

               Skills       Hospitality,   One
               training     computer       month
                            skills,
                            sewing
Source: Li 2012.

The training programs usually comprise a combination of orientation, covering basic legal and
policy knowledge, and skill-specific courses. Trainings conducted by the Department of
Agriculture tend to focus on agriculture-related technical skills, whereas Poverty Alleviation
Office and HRSS programs offer a variety of courses geared toward urban employment, mostly
in the construction and service sectors. An important characteristic of rural training is the form of
delivery: although many sessions are conducted in county vocational schools and other
accredited training agencies, training sessions are also conducted in local villages and sometimes
at home.

The content of training provided through HRSS is based on a standard framework that leads to
skills certification. Trainees who receive training through HRSS are required to pay for the final
certification tests. A higher proportion of those who receive training through HRSS obtain skills
certificates compared to those in other training programs. Those who fail the tests still obtain
module attendance recognition, which entitles them to work toward final certification in the
specific technical area. In comparison, programs organized by the Poverty Alleviation Office
often do not lead to certification. The Department of Agriculture generally provides short-term
agriculture courses, which also do not lead to any certification.

Impact of Rural Training: Evidence from Honghe Prefecture

Despite the increased investment made by the government at various levels, there has been very
little analysis of the impact of rural training. Honghe Prefecture is one of the target areas
participating in the national and provincial rural labor training projects. Given the low socio-
economic level and diverse demographic and geographic features, Honghe is an ideal site for
studying the effects of rural labor training in Yunnan. Fortunately, we are able to access data
from a longitudinal household survey conducted from 2006 to 2010 in rural Honghe Prefecture
in Yunnan. The unique data allow us to conduct a quantitative analysis to investigate the impact
of rural training on household incomes.
103


The Honghe Hani and Yi Autonomous Prefecture (“Honghe�) is located in southeast Yunnan,
bordering Kunming, Yuxi, Qujing, and Vietnam. Honghe has an area of 32,931 square
kilometers, 85 percent of which is mountainous area. As of the end of 2006, Honghe’s registered
population totaled over 4 million, 57 percent of whom belong to ethnic minorities. The largest
ethnic minority groups in Honghe are Yi and Hani, accounting for 24 percent and 17 percent,
respectively, of the total population. In 2006, Honghe had six national-level poverty counties and
one provincial-level poverty county. Population below the poverty line numbers 734,000, nearly
18 percent of its total registered population.

A description of the Honghe Rural Household Survey and the sample restrictions used in this
study can be found in box 5.2. In our final analytic sample, 1,260 members from 571 households
participated in training in 2006. The vast majority attended training organized by government,
while a small number sought training programs on their own or attended training organized by
enterprises. Appendix table C.1 further tracks our treatment and control groups in the years
following 2006.

                       Box 5.2. The Honghe Rural Household Survey and Analytic Sample

This study draws on data from the Honghe Rural Household Survey conducted by the Honghe Statistics
Bureau and the Statistical and Mathematical College at the Yunnan University of Finance and Economics.
The sample of 3,000 households was selected randomly using two-stage cluster sampling from 295
villages in 13 counties. The surveyed households record their income from different sources and
expenditures on a daily basis. In addition, the survey collects information on household members’
education, training, migration, and work background. We include in our analytic sample 2,143 households
who participated in the survey each year from 2006 through 2010. We analyze the trends for labor
transfer across industries, migration, and training participation over the five years.

In order to estimate the effects of training on income, we further restrict our sample to include (1)
households with at least one member attending training in 2006 (the treatment group) and (2) households
with no members attending training from 2006 through 2008 (the control group). We measure the net
income per person in the household in 2007 and 2008. Incomes are adjusted using 2006 as the base year,
according to the price indices provided by the Honghe Statistics Bureau. Table 5.3 lists the three major
training projects implemented in rural Honghe since 2005.

Table 5.3. Training in Rural Honghe

 Organizer      Starting      Types             Spending per student          Duration    Number of
                  date                                   (Y)                             Participants
Honghe          2005     Guidance training       2005–07: 100            2005–07: 7 days 40,000
Poverty                                         2008 to date: 200        2008 to date: 3
Alleviation                                                              days
Office: Rain
and Dew                    Skills training     2005–07: 500              One month
                                               2008 to date: 800
Honghe          2006       Guidance training 2006–08: 500                At     least    one 60,000
Bureau     of              and skills training 2009 to date: 500         month;       cook’s
HRSS                       leading          to (400 training fee and     training
                           certification       100 certification fee)    approximately 45
                                                                         days.
104

Honghe         2006     Guidance training; 400                      7–16 days           60,000
Bureau     of           agricultural skills
Agriculture:            training
Sunshine
Project
Source: Li 2012.

Households members participated in training in 2006 differ significantly in both village-level and
household-level characteristics from those whose members did not (appendix table C.2). On the
village level, as many as 84 percent of households whose members did not participate in training
are from villages with a high concentration of ethnic minorities. Moreover, over 80 percent of
them are located more than 20 kilometers away from county towns or in mountainous areas.
Among households receiving training, however, about half are from non-ethnic-minority villages,
and only 43 percent live in mountainous areas. These findings imply that geographic barriers
might have prevented rural laborers from receiving training. During the survey, when asked why
they did not participate in training, “training is not available locally� is the second- most-cited
reason. In addition, households whose members participated in training tend to come from
villages with high participation rates.

On the household level, households that did not participate in training seem to be characterized
by much lower literacy rates: in 7 percent of those households, all members are illiterate, much
higher than the 1 percent among households whose members participated in training. Similarly,
in only 13 percent of the households without training, are there members with education levels
beyond high school, compared to 26 percent of the households with training. Some of the
training programs require participants to have attained at least junior secondary level education,
which prevents many less-educated and uneducated rural laborers from receiving training.
Furthermore, households whose members participated in training had higher total net income and
agricultural income than those households without training. Finally, 17 percent of the households
who participated in training had cadres, as compared to only 9 percent of the nonparticipating
households.

Using propensity score matching estimators, we match households whose members participated
in training with similarly situated households without training. 1 We then used the matched
sample to estimate the effects of training. The results are presented in table 5.4. Overall, we find
a positive and significant effect of training on the net income of family members. Average
treatment effect on the treated (ATT) tells us the extent to which the households whose members
attended training benefit from it. Households whose members participated in training in 2006
showed an average increase in their 2007 net income per person of Y1,471, an increase that is 26
percent higher than if they did not attend training. The effects are slightly smaller on 2008
income, but still positive and significant. In contrast, as shown by the average treatment effect on
the controls (ATC), households in the control group would have increased their income by
Y1,041 (24 percent) in 2008 had they participated in training in 2006. On average, having
member(s) receiving training in 2006 increased the real net income per person in the household
by Y996 (22 percent) in 2007 and Y962 (20 percent) in 2008, as indicated by the average
treatment effect (ATE). We have tested the robustness of the results employing other methods
and models. These results are available upon request.
105

Table 5.4. Estimated Effects of Training on Net Income per Person, 2007 and 2008

                                        2007                               2008
                             Income         log(income) income                log(income)
                ATT          1471.2***      0.259***        761.08*           0.10912
                             (416.21)       (0.06)          (387.41)          (0.07)
                ATE          996.3**        0.22395**       962.08*           0.20404*
                             (379.79)       (0.08)          (412.90)          (0.08)
                ATC          810.25         0.21021         1040.8*           0.24123*
                             (429.53)       (0.11)          (515.32)          (0.11)
               Source: Honghe Household Income Survey
               Abadie-Imbens (AI) standard errors in parenthesis.

Summary and Policy Implications

As Yunnan’s economic development and urbanization advance, a significant number of rural
unskilled laborers need to be transferred to the secondary or tertiary sector. Those who remain in
the agricultural sector need training to upgrade their skills to improve productivity. Both the
central and provincial governments have introduced training schemes targeted at rural surplus
laborers.

Using data collected in Honghe, we were able to evaluate the impact of training by estimating
the effects of household members who received training in 2006 on the net income per person in
the household in 2007 and 2008. The propensity score matching estimators illustrate a positive
and significant effect of training: Households whose members participated in training in 2006
earned 26 percent more income per person in 2007, and 11 percent more income per person in
2008 than households that did not participate. On average, households whose members attended
training showed an increase in their net income per person in 2007 and 2008 of over 20 percent.
Our results provide evidence of the economic returns of training, and further highlight the value
of rural labor training. It is particularly worth noting that although training is given to individual
household members, the training effect goes beyond the individual and extends to the net income
of the entire family.

Despite the positive economic effects, our analysis and interviews in the field have revealed
several areas of potential improvement in the effectiveness and efficiency of rural worker
training programs. The participation rates in rural training are low, as reflected by official
statistics and our survey data. Rural laborers face geographic barriers that prevent them from
attending training projects. Many rural laborers do not have the minimum education level
(usually completion of middle school) required as a starting point by certain training programs.
Some are unaware of the value of training, and some are concerned that the training might take
too much of their time away from paid agricultural activities. Households in ethnic minority
villages are also less likely to participate. Training materials should be tailored to the specific
needs of the rural training participants, bridging the gaps between their level of previous
education and knowledge, and the training curriculum. Extra effort should also be made in
outreach and marketing in the field.
106


There are also issues concerning the quality of training, especially its effectiveness in helping
participants migrate and obtain employment. Because training programs are relatively short, they
can provide only limited skills development. Many participants need retraining after they start
work. Likewise, the current guidance training programs, accounting for the majority of the
training programs, provide only basic work- and living-related orientation for rural workers
before migration. Only a quarter of all the training programs actually equip participants with
practical skills.

The quality issue reflects in large part insufficient funding, which has prevented training
providers and instructors from delivering quality training. Because funding for rural workers’
training programs comes from multiple resources and is managed by a range of government
departments, training efforts are not coordinated. Experience from other countries has shown
how a national training fund can consolidate multiple funding sources and allocate resources
across a range of training projects (Box 5.3).

Further, accountability measures need to be in place to assure the quality of training and improve
efficiency. In addition to the prescreening of training providers, requirements for regular
auditing of the funds and monitoring of training need to be implemented. Moreover, outcome-
based funding mechanisms could be adopted, linking public fund allocation to employment and
certification rates. This would further require a skills certification system with improved
relevance and credibility.

                            Box 5.3. Consolidating Funds for Non-formal Training

The Asian Development Bank–funded Employment-Oriented Skills Development Project in Papua New
Guinea established a Skills Development Trust Fund to provide a sustainable source of financial support
for informal-sector training. By requiring individual provinces to contribute to the fund’s capital in order
to qualify for its resources, the Fund increased the financial commitment to skills training at the
provincial level. Moreover, the Fund adopted cost sharing for both providers and participants, resulting in
broad-based ownership. The Fund also became a catalyst for mobilizing additional resources for informal
sector training. Furthermore, the fund’s capital (about K50 million) was invested and managed by an
experienced professional. The accrued interest from the invested capital was, in turn, used to cofinance
short-term employment-oriented skills training.
In some countries, the National Training Fund is allocated according to predetermined quotas for funding
“windows.� For example, in 2002, Chile’s National Training Fund financed twelve different programs of
widely varying sizes, such as apprenticeship subsidies for youth, trade training for low income youth,
short courses for those without qualifications, youth job retention, individual training scholarships, and
training for those in the military and security services. When funding windows are predetermined, it is
important to set a reasonable number of windows within available administrative capacity.
The World Bank-funded Labor Force Training Support Project in Côte d’Ivoire took a different approach.
A sponsor (such as a mayor, nongovernmental organization, or company) would send a training proposal
to the national Vocational Training Development Fund. Training advisers for the Fund would then work
with the sponsor along with representatives of the beneficiary group to develop the idea into a training
program design. The proposed training program is then sent to 3-6 prequalified training providers for
competitive bidding. Such allocations are open-ended and responsive to applications, allowing
stakeholders to generate their own proposals. However, technical assistance is needed to convert a
training request into a complete training program design, including objectives and specifications that
address beneficiaries’ needs.
107


Source: Johanson 2009.

References
Johanson, Richard. 2009. “A Review of National Training Funds.� SP Discussion Paper No.
0922, World Bank, Washington, DC.

HRSS (Yunnan Department of Human Resources and Social Security). 2011. Yunnan Rural
Transfer Labor Employment Policy Study (Working Paper),云��农�劳动力转移就业对策
研究（�求��稿）

———. 2012: 2011 年云��人力资�和社会�障事业�展统计公报, 2011 Yunnan Human
Resources       and       Social         Securities        Development         Statistical
Announcement,http://www.ynhrss.gov.cn/readinfo.aspx?InfoId=df0ff02e5aa74c4aaed19e6e8c0
3c978

Li, Xingxu. 2012. Background study commissioned for this report.

National          Bureau           of       Statistics     of              China.          2008.
http://www.stats.gov.cn/tjfx/dfxx/t20080220_402463674.htm.

Yunnan Bureau of Statistics. 2011. Yunnan Statistical Yearbook. Beijing: China Statistics Press.


Notes

1
    A detailed account of the method is presented in appendix C.
108


Chapter 6       Work-Based Training in Urban Yunnan


In addition to the formal Technical and Vocational Education and Training system that
primarily train young students before they enter the labor force, and the non-formal
training that targets rural worker and the unemployed, enterprises play an important role
in training and upgrading the skills of the labor force. Globally available evidence shows
that investment in work-based training is correlated with firm productivity. In this
chapter, we provide a snapshot of work-based training in Yunnan’s urban areas and
examine the challenges encountered by small and medium-sized enterprises (SMEs) in
providing training. We also propose potential approaches to improving the quality and
relevance of work-based training.

Data and research are scarce on work-based training in China. Our analysis draws on
official statistics as well as survey data collected during 2004 and 2005. To supplement
the data, qualitative interviews and observations were conducted at the end of 2011 at six
enterprises in Yunnan. The Skills Toward Employment and Productivity (STEP)
Employer survey conducted in 2012 also provides insights based on data from 300 non-
public enterprises in Kunming.

Demand for and Supply of Work-Based Training in Urban Yunnan

Yunnan has a total of six million employed workers in the urban areas. For them, training
is often delivered through staff orientation, on-the-job training, training to upgrade skills,
and licensing training. Training can be outsourced to external training providers or
delivered in-house by a trainer or fellow worker. Training takes place in or outside the
workplace, during or outside working hours. Figure 6.1 demonstrates the demand for and
supply of work-based training in urban Yunnan.

Figure 6.1. Demand for and Supply of Work-Based Training in Urban Yunnan




  Demand for Training                                   Training Providers
  •Employed: orientation,on-the-job                     •External:
   training, skill upgrade, training for                 •Skilled worker school
   certificate and license, etc                          •Employment training center
  •Unemployed/laid-off: training for re-                 •private agencies
   employment, entrepreneurship                         •In-house: enterprise
109


Our analysis focuses on the work-based training financed, organized, or provided by
enterprises that targets currently employed workers. It does not include public training
programs, such as reemployment training and entrepreneurship training, provided or
subsidized by the government that targets unemployed and laid-off workers. We also do
not address diploma-oriented adult education programs.

Policy Framework

The government’s role in work-based training has shifted from one of planning and
commanding to one of providing guidance. A regulatory system for work-based training
has been gradually taking shape. In 1994, the Labor Law of the People’s Republic of
China was issued, which, for the first time, guaranteed to laborers, as a matter of law, the
right to professional skills training and established the obligation of employers to provide
it. In June 1995, the then–Department of Labor issued Regulations on Compulsory Pre-
Service Training for Technical Workers, which made training compulsory for all
technical workers. In 1996, the Law of Vocational Education and the Regulations for
Enterprise Employee Training successively specified the types of trainees, and the
sources and uses of training funds, as well as training methods. Beginning in 2001,
managers of enterprises were accredited for their professional qualifications. Training for
employees at both the managerial and operational levels has been guaranteed ever since.

In 2008, Suggestions on Bringing into Full Play of the Function of Industries and
Enterprises in Vocational Education specifically required that the enterprises retain 1.5 to
2.5 percent of their wage bill and compensation fund as designated employee training
funds. In 2010, after soliciting public opinion, the State Council issued Opinions on
Enhancing Vocational Training and Boosting Employment, specifying that no less than
60 percent of the total employee training funds be spent on frontline/operative workers.
In December 2010, the Central Government and the State Council announced the Outline
of the National Medium and Long Term Development Plan for High-Skilled Workers
2010–20, which presents some groundbreaking policies, including the proposal that high-
skilled workers in private enterprises, institutions, and organizations should receive the
same training subsidies and bonuses for starting a new business as those in the public
sector.

To enforce the Central Government’s regulations and policies, the Yunnan provincial
government has implemented policies adapted to its local needs. For example, the
Yunnan government announced the adoption of the Implementing Special Vocational
Training Plan in 2010, detailing the allocation of subsidies to support in-service and
transfer training for employees, migrant employee training, and urgently needed talent
training. (See box 3.1.)

Participation in Work-Based Training

There is little data on the participation in and efficiency of training organized, financed,
or provided by employers. According to the 2005 World Bank Investment Climate
Assessment, 91 percent of the 300 firms surveyed in Yunnan provided formal training in
2004, higher than the average in China. Among the three cities in Yunnan, Yuxi has the
          110


          highest proportion of firms providing training (96 percent), followed by Kunming (92
          percent), and Qujing (85 percent).

          Although over 90 percent of the firms surveyed in Yunnan provided training, only a
          small proportion of employees received training. As shown in figure 6.2, the distribution
          is skewed toward the high and low extremes, with the majority of firms providing
          training to fewer than 10 percent of their employees. This is particularly true in Yuxi and
          Qujing. In comparison, in cities like Guangzhou most firms train the majority of their
          workers.

          Figure 6.2. Percentage of Employees Receiving Training, China and Various Cities

          a. China

                                                   China
           .08
           .06
Density


           .04
           .02
                 0




                     0         20            40            60             80        100
                             percentage of employees receiving formal training

          b. Yunnan

                                                  Yunnan
           .05
           .04
           .03
Density


           .02
           .01
                 0




                         0       20           40             60             80       100
                             percentage of employees receiving formal training in
111




c. Various cities




Source: World Bank 2005

The 2012 Skills Toward Employment and Productivity (STEP) Employer Survey
conducted among enterprises in Kunming also revealed that about 91 percent of firms
have provided some form of training on the premises of the workplace for both their type
A and type B employees. As shown in figure 6.3, most workers have received on-the-job
training, either through an apprenticeship or by learning as they worked. Many are
trained by a firm’s managers, technical personnel or peers, or by the firm’s dedicated
trainers. Only a small proportion of employees have received training on a firm’s
premises with external trainers. Similarly, on average, employees receive 15 days of on-
the-job training per year, and an additional 11 to 12 days of training by firms’ managers,
technical personnel, or peers. Training by an external trainer on the firm’s premises
accounts for only 2 days per year.
112

Figure 6.3. Provision and Duration of Training on the Premises of the Workplace, Kunming

a. percent of workers by training provider


                               % of workers
                                    TYPE A   TYPE B

   By firm's managers, technical persons,                                45
                peers, etc.                                              45

                       On the job training                               45
                                                                              52

              By firm's dedicated trainers                             42
                                                                        44

                                    Other                     21
                                                    11

 On firm's premises with external trainers               15
                                                6



b. days of training per year by training provider


                        Number of days/year
                                    TYPE A   TYPE B

   By firm's managers, technical persons,                                45
                peers, etc.                                              45

                       On the job training                               45
                                                                              52

              By firm's dedicated trainers                             42
                                                                        44

                                    Other                     21
                                                    11

 On firm's premises with external trainers               15
                                                6

Source: STEP Employer Survey 2012.
Note: Type A occupations include managers, professionals, technicians and associate
professionals; type B occupations include clerical support workers, service workers, sales
workers, skilled agricultural/forestry/fishery workers, construction/craft and related trades
workers, plant and machine operators and assemblers, and elementary occupations.

Although the provision of in-house training is relatively consistent across levels of
occupations, as shown in figure 6.4, on average, only 41 percent of type A workers and
38 percent of type B workers have received training through private training providers,
113


the most common form of external training. Public Technical and Vocational Education
and Training schools and equipment suppliers are also common providers of outside
training. Overall, employers have provided formal training outside the workplace to a
higher proportion of type A workers than type B workers.

Figure 6.4. Percent of Workers Receiving Employer-Organized Formal Training Outside the Workplace


                                % of workers
                                    TYPE A       TYPE B


                                                                            41
            Private training providers
                                                                       38

                                                                22
                  Public TVET school
                                                      13

                                                               20
                Equipment suppliers
                                                          15

                                             3
 NGO's or international organizations
                                             4


Source: STEP Employer Survey 2012.
Note: Type A occupations include managers, professionals, technicians and associate
professionals; type B occupations include clerical support workers, service workers, sales
workers, skilled agricultural/forestry/fishery workers, construction/craft and related trades
workers, plant and machine operators and assemblers, and elementary occupations. TVET =
Technical and Vocational Employment and Training.

Financial Input in Work-Based Training

Law and regulations have specified the amount of training funds that should be derived
from enterprise revenues. For example, the Labor Law states that all employers should
retain and allocate training funds, “according to the reality of the enterprise.� For
employees in technical positions, “pre-work training should be provided.� A related tax
regulation requires enterprises to allocate 1.5 to 2 percent of their total payroll for
employee training. If training costs are below the allocated amount, the enterprise should
pay the costs in full. If training costs are above the allocated amount, the enterprise may
pay the excess or negotiate with employees to pay the extra costs themselves.

Yet the law and regulations have in the past left room for employers to avoid compliance.
For example, an employer might retain the 1.5 to 2.5 percent of payroll required (or even
exceed the higher limit), but use the funds solely for training at the managerial level. The
most recent regulation issued by the State Council prohibits this by setting a minimum
requirement that 60 percent of the employee training funds be spent on
frontline/operative workers.
114


Despite the government’s continuing efforts to ensure the enforcement of the laws and
regulations, actual investment in training varies greatly among employers. A previous
survey (Guo and Lamb 2010) of 18 enterprises in Yunnan in 2004 reveals that 22 percent
of employers spent less than 1 percent of their payroll on training. Only 33 percent of the
employers surveyed spent more than 2 percent of their payroll on employee training and
education. Thirty-seven percent of the employers indicated that their employees had to
bear on average 35 percent of the total training cost.

The most recent STEP employer survey also provides information on how much
enterprises pay outside training providers. As illustrated above in figure 6.3, most
training takes place on the premises of the enterprise as on-the-job training or training by
personnel from within the firm. Figure 6.5 shows the distribution of the average total
amount paid to outside training providers as a percentage of the total wage bill and
compensation fund. Almost half of the firms paid nothing to outside training providers.
On average, employers in Kunming spent about 0.8 percent of their total wage bill and
compensation fund on external training providers. Seventeen percent of the firms spent
up to 0.5 percent of their total wage bill and compensation fund on external training
providers. Only about 5 percent of the firms spent more than 2 percent of their total
payroll on external training providers.

Figure 6.5. Percent of Wage Bill and Compensation Fund Spent by Firms on Outside Training Providers


                                           % OF FIRMS
 60.00%
 50.00%
 40.00%
 30.00%
                                                                                              % OF FIRMS
 20.00%
 10.00%
  0.00%
            0.00%    up to 0.5   0.5-1    1-1.5     1.5-2     2-2.5    2.5-5      5+

Source: STEP Employer Survey 2012.

Only about 47 percent of the firms made any payments to outside training providers for
training type A workers, and even fewer, 38 percent, paid for any outside training of type
B workers (figure 6.6). On average, firms spent Y866 for each type A worker and Y265
for each type B worker. Moreover, although 16 percent and 13 percent of the firms spent,
respectively, Y500 to Y1000 and Y1000 to Y5000 per type A worker on external training,
20 percent of the firms spent no more than Y100 per type B worker on external training.
These numbers show that investment in formal training for all types of workers, and
particularly for type B workers, is insufficient.
115

Figure 6.6. Average Amount Spent per Employee on External Training Providers

 70.00%
 60.00%
 50.00%
 40.00%
                                                                                                a
 30.00%
                                                                                                b
 20.00%
 10.00%
  0.00%
                0            100          100-500      500-1000      1000-5000   5000+

Source: STEP Employer Survey 2012.
Note: Type A occupations include managers, professionals, technicians and associate
professionals; type B occupations include clerical support workers, service workers, sales
workers, skilled agricultural/forestry/fishery workers, construction/craft and related trades
workers, plant and machine operators and assemblers, and elementary occupations.

At the end of 2011, we interviewed personnel at a privately owned coal mine. All the
technical workers were reaccredited and their vocational qualifications revalidated every
two or three years. The coal mine employs more than 600 workers at three different mine
sites, and about half of them are technical workers. The mine owner spent about Y3000
per employee for a 15-day training program administered by training institutions
established by the provincial Department of Coal Industry. In contrast, at an automobile
maintenance business with 600 employees, the owner said that the company spent only
about Y20,000 annually (about Y33 per employee) for an external trainer to come once a
week to train the employees after working hours. Although many large enterprises are
able to meet the requirements for employee training, others, especially SMEs, do not
have sufficient revenue to meet the requirement of 1.5 to 2.5 percent of payroll for
employee training funds. Moreover, employees told us of the many restrictions on
accessing training funds and getting reimbursement for training fees. One employee,
when asked who paid for his training for certification, said, “My parents and I had to pay
for my training for certification. It is difficult to get the employer to pay for them. The
employer only gives us a free one-hour training session before or after work hours.�

Training Implementation in SMEs

Eighty percent of the enterprises in Yunnan are SMEs. For these enterprises,
implementation of work-based training is hampered by insufficient resources. At the end
of 2011, we interviewed six SMEs about their work-based training and made several
important findings.

First, training in many cases is little more than a formality. For example, an automobile
maintenance business and a tourism service business invite an external trainer or an
employee to deliver 1–2 hours of training per week outside working hours. This approach
is far less costly than providing training organized by accredited training institutions, but
116


it lacks formal structures and methods for delivery, as well as formal evaluation and
monitoring processes.

Second, SMEs rarely take a systematic approach to training that includes training
management, needs assessment, planning, design, and evaluation. Few have full-time
managers specifically in charge of training; most managers have other obligations and
training management is only a small proportion of their duties. Most of the SMEs do not
conduct any training needs assessment. As a result, the contents of training seldom align
with the development goals of the enterprises. There are no work-based training materials
tailored to the needs of specific local industries or to the educational and social
backgrounds of the participants.

In some firms we interviewed, neither the employer nor the employees seemed
enthusiastic about training: the employer is unwilling to sacrifice work hours for training
and the employees resent being asked to sacrifice their time outside working hours. When
asked about training participation, one of the employees of the automobile maintenance
business said, “If I have to participate in the training, I have to. If I had a choice, I would
not have participated, because I have lots of family commitments outside work and
training always takes place after my work hours.�

Finally, the high costs of training make it difficult for SMEs to deliver adequate and high-
quality training or to access the resources made available by the government. If subsidies
are limited, and employers are unable to fully fund training costs, employees are forced to
pay for all or part of their training and accreditation or find training opportunities on their
own. Although both the central and provincial governments have made great efforts to
provide support and guidance, these efforts have not been effective for many SMEs. For
example, the government provides enterprises with free online resources about training
planning, needs assessment, and course design. However, many SMEs find it difficult to
access and utilize these resources. As put by one manager, “We are all overloaded with
daily tasks; there is no time and energy to look into that matter.�

Summary and Policy Implications

The Vocational Education Law of 1996 stipulated that enterprises devote 1.5 to 2.5
percent of their total wage bill and compensation fund for employee training. However,
compliance with the law and the details of its implementation were left to the
determination of the enterprises themselves. As a result, how and the extent to which
work-based training is provided varies greatly among enterprises.

Survey results indicate that more than 90 percent of the firms in both Yunnan and China
have provided training. However, on closer observation, we found that the majority of
enterprises in Yunnan have provided training to fewer than 10 percent of their employees.
Most training is conducted in-house by firm managers, technical personnel or peers, or
through on-the-job learning or apprenticeships. Very few firms have paid external
trainers to conduct training: only 41 percent of type A workers and 38 percent of type B
workers have received training through private training providers, the most common form
of external training. Such patterns of training-resource allocation are not unique to
117


Yunnan. As Professor Thomas A. Kochan of the Massachusetts Institute of Technology’s
Sloan School of Management best put it:

        It’s a gross misallocation of resources to spend millions on executive education
        for middle managers or senior executives while systematically underinvesting in
        people who develop the products, manufacture them, and service them. These
        middle skills are eroding because firms have given up on training. They expect to
        be able to buy all these skills on the market, and don’t want to invest in training
        because they’re afraid if they do their competitors will hire these newly minted
        high-skills employees away. Employers need to reinvest by engaging each other
        in their industry, in their location, sharing the cost and the benefits.

Firms of all sizes face low incentives to train their workers, and the relevance and value
of training are constantly called into question. Firms that do not believe in the value of
skills certifications will be unwilling to spend the funds needed to enable their employees
to obtain them. Diverse financing mechanisms and more effective incentive schemes for
work-based training could be used to address these issues. The Republic of Korea’s
Employment Insurance System (box 6.1) and Singapore’s Skills Development Fund (box
6.2) offer two excellent examples.

                         Box 6.1. The Korean Employment Insurance System

Since 1962, Korea’s vocational training schemes evolved in response to the demands of the
economy and the changing supply of skills (figure B6.1a). In 1995, an Employment Insurance
System (EIS) was introduced to address the demand from the knowledge-based economy for
continuing training to upgrade the knowledge and skills of incumbent workers, as well as the
need for an active labor market policy. EIS consists of three components: (1) an employment
security program, which supports the training costs of employees who plan to change their
careers; (2) an unemployment benefits program, which provides living expenses for the
unemployed; and (3) a vocational competence development program, which provides financial
assistance to employers for vocational training of their employees, and to individual employees
and the unemployed for their vocational training.

The Employment Insurance Fund is collected differently for each program. For the
unemployment benefit program, the employer and the employee each pay an insurance fee of
0.45 percent of an employee’s annual total wage. For the employment security and the vocational
competence development program, employers pay insurance fees of between 0.25 percent and
0.85 percent of their employees’ payroll, depending on the number of employees.

With EIS, vocational training in Korea has experienced a significant paradigm shift (figure
B6.1b) specifically: (1) the focus of vocational training changed from initial training to
continuing training (life-long training); (2) the training market became open to all public and
private vocational training institutes, forcing them to compete with each other to attract trainees—
which is expected to improve the quality of training and overall efficiency of the system; (3) the
operational mode of the training system changed from a government-controlled and supply-
driven approach to a demand-driven and incentive-based approach. The vocational competence
development program is a levy-rebate program under which all enterprises pay training levies as
part of their insurance fees, but are left free to decide whether to train and what types of training
to provide (If they carry out training, the expenses are reimbursed from the Employment
118

Insurance Fund. To provide greater incentives for SMEs, the level of reimbursement for large
firms is about 80 percent of the paid training levies, while SMEs receive rebates of up to 270
percent of paid training levies.); and (4) the training area was expanded from the manufacturing
sector to all sectors, including the service sector. The EIS played a crucial role in providing
vocational training for the unemployed, whose number increased rapidly during the foreign
currency crisis (1997–99), and has successfully responded to an ever-increasing demand for
continuing training of employees.

Figure B6.1a Evolution of Vocational Training Funding Schemes in Korea



 1967-1971                    1972-1976                 1977-1994                  1995 to present

      •Direct subsidy             •Compulsory               •Train, or pay                •Levy-rebate
       to large                    in-plant                  levy                          programs
       enterprises for             training
       in-plant
       training




Figure B6.1b Transformation of Korea’s Vocational Training System


  Before (1977-1998)                                 Now (since 1995)
  •Compulsory Training system                        •Vocational Competence Development
  •Government controlled/Supply-oriented              Program in EIS
  •Training Market closed                            •Demand-oriented/Incentive-based
  •Vocational Training Promotion Fund                •Training Market Open
  •Initial Training                                  •Employment Insurance Fund
  •Mainly for Manufacturing                          •Continuing Training
                                                     •For all industries

Source: Professor Sung Joon Paik, Korea Development Institute School of Public Policy and
Management; Lee 2006.
119


                         Box 6.2. Singapore’s Skills Development Fund (SDF)

Singapore established its Skills Development Fund (SDF) in 1979 to encourage employers to
upgrade the skills of their employees. Employers in Singapore are required to contribute a Skills
Development Levy for all employees of up to 0.25 percent of the first $4,500 of gross monthly
remuneration, or a minimum of $2, whichever is higher. The payment is added to the Skills
Development Fund, which offers grants supporting up to 80 percent of companies’ direct training
costs. The SDF provides not only a financial incentive for employers to provide training, but also
operates on a cost-sharing principle: to be eligible for the Fund, companies must pay the
unfunded portion of the training fees.

The Fund is managed by the Workforce Development Authority, composed mainly of industry
representatives (company chairmen and vice-chairmen), along with representatives of
government and the labor force. The Fund supports a diverse range of programs, but the selection
criteria and review processes incorporate Singapore’s development strategies, targeting specific
types of training, population, and sectors. For example, the Fund encourages training in strategic
sectors, promotes special training programs for upgrading skills, assists SMEs with training in
information technology, and targets specific populations such as older workers and low-skilled
workers.

The Fund also provides special incentives to encourage smaller firms to offer training. For
example, the SDF grants encourage smaller enterprises to hire external consultants to conduct
training needs analysis before developing a training plan— giving them access to the specialized
resources needed to assess training needs and design appropriate training programs. The Fund
also provides preapproved public courses for companies by subscription. This program attracts
small companies that have neither the expertise nor the critical mass to create such programs on
their own.

Source: Hirosato 1997.

SMEs find it particularly difficult to comply with the requirement that 1.5 to 2.5 percent
of payroll be spent on training. Limited capacity appears to be another constraint.
Organizing SMEs into consortiums and providing technical and institutional assistance,
as is done in Korea, might be an effective alternative approach to solving these problems
(Box 6.3).

  Box 6.3. Providing Institutional and Technical Assistance to SMEs: Korea’s SME Training Consortium

Although Korea’s employment insurance system proved to be effective overall, SMEs appear to
have benefitted less from its provisions. Specifically, although both large enterprises and SMEs
pay training levies, many fewer SMEs actually recover the levies by providing training. SMEs are
faced with higher training costs per worker, higher labor mobility, and more limited financial
resources and capacity in designing and organizing training.

In response to this situation, the Korean government launched the SME training consortium
projects in 2001. Each local chamber of commerce and industry helps a group of some 50 SMEs
organize into a training consortium and finances two training managers for each consortium. Each
training consortium forms an operating committee composed of representatives from member
SMEs, the local chamber of commerce, the Ministry of Labor field office, and training experts.
120

The committee meets periodically for the planning and management of the training needs of the
member SMEs.

The two training managers play key roles in the training consortium. They establish an
information network among member SMEs, conduct a training-needs survey of each member
SME, plan and program training activities, contract outside training institutions or collaborate
with them to develop training programs, monitor training activities, and conduct an evaluation
upon completion of major training courses on behalf of the member SMEs.

With the institutional and technical assistance provided by the local chamber of commerce
through the training consortium, more SMEs have been able to claim rebates of their training
levies. The pilot SME training consortium project has achieved significant results and has been
replicated both nationally and internationally.

Source: Lee 2006.

References
Guo, Z., and S. Lamb. 2010. International Comparisons of China's Technical and
Vocational Education and Training System. Dordrecht: Springer.

Hirosato, Y. 1997. “Skills Development Fund: A Preliminary Assessment of a Financing
Alternative for Enterprise-Based Training in the Context of APEC.� APEC Study Center,
Graduate School of International Development, Nagoya University, Nagoya, Japan.

Kochan, Thomas A. 2012. “Interview: Can America Compete?: The Workforce.�
Harvard Magazine (September–October 2012): 35–38.

Lee, K. W. 2006. “Effectiveness of Government's Occupational Skills Development
Strategies for Small- and Medium-Scale Enterprises: A Case Study of Korea.�
International Journal of Educational Development 26(3): 278–94.

Skills Toward Employment and Productivity Employer Survey. 2012.

World Bank. 2005. Investment Climate Assessment. Washington DC: World Bank.
121


Chapter 7        Developing Cognitive Skills in Schools: Lessons from PISA
                 Shanghai 2009


In the previous chapters, we have reviewed the formal Technical and Vocational
Education and Training system, non-formal training for rural workers, and work-based
training provided by employers, and examined their role in developing technical and
vocational skills. In this chapter, we turn our attention to the role of general education in
developing student cognitive skills.

The Programme for International Student Assessment (PISA) measures the cognitive
skills of 15-year-olds in math, reading, and science. PISA scores closely predict students’
academic paths and ultimate success in the labor market. Shanghai is the first Chinese
province to fully participate in PISA. In the most recent PISA given in 2009, Shanghai
placed as the top performer in math, reading, and science among 65 countries and
economies.

We draw lessons from the Shanghai experience to infer how general education might
contribute to the development of cognitive skills among 15-year-olds. We analyze the
2009 Shanghai PISA data to determine what school-level characteristics correspond to
higher math, reading, and science skills. The analysis is followed by a comparison of the
education systems in Yunnan and Shanghai, with a discussion on how best to apply
Shanghai’s experience in the context of Yunnan.

PISA and Skills

PISA tests the knowledge and skills of 15-year-old students using extensive and rigorous
surveys. Every three years, PISA compares the scores of 15-year-old students around the
world on measures of reading literacy, mathematics, and science, as defined in box 7.1
below. PISA assesses not only how well students have mastered a particular curriculum
but, more importantly, whether they can apply the knowledge and skills they have
acquired to solving real-world issues.

                              Box 7.1. Definitions of the PISA Domains

Reading literacy: An individual’s capacity to: understand, use, reflect on and engage with written
texts, in order to achieve one’s goals, to develop one’s knowledge and potential, and to participate
in society.

Mathematical literacy: An individual’s capacity to identify and understand the role that
mathematics plays in the world, to make well-founded judgments and to use and engage with
mathematics in ways that meet the needs of that individual’s life as a constructive, concerned and
reflective citizen.

Scientific literacy: An individual’s scientific knowledge and use of that knowledge to identify
questions, to acquire new knowledge, to explain scientific phenomena, and to draw evidence-
based conclusions about science-related issues, understanding of the characteristic features of
science as a form of human knowledge and enquiry, awareness of how science and technology
122

shape our material, intellectual, and cultural environments, and willingness to engage in science-
related issues, and with the ideas of science, as a reflective citizen.

Source: OECD 2009.

Not only does PISA provide a tool for benchmarking different education systems across
the world, but PISA scores have also been shown to link directly to students’ future
academic progress and indirectly to the school-to-work transition. For example,
longitudinal studies in Canada (OECD 2010) have shown that high PISA scores were
strongly correlated with the completion of secondary school, participation in post-
secondary education, and choice of field of study at university, even after taking students’
other background characteristics into account. Figure 7.1 shows the increasing likelihood
of university participation associated with higher PISA reading efficiencies according to
their PISA reading proficiency levels.

Figure 7.1. University Participation at Age 21 by PISA Reading Proficiency at Age 15 in Canada, 2009


                                   PISA reading proficiency

                              Level 5

                              Level 4

                              Level 3

                              Level 2

                                        0    5     10    15     20    25
Source: OECD 2010a.
Note: 1. After accounting for school engagement, mother tongue, place of residence, parental
education and family income. 2. The reference group for the PISA reading proficiency levels is
Level 1, and for the marks in reading it is the group that obtained less than 60 percent.

Another study (Bertschy, Cattaneo & Wolter 2009) using a sample of vocational students
in Switzerland confirmed that higher PISA scores are associated with more intellectually
demanding vocational training, and students who pursued vocational training with higher
intellectual demands had a higher probability of finding an adequate job.

PISA Shanghai 2009

Shanghai participated in PISA for the first time in 2009 and impressed the world with top
performances in all three subjects: mathematics, science, and reading (table 7.1). Students
in Shanghai scored significantly higher than students from the OECD countries, even
though Shanghai’s gross domestic product (GDP) per capita is lower than that of the
OECD countries. In addition, 33 percent of Shanghai’s students come from
disadvantaged families, a percentage higher than more than a majority of the OECD
countries (figure 7.2).
123

Table 7.1. Mean PISA Scores, Shanghai and OECD, 2009

                                                               Mean PISA score
                                                Math               Science         Reading
                 Shanghai                   600     (2.8)      575       (2.3) 556       (2.4)
                 OECD average               497     (0.5)      501       (0.5) 494       (0.5)
                 Shanghai ranking           1                  1               1
                Source: OECD 2010b.
                Note: Standard errors in parentheses.



Figure 7.2. PISA Reading Performance of Shanghai and OECD Countries, 2009

      a. PISA score by per capita GDP

                580
                560             Shanghai
                540
                520
   PISA score




                500
                480
                460
                440                                                       y = 0.0002x + 489.33
                                                                               R² = 0.0101
                420
                400
                      0   10 000   20 000   30 000   40 000   50 000   60 000   70 000   80 000   90 000
                                            GDP per capita (USD in ppp terms)

   Source: OECD 2009; OECD 2011.
124


b. PISA reading performance by share of socio-economically disadvantaged students

 600
                                                      Shanghai
 500

 400

 300                                                           y = -0.9121x + 509.24
                                                                     R² = 0.2598
 200

 100

   0
       0.0         10.0         20.0          30.0         40.0          50.0          60.0         70.0
             Share of students whose PISA index of economics, social and cultural status is below
                                                     -1

Source: OECD 2009.

a. Calculated in purchasing power parity terms.

Not only is Shanghai the highest performing education system among all 65 participating
countries and economies, but it has also been shown to be one of the most egalitarian.
Nearly 19 percent of the students are resilient achievers,i and as few as 0.3 percent are
disadvantaged low achieversii (figure 7.3). There is a weak relationship between socio-
economic background and student performance.iii
125

Figure 7.3. Proportion of Resilient Achievers and Disadvantaged Low Achievers


      Shanghai-China
     Hong Kong-China
                   Korea
         Macao-China
            Singapore
                 Finland
                   Japan
                  Turkey
               Portugal
                Canada
       Chinese Taipei
                  Poland
         New Zealand
                    Spain
        Liechtenstein
                 Estonia
          Netherlands
                      Italy
          Switzerland
                   Latvia
              Australia
        OECD average
                  France
               Belgium
                 Ireland
                 Iceland
                 Mexico
        United States
                 Greece
               Thailand
                 Croatia
                 Tunisia
                Norway                                                                 disadvantaged
               Hungary
                Sweden                                                                 resilient
               Slovenia
             Indonesia
              Denmark
                     Chile
      United Kingdom
                    Israel
             Colombia
              Germany
                    Brazil
       Czech Republic
      Slovak Republic
         Luxembourg
              Lithuania
                 Austria
  Russian Federation
 Trinidad and Tobago
               Uruguay
                   Serbia
                  Jordan
                Albania
             Argentina
          Dubai (UAE)
              Romania
               Bulgaria
               Panama
         Montenegro
           Kazakhstan
                     Peru
            Azerbaijan
                   Qatar
           Kyrgyzstan
                              0.0   5.0    10.0        15.0       20.0          25.0

Source: OECD 2010c.

Shanghai’s top performance has drawn interest from researchers, policymakers, and
educators around the world. What are the strengths of Shanghai’s education system that
have contributed to the high performance on PISA? Because Shanghai and Yunnan are
both situated within the Chinese educational system, lessons from the Shanghai PISA
126


will be particularly relevant to Yunnan’s efforts to improve its own secondary education
system.

We analyze the Shanghai PISA data to explore the question: “What school -level
characteristics correlate with higher levels of student skills as measured by PISA?� Box
7.2 offers an overview of the Shanghai PISA sample. We study separately lower
secondary school (middle school) students, upper secondary general school (high school)
students, and upper secondary vocational school (vocational school) students.

                                    Box 7.2. PISA Shanghai 2009

The main focus of PISA 2009 was reading, although the survey also assessed performance in
mathematics and science. Students answered a questionnaire focused on their background,
learning habits, attitudes toward reading, and their involvement and motivation. School principals
completed a questionnaire about their school that included demographic characteristics and an
assessment of the quality of the learning environment at school.

In Shanghai, 5,115 students from a total of 152 schools participated in PISA 2009, including 52
middle schools, 27 mixed secondary schools, 40 general high schools, and 33 upper secondary
vocational schools (appendix table D.1). After excluding observations with missing values, our
final analytic samples include (1) 2,041 middle school students, (2) 1,376 general high school
students (excluding students who attend upper secondary programs in vocational schools or
mixed secondary schools), and (3) 1,063 vocational school students.

PISA scores vary by school type within Shanghai, largely due to the use of high school
entrance examinations for admission to the different upper secondary schools (table
7.2).iv Students in model and experimental high schools, which take the highest-scoring
students on the high school entrance exam, perform the best in PISA. Vocational school
students, often lower-performing students on the entrance exam, score lower than the
middle school students on PISA; their scores are around the OECD average.

Table 7.2. PISA Scores of Schools in Shanghai and OECD Average

                                                             Math      Science    Reading
 Upper secondary, general, and model and experimental        699       646        631
 Upper secondary, general                                    664       618        603
 Upper secondary, general and ordinary                       641       598        583
 Shanghai total                                              600       575        556
 Lower secondary                                             588       568        544
 Upper secondary and vocational                              518       515        500
 OECD average                                                497       501        494
Source: OECD 2010b.

Student characteristics and the family backgrounds of students are significantly
associated with their selection into different schools on the upper secondary level
(appendix table D.3). Students in general high schools tend to come from families that are
relatively wealthy, with more home education resources and cultural possessions, and
127


higher levels of parental education, compared to those attending vocational high schools.
Moreover, a much higher proportion of general high school students than vocational
school students have attended preschool for at least a year (91 percent vs. 82 percent).
Among general high school students, those attending model and experimental schools
have significantly higher levels of family wealth, home education resources, cultural
possessions, and parental education (appendix table D.4).

Comparing School Characteristicsv
As many as 19 percent of middle schools in Shanghai are private, a larger percentage
than that for upper secondary schools (appendix table D.5). Moreover, middle schools
show the strongest accountability: 87 percent report achievement data to parents, and the
achievement data of 84 percent are tracked over time. Compared to general high schools,
middle schools tend to have more autonomy over resource allocation and curriculum. On
average, however, fewer middle school teachers have tertiary qualifications (86 percent),
less than the 97 percent and 92 percent in general high schools and vocational schools,
respectively. At the same time, more middle schools are faced with teacher shortages,
especially in science and math.

At the upper secondary level, compared to general high schools, vocational schools are
characterized by much higher student-teacher ratios, worse teacher-student relations,
more disruptive student behavior, and less accountability, but higher levels of autonomy
over resource allocation and curriculum. All vocational schools surveyed are public and
40 percent of them do not consider academic performance as a criterion for admission.
To further distinguish between the more competitive, and presumably better quality,
general high schools and the less competitive ones, we compare the model and
experimental schools with ordinary high schools (appendix table D.6). Although the two
types of schools are similar on most measures, the major difference lies in the resources
the schools enjoy: model and experimental schools have a higher quality of educational
resources and offer more extracurricular activities. Students in model and experimental
schools also report better student-teacher relations. More principals of model and
experimental schools are evaluated using the school’s performance data. In addition,
model and experimental schools report a higher level of teacher participation in school
governance.

School Characteristics and Student Performance

How do different school characteristics relate to student performance? We run OLS
regressions on the three different samples. A Chow’s t-test suggests that the relationship
between school characteristics and performance varies across the three types of schools,
justifying our analytic approach. Findings are presented below for each sample.

Middle Schools
After controlling for background characteristics and grade levels, a number of middle
school characteristics are significantly correlated with student performance on PISA
(table 7.3). First, private middle school students performed 29 points better than public
middle school students. The better performance seems more indicative of the quality of
schools than of the students’ prior performance; whether or not the school adopts
128


academic performance as an admission criterion is not significant after controlling for
other factors. Within-school tracking seems to be related to performance: middle schools
that group students by ability perform worse than those that do not, and the differences
are statistically significant for mathematics and science. Third, teacher quality plays a
role. Better student-teacher relations are significantly related to better performance. A
higher percentage of teachers with tertiary qualifications is related to higher performance
in mathematics, while schools that reported a teacher shortage in science scored 19 points
lower than those that did not. Interestingly, schools whose principals are evaluated based
on student achievement tend to be lower-performing schools. Finally, middle schools
with more autonomy over curriculum seem to be lower performing; in particular, the
association is significant for science.

Among background control variables, preschool attendance is associated with as much as
23- to 37-point higher PISA scores, and the association is statistically significant across
all three subjects. Family background, including educational resources, cultural
possessions and parental education, also matter for middle school students’ performance.
In addition, girls perform significantly better than boys in reading but worse in
mathematics and science. These results hold cross-nationally in the PISA sample.

Table 7.3. Characteristics and Performance of Middle Schools

                                                Math                  Reading               Science
School characteristics
Organizational
  Student-teacher ratio                1.69        (1.60)      2.86*       (1.10)   1.48         (1.21)
                                                   (13.29                                        (10.83
  Public                               -28.94*     )           -15.81      (9.91)   -17.23       )
                                                   (15.16                  (11.97                (12.68
  Competition from other schools       15.76       )           9.02        )        17.04        )
Policy
  Academic admissions criteria         4.05        (8.13)      3.96        (5.95)   -0.22        (6.58)
  Tracking                             -18.82**    (6.52)      -9.75       (5.21)   -13.47*      (5.73)
Teachers
                                                               10.12**              10.40**
  Student-teacher relations            12.85***    (2.37)      *           (1.66)   *            (2.02)
  % with tertiary qualifications       0.90*       (0.44)      0.60        (0.33)   0.27         (0.33)
                                                   (10.96                           -
  Teacher shortage                     -10.53      )           -3.31       (7.19)   18.95**      (6.92)
Climate
  Teacher behavior                     2.04        (6.51)      0.96        (4.63)   -2.54        (5.31)
  Student behavior                     3.13        (4.84)      2.53        (3.45)   2.42         (4.16)
                                                   (11.34
  Pressure from parents                5.14        )           4.92        (9.23)   14.02        (9.24)
Resources
  Educational resources                0.39        (5.09)      2.77        (4.21)   0.08         (4.27)
  Extracurricular activities           -2.88       (4.44)      -2.86       (2.68)   0.40         (2.89)
Accountability
  To parents                           13.00       (11.87      18.01       (10.03   8.45         (10.90
129

                                               )                  )                   )
                                               (16.58             (12.10              (14.33
   Achievement tracked              -8.18      )        -7.71     )        -18.28     )
   Achievement used to evaluate                         -                  -
     principal                      -21.31*    (8.32)   17.89**   (6.53)   18.85**    (6.21)
Governance
   Autonomy over resource
allocation                          7.16       (4.62)   2.32      (3.32)   3.16       (4.04)
                                                                           -
  Autonomy over curriculum          -8.54      (5.51)   -6.18     (3.45)   10.82**    (3.79)
  School leadership                 -2.55      (8.46)   -2.55     (3.92)   -3.66      (4.75)
  Teacher participation             -0.94      (3.49)   0.74      (2.13)   -0.49      (2.61)
Background controls
                                                        28.97**            -
Female                              -11.25*    (4.30)   *         (3.44)   10.81**    (3.36)
Wealth                              -4.40      (2.29)   -3.26     (2.05)   -4.08*     (1.94)
Home education resources            8.34**     (2.48)   8.74***   (2.12)   9.66***    (2.27)
                                                                           10.39**
Home cultural possessions           12.26***   (2.39)   8.63***   (1.84)   *          (1.83)
Parental education                  3.11**     (0.98)   2.67***   (0.72)   2.34**     (0.77)
                                                        25.65**            22.95**
Preschool for at least one year     36.54***   (6.90)   *         (5.54)   *          (5.67)
                                    -                   -
Remedial                            29.69***            12.00**
                                                   (5.13)              (3.85) -16.29* (6.35)
Enrichment                          17.63***            1.52
                                                   (4.16)              (4.45) 9.11    (7.39)
R2                                  0.36                0.36                  0.35
N                                   2041                2041                  2041
Source: OECD 2010b.
Note: Standard errors in parentheses. All models control for grade level, student background
characteristics, and participation in enrichment and remedial classes.

General High Schools
Similar analysis using the high school student sample generates interesting results (table
7.4). The most important finding is that school resources are significantly related to
performance: higher quality educational resources, as well as provision of extracurricular
activities, is related to higher performance in all three subjects even after controlling for
other individual and school-level factors. As is true for middle schools, better student-
teacher relations are correlated with higher scores. Better performing schools also tend to
have a higher level of teacher participation in school governance. Although in our initial
comparison, experimental and model schools are more likely to have their principals
evaluated using achievement data, this accountability policy does not seem to be related
to student performance after controlling for other factors. Our analysis also reveals
several interesting findings unique to high schools. For example, higher performing
students tend to go to schools with higher student-teacher ratios. Although tracking
within schools mattered for middle schools, it does not have a significant correlation with
high school students’ performance. Despite the overall high percentage of teachers with
tertiary qualifications among high schools, schools at the higher end of the range actually
perform slightly worse than those with lower percentages of teachers with tertiary
   130


   qualifications. Teacher behavior issues reported by principals is negatively linked to
   performance, possibly because principals in better performing schools might be more
   aware of and thus more likely to report problematic teacher behavior.

   Unlike middle school students, among students who have gotten into general high
   schools family background is not significantly associated with performance, except that
   family cultural possessions appear to be associated with higher math scores. Preschool
   attendance does not seem to be correlated with high school students’ PISA performance.

   Table 7.4. Characteristics and Performance of General High Schools

                                            Math                      Reading                 Science
School characteristics
Organizational
   Student-teacher ratio          2.04             (1.92)    4.23**        (1.45)    4.89*         (1.90)
   Public                         86.68**          (29.58)   81.52**       (25.42)   78.01***      (19.74)
   Competition from other
      schools                     0.79             (13.24)   -12.16        (8.93)    -20.35*       (8.93)
Policy
   Academic        admissions
criteria
   Tracking                       -8.2             (8.28)    5.64          (6.86)    -4.8          (6.66)
   Teachers
   Student-teacher relations      5.24*            (2.43)    3.48          (1.99)    1.96          (2.08)
   %        with       tertiary
qualifications                    -5.8*            (2.11)    -6.25***      (1.63)    -6.59***      (1.57)
   Teacher shortage               -6.72            (9.29)    -15.24        (10.86)   -2.37         (8.62)
Climate
   Teacher behavior               -22.1***         (6.23)    -14.9*        (5.83)    -14.17*       (5.68)
   Student behavior               5.56             (5.19)    0.2           (3.79)    1             (4.65)
   Pressure from parents          18.43            (11.74)   14.11         (10.10)   5.98          (9.57)
Resources
   Educational resources          17.58**          (5.18)    12.57***      (3.37)    13.9**        (3.90)
   Extracurricular activities     10.41            (5.34)    10.07**       (3.45)    14.91***      (3.40)
Accountability
   To parents                     3.15             (8.72)    -2.67         (10.09)   5.95          (9.22)
   Achievement tracked            11.22            (12.15)   -2.01         (11.58)   -7.26         (13.44)
   Achievement used to
      evaluate principal          -5.06            (8.81)    1.81          (7.45)    1.92          (6.99)
Governance
   Autonomy over resource
allocation                        -5.08            (4.68)    0.74          (3.96)    3.48          (3.81)
   Autonomy over
      curriculum                  5.47             (4.97)    1.01          (3.43)    5.67          (4.21)
   School leadership              -17.21           (12.89)   -15.72        (8.94)    -14.05        (9.36)
   Teacher participation          7.91*            (3.00)    3.47          (2.29)    5.03*         (1.90)
Background controls
Female                            -11.43**         (3.28)    23.89***      (2.45)    -14.06***     (3.09)
Wealth                            -0.98            (2.14)    -1.69         (1.88)    -2.19         (1.95)
  131

                                           Math                  Reading                   Science
Home education resources         -0.85          (2.93)    2.88          (2.32)    2.25           (2.29)
Home cultural possessions        7.07**         (2.51)    4.31          (2.52)    3.80           (2.20)
Parental education               0.32           (0.89)    0.96          (0.74)    -0.15          (0.75)
Preschool for at least one
year                             9.96           (8.07)    2.40          (5.31)    5.25              (6.45)
Remedial                         -23.76***      (3.63)    -18.84**      (5.66)    -3.96             (5.43)
Enrichment                       0.41           (5.34)    -2.53         (5.88)    12.93*            (5.61)
R2                               0.25                     0.28                    0.26
N                                1376                     1376                    1376
   Source: OECD 2010b.
   Note: Standard errors in parentheses. All models control for grade level,     student background
   characteristics, and participation in enrichment and remedial classes.

  Vocational High Schools
  Results from the vocational school sample are relatively less consistent across subjects,
  most likely because vocational schools often offer specialized curriculums and have
  specific admissions criteria that either do not necessarily take into account academic
  performance or consider only some subjects (table 7.5). Our results confirm that
  vocational schools that admit students based on academic performance score higher on
  PISA in all three subjects. Nonetheless, the data generate several interesting findings. For
  example, lower-performing vocational schools face constant pressure from parents.
  Principals of better-performing vocational schools are more likely to report disruptive
  teacher behavior, but less likely to report problematic student behavior. Schools that have
  higher performance in mathematics tend to have higher quality educational resources and
  higher levels of autonomy over resource allocation. Their performance data are more
  likely to be tracked. There is also evidence that teacher participation in school
  governance might be related to better performance.

  Among background characteristics, the presence of home educational resources and
  cultural possessions correlate with higher scores among vocational school students, but
  family wealth, interestingly, is associated with lower performance. Preschool attendance
  is positively correlated with students’ performance in vocational school, and the
  association is statistically significant for reading scores.

  Table 7.5. Characteristics and Performance of Vocational Schools

                                       Math                     Reading                   Science
  School characteristics
  Organizational
     Student-teacher ratio     0.22         (0.35)     0.02          (0.22)      0.44          (0.36)
     Competition from
  other schools                1.38         (11.41)    23.63*        (9.91)      15.83         (14.99)
  Policy
     Academic admissions
  criteria                     26.28***     (7.28)     22.39**       (6.72)      11.02         (9.14)
     Tracking                  -3.2         (7.24)     -1.79         (6.76)      -4.77         (8.68)
  Teachers
132

   Student-teacher
relations                  1.77       (2.91)    0.81        (1.87)      -0.62      (2.45)
   % with tertiary
qualifications             -0.01      (0.49)    0.14        (0.30)      0.37       (0.60)
   Teacher shortage        49.65***   (8.65)    12.68       (10.21)     17.82      (9.91)
   Climate
   Teacher behavior        -9.43*     (4.01)    -3.78       (3.57)      -2.32      (4.85)
   Student behavior        7.75*      (3.28)    0.89        (2.54)      2.62       (3.01)
                           -
   Pressure from parents   66.39***   (12.04)   -43.48**    (13.29)     -40.41**   (13.11)
Resources
   Educational resources   11**       (3.60)    1.67        (2.72)      -2.74      (3.75)
   Extracurricular
activities                 4.53       (3.11)    0.96        (3.29)      -1.97      (3.43)
Accountability
   To parents              -2.15      (7.69)    -4.29       (5.36)      8.52       (6.09)
   Achievement tracked     19.25**    (5.98)    6.12        (8.49)      3.32       (8.80)
   Achievement used to
evaluate principal         2.37       (9.84)    -6.4        (10.16)     -5.69      (9.99)
Governance
   Autonomy over
resource
     allocation            10.28*     (4.41)    0.98        (2.77)      -0.85      (5.49)
   Autonomy over
curriculum                 3.9        (3.15)    3.39        (2.50)      -0.21      (3.84)
   School leadership       5.63       (4.32)    7.54        (4.12)      3.86       (3.71)
   Teacher participation   0.11       (2.35)    3.64        (2.21)      5.12*      (2.12)
Background controls
                              -
Female                        17.51*** (5.09)       31.33***      (4.68)  -7.25     (4.11)
Wealth                        -9.70**      (2.95)   -8.04**       (2.36)  -9.93*** (2.52)
Home education
resources                     7.24*        (2.67)   9.71***       (2.03)  12.78*** (2.29)
Home cultural
possessions                   9.99***      (2.76)   8.55***       (2.39)  8.64***   (2.42)
Parental education            -0.09        (1.20)   0.65          (0.78)  0.74      (1.11)
Preschool for at least one
year                          12.21        (6.50)   9.28*         (4.61)  8.45      (6.11)
Remedial                      -22.06*      (8.17)   -21.98*       (8.92)  -12.60    (15.99)
Enrichment                    12.05        (9.70)   -18.54        (12.91) -12.06    (13.28)
R2                            0.18                  0.27                  0.15
N                             1063
Source: OECD 2010b.
Note: Standard errors in parentheses. All models control for grade level, student background
characteristics, and participation in enrichment and remedial classes.
133


Lessons from Shanghai

Shanghai’s experience demonstrates that schools play an important role in developing
youth cognitive skills. Most importantly, the quality of the educational resources at a
school appears to be the key characteristic that distinguishes higher-performing schools
from lower-performing schools, especially at the upper secondary level. The experience
of Shanghai has not only demonstrated the importance of abundant investment in
educational resources in schools, but more importantly, it has pinpointed the specific
areas in which resources can most efficiently be put into use. Resources could be invested
in quality instructional resources and facilities, such as laboratories, teaching materials,
computers and the Internet, and libraries. Additional investments could be made in
organizing extracurricular activities, ranging from student choirs and volunteering
activities to arts and sports.

Other critical elements of school resources include abundant teacher resources, high
teacher qualifications, and supportive student-teacher relations. Resources could be
invested in hiring qualified teachers and in providing teacher training. In addition,
Shanghai’s experience also suggests that teacher involvement in decision-making about
school governance, including resource allocation and curriculum, might be beneficial for
both general high schools and vocational schools.

Privatization could be an option to mobilize resources and improve quality even at the
middle school level. Private middle school students in Shanghai perform better on PISA
than public middle school students, even after controlling for family background and
school admission criteria. The better performance measured near the end of middle
school suggests the better quality of private middle schools in Shanghai, rather than the
competencies of students prior to entering middle school.

PISA Shanghai also suggests the benefits of pre-primary education on students’ later
academic paths and cognitive skills development. The positive effects of preschool
attendance remain significant even after accounting for school-level characteristics. In
particular, middle school students who have attended pre-primary school for at least one
year do consistently better in all subjects than those who have not. As middle school
performance predicts student stratification into the general or vocational track and into
the higher- or lower-performing high schools, early childhood development is critical in
providing an important foundation for cognitive skill development.

In light of lessons from Shanghai PISA, the low levels of resource input and limited
coverage of pre-primary education in Yunnan might put Yunnan’s children at a
disadvantage for the later development of important skills. In 2009, the 3,790
kindergartens in Yunnan provided pre-primary education to nearly one million children,
more than double the enrollment in Shanghai of 400,000. However, with a per student
expenditure of Y2,031, Yunnan has the lowest level of spending on pre-primary
education after Guangxi, Hebei, and Anhui (figure 7.4). Although Shanghai has
popularized pre-primary education for children aged three to six, enrollment rates in pre-
primary schools remain low and vary greatly within Yunnan. Half of the 16 cities and
134


prefectures have enrollment rates lower than the provincial average of 60 percent (figure
7.5). Kunming has the highest enrollment with nearly 88 percent.

Figure 7.4. Per Student Expenditure at the Pre-primary Level, Yunnan

 18000
 16000
 14000
 12000
 10000
  8000
  6000
                                                                                                                                                                  Non-budgetary
  4000
  2000                                                                                                                                                            Budgetary
     0
                          qinghai




                             gansu
                         xinjiang


                           jiangsu



                             hubei
                            tianjin




                            jiangxi
                      chongqing
                           hainan



                         guizhou
                        zhejiang




                       shandong




                         guangxi
                         liaoning




                          shaanxi
                              tibet

                 inner mongolia



                                jilin




                             anhui
                             fujian




                           ningxia


                            henan


                          yunnan
                    heilongjiang




                          sichuan
                            hunan
                     guangdong




                             hebei
                       shanghai




                           shanxi
                          beijing




Source: China Education Expenditure Statistical Yearbook 2010.

Figure 7.5. Enrollment Rates in Pre-primary Schools, Yunnan

      87.72
              78.17                                                            71.168.11
          63.83   68.67                                               63.34            60.21
  59.55                                              54.65   52.4854.3
                                                         49.54            43.73
                                                 31.39                                     39.24
                                                                                               30.38
                                                                                         Chuxiong
    Yunnan



                       Qujing




                                                             Lijiang



                                                                               Lincang




                                                                                                             Wenshan




                                                                                                                                                         Diqing
                                                  Zhaotong




                                                                                                    Honghe



                                                                                                                       Banna
             Kunming




                                                                                                                               Dali

                                                                                                                                      Dehong

                                                                                                                                               Nujiang
                                Yuxi

                                       Baoshan




                                                                       Pu'er




Source: Yun, Liu and Li 2011.

References
Bertschy, Kathrin, M.A. Cattaneo, and S.C. Wolter. 2009. “PISA and the Transition into
the Labour Market.� LABOUR 23 (Supplement s1): 111–37.

China Education Expenditure Statistical Yearbook 2010. Beijing: China Statistics Press.

OECD (Organisation for Economic Co-operation and Development). 2009. PISA 2009
Assessment Framework. Paris: OECD.

———. 2010a. Pathways to Success: How Knowledge and Skills at Age 15 Shape Future
Lives in Canada. Paris: OECD.

———. 2010b. PISA 2009 Results: What Students Know and Can Do—Student
Performance in Reading, Mathematics and Science. Vol. I. Paris: OECD.
135


———. 2010c. PISA 2009 Results: Overcoming Social Background – Equity in
Learning Opportunities and Outcomes. Vol. II. Paris: OECD.

———. 2011. Lessons from PISA for the United States, Strong Performers and
Successful Reformers in Education. Paris: OECD.

Yun, Bo, Haijun Liu, and Yaqin Li. 2011. “云���业技术教育�展研究报告.�


Notes

1
    A student is classified as resilient if he or she is in the bottom quarter of the PISA index of economic,
social and cultural status (ESCS) in the country of assessment and performs in the top quarter across
students from all countries after accounting for socio-economic background.
ii
    A student is classified as disadvantaged low achiever, if he or she is in the bottom quarter of the PISA
index of economic, social and cultural status (ESCS) in the country of assessment and performs in the
bottom quarter across students from all countries after accounting for socio-economic background.
iii
    As measured by the slope and strength of the socio-economic gradient. See OECD 2010c.
iv
    A major limitation in using the PISA data is the selection bias. Because students were stratified into
different upper secondary schools based on their performance on the high school entrance examination
before the PISA was administered, the score variance we observe in PISA on the upper secondary level is
more likely an indication of the students’ prior performance than the result of the effects of the upper
secondary schools that the students attend. Nonetheless, because, in general, higher performing schools
admit better performing students, our data still allow us to analyze the more competitive schools that attract
higher performing students. The same issue affects the middle school sample relatively less as the PISA test
was administered near the end of middle school, and admissions into middle schools in Shanghai are based
on neighborhood attendance with a few exceptions.
v
    Definitions and measurements of variables used in the analysis are presented in appendix table D.2.
136


Chapter 8      Conclusions and Policy Implications


As in many parts of the world, Yunnan, China’s economy and labor market exhibit an
increasing demand for both strong, high-level technical skills and non-cognitive skills,
such as communication, problem solving, and teamwork. Upper secondary and tertiary
level education are critical to ensuring that Yunnan’s workers gain these skills.

The demand is driven by various factors. First, the human development stock in Yunnan
is relatively weaker in these skills than in the rest of China. The relatively short supply of
well-educated and high-skilled labor is creating pent-up demand, in particular in the
context of the rapid economic development and increasing globalization of the last three
decades.

At the same time— and unique to China— the central and provincial economic
development strategies and their associated investments in key economic sectors have
become strong drivers for the acquisition of more and better worker skills. China is
determined to transform itself from a “world-factory� into a “high-tech, high-quality�
producer of goods and services. As such, it recognizes the importance of upgrading
worker skills in ensuring a better-skilled labor force. Yunnan, as a strategic border
province, benefitted from the central government’s Bridgehead Strategy and related
investments. Strong demand for higher skills is also driven by technological changes
associated with the modern production process and the expanding service sector.

In addition to the skills shortage, albeit to a lesser degree, Yunnan also exhibits skills
mismatches, as evidenced by unfilled job vacancies in the midst of unemployment, as
well as by feedback from employer surveys. It is important to keep in mind that skills
gaps and mismatches result from an imbalance in the demand and supply of skills and are
indeed normal. Skills gaps and mismatches tend to exist in almost all countries,
developing and developed. What is most important is to recognize is the need to reform
education and training systems so that they are responsive to the ever-changing demands
from the economy and society at large.

This report does not attempt to analyze factors that cause bottlenecks in the labor market
itself, such as hukou (household) registration, minimum wage policy, and pension
portability, all of which may contribute to the mismatch of skills. Rather, the report
closely examines Yunnan’s education and training system and its capacity to be
responsive to the labor market and supply sufficient quantity and quality skilled labor.

Both the Chinese Government and the Provincial Government of Yunnan demonstrate a
strong commitment to education and training. Our analysis reveals the remarkable
progress Yunnan has made in the last three decades in improving various aspects of its
education and training system, including formal TVET education, non-formal training,
work-based training, and general education. A relatively higher proportion of gross
domestic product and the government budget has been spent on education and training in
Yunnan than in most other provinces.
137


Technical and Vocation Education and Training (TVET) is one of the most important
means of providing sought-after skills to Yunnan’s workforce. The most recent vision
and policy direction outlined in the national and provincial Medium- and Long-Term
Education Development Plan calls for: establishing closer partnerships between industry
and TVET; improving teacher quality; creating pathways that allow students more
flexibility to move between TVET and general education— thus improving the
attractiveness of TVET; and emphasizing rural skills development. These are all
important steps that will help create a more conducive environment for effective TVET.

Our analyses reveal certain policy gaps that may have hampered Yunnan’s system from
operating at its full potential. Admittedly the Chinese education and training system is
vast and complicated. Any reform effort, such as curriculum reform and teacher training,
will have enormous resource implications. Developing and implementing a set of
centralized policies poses enormous challenges, especially for a resource-constrained
province like Yunnan.

Globally, nations and sub-national governments (regions, provinces, and so on.), are
responding to the skills demands and challenges in various ways, both within and beyond
the education and training sector. Drawing on global trends, we offer the following policy
suggestions to help Yunnan strengthen its current efforts in skills development and
achieve a demand-driven, high-quality education and training system.

Basic cognitive skills are the foundation for higher-level technical skills and socially
important values. For many, particularly those in lower-income countries with limited
access to primary education, the key to strengthening skills development may lie in
improving the quality of basic education. In the Chinese context, however, we take for
granted that there has been near 100 percent enrollment in nine-year basic education.
Although there continues to be great variation in quality, in general the Chinese system
has achieved marvelous results in basic education. Internationally, Shanghai is
recognized as the number one system, producing the highest average Programme for
International Student Assessment (PISA) score with the least variation. Our
recommendations for strengthening the skills supply system therefore focus more on
improving the skills development system beyond basic education, specifically formal
TVET, non-formal education, and work-based training.

We also identify pre-primary education as another area worthy of policy attention for
Yunnan, not only because it is currently not yet part of compulsory education, but, more
importantly, because evidence suggests that investment in pre-primary education has a
significant payoff in cognitive skills later in life, as our analysis of Shanghai PISA
demonstrates.

Some policy areas, such as governance and coordination, industry participation, financing,
and quality assurance cut across sectors. They pertain not just to the formal TVET system,
but also to non-formal education, and even to work-based training. Sector-specific
recommendations for non-formal training, work-based training— especially for small and
medium-sized enterprises— and pre-primary education follow.
138


Implications for Cross-Subsector Issues

1. Governance and Coordination
The effective functioning of all subsectors could be improved by greater coordination
among the relevant stakeholders in government, education, and industry.

In the short- and medium-term, it is feasible to develop and strengthen the role of existing
coordination mechanisms, such as the Interdepartmental TVET Coordination Committees.
Areas such as policy development, planning, and service delivery at both the provincial
and local levels should be addressed. It is important to revitalize Yunnan’s committee and
regularize its operations.

In the long term, Yunnan might consider consolidating policymaking, planning, financing,
and service delivery of TVET into a central government agency, such as a new Skills
Development Authority, or into one of the existing ministries. This would eliminate the
distinction between schools governed by the Department of Education and those
governed by the Department of Human Resources and Social Security, as Shanghai has
already done for its secondary TVET institutions.

Consolidated governance and management may be even more pertinent for non-formal
training, which currently appears even more fragmented and less structured. The strongly
vested interests of various stakeholders in this sector may mean that consolidation will be
a long-term effort.

2. Industry Participation
Industry and employer involvement can play a key role in ensuring that the educational
and training system, especially at the TVET level, is responsive to the demand side of the
skills equation. Though emerging, industry participation has yet to be systematically
incorporated into the governance structure of any of the education and training sectors.
Industry involvement should be further institutionalized and expanded to include a
diverse range of issues related to skills development, including setting skills development
priorities, developing competency standards for skills certification, allocating resources,
and making policies regarding work-based training. In many countries, employer or
industry-based bodies participate at different levels, providing advice, taking part in
setting development strategies, voicing the needs and demands of the industry, and even
exerting decision-making power.

At the school-level, Yunnan could introduce a legal framework and implementation
guidelines to promote school-industry collaboration. The framework and regulations
should establish the rights and obligations of employers, schools, teachers, students, and
all relevant stakeholders in the school-industry partnership.

Similar principles are applicable in the context of non-formal education and work-based
training. For example, industry and employers could be more involved in managing
work-based training funds, designing training programs, and assessing training
participants.
139


3. Quality Assurance
Under an effective qualifications framework, skills certifications are demanded by
employers and sought by students and employees. Certifications are accepted measures of
competency and substantive learning. A well-developed framework could address both
accreditation of providers and certification of learners.

The existing qualifications framework needs to be updated with standards and
competencies that reflect the demands of the labor market. Competency standards should
be based on input from industry, employers, education and training providers, and other
stakeholders. An updated qualifications framework can also facilitate recognition of non-
formal learning, integrating different kinds of non-formal education and training and
improving their quality and efficiency. The qualifications framework should respect all
forms of learning. It should assess learners based on competencies rather than credentials,
and should not disadvantage one type of learning against another.

At the operational level, Yunnan must consider several important parameters as it
develops its qualifications framework. For example, to what extent will Yunnan rely on
the existing national framework in developing one specific to its provincial context? How
will the coordinating entity be chosen and how will industry input be systematically
incorporated into the development of the framework and each of the underlying standards?
To what extent should the framework take into consideration regional and even
international standards? It might be most feasible to develop a framework first in key
industries with a mature, organized private sector— before extending it to all sectors and
occupations. Good examples can be found in Turkey, Australia, New Zealand, Hong
Kong SAR, China, Singapore, and the European Union. Ultimately what is most
important is the development of a consensus specific to the needs of Yunnan on the scope
and depth of the framework and on the operational details.

4. Financing TVET
The targeted and efficient use of both public and private resources is critical in financing
TVET. School resources appear insufficient in the current system, especially for
secondary TVET— even though public expenditures are high when measured as a
percentage of Yunnan’s provincial gross domestic product and total government spending.
Yunnan should explore ways to use its existing public resources more efficiently.

Returns to private financing of upper secondary and tertiary education are high, however,
and continue to increase. Yunnan could mobilize more private resources to increase the
level of investment in education and training, leading to even greater returns. Yunnan
should be able to increase the relatively low level of private financing and provision in
the current system, letting high-income households share the cost of education and
training, while further diversifying and freeing up other sources of revenue for education
and training. The Yunnan government’s recent announcement of regulations and
establishment standards for private TVET institutions is an important first step.

At the same time, more public resources could be directed toward promoting equity of
financing, and bridging gaps between rural and urban areas and between schools. With
140


the current resource allocation, huge disparities exist in secondary TVET across cities
and prefectures, and between urban and rural areas within Yunnan. Special funds for
TVET tend to disproportionally benefit already stronger schools. Public finance should
be targeted at helping TVET schools in disadvantaged localities, low-income families,
and poor-performing schools with more resource constraints.

5. Monitoring and Evaluation
Monitoring and evaluation provide feedback on the extent to which policies have been
implemented across the system, what outcomes have been achieved, and what
interventions have been effective or ineffective in achieving the expected outcomes. The
very fact that the monitoring system exists and works creates an incentive for local
authorities and schools to implement required policies and adhere to standards.
Evaluation findings can also be used to promote good practices within the system and
replicate them on a wider scale.

Although our analyses identify certain policy gaps in Yunnan’s skills development
system and propose broad policy suggestions, to a large extent, many of the challenges
and issues arise not from a lack of central policies but from a lack of consistent and full
implementation of these policies, particularly in rural and disadvantaged areas. This
suggests the need to invest further in the monitoring and evaluation of existing policies
and programs. Commissioned special studies and ad hoc impact evaluations currently
exist, but conducting routine monitoring and evaluation in a more systematic manner and
with more sophisticated methodologies would be a big step forward.

In addition, the use of pilot and experimental programs, and an analysis of their results, is
critical to ensuring the long-term health of the education and training system. China has
been learning from other TVET systems around the world, but its unique social, cultural,
and economic context makes the adoption of any imported good practice a challenge.
Even within China, practices differ greatly among provinces. Policies and practices that
are effective in Shanghai cannot be easily transplanted to Yunnan. Developing a Yunnan-
style TVET system is the only choice, and the willingness to experiment and evaluate is
key.

Implications for Subsector-Specific Issues

1. Non-formal Training
China’s industrial sector growth and economic development strategies have led to the
need to retrain large numbers of rural workers outside the formal educational system for
transfer to the secondary and tertiary sectors. In Yunnan, rural surplus labor amounts to a
third of the total rural labor force. Yunnan’s rural working-age population averages only
seven years of education, and more than 85 percent of rural workers have received no
training. Those who remain in the agricultural sector also need training in both higher-
value-added agricultural activities, and in technology to improve their productivity. Non-
formal training is a critical means of implementing the necessary rural training.
141


Yunnan faces challenges in improving both the delivery and content of, and access to,
non-formal training. The design and implementation of non-formal training needs to be
better informed by a needs assessment and demand analysis. The forms of delivery and
the content of training should be tailored to the specific needs and circumstances of
training participants, such as rural laborers with relatively low levels of education and
literacy, and the employers in the industrial sectors in which they may ultimately work.

An improved qualifications framework would provide a mechanism that enables non-
formal training programs to lead to skills certification. Participation in non-formal
training is hampered by barriers, such as the requirement of a minimum education level, a
lack of awareness of the value of training, and the quality and relevance of the training
itself. Encouraging both enterprises and workers to value skills certifications, and adding
flexibility that accounts for prior work and life experience, can help overcome these
barriers.

Public, private, and industry funding for non-formal training could be consolidated into a
single fund. A set of transparent funding criteria for the disbursement of training funds
would be necessary. To assure quality of training, the funding criteria should be outcome-
based rather than input-based. Training vouchers would be another option, creating a
competitive market for non-formal training while giving participants more choices.

2. Incentives and Technical Assistance for Work-Based Training
China’s ambitious economic goals require a well-educated and highly-skilled labor force.
Work-based training will be the means by which enterprises improve the skills and
performance of their existing employees.

Employers are currently required to dedicate 1.5 to 2.5 percent of their wage bill to
training, and Yunnan must ensure the effective implementation of this requirement. Other
mechanisms that provide more incentives and quality assurance for work-based training
should also be considered. Korea’s Employment Insurance Scheme is a good example.

In addition to financial assistance, the government should provide technical and
institutional assistance to facilitate the provision of work-based training, especially by
small and medium-sized enterprises. Technical assistance is particularly needed in the
areas of needs assessment, training design and implementation, and monitoring and
evaluation. Institutional support might include organizing groups of small and medium-
sized enterprises to reduce the cost of training design and delivery. Government can also
systematically help firms build partnerships with training providers.

3. Pre-primary Education
Yunnan could invest more heavily in and help popularize pre-primary education. Ample
research demonstrates that early childhood development programs, such as nutrition and
health services, parenting interventions, and center-based programs, can improve
children’s physical and cognitive development, as well as their social and emotional
wellbeing. Data from the Shanghai PISA and other cross-national evidence show the
benefits of early childhood education for children’s later cognitive development.
142


Currently, government investment and enrollment rates in pre-primary school and early
childhood development programs in Yunnan are low. Investing in pre-primary education
might be a relatively cost-effective way to improve the skill levels of a young generation.
    143



                                              Appendix A. Returns to Education and Skills

    Table A.1 Estimates of Rates of Return to Education in Urban China

                                1998                             1995                      2002                        2007
                                0.0324***                        0.0432***                 0.0705***                   0.0610***
Years of education              (0.00132)                        (0.00190)                 (0.00287)                   (0.00318)
Education level (reference group = junior secondary school)
                                               -0.161***                     -0.286**                    -0.330*                      0.156
Less than primary school                       (0.0248)                      (0.0942)                    (0.158)                      (0.222)
                                               -0.0836***                    -0.130***                   0.000552                     -0.0627
Primary school                                 (0.0114)                      (0.0240)                    (0.0440)                     (0.0580)
                                               0.0707***                     0.0987***                   0.116***                     0.111***
High school                                    (0.00812)                     (0.0130)                    (0.0179)                     (0.0237)
                                               0.117***                      0.211***                    0.260***                     0.280***
Secondary TVET                                 (0.0105)                      (0.0148)                    (0.0218)                     (0.0303)
                                               0.172***                      0.283***                    0.422***                     0.455***
Tertiary, non-university                       (0.0129)                      (0.0158)                    (0.0203)                     (0.0263)
                                               0.263***                      0.361***                    0.574***                     0.676***
College and above                              (0.0140)                      (0.0201)                    (0.0255)                     (0.0298)
                                0.0306***      0.0299***         0.0492***   0.0482***     0.0539***     0.0596***     0.0161***      0.0179***
Potential experience            (0.00104)      (0.00107)         (0.00170)   (0.00173)     (0.00229)     (0.00235)     (0.00273)      (0.00269)
                                -              -                 -           -
                                0.000227*** 0.000222***          0.000724*** 0.000712***   -0.00100***   -0.00114***   -0.000274***   -0.000295***
Potential experience squared    (2.31e-05)     (2.39e-05)        (3.81e-05)  (3.93e-05)    (5.13e-05)    (5.28e-05)    (6.00e-05)     (6.00e-05)
                                -0.0651***     -0.0641***        -0.131***   -0.130***     -0.185***     -0.183***     -0.226***      -0.224***
Female                          (0.00625)      (0.00626)         (0.00954)   (0.00953)     (0.0124)      (0.0124)      (0.0157)       (0.0153)
                                0.0152         0.0142            -0.0391     -0.0247       0.00177       -0.00429      -0.0565        -0.0675
Minority                        (0.0166)       (0.0165)          (0.0241)    (0.0240)      (0.0334)      (0.0333)      (0.0740)       (0.0722)
Ownership
                                -0.0644*       -0.0675*          -0.0152     -0.0161       0.0248        0.0260        0.0607*        0.0633*
State-owned enterprise          (0.0313)       (0.0313)          (0.0252)    (0.0251)      (0.0151)      (0.0151)      (0.0299)       (0.0292)
                                -0.142***      -0.145***         -0.263***   -0.252***     -0.180***     -0.180***     -0.0443        -0.0481
Urban collective                (0.0320)       (0.0320)          (0.0273)    (0.0273)      (0.0247)      (0.0245)      (0.0518)       (0.0506)
Government                                                                                                                            0.228***
     144

                                                                                                                                      (0.0324)
                                                                                                                                      0.0451*
Institution                                                                                                                           (0.0207)
                             6.008***       6.663***         7.456***      7.787***       7.751***       8.334***      9.265***       10.02***
Constant                     (0.215)        (0.214)          (0.143)       (0.141)        (0.0741)       (0.0644)      (0.0994)       (0.320)
Observations                 6,788          6,788            11,137        11,137         9,147          9,147         5,695          5,694
 2
R                            0.503          0.505            0.261         0.265          0.327          0.335         0.286          0.321
Number of provinces          17             17               11            11             12             12            9              9
    Source: CHIP 1988, 1995, 2002, RUMiC 2007.
    Note: All models include fixed effects for sectors, cities and provinces. Standard errors in parentheses. TVET =   Technical and Vocational
    Education and Training.
145

Table A.2 Estimates of Rates of Return to Education in Urban Yunnan

                                                    1988                           1995                          2002
                                                   0.0337***                       0.0399***                     0.0717***
 Years of education                                (0.00302)                       (0.00535)                     (0.00784)
 Education levels (reference group = junior secondary school)
                                                                      -0.192**                    -0.188                       -0.948*
   Less than primary school                                           (0.0623)                    (0.294)                      (0.446)
                                                                      -0.0778**                   -0.0915                      -0.0348
   Primary school                                                     (0.0241)                    (0.0566)                     (0.0880)
                                                                      0.0462*                     0.0876*                      0.174**
   High school                                                        (0.0227)                    (0.0398)                     (0.0531)
                                                                      0.137***                    0.171***                     0.224***
   Secondary TVET                                                     (0.0229)                    (0.0381)                     (0.0521)
                                                                      0.202***                    0.213***                     0.430***
   Tertiary, non-university                                           (0.0323)                    (0.0446)                     (0.0560)
                                                                      0.265***                    0.320***                     0.584***
   College and above                                                  (0.0311)                    (0.0590)                     (0.0670)
                                                     0.0143***        0.0133***    0.0570***      0.0549***      0.0641***     0.0668***
 Potential experience                                (0.00243)        (0.00247)    (0.00521)      (0.00538)      (0.00742)     (0.00755)
                                                     0.000147**       0.000156**   -0.000868***   -0.000827***   -0.00108***   -0.00115***
 Potential experience squared                        (5.36e-05)       (5.47e-05)   (0.000117)     (0.000124)     (0.000164)    (0.000168)
                                                     -0.0373*         -0.0347*     -0.0883***     -0.0853**      -0.0624       -0.0610
 Female                                              (0.0156)         (0.0156)     (0.0261)       (0.0265)       (0.0318)      (0.0319)
                                                     0.0260           0.0209       -0.0454        -0.0416        0.0730        0.0711
 Minority                                            (0.0307)         (0.0305)     (0.0371)       (0.0373)       (0.0440)      (0.0441)
 Ownership
                                                     0.0777           0.0596       -0.0750        -0.0865        -0.0866*      -0.0926*
   State-owned enterprise                            (0.130)          (0.130)      (0.0807)       (0.0813)       (0.0386)      (0.0386)
                                                     -0.0609          -0.0748      -0.247**       -0.258**       -0.336***     -0.339***
   Urban collective                                  (0.133)          (0.132)      (0.0894)       (0.0900)       (0.0643)      (0.0645)
                                                     6.106***         6.449***     7.467***       7.809***       7.713***      8.349***
 Constant                                            (0.259)          (0.255)      (0.145)        (0.128)        (0.281)       (0.266)
 Observations                                        893              893          1,082          1,082          833           833
 R2                                                  0.471            0.481        0.262          0.257          0.376         0.384
Source: CHIP 1988, 1995, and 2002.
146

Note: All models include fixed effect for sector and city. Standard errors in parentheses. TVET = Technical and Vocational Education and
Training.
147




Table A. 3 Description of the weighted analytic sample from STEP household survey, Kunming 2012

                                                                  Proportion Standard
                                                                     (%)     errors
         Gender
         Male                                               51.99              0.0186951
         Female                                             48.01              0.0186951
         Level of education
           Less than primary school                         0.37               0.0021799
           Primary school                                   6.55               0.0105317
         Middle School                                      27.73              0.0236402
           High school                                      16.47              0.0145356
           Secondary TVET                                   14.16              0.0121012
           Tertiary, non-university                         19.15              0.0168621
           College and above                                15.58              0.0191622
         Sector
         Primary                                            1.26               0.0044776
         Secondary                                          18.79              0.0218992
         Tertiary                                           79.95              0.0223624
         Occupation
         Managers                                           3.61               0.0064444
         Professionals                                      16.38              0.0171537
         Technicians and associate professionals            7.35               0.0131564
         Clerical support workers                           20.89              0.0148527
         Service workers                                    24.99              0.0173865
         Skilled agricultural, forestry and fishery workers 0.61               0.0025663
         Craft and related trades workers                   7.07               0.0101211
         Plant and machine operators, and assemblers,
         Drivers                                            7.09               0.0100183
         Elementary occupation                              12.02              0.0154774
        Source: STEP Household Survey 2012.
        Note: TVET = Technical and Vocational Education and Training.
148

Table A.4 Estimates of Rates of Return to Education, Kunming in 2012

                                                         Years of         Level of
                                                        education        education
                                                       0.0802***
            Years of Education                         (0.00793)
                                                                        -0.0903
               Less than primary                                        (0.0958)
                                                                       -0.0349
               Primary                                                 (0.0737)
                                                                        0.181***
               Upper secondary (general)                                (0.0463)
                                                                        0.222***
               Upper secondary (TVET)                                   (0.0551)
                                                                        0.457***
               Tertiary, associate degree                               (0.0582)
                                                                        0.754***
               Tertiary, Bachelor's and above                           (0.0538)
                                                       0.00733          0.0161**
               Potential experience                    (0.00463)        (0.00499)
                                                                        -
                                                       -0.000187        0.000398***
               Potential experience squared            (0.000102)       (0.000109)
                                                       -
                                                       0.107**          -0.0145
               Female                                  (0.0319)         (0.0410)
                                                       0.182***         0.172***
               Government agencies                     (0.0471)         (0.0461)
                                                      0.153***         0.144***
               State-owned enterprises                (0.0392)         (0.0390)
                                                       8.880***         9.522***
            Constant                                   (0.108)          (0.0638)
            Observations                              1,026            1,026

                                                     0.279            0.306
            R2
          Source: STEP Household Survey 2012.
          Note: All models control for sector-level fixed effect. Data are weighted.
          Standard errors in parentheses. TVET = Technical and Vocational
          Education and Training.
149

Table A.5 Rates of Return to Education by Gender, Ownership, and Ethnicity

               China                1988             1995          2002          2007
         Female×year            0.00855***       0.0200***      0.0286***    0.0104*
                                (0.00)           (0.00)         (0.00)       (0.00)
         Minority×year          -0.00318         0.00650        -0.00123     0.00868
                                (0.01)           (0.01)         (0.01)       (0.03)
         Ownership (reference group = other)
          SOE×year              -1.81%           -0.0407***     -0.0157**    -0.0404***
                                (0.01)           (0.01)         (0.01)       (0.01)
          Collective×year       -0.0188          -0.0332***     -0.0251*     -0.00286
                                (0.01)           (0.01)         (0.01)       (0.02)
          Government×year n.a.                   n.a.           n.a.         -0.0246**
                                                                             (0.01)
           Institution×year       n.a.           n.a.           n.a.         -0.0185**
                                                                             (0.01)
               Yunnan                 1988           1995          2002           2012a
         Female×year              -0.00317       0.0208*        0.0182        0.00220
                                  (0.01)         (0.01)         (0.01)        (0.00945)
         Minority×year          0.00734          0.0147         -0.0253      n.a.
                                (0.01)           (0.01)         (0.01)
         Ownership (reference group = other)
          SOE×year              0.0541           -0.0238        -0.00626      -0.0130
                                (0.05)           (0.03)         (0.02)        (0.0153)
           Collective×year        0.0437         -0.00448       -0.0318      n.a.
                                  (0.05)         (0.03)         (0.03)
           Government×year        n.a.           n.a.           n.a.       0.0316*
                                                                           (0.0137)
       Source: Author’s calculations using CHIP 1988, 1995, 2002, RUMiC 2007 and STEP
       Household Survey 2012.
       Note: Standard errors in parentheses. SOE = state-owned enterprise.
       a. STEP Household Survey 2012 was administered only in Kunming.
150

Table A.6 Definitions of variables used in the analysis, STEP Household Survey 2012

                Variable                                            Definition
Cognitive Skills
   Lengths of reading                       Size of the longest document normally read at work
Never read at work
2 to 10 pages
11 to 25 pages
More than 25 pages
 Numeracy                                   Level of numeracy skills used at work
Never use
Easy                                        measure or estimate, calculate prices, multiplication,
                                            division
Intermediate                                fractions, decimals or percentages
Advanced                                    Algebra, geometry, trigonometry
   Lengths of writing                       Size of the longest document normally written at work
Never write
2 to 10 pages
11 to 25 pages
More than 25 pages
Think                                       Frequency of undertaking tasks that require at least 30
                                            minutes of thinking at work
Never
Less than once a month
Less than everyday
Everyday
Technical Skills
 Contact                                    Contact with people other than co-workers as part of
                                            the work
                                            1=yes 0=no
 Computer                                   Use computer at work
                                            1=yes 0=no
 Repair                                     Repair/maintain electronic equipment as part of the
                                            work
                                            1=yes 0=no
 Operate                                    Operate or work with any heavy machines or industrial
                                            equipment
                                            1=yes 0=no
 Freedom                                    On a scale of 1 to 10, how much freedom to decide
                                            how to do your work in your own way
                                            0 if self-rating is equal to or less than 2; 1 if greater
                                            than 2
 Fluent in a foreign language               yes = 1 no = 0
Non-cognitive skills                        Average of the items listed on the scale:
                                            1 = almost never, 2 = some of the time, 3 = most of the
                                            time, 4 = almost always
  Extraversion                              Are you talkative?
                                            Do you like to keep your opinions to yourself? Do you
                                            prefer to keep quiet when you have an opinion?
151

                Variable                                           Definition
                                            (reversed)
                                            Are you outgoing and sociable, for example, do you
                                            make friends very easily?
  Conscientiousness                         When doing a task, are you very careful?
                                            Do you prefer relaxation more than hard work?
                                            (reversed)
                                            Do you work very well and quickly?
  Openness                                  Do you come up with ideas other people haven't
                                            thought of before?
                                            Are you very interested in learning new things?
                                            Do you enjoy beautiful things, like nature, art, and
                                            music?
  Emotional stability                       Are you relaxed during stressful situations? (reversed)
                                            Do you tend to worry? (reversed)
                                            Do you get nervous easily? (reversed)
  Agreeableness                             Do you forgive other people easily?
                                            Are you very polite to other people?
                                            Are you generous to other people with your time or
                                            money?
Source: STEP Household Survey 2012.

Table A.7 Estimates of wage returns to cognitive and technical skills, Kunming 2012

                               Model 1      Model 2       Model 3            Model 4    Model 5
 Cognitive Skills
 Lengths of reading (reference=never read at work)
 2 to 10 pages               0.0898*      0.0972*       0.0949*            0.103*      0.0843
                             (0.0449)     (0.0435)      (0.0426)           (0.0436)    (0.0431)
 11 to 25 pages              0.127*       0.126*        0.138*             0.132*      0.124*
                             (0.0565)     (0.0563)      (0.0540)           (0.0563)    (0.0547)
 More than 25 pages          0.140*       0.103         0.105              0.103       0.0919
                             (0.0651)     (0.0634)      (0.0615)           (0.0626)    (0.0619)
 Numeracy (reference=do not use at work)
 Easy                        0.0804       0.103*        0.0742             0.0989*     0.0768
                             (0.0457)     (0.0449)      (0.0446)           (0.0445)    (0.0451)
 Intermediate                0.0175       0.0162        0.0193             0.0112      0.0149
                             (0.0488)     (0.0461)      (0.0440)           (0.0462)    (0.0449)
 Advanced                    0.114        0.0671        0.0602             0.0361      0.0297
                             (0.0649)     (0.0594)      (0.0574)           (0.0603)    (0.0575)
 Lengths of writing (Reference group = do not write at work)
 2 to 10 pages               0.0518       -0.00738      -0.0269            -0.0126     -0.0292
                             (0.0456)     (0.0449)      (0.0423)           (0.0442)    (0.0408)
 11 to 25 pages              0.0776       -0.0220       -0.0327            -0.0258     -0.0356
                             (0.0676)     (0.0650)      (0.0611)           (0.0640)    (0.0596)
 more than 25 pages          0.355***     0.219*        0.181*             0.209*      0.173*
                             (0.0934)     (0.0839)      (0.0810)           (0.0840)    (0.0777)
 Think (reference=never)
 less than once a month      0.0384       0.0106        0.0105             0.00328     0.0122
152

                                 Model 1       Model 2        Model 3         Model 4     Model 5
                               (0.0371)      (0.0360)       (0.0354)        (0.0358)    (0.0353)
 less than everyday            0.104*        0.0722         0.0559          0.0592      0.0518
                               (0.0443)      (0.0418)       (0.0416)        (0.0410)    (0.0405)
 everyday                      0.254**       0.217**        0.180*          0.217**     0.199**
                               (0.0883)      (0.0751)       (0.0702)        (0.0762)    (0.0741)
 Technical Skills
 Contact                       -0.202***     -0.181***      -0.197***       -0.161**    -0.184**
                               (0.0568)      (0.0509)       (0.0517)        (0.0525)    (0.0551)
 Computer                      0.212***      0.0973*        0.158***        0.108**     0.160***
                               (0.0466)      (0.0413)       (0.0409)        (0.0412)    (0.0404)
 Repair                        0.00967       0.00345        -0.0192         0.0124      -0.00890
                               (0.0444)      (0.0406)       (0.0396)        (0.0408)    (0.0412)
 Operate                       0.120         0.123          0.0890          0.0754      0.0620
                               (0.0654)      (0.0634)       (0.0705)        (0.0623)    (0.0672)
 Freedom                       0.0313***     0.0280***      0.0244***       0.0287***   0.0226***
                               (0.00623)     (0.00575)      (0.00549)       (0.00571)   (0.00542)
 Foreign language              0.185***      0.107*         0.101*          0.115*      0.0917
                               (0.0535)      (0.0482)       (0.0484)        (0.0491)    (0.0481)
 Controls
 Level of Education                      Yes          Yes                   Yes         Yes
 Sector                                                                     Yes         Yes
 Occupations                                          Yes                               Yes
 Non-cognitive skills                                                                   Yes
 Potential experience        Yes         Yes          Yes                   Yes         Yes
 Potential      experience
 squared                     Yes         Yes          Yes                   Yes         Yes
 Female                      Yes         Yes          Yes                   Yes         Yes
 Observations                1,007       1,007        1,007                 1,007       1,007
 R-squared                   0.287       0.362        0.408                 0.368       0.420
Source: Author’s calculations using STEP Household Survey 2012.
Note: Standard errors in parentheses.

Table A.8 Estimates of wage returns to non-cognitive skills, Kunming 2012

                                Model 1       Model 2        Model 3        Model 4     Model 5
 VARIABLES
 Non-cognitive Skills
 extraversion                   -0.017        0.0192         0.0294         0.026       0.0175
                                -0.0358       -0.0319        -0.0309        -0.0311     -0.0283
 conscientiousness              0.0827        0.0719         0.0473         0.069       0.0515
                                -0.0502       -0.0445        -0.0395        -0.0442     -0.0404
 openness                       0.174***      0.0805*        0.0810*        0.0843*     0.0666*
                                -0.038        -0.0327        -0.0324        -0.0325     -0.0306
 emotional stability            -0.00502      0.0193         0.0139         0.0222      0.00809
                                -0.0504       -0.0416        -0.0385        -0.0422     -0.0363
 agreeableness                  0.0377        -0.0162        -0.00803       -0.0185     -0.0133
                                -0.042        -0.0367        -0.0339        -0.0359     -0.0322
 Controls
153

 Level of Education                       Yes          Yes        Yes     Yes
 Sector                                                           Yes     Yes
 Occupations                                           Yes                Yes
 Cognitive and technical skills                                           Yes
 Potential experience           Yes       Yes          Yes        Yes     Yes
 Potential        experience
 squared                        Yes       Yes          Yes        Yes     Yes
 Female                         Yes       Yes          Yes        Yes     Yes
 Observations                   1007      1007         1007       1007    1007
 R-squared                      0.13      0.3          0.355      0.311   0.42
Source: Author’s calculations using STEP Household Survey 2012.
Note: Standard errors in parentheses.
154



         Appendix B. Education and Training System in Yunnan

Table B.1 Number of Students Admitted, Enrolled, and Graduated in Yunnan, 2010

          Type of school                       Admissions           Enrollment         Graduates
Kindergarten                             n.a.                    986,900            n.a.
Primary school                           669,300                 4,352,084          736,946
General middle school                    706,567                 2,073,500          641,624
Vocational middle school                 1,517                   6,908              2,875
General high school                      228,997                 632,812            184,415
Specialized and vocational high
schoola                                  169,557                 575,301            128,904
Secondary skilled worker school          38,215                  93,835             28,908
Regular higher education institution     140,966                 439,042            95,379
  Bachelor’s degree                      77,755                  255,788            46,727
  Associate degree                       63,211                  183,254            48,652
Source: National Bureau of Statistics 2011; Yunnan Statistical Bureau 2011.
Note: n.a. = not applicable.
a. Specialized schools and vocational high schools open admissions not only to middle school
graduates of the current year, but also to anyone who has a middle school diploma or equivalent.
Admissions here include only middle school graduates of the current year. Admissions of all
students totaled 279, 015.

Table B.2 Training Duration, Certification Rates, and Employment Rates by Training Providers in Yunnan,
2009

                                     Skilled-worker        Employment training        Non-public training
                                         schools                 centers                   agencies
   Number of agencies            36                       94                         676
   Total number of trainees      106,570a                 143,231                    477,188
   Duration of training
                               number         percent     number        percent      number        percent
    Less than 6 months         n.a.                       110,735       77           39,355        8
    6–12 months                n.a.                       3,456         2            9,594         2
    Over 12 months             n.a.                       6,990         5            10,751        2
  Graduates                    69,645         100         121,181       100          427,415       100
  Total certified              69,645         100         98,542        81           386,772       90
    Junior (level 5)           38,140         55          90,265        74           204,196       48
    Medium (level 4)           20,916         30          6,211         5            149,748       35
    Senior (level 3)           8,358          12          1,936         2            27,113        6
    Technician and senior      2,231          3           130           0.11         5,715         1
     technician (Levels 1
     and 2)
  Employment                   n.a.           n.a.        91757         76           325,662       76
Source: Yunnan Statistical Bureau 2010.
a. Does not include students advancing from middle school.
Table B.3 Formal Technical and Vocational Education and Training Provision in Yunnan, 2010
155

                 Level of Education                    Number of Institutions         Enrollment
      Lower secondary
      Junior vocational                                      14                            6,900
      Upper secondary school
      Specialized secondary                                  91                            290,000
      Skilled-                                               36                            93,800
      worker School
      Vocational                                             184                           279,600
      high school
      Tertiary
      Tertiary Specialized College                           35
      Tertiary technical and vocational                      26
      college
      Total Tertiary                                                                       183,300
    Source: Yun, Liu and Li 2011.

Table B.4 Sources of Financing for Education in Yunnan and China, 2009

                                                         China                          Yunnan
                                               Y, millions    percent           Y, millions percent
Total                                          1,650,271      100.00            44,132       100
     Fiscal allocation for education           1,223,109      74.12             37,283       84
      Percent of GDP                           4                                6
      Budgetary allocation                     1,141,930      69                36012        82
             Central                           117,169        7                 47           0.11
             Local                             1,024,762      62                35,965       82
     Education surcharges at all levels
     of local government                       73,738             4             1,116         3
             Education surcharges              55,567             3             1,023         2
             Local education surcharges        13,506             1             93            0.21
             Local education fund              4,664              0.28          ..            ..
     Allocation by enterprise for
     enterprise-run schools                    4,413              0.27          46            0.11
      School-generated funds                   3,028              0.18          110           0.25
  Investment from sponsors of private
  schools                                      7,498              0.45          174.          0.39
   External donations                          12,550             0.76          169           0.38%
   Income from undertakings                    352,759            21            5,641         13
             Miscellaneous fees portion        251,560            15            4,296         10
   Other income                                54,354             3             864           2
Source: National Bureau of Statistics 2010.
156

Table B.5 Recurrent Budgetary Expenditures per Student by Level of Education, 2010

                                                                              Regular
                               General
                 Primary                      General        Secondary        higher
                               middle
                 school                       high school    TVETa            education
                               school
                                                                              institution
 China
 Personnel     3,083         3,800          3,438      3,374                  5,227
 Public        930           1,414          1,072      1,468                  4,363
   Total       4,012         5,214          4,510      48,420                 9,590
 Yunnan
 Personnel     2,484         3,187          3,289      2,875                  3,909
 Public        803           1,162          1,027      1,854                  46,064
   Total       3,286         4,349          4,316      4,728                  8,515
Source: National Bureau of Statistics 2010.
Note: TVET = Technical and Vocational Education and Training.

a. Includes secondary specialized schools, vocational high schools, and skilled-worker
schools.


Table B.6 Total and Rural Budgetary Expenditure per Student in Yunnan, 2009

                   Primary          General middle          General high          Vocational high
                    school             school                 school                  school
Rural
Recurrent        2,780           3,599                   3,626                 2,334
   Public        587             9256                    800                   398
   Personnel     2,193           2,673                   2,826                 1,936
Infrastructure   160             558                     78                    n.a.
Total            2,940           4,157                   3,704                 2,334

Yunnan
Recurrent      2,773            3,716                    3,898                 3,770
   Public      585              995                      877                   1,097
   Personnel 2,189              2,721                    3,021                 2,673
Infrastructure 154              595                      96                    190
Total          2,928            4,311                    3,994                 3,960
Source: National Bureau of Statistics 2010.
157



Table B.7 Per Student Expenditure in Yunnan and China, 2009

                                                                        General high Secondary                         Associate
                 Pre-primary        Primary school      Middle school   school       TVET           Bachelor's         degree
                 yuan        %      yuan        %       yuan       %    yuan     % yuan        %    yuan          %    yuan      %
  Yunnan
  Budgetary      1,208     60       2,928   81     4,311         89     3,993   64    4,438   61    9,859        58    5606     55
  Non-
  budgetary      823       40       688     19     554           11     2,227   36    2,871   39    7,123        42    4,519    45
  Total          2,031     100 3,616        100 4,865            100    6,220   100   7,309   100   16,982       100   10,125   100
  China
  Budgetary      2,240    57        3,425   82     4,538         82     3,912   55    4,548   57    10,561       50    5,411    45
  Non-
  budgetary      1,662    43        747     18     1,026         18     3,148   45    3,440   43    10,709       50    6,498    55
  Total          3,902    100       4,172   100 5,564            100    7,060   100   7,988   100   21,270       100   11,909   100
Source: National Bureau of Statistics 2010.
Note: TVET = Technical and Vocational Education and Training.

References
National Bureau of Statistics. 2010. China Education Finance Statistical Yearbook 2010. Beijing: China Statistics Press.

———. 2011. China Statistical Yearbook 2011. Beijing: China Statistics Press.

Yun, Bo, Haijun Liu, and Yaqin Li. 2011. “云���业技术教育�展研究报告.�

Yunnan Statistical Bureau. 2010. Yunnan Labor Statistical Yearbook 2010. Beijing: China Statistics Press.

———. 2011. Yunnan Statistical Yearbook 2011. Beijing: China Statistics Press.
158




                             Appendix C. Honghe Rural Labor Training Impact Evaluation

Table C.1 Tracing Households With and Without Training in 2006

                                       2006                    2007             2008             2009             2010
                                 number percent           number percent   number percent   number percent   number percent
Total number of     Training    571       100             571     100      571     100      571     100      571     100
households          No training 1452      100             1452    100      1452    100      1452    100      1452    100
Households with     Training    79        14              72      13       98      17       128     22       95      17
transferred laborer No training 101       7               111     8        147     10       160     11       151     10
   To secondary     Training    18        3               10      2        24      4        25      4        16      3
sector              No training 37        3               29      2        52      4        59      4        59      4
   To tertiary      Training    60        11              44      8        60      11       68      12       38      7
sector              No training 55        4               35      2        52      4        67      5        46      3
                    Training    1         0               2       0        3       1        6       1        1       0
   Inter-provincial
                    No training 1         0               4       0        5       0        6       0        9       1
                    Training    9         2               19      3        20      4        38      7        41      7
   Intra-provincial
                    No training 14        1               54      4        54      4        47      3        48      3
 Source: Honghe Household Income Survey 2006–10.
159

Table C.2 Selected Characteristics of Households With and Without Training

                                                                    No training   Training
                                                                     percent      percent    Chi2
 Ethnic minority village                                                                     ***
 No                                                                16             51
 Yes                                                               84             49
 Village distance to county town                                                             ***
 Less than 2 km                                                    1              0
 2–5 km                                                            4              4
 5–10 km                                                           4              16
 10–20 km                                                          11             23
 More than 20 km                                                   81             57
 Village geographic features                                                                 ***
 Plain                                                             8              30
 Hilly                                                             7              26
 Mountainous                                                       85             43
 Highest education attainment of household members                                           ***
 Illiterate                                                        7              1
 Primary                                                           34             16
 Junior secondary                                                  46             56
 Senior secondary                                                  12             24
 Tertiary and above                                                1              2
 Having household members who are village cadre
                                                                                             ***
 No                                                                91.39          83.01
 Yes                                                               8.61           16.99
                                                                                             t-
                                                                   Mean           Mean       test
 Proportion of households in the village participating in
 training                                                 0.05                    0.83       ***
 2006 net income per person in the household (Y)          3061.15                 4293.59    ***
 2006 household monthly agricultural net income (Y)       208.71                  284.61     ***
 Proportion of young labor (aged 21–40) in the household  0.56                    0.50       **
 N                                                        1452                    571
Source: Honghe Household Income Survey 2006–10.

Technical Notes on Propensity Score Matching
Households with one or more members participating in training and those with none
might differ in various ways that affect their income. Therefore, simply comparing the
outcomes of the two groups might yield biased estimates of the training effect. In order to
account for the nonrandom selection into training of participating households, we use
propensity score matching estimators to control for the observable characteristics that
might confound the relationship between training participation and income.

To implement propensity score matching, we first fit a logit model to estimate the
propensity score for each household. The model includes all of the pretreatment
covariates listed in appendix table C.2, as well as quadratic term incomes, interaction
160


terms of incomes with variables indicating whether the household is in a minority village,
distance to county town, whether the household has a cadre member, and the highest
educational attainment of the household, and interaction terms of proportion of young
laborers with variables indicating whether the household is in a minority village, distance
to county town, whether the household has a cadre member, and highest educational
attainment.

Then we perform one-to-one matching with replacement using Matching, a package in R
(Sekhon 2008). We find matches for all 571 households in the treatment group, indicating
a wide and representative common region of support. Results from balance check after
matching is available upon request.

Matching then produces estimates for (1) average treatment effect for the treated
(ATT)—the average training effect among the households who attend training; (2)
average treatment effect (ATE)—the average training effect over the total population
represented by the analytic sample; and (3) average treatment effect for the controls
(ATC)—the extent of benefit that households with no training would have experienced
had they attended training.

References
Sekhon, Jasjeet S. 2011. "Multivariate and Propensity Score Matching Software with
Automated Balance Optimization: The Matching Package for R." Journal of Statistical
Software 42(7): 1–52.
161



                              Appendix D. PISA Shanghai

Table D.1 PISA Shanghai 2009 School Sample Description

                                  Program                               Total
                       Middle school
                          Middle school                     1,716
                          Mixed secondary school            448
                          Upper secondary/general           4a
                          Total middle school                          2,168
                       Upper secondary general
                          Upper secondary                   1,380
                          Vocational                        22b
                          Mixed secondary school            475
                          Total upper secondary                        1,877
                       Upper secondary vocational           1,070      1,070
                       Total                                           5,115
                      Source: OECD 2010.

a. Students attend lower secondary programs in an upper secondary school.

b. Students attend general upper secondary program in a vocational school.

Table D.2 Variables and Definitions Used in the PISA Analysis

           Variables                                            Definitions
School characteristics
Organizational
   Student-teacher ratio           Student-teacher ratio
   Public                          1 = public 0 = private
   Competition from other          1 = at least one other school in this area that competes for
schools                            students
Policy
   Academic          admissions    1 = always consider academic performance or feeder-school
criteria                           recommendation
   Tracking                        1 = ability grouping in all or some subjects
Teachers
   Student-teacher relations       Scale.a
   Tertiary qualifications         Percent of teachers with tertiary qualification
   Teacher shortage                1 = teacher shortage in the designated subject
Climate
   Teacher behavior                Scale.
   Student behavior                Scale.
   Pressure from parents           1 = school faces constant pressure from parents
Resources
   Educational resources           Scale.
   Extracurricular activities      Scale.
162

         Variables                                           Definitions
Accountability
  To parents                      1 = school provides information to parents on their child's
                                  academic performance relative to other students in school,
                                  relative to national or regional benchmarks, or relative to
                                  students in the same grade in other schools
  Achievement tracked             1 = achievement data tracked over time by an administrative
                                  authority
   Achievement used to            1 = achievement data used in evaluation of principal's
evaluate principal                performance
Governance
   Autonomy on resource           Scale.
allocation
   Autonomy on curriculum         Scale.
   School leadership              Scale.
   Teacher participation          Scale.
Background controls
   Female                         1 = female
   Wealth                         Scale.
   Home education resources       Scale.
   Home cultural possessions      Scale.
   Parental education             Scale.
   Preschool for at least one 1 = attended pre-school for at least one year
year
   Remedial                       1 = attended remedial classes in the designated subject
   Enrichment                     1 = attended enrichment classes in the designated subject
Source: OECD 2010.
Note: School characteristics are reported by school principals, except for student-teacher relations,
which is reported by students. Background controls are reported by students.

a. The scale of choices that was constructed for each variable can be found in OECD
2012.
   163

   Table D.3 Comparison of Student Backgrounds by Type of School

                                                                                Vocational
                                       Middle school        High school           school
                                       mean     s. e.      mean     s. e.      mean     s. e.       t-test
Background characteristic
Female                                0.47      (0.01)    0.55      (0.01)    0.48         (0.03)   ***
Wealth                                -1.33     (0.06)    -1.11     (0.04)    -1.45        (0.04)   ***
Home education resources              -0.32     (0.06)    0.00      (0.03)    -0.30        (0.05)   ***
Home cultural possessions             0.35      (0.04)    0.52      (0.02)    -0.03        (0.05)   ***
Parental education                    12.47     (0.17)    13.23     (0.10)    11.67        (0.15)   ***
Preschool for at least one year       0.86      (0.02)    0.91      (0.01)    0.82         (0.02)   ***
Enrichment in math                    0.36      (0.02)    0.28      (0.01)    0.12         (0.01)   ***
Enrichment in reading                 0.19      (0.01)    0.08      (0.01)    0.07         (0.01)   ***
Enrichment in science                 0.09      (0.01)    0.11      (0.01)    0.06         (0.01)   **
Remedial math                         0.50      (0.02)    0.35      (0.02)    0.16         (0.02)   ***
Remedial reading                      0.30      (0.01)    0.08      (0.01)    0.08         (0.01)   ***
Remedial science                      0.08      (0.01)    0.07      (0.01)    0.03         (0.01)   **
   Source: OECD 2010.

   Table D.4 Comparison of Student Backgrounds in Model and Experimental High Schools and General
   High Schools

                                                 Model and
                                                 experimental       General
    Background characteristic
                                                 mean      s. e.    mean      s. e.    t-test
    Female                                       0.58      (0.02)   0.54      (0.01)   *
    Wealth                                       -0.92     (0.07)   -1.32     (0.07)   ***
    Home education resources                     0.18      (0.03)   -0.19     (0.08)   ***
    Home cultural possessions                    0.69      (0.04)   0.38      (0.04)   ***
    Parental education                           14.13     (0.17)   12.41     (0.17)   ***
    Preschool for at least one year              0.91      (0.01)   0.91      (0.01)
    Enrichment in math                           0.31      (0.03)   0.22      (0.02)
    Enrichment in reading                        0.07      (0.01)   0.07      (0.02)
    Enrichment in science                        0.17      (0.03)   0.08      (0.02)   *
    Remedial math                                0.29      (0.02)   0.37      (0.03)   *
    Remedial reading                             0.06      (0.01)   0.08      (0.02)
    Remedial science                             0.08      (0.01)   0.06      (0.02)
   Source: OECD 2010.
 164

 Table D.5 Comparison of School Characteristics

                                                                           Vocational
                                         Middle school   High school       school
                                         mean    s. e.   mean    s. e.     mean    s. e.   t-test
Organizational
Student-teacher ratio                    11.62    0.36   11.23    0.26     23.33    1.84   ***
Public                                   0.81     0.02   0.93     0.01     1.00     0.00   ***
Competition from other schools           0.83     0.05   0.82     0.06     0.92     0.04
Policy
Academic admissions criteria             0.29     0.05   0.90     0.03     0.59     0.08   ***
Tracking                                 0.56     0.07   0.69     0.06     0.65     0.08
Teachers
Student-teacher relation                 0.19     0.03   0.30     (0.02)   0.09     0.03   ***
Tertiary qualifications (%)              86.35    0.02   97.47    0.01     92.71    0.01   ***
Science teacher shortage                 0.48     0.06   0.29     0.05     0.19     0.04   ***
Math teacher shortage                    0.43     0.06   0.27     0.05     0.23     0.06   *
Reading teacher shortage                 0.37     0.06   0.26     0.06     0.31     0.08
Climate
Teacher behavior                         -0.65    0.18   -0.49    0.19     -0.70    0.14
Student behavior                         0.30     0.23   0.51     0.20     -0.89    0.12   ***
Pressure from parents                    0.21     0.05   0.21     0.05     0.05     0.04   *
Resources
Educational resources                    0.04     0.17   0.21     0.13     0.27     0.18
Extracurricular activities               0.39     0.10   1.53     0.11     1.02     0.16   ***
Accountability
To parents                               0.87     0.04   0.86     0.04     0.52     0.07   ***
Achievement tracked                      0.84     0.05   0.76     0.06     0.26     0.09   ***
Achievement used to evaluate
principal                                0.49     0.07   0.51     0.06     0.25     0.09
Governance
Autonomy over resource allocation        0.81     0.09   0.63     0.13     1.26     0.14   **
Autonomy over curriculum                 -0.16    0.13   -0.46    0.11     0.59     0.17   ***
School leadership                        0.16     0.11   -0.02    0.07     -0.17    0.12
Teacher participation                    0.91     0.22   0.87     0.18     0.84     0.31
  Source: OECD 2010.
165

Table D.6 Comparison of School-Level Characteristics of Model and Experimental vs. General High
Schools

                                          Model and experimental Ordinary
                                          mean         s. e.     mean s. e.       t-test
 Organizational
 Student-teacher ratio                    11.05        0.31        10.68 0.56
 Public                                   1.00         0.00        0.91 0.01 ***
 Competition from other schools           0.80         0.10        0.74 0.12
 Policy
 Academic admissions criteria             1.00         0.00        1.00    0.00
 Tracking                                 0.85         0.09        0.50    0.13 *
 Teachers
 Student-teacher relations                0.42         0.44        0.18    0.04 ***
 Tertiary qualifications (%)              99.62        0.16        98.81   0.68
 Science teacher shortage                 0.15         0.09        0.35    0.09
 Math teacher shortage                    0.10         0.07        0.35    0.11
 Reading teacher shortage                 0.15         0.09        0.29    0.10
 Climate
 Teacher behavior                         0.77         0.40        0.31    0.34
 Student behavior                         -0.28        0.39        -0.57   0.29
 Pressure from parents                    0.25         0.09        0.14    0.08
 Resources
 Educational resources                    0.93         0.25        -0.33   0.17 ***
 Extracurricular activities               2.14         0.20        1.06    0.15 ***
 Accountability
 To parents                               0.90         0.07        0.79    0.07
 Achievement tracked                      0.70         0.12        0.83    0.09
 Achievement used to evaluate principal   0.70         0.23        0.47    0.24 *
 Governance
 Autonomy over resource allocation        -0.28        0.18        -0.58   0.20
 Autonomy over curriculum                 -0.15        0.11        -0.05   0.12
 School leadership                        0.40         0.12        0.73    0.09
 Teacher participation                    1.42         0.35        0.28    0.26 *
Source: OECD 2010.

References
OECD (Organisation for Economic Co-operation and Development). 2010. PISA 2009
Results: What Students Know and Can Do—Student Performance in Reading,
Mathematics and Science. Vol. I. Paris: OECD.

———. 2012. PISA 2009 Technical Report. Paris: OECD.
166