TECHNICAL REPORT CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW TECHNICAL REPORT CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW Wendy Cunningham and Claire H. Hollweg with Johanne Buba, Marong Chea, Dilaka Lathapipat, Une Lee, Anne Lopez, Sodeth Ly, Lan Nguyen, Miguel Sanchez, and Maheshwor Shrestha Foreword Jobs have been a fundamental part of Cambodia’s transformation to a modernizing, globally integrated middle-income country. The opening of markets to international trade and foreign investment not only changed the structure of Cambodia’s economy, but it also had deep implications for jobs, which themselves contributed to the economic development process. Although Cambodia’s current economic model is still delivering new jobs, several emerging mega-trends will alter how external forces shape Cambodia’s jobs picture in the future. The growth of the consumer class in the region, shifting trade partners and patterns, and increasing digital technologies in the workplace could threaten further job growth. Or they could offer opportunities. At the invitation of the Royal Government of Cambodia, the World Bank produced the report Cambodia’s Future Jobs: Linking to the Economy of Tomorrow. It takes a broad approach by considering the elements of Cambodia’s overall development strategy, and identifies policies that could catalyze the creation of more and better jobs within it. The report defines how firms, workers, public and private institutions, and macroeconomic policymakers working together, can create the economy and the jobs of tomorrow. This work is closely aligned with the World Bank Group’s FY19-FY23 Country Partnership Framework, which emphasizes boosting private sector development and fostering human development, among other topics. It is also consistent with the policies defined in the Fourth Rectangular Strategy for Growth, Employment, Equity and Efficiency. The report concludes that good job growth is best spurred by a four-pronged strategy. First, better leverage the foreign sector as a source of jobs through policies that diversify exports and FDI into higher value-added segments of global value chains. Second, increase the quality of jobs in domestically owned firms and household enterprises through policies that support entrepreneurs to create and grow their firms. Third, deepen the linkages between domestic enterprises and the FDI sector. Fourth, build a skills development system that will attract higher-value FDI and increase productivity across the economy while facilitating labor mobility to the highest-value jobs. We hope that Cambodia’s Future Jobs: Linking to the Economy of Tomorrow serves to inspire and connect policymakers, the private sector and development partners to create the conditions for Cambodia’s rms and workers to thrive in tomorrow’s jobs. Inguna Dobraja Country Manager Cambodia Country Of ce CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW v TECHNICAL REPORT Acknowledgements The report was written by a World Bank team lead by Dr. Wendy Cunningham and Dr. Claire H. Hollweg. The team included Sodeth Ly, Marong Chea, Miguel Sanchez, Maheshwor Shrestha, Johanne Buba, Une Lee, Lan Nguyen, and Anne Ong Lopez. Excellent research assistance was provided by Audrey Stienon and Anam Rizvi and administrative support provided by Da Lin, Socheat Ath, and Corinne Bernaldez. The report was prepared under the guidance of Victoria Kwakwa, Vice President for East Asia and the Pacific; Ellen Goldstein, Country Director for Cambodia, Myanmar, and Lao PDR; Inguna Dobraja, Country Manger for Cambodia; and Philip O’Keefe and Jehan Arulpragasm, Social Protection Sector Managers at the World Bank. The report benefitted from detailed peer review guidance provided by Obert Pimhidzai, Margo Hoftijzer, Julian Clarke, Tom Farole, and many colleagues in the World Bank who provided insights, suggestions, and improvements to the report process and the final document. The team wishes to thank H.E. Vongsey Vissoth (Permanent Secretary of State, MEF), H.E. Hem Vanndy (Under Secretary of State, MEF), Mr. Chea Kok Hong (Director of Macroeconomic and Fiscal Policy Department, MEF), H.E. Laov Him (Director General of TVET, MLVT), H.E. Hong Cheoun (Director General, NEA), and H.E Sisowath Doung Chanto (President of National Productivity of Cambodia) for their constant guidance and support throughout the process. The team is also grateful to participants in numerous consultations meetings with the Ministries of Economy and Finance; Labor and Vocational Training; Commerce; Education, Youth and Sport; and Tourism, as well as with representatives from the private sector, NGO community, and the development partners. Many thanks to the Korean Government for their generous financing of this project through the Korea World Bank Partnership Facility (KWPF). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW vi TECHNICAL REPORT Contents Foreword v Acknowledgements vi List of Boxes, Figures and Tables ix Currency Equivalents xiv Acronyms and Abbreviations xiv Executive Summary xv CHAPTER 1: An Introduction to Jobs in Cambodia 1 Wendy Cunningham, Claire H. Hollweg and Maheshwor Shrestha 1.1 Introduction 1 1.2 Cambodia’s Labor Market in the Past 4 1.3 Cambodia’s Jobs Today 6 1.4 Job Quality: A Review of Job Type, Occupation, Wages and Labor Productivity 7 1.5 Inclusive Jobs: Women and Older Workers Have Lower-Quality Jobs 16 1.6 Cambodia’s Future Jobs 20 1.7 References 23 Annex 1.1: Data Source and Definitions 24 Annex 1.2: Additional Figures and Tables 26 CHAPTER 2: Macroeconomic Policies Conducive for Creating Good Jobs 33 Dilaka Lathapipat, Sodeth Ly and Miguel Eduardo Sánchez 2.1 Introduction 33 2.2 Structural Transformation Supported Job Creation in Cambodia 33 2.3 Investments are Needed for Cambodia to Capitalize on its Demographic Dividend 38 2.4 The Current Macroeconomic Situation Presents Challenges to Jobs 42 2.5 Macroeconomic Policies for Sustaining Strong Growth and Job Creation 44 2.6 References 47 Annex 2.1: Methodology for Economic Growth Projections 48 CHAPTER 3: Creating More and Better Jobs through Private Sector Development 53 Johanne Buba and Maheshwor Shrestha 3.1 Introduction 53 3.2 The Private Sector is an Important Contributor to Job Creation 55 3.3 Foreign Direct Investment Also Contributes to Domestic Job Creation 62 3.4 The Domestic Private Sector Could Create Many More Jobs, but is Struggling To Do So 66 3.5 Household Enterprises are An Attractive Alternative to Many 71 3.6 Policies for More and Better Private Sector Job Creation 75 3.7 References 83 Annex 3.1: Household Enterprises (Regressions) 85 CHAPTER 4: Jobs and the Export Sector 89 Claire H. Hollweg and Anne Ong Lopez 4.1 Introduction 89 4.2 How Do Cambodian Exports Support Wages and Jobs? 90 4.3 Policies for Fostering a Job-friendly Export Agenda for Cambodia 109 4.4 References 115 Annex 4.1: Labor Sophistication of Exports 116 Annex 4.2: Labor Content of Exports 117 Annex 4.3: Additional Figures and Tables 119 CHAPTER 5: Workforce Skills for More, Better, and More Inclusive Jobs 129 Maheshwor Shrestha and Une Lee 5.1 Introduction 129 5.2 Most Cambodian Workers See Low Economic Benefits from Education 130 5.3 Workers are poorly equipped to supply the skills that employers demand 136 5.4 The Tasks Required in Common Cambodian Occupations are Changing 137 5.5 A Gender Wage Gap Exists in the Cambodian Labor Market 145 5.6 Policies to Improve the Skills of Cambodian Workers 148 5.7 References 153 Annex 5.1: Additional Figures and Tables 155 CHAPTER 6: Policy for Future Jobs Consistent with Cambodia’s Vision 2050 163 Wendy Cunningham and Claire H. Hollweg 6.1 Introduction 163 6.2 Seven Sets of Policies for More, Better, and More Inclusive Jobs 165 1. Diversify Exports and FDI into Higher Value-added Value Chains (or Segments of Value Chains) 165 2. Streamline Procedures and Reduce the Costs of Establishing and Expanding SMEs 165 3. Support Household Enterprises to Enhance Their Productivity 167 4. Support the Development of Links between FDI Firms and Domestic Input-supplying Firms 168 5. Build a Skills Development System that Will Attract Higher-value FDI and Increase Productivity across the Economy 169 6. Promote Efficient Labor Mobility and Job Matching 170 7. Regain Macroeconomic Independence and Exchange Rate Flexibility 170 6.3 Policy implementation 171 6.4 References 173 List of Boxes, Figures and Tables BOXES Chapter 1 Box 1.1: Jobs Definitions 3 Box 1.2: Opportunities for Entrepreneurship and Self-Employment in the Global Digital Economy 21 Chapter 2 Box 2.1: The Peru De-Dollarization Experience and its Policy Implications for Cambodia 46 Annex Box 2.1: Assumptions of the Model 50 Chapter 3 Box 3.1: Minimum Wage in the Garment Sector 64 Box 3.2: The Importance of Good Business Practices for Sales Growth and Employment 77 Box 3.3: Strengthening the Entrepreneurial Ecosystem—The Jobs Fund in South Africa 78 Box 3.4: YouWin! In Nigeria: Generating Jobs through a Business Plan Competition that Identifies a Pool of High-Growth Firms 79 Box 3.5: Using Mobile Phones to Deliver Simplified Financial Advice in the Dominican Republic, India, and the Philippines 81 Box 3.6: Key Characteristics of Successful Programs for Integrating Household Enterprises into the Economy 82 Chapter 4 Box 4.1: Concepts and Measures 90 Box 4.2: Labor Intensity in Cambodia’s Garments Sector 96 Box 4.3: International Experience in Backward Linkages Programs 111 Box 4.4: The Role Played by Joint Ventures in Upgrading the Sri Lankan Apparel Sector 112 Box 4.5: Export Promotion in Jamaica 113 Chapter 5 Box 5.1: Methodology Used to Create the Task Classification for Cambodian Occupations 139 Chapter 6 Box 6.1: The Stages of Global Integration for Development and Jobs 163 Box 6.2: Policies to Move Up the Value Chain in Cambodia’s Tourism Sector 166 Box 6.3: Policy Implications for Inclusive Jobs 172 CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW ix TECHNICAL REPORT FIGURES Chapter 1 Figure 1.1: Growth and Poverty in Cambodia 1 Figure 1.2: Sectoral Contribution to GDP, 1994-2017 2 Figure 1.3: Sectoral Contribution to Jobs, 2004-2015 2 Figure 1.4: Exports of Goods and Services and Net FDI Inflows, 1991-2016 4 Figure 1.5: Trends in Rubber and Rice Prices, and Remittances 5 Figure 1.6: Net Effect of High-Skilled Emigration on the Share of High-skilled Workers in ASEAN Source Countries 6 Figure 1.7: Labor Force Participation Rates and GDP per capita, global, 2016 7 Figure 1.8: Distribution of Job Types in Cambodia, 2014 8 Figure 1.9: Distribution of Jobs by Type 9 Figure 1.10: Distribution of Jobs by Occupation, 2014 10 Figure 1.11: Evolution of Wages, by wage quintile, and Share of Wages Below the Poverty Line, 2007-2015 11 Figure 1.12: Distribution of Wages in Agriculture and Non-Agricultural Jobs, by worker education 11 Figure 1.13: Impact of Minimum Wage Increase 12 Figure 1.14: Productivity Trends, by Sector 13 Figure 1.15: Average Unit Labor Costs of Apparel Firms Across Comparator Countries 14 Figure 1.16: Global Secondary and Tertiary School Completion Rates and GDP 15 Figure 1.17: Women’s Labor Force Participation and Per Capita GDP in Cambodia and Comparator Countries 16 Figure 1.18: Vertical Gender Segregation within Industries 17 Figure 1.19: Returns to Education by Gender 17 Figure 1.20: Gender Gap in Hourly Wages 18 Figure 1.21: Source of Employment, by Age 19 Figure 1.22: Occupation and Sectoral Composition of Employment by Age 19 Annex Figure 1.1: Garments and construction-related occupations boomed during 2007-2015 26 Annex Figure 1.2: Sector Employment and Age, 2014 27 Annex Figure 1.3: Occupation and age, 2014 28 Chapter 2 Figure 2.1: Growth in GNI per capita and GDP, 1994-2016 34 Figure 2.2: Cambodia GDP Growth Decomposition 36 Figure 2.3: Labor Productivity Growth Decomposition 36 Figure 2.4: Productivity in Agriculture, Industry and Services 37 Figure 2.5: Real Output Share Weighted Average Components of Labor Productivity Growth 38 Figure 2.6: Projected Population Growth and Dependency Ratios in Cambodia 39 Figure 2.7: Estimated Percent of Population with At Least an Upper Secondary Education, Baseline and Reform Scenarios 39 Figure 2.8: Investment, Savings, and Rate of Return to Capital 40 Figure 2.9: Projected Evolution of Key Economic Variables Under Baseline and Reform Scenarios 41 CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW x TECHNICAL REPORT Figure 2.10: Export Growth to the EU and EUR/USD Exchange Rate 42 Figure 2.11: Equipment and Construction in Total Investment in Cambodia 43 Figure 2.12: Investment, Savings and Infrastructure in Cambodia 44 Chapter 3 Figure 3.1: Jobs Contributed by the Private Sector 55 Figure 3.2: Number of Firms by District 56 Figure 3.3: Total Jobs by District (quartile, by job density within the district) 56 Figure 3.4: Employment Elasticity to GDP Across Countries 57 Figure 3.5: Employment Elasticity to Sales Growth for Apparel Firms 58 Figure 3.6: Monthly Wages in Private Firms, 2014 58 Figure 3.7: Number of Hours Worked in Main Wage Job, 2014 58 Figure 3.8: Distribution of Monthly Wages by Educational Achievement 59 Figure 3.9: Prevalence of Non-farm Household Enterprises in Cambodia 60 Figure 3.10: Number of Household Members Working in Household Enterprises 61 Figure 3.11: Percent of Garment Firms’ Inputs Sourced Locally 62 Figure 3.12: Correlation between Wage Employment Rates and the Number of Non-farm Enterprises 63 Figure 3.13: Value of Fixed Assets per Full-time Permanent Worker, 2010 64 Figure 3.14: Average Wages and Unit Labor Costs in Apparel Firms 65 Figure 3.15: Lifecycle of Firms in Cambodia and Selected Comparator Countries 66 Figure 3.16: Share of Young Firms in the Economy 66 Figure 3.17: Domestic Firm Employment and Sales, 2010 67 Figure 3.18: Biggest Perceived Obstacles to Sales Growth Among Cambodian Firms, by Size 68 Figure 3.19: Objective Measures of Possible Barriers to Sales Growth 69 Figure 3.20: Wages in Cambodia by Type of Firm, 2014 70 Figure 3.21: Earnings for Household Enterprise Workers and Wage Workers 71 Figure 3.22: Composition of Household Employment in Cambodia 72 Figure 3.23: The Role Played by Education in the Decision to Hire an External Worker in a Household Enterprise 73 Chapter 4 Figure 4.1: Labor Content of Cambodian Exports, 2004-2014 91 Figure 4.2: Exports of Goods and Services and Net FDI Inflows, 1991-2016 91 Figure 4.3: CAGR of Exports and Labor Content of Exports and Output, 2004-2014 92 Figure 4.4: Total Labor Value-Added of Cambodian Exports (backward linkages) 92 Figure 4.5: Share and Value of Cambodia’s Total Exports to the World, 2000-2016 93 Figure 4.6: Herfendehal Hershman Index (HHI) of Product Concentration 94 Figure 4.7: Labor Intensity of Exports, 2014 95 Figure 4.8: Labor Intensity in Apparel Exports Across Countries 95 Figure 4.9: Total Labor Content in Gross Exports across Sectors and Countries, 2014 96 Figure 4.10: Export Intensity and Labor Intensity Across Sectors, 2014 97 Figure 4.11: Labor Sophistication of Cambodia’s Exports Relative to Its Peers 98 Figure 4.12: Labor Sophistication of Cambodia’s Exports by Destination Market 99 Figure 4.13: Cambodia’s Integration in GVCs 100 Figure 4.14: Direct, Indirect, and Total Labor Content of Cambodian Exports 102 CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW xi TECHNICAL REPORT Figure 4.15: Labor Value-Added in Cambodian Exports, 2014 (Forward Linkages) 102 Figure 4.16 Firms Indirectly Exporting At Least 10 Percent of Their Sales 103 Figure 4.17: Share of Inputs Sourced Domestically 103 Figure 4.18: Composition of Services Exports, 2016 104 Figure 4.19: Total Contribution of Tourism to Employment, 1995-2016 105 Figure 4.20: Productivity and Firm Size of Exporters versus Non-exporters 106 Figure 4.21: Characteristics of Exporters versus Non-exporters 107 Figure 4.22: Biggest Obstacles Faced by Exporters and Non-exporters 108 Annex Figure 4.1: Cumulative Distributions of Cambodia’s Export Basket to the World, 2000-2016 120 Annex Figure 4.2: Cumulative Distribution of Cambodia’s Export Basket to the EU28, EAP, and the US, 2016 121 Annex Figure 4.3: Share and Value of Cambodia’s Total Exports to EU27, China, the US, and EAP 122 Annex Figure 4.4: Composition of Cambodia’s Export Basket 123 Annex Figure 4.5: Intensive and Extensive Margins 124 Annex Figure 4.6: Labor Intensity by Sector 124 Annex Figure 4.7: Labor Intensity of Exports across Sectors and Years 125 Annex Figure 4.8: Labor Intensity of Exports across Sectors and Countries 125 Annex Figure 4.9: Herfindahl Hershman Market Concentration Index in All Sectors 126 Annex Figure 4.10: Her ndahl Hershman Market Concentration Index in Selected Sectors 126 Annex Figure 4.11: Labor Intensity of Wearing Apparel Exports, 2004-2014 127 Chapter 5 Figure 5.1: Educational Attainment of the Working Age Population, 2004 and 2014 130 Figure 5.2: Secondary and Tertiary Completion Rates and GDP 130 Figure 5.3: Age and Education Profile of Cambodian Workforce 131 Figure 5.4: Occupation Choices by Years of Schooling 132 Figure 5.5: Returns to Education by Years of Schooling 132 Figure 5.6: Quality-Adjusted Years of Education 134 Figure 5.7: Skill Shortages in the Cambodian Labor Market 135 Figure 5.8: Opinions of Large Employers about Which Key Skills Their Workers Lack 136 Figure 5.9: Skill-Task Intensity of the Ten Largest Occupations 140 Figure 5.10: Skill-Task Intensity of the Ten Fastest-Growing Occupations 141 Figure 5.11: Task Intensity by education level 142 Figure 5.12: Task Intensity and Age of Workers 143 Figure 5.13: Evolution of Task Intensity, 2009-2015 144 Figure 5.14: Decomposition of the Gender Wage Gap, 2007-2015 147 Figure 5.15: Sample Workforce Development SABER Results 151 Annex Figure 5.1: Occupation and Age of Cambodia’s Labor Force, 2014 155 Annex Figure 5.2: Occupation and Education by Sub-group, 2014 156 Annex Figure 5.3: Returns to Education in Wage Jobs, 2007-2015 157 Chapter 6 Figure 6.1: Seven Policies for More and Better Jobs 164 CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW xii TECHNICAL REPORT TABLES Chapter 1 Table 1.1: Fastest Growing and Declining Occupations, 2007-2015 10 Annex Table 1.1: Profile of the Cambodian Labor Force, 2014 29 Annex Table 1.2: Employment of the Working-age Population 30 Annex Table 1.3: Number of Wage Workers, in thousands 30 Annex Table 1.4: Largest Occupations in 2009 and 2014 31 Annex Table 1.5: Real Wage Earnings, 2007-2015 31 Chapter 2 Table 2.1: Contributions to GDP Growth by Sector 34 Table 2.2: Net Employment Creation by Sector 35 Annex Table 2.1 OLS Estimates of Human Capital Externalities from Average Years of Schooling, Years of Secondary Schooling, and Years of Post-Secondary Schooling at the Provincial Level, 2009-2014 51 Chapter 3 Table 3.1: Wage Workers in Private Firms (Formal and Informal), by Sector (‘000) 56 Table 3.2: Size of Regional Markets 71 Table 3.3: Probability of Hiring a Wage Employee 74 Table 3.4: Interventions Targeting Self-employed Workers and Micro-enterprises 80 Annex Table 3.1: Determinants of Productivity and Selection into Household Enterprises 85 Annex Table 3.2: Determinants of Hiring a Wage Employee 86 Annex Table 3.3: Employment in the Wage Sector and the Number of Households with Enterprises 87 Chapter 4 Table 4.1: Composition of Indirect Exporters 104 Table 4.2: Characteristics of Exporters verses Non-Exporters, Manufacturing Sector 107 Chapter 5 Table 5.1: Definitions of Skill-Task Categories 138 Table 5.2: Decomposition of the Gender Gap in Wage Work, 2014 146 Table 5.3: A Broad Set of Actors for Lifelong Learning 149 Annex Table 5.1: Profile of the Cambodian Labor Force, 2014 158 Annex Table 5.2: Employment of the Working Age Population, thousands 159 Annex Table 5.3: Number of Wage Workers, thousands 159 Annex Table 5.4: Largest Occupations, 2009 and 2014 160 Annex Table 5.5: Real Wage Earnings, 2007-2015 160 Annex Table 5.6: Determinants of Task Content of Jobs, 2014 161 CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW xiii TECHNICAL REPORT Currency Equivalents Exchange Rate Effective July 22, 2019 Currency Unit = Cambodian Riels (KHR) USD 1 = KHR 4,081.80 Acronyms and Abbreviations ASEAN Association of Southeast Asian Nations NGO Non-Governmental Organization BEEPS Business Environment and NSSF National Social Security Fund Enterprise and Performance Survey OECD Organization for Economic CAGR Compound Annual Growth Rate Co-operation and Development CAL Computer Aided Learning PIM Public Investment Management CDC Council for Development of Cambodia PISA Programme for International Student CIT Corporate Income Text Assessment CSES Cambodia Socio-Economic Survey PPP Purchasing Power Parity EAP East Asia Pacific QIP Qualified Investment Projects EBA Everything But Arms R&D Research and Development EU European Union RBF Results-Based Financing FDI Foreign Direct Investment RCEP Regional Comprehensive Economic FTA Free Trade Agreement Partnership GDP Gross National Product RGC Royal Government of Cambodia GNI Gross National Income SABER Systems Approach for Better GSP Generalized System of Preferences Education Results GVC Global Value Chain SEZ Special Economic Zone HHE Household Enterprise SME Small and Medium Enterprise HHI Herfendehal Hershman Index STEM Science Technology Engineering Math ICT Information Communication Technology TFP Total Factor Productivity ILO International Labor Organization TIFA US-Cambodia Trade and Investment IMF International Monetary Fund Framework Agreement IT Information Technology TVET Technical and Vocational Education LDC Least Developed Country UNTAC United Nations Transitional Authority LFP Labor Force Participation in Cambodia MFN Most Favored Nation VAT Value-Added Tax MOOC Massive Open Online Course WDI World Development Indicator NEA National Employment Agency WTO World Trade Organization Regional Vice President Victoria Kwakwa Practice Senior Director Michal Rutkowski Practice Manager Philip O’Keefe Task Team Leaders Wendy Cunningham and Claire H. Hollweg CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW xiv TECHNICAL REPORT Executive Summary Cambodia has succeeded in implementing an family farming or household enterprises, which are lagging effective FDI-led development strategy. Its 7 percent behind the rest of the economy in terms of productivity GDP growth rates have been accompanied by rapidly falling and growth. Fully 94 percent of jobs are in low-skilled poverty rates. Over 1 million workers are now employed occupations, with managers and professionals accounting for in the garment sector, and the share of jobs in industry less than 5 percent of all jobs. While wages have increased continues to grow. in recent years, particularly for those near the bottom of the income distribution, more than half of jobs operate outside Cambodia’s exposure to the global economy of the modern economy. Family farming and household combined with the government’s outward-looking enterprises primarily trade amongst themselves, which limits economic policies have yielded good results. First, by their growth and profits. The economy provides markedly opening its borders to international trade and investment, fewer opportunities for women and older workers than it does Cambodia attracted FDI, particularly in the garment sector, for men and younger workers. because of its low trade barriers, stable government, and relatively hassle-free business practices. Second, the spike The current model is further at risk due to several in international commodity prices was a boon to a still largely emerging mega-trends that will alter how external agricultural economy, thus alleviating pressure to make a forces shape Cambodia’s jobs picture in the future. quick transition out of agriculture. Third, a further reduction Changing consumption patterns resulting from a growing of trade costs and the existence of an abundance of highly consumer class within Cambodia and across Asia will change mobile, low-skilled, and, thus, low-cost, labor provided a the composition of both domestic and export demand. New further impetus to investments in the garment sector and a trade agreements and technologies will further shift global continuation of the structural transformation process. trade patterns. Automation has the potential to create more, better, and more inclusive jobs, but also to make others Although Cambodia’s current economic model is disappear. The advent of new technologies means that still delivering good new jobs, its success may be workers will require a more complex set of skills than in slowing down. The garment sector that has been the the past. main driver of Cambodia’s jobs and growth is showing some weaknesses. Job creation is slower than in the Cambodia can manage these trends by taking the same sector in other countries. Its labor productivity steps necessary to move into the next stage of FDI- has stagnated, and unit labor costs are higher than those led development: diversi cation and densification. in competitor countries, thereby negatively affecting This would mean pursuing higher-value FDI and fostering Cambodia’s competitiveness. The potential loss of trade links between exporting firms and domestic input-supplying preference with the EU and the US in the near term firms. However, this will only be possible if Cambodia lays will also impede the country’s competitiveness in key the foundations by reforming its own policies to attract and export markets. The garment sector has few links with retain more, better, and more inclusive jobs. domestically owned small and medium-sized enterprises (SMEs) as it relies less on local sourcing than do garment We propose seven jobs-speci c policies that, in firms in other countries—such as Vietnam, Turkey or conjunction with broader sectoral development India—or other export-oriented industries in Cambodia. strategies, have the potential to generate better The labor force does not have the skills to attract jobs in Cambodia in the future. These policies are in higher-value FDI, nor do they have the means to acquire alignment with the principles and recommendations of the those skills. Royal Government of Cambodia’s Rectangular Strategy- Phase IV. We present these seven policy actions in one Moreover, a large share of Cambodia’s current jobs document to show how they can improve the quality and may not be of the kind that the country will aspire to increase the inclusiveness of Cambodia’s jobs. The specific create in the future. More than half of all jobs are still in policy recommendations are as follows: CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW xv TECHNICAL REPORT First, diversify exports and FDI into higher value-added Fifth, build a skills development system that will value chains or segments of value chains. Most current attract higher-value FDI and increase productivity jobs are in low-value segments of global value chains (GVCs) across the economy. While a solid education system is in which Cambodia’s exports sectors have the lowest median the foundation of any future workforce, today’s workforce value-added among those of their peer countries. Simplifying in Cambodia is getting by with only 6.3 years of education processes, providing incentives to foreign investors, and on average, while having few opportunities to upgrade their creating quality assurance facilities will encourage the skills. This points to the need for a dual strategy. First, it diversification of exports and FDI into higher value-added value would be advisable to reform today’s education system to chains (or segments of value chains). Signing, finalizing, and help the next generation of workers to acquire the broad enforcing international trade and investment agreements with range of skills needed to work in a knowledge-intensive key trade partners will also be vital to mitigating the risk of economy. Second, policymakers should consider developing losing special access to markets such as the European Union. a strategy (including incentives) for engaging enterprises in the design, financing, and support of a modernized technical Second, streamline procedures and reduce the costs and vocational education and training system (TVET). of establishing and expanding SMEs, which have considerable potential to create jobs. It is harder for firms Sixth, promote efficient labor mobility and job to expand in Cambodia than in other countries. Young firms matching. A mis-match between skills and jobs stifles in Cambodia —defined as those that have been operating for productivity and frustrates workers’ sense of well-being. less than five years—create little employment, even though Helping people to find and acquire the right jobs will reduce this stage is critical for a firm’s growth. Establishing a range of these inefficiencies. First, policymakers should systematically institutions that can support entrepreneurs during the various produce and disseminate information about jobs inside and stages of a firm’s development, including ideation, incubation, outside of Cambodia to help students, jobseekers, education and acceleration, will help SMEs to grow and create jobs at and training institutes, and employers to make skills the various stages of their lifecycles. There is also a need to development choices that are aligned with rapidly evolving reduce the costs that they incur in doing business as well as market demand. Second, they should continue opening to enhance their access to finance to be able to compete on formal international migration channels while supporting both the domestic and international markets. programs that encourage circular migration. Third, help household enterprises enhance their Seventh, regain macroeconomic independence productivity and create better jobs. Nearly one in five and exchange rate flexibility. US dollar fluctuations jobs in Cambodia are in household enterprises, and this significantly impact Cambodia’s trade and commodities share is projected to grow as a consequence of more rapid sectors, which are responsible for most of the country’s urbanization. Thus, it will be necessary for policymakers to jobs. As Cambodia begins to export to a broader range have a much stronger focus on this often-forgotten segment of countries, macroeconomic and fiscal stability will help of the economy to enable these very small enterprises to shield existing jobs from factors related to the US dollar. increase their revenues and, hence, to create better jobs. Information technology can help household enterprises If Cambodia can successfully transition to the next to improve their basic business practices and to access stage of an FDI-driven development strategy, this is broader markets. likely to produce more and better jobs. Achieving this goal will be a complex challenge and will require strong Fourth, support the development of links between leadership by a coordinating body tasked with championing exporting FDI firms and domestic input-supplying the jobs issue and guiding the many actors from different firms. Government policies can provide foreign firms with sectors toward a shared future jobs vision. While the world incentives to source their inputs from local SMEs. At the is changing fast, Cambodia still has an opportunity to make same time, the government can help SMEs to understand this transition and improve the quality of jobs for all of its and meet the quality standards of FDI firms and provide citizens. them with financing to upgrade their production processes. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW xvi TECHNICAL REPORT CHAPTER 1: An Introduction to Jobs in Cambodia Wendy Cunningham, Claire H. Hollweg and Maheshwor Shrestha 1.1 Introduction Rapid and sustained economic growth has transformed Cambodia into a world leader in poverty reduction and elevated it to lower-middle-income status. The country’s GDP grew by an average annual rate of 7.7 percent between 2000 and 2015 (Figure 1.1(a)), which put it among the top 15 economies in the world in terms of economic growth. Its per capita GDP increased fivefold from USD 229 in 1993 to around USD 1,423 in 2017 (based on official numbers). Between 2004 and 2012, poverty incidence under the national poverty line declined from 50.2 percent to 17.7 percent of the population. Notably, the real consumption growth of those in the bottom 40 percent of the income distribution was larger than that of those in the top 60 percent, leading to a decrease in inequality, with the Gini coefficient falling from 32 to 28 between 2008 and 2012 (Figure 1.1(b)).1 Figure 1.1: Growth and Poverty in Cambodia (a) GNI Per Capita and GDP Growth 14% 12% 10% Annual Growth 8% 6% 4% 2% 0% GNI Per Capita -2% GDP 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 (b) Poverty and Inequality Population Below Poverty Line 50% 50 40% 40 Gini Coeffcient 30% 30 20% 20 Poverty 10% 10 Incidence 0% 0 Inequality 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 (WB Estimate) 1 Cambodia reached lower middle-income status in 2015, though it is still one of the poorest countries in the Southeast Asia region. Despite this progress, the vast majority of the families that rose above the poverty line did so by a small margin, which still leaves them at risk of being negatively affected by an adverse shock. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 1 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Cambodia’s economy changed significantly during shifts in economic activity were a reflection of the movement this high-growth period, along with the sectors that of labor from lower- to higher-productivity sectors as the drove the country’s growth. While 48 percent of GDP majority of new jobs were created in the more productive growth was attributable to agricultural production in 1995, sectors due to the big influx of foreign investment as part of this had fallen to 5 percent in 2017 (at constant prices). Cambodia’s integration into the global economy.2 Not only did Meanwhile, the services sector and, more notably, the workers move to jobs where productivity was higher, they also manufacturing sector expanded. Exports of goods and vacated low-productivity jobs, thereby increasing productivity services—mainly garments, rice, and tourism—played a in those sectors. While 77.5 percent of Cambodia’s jobs were big role in Cambodia’s growth, increasing by an average in the agricultural sector in 1995, this had fallen to 40.1 rate of more than 19 percent annually between 1994 and percent by 2017, a similar magnitude of change experienced 2015. The end result was a classic picture of an economy’s by Vietnam during the same period. The employment share of structural transformation out of agricultural production the services sector rose from 26.5 percent to 33.0 percent and into manufacturing and services (Figure 1.2). In 1995, while the share of the industrial sector rose from 15.9 percent agriculture accounted for 44.4 percent of Cambodia’s GDP, to 25.5 percent (Figure 1.3). These shifts were accompanied but by 2017, this contribution had dropped to 19 percent. by an increase in earnings and thus consumption. During the same period, the industrial sector’s share of the economy (including construction and manufacturing) The changing nature of the global economy may increased from 14 percent to 31 percent. disrupt Cambodia’s structural transformation process in the future. The global economy is becoming increasingly Changes in Cambodia’s labor market played a integrated, with more and more countries crowding significant role in its strong economic growth and into global value chains, the number of middle-income poverty reduction during this transitional period. The consumers growing across the globe, and the sophistication Figure 1.2: Sectoral Contribution to GDP, 1994-2017 50% 40% Percent of GDP 30% 20% Services 10% Industry 0% Agriculture 94 95 96 97 98 99 00 01 20 2 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 0 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Source: Data from the World Bank’s World Development Indicators. Figure 1.3: Sectoral Contribution to Jobs, 2004-2015 Percent of Total Employment 60% 50% Agriculture 40% Garments 30% Construction and Real Estate 20% Hotels and 10% Restaurants 0% Other 2004 2007 2008 2009 2010 2011 2012 2013 2014 2015 (incl. trade) Source: World Bank staff estimates based on Cambodia Socioeconomic Surveys (CSES), several years, and ILO employment figures. 2 Although the Cambodian economy began to shift towards services and manufacturing in the 1990s, the spike in agricultural prices in the 2000s resulted in a sizeable return of workers to agriculture. When these prices declined again in the late 2000s, labor shifted back to manufacturing and services. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 2 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA of production processes and outputs increasing due to both rectangles being sector-specific.3 In contrast, this report new knowledge and technology. Therefore, simply moving looks at jobs across all sectors and develops a broad range labor from low-skilled agriculture to low-skilled services and of policy recommendations that align with the Rectangular manufacturing may not be enough to enable Cambodia to Strategy and that collectively are likely to create more, better, reach middle-income status in the next generation. Instead, and more inclusive jobs. Second, while the government’s the Cambodian economy may need different businesses, “National Employment Policy: 2015-2025” provides general different workers, and different jobs to meet these new direction, this report attempts to provide more detailed jobs challenges. policy options based on international evidence and new data analysis that should be considered priorities for improving In this report, we posit that adopting policies to jobs in Cambodia. increase the quality and inclusiveness of jobs will reinforce existing strategies for greater economic The report has five sections. This Introduction provides growth by making it easier for that growth to occur. the context for the rest of the report. It presents a brief history Improving the quality of jobs will drive economic growth of jobs in Cambodia, maps out the current jobs picture, and and will raise the population’s well-being, with quality being discusses emerging mega-trends that may present new defined as a measure of labor productivity, wages and job opportunities to upgrade Cambodia’s jobs. The following benefits, and work conditions. Integrating the full working chapters look at different segments of the jobs picture. Chapter population into the economy in a way that best uses their 2 describes the recent economic growth in Cambodia and its talents will also help Cambodia achieve its economic and impact on job creation and productivity, and identifies potential social goals. In this context, this report explores Cambodia’s macro-level challenges to job creation and productivity going current job structure and recommends some policies aimed forward. Chapter 3 identifies the micro-level constraints at improving the quality and increasing the inclusiveness faced by Cambodia’s firms and household enterprises that of jobs, with the final goal of ensuring greater economic prevent the private sector from maximizing the creation of growth and prosperity. more and better jobs. Chapter 4 looks more closely at how Cambodia’s export sector has supported jobs and incomes to The analytical findings of and policy directions inform how export diversification and upgrading could create presented in this report are intended to contribute to more, better and more inclusive jobs in Cambodia. Chapter the policymaking process in two ways. First, they focus 5 focuses on the workers themselves, including skills gaps on jobs. “The Rectangular Strategy for Growth, Employment, and mismatches. Migration is another important factor in Equity and Efficiency: Building the Foundation Toward Cambodia’s jobs narrative, but no analysis on this subject is Realizing the Cambodia Vision 2050 – Phase IV”, the Royal included in this report due to a lack of relevant data. Within Government of Cambodia’s newest strategy, is comprehensive each chapter, we identify short- and medium-term policy by design and addresses a wide range of policy objectives, reforms that have the potential to increase the number and with most of the strategic sides of the respective strategic improve the quality of Cambodia’s jobs. Box 1.1: Jobs Definitions For the purposes of this report, a “job” will be defined as any income-earning activity that is not illegal. Thus, a job may be as simple as the employment of a full-time worker in a factory who is earning a wage or as complex as the range of income-earning activities carried out by a household in various sectors at different times of the year. A “good” job is a subjective concept that depends on the economic structure, culture, and aspirations of a country, its industry, and its people. In the case of Cambodia at this time, the report will assume that a good job is characterized by high productivity, a wage commensurate with productivity levels, safe work conditions, and job security. An “inclusive” job can be defined as one that is open to any Cambodian citizen regardless of gender, age, or ability, as well as one that integrates different segments of the economy into a single market. 3 The Rectangular Strategy is depicted as four strategic “rectangles” that encompass four aspects of development: (i) human resources, (ii) economic diversification, (iii) the private sector and job development, and (iv) inclusive and sustainable development. Each “side” of each rectangle is a (generally) sector-specific priority within the respective rectangle. Each side of the rectangle is underpinned by several policies. The four rectangles surround four “environment” concepts, and the core of the model is governance reform. The Human Resource Development strategic rectangle is the priority for Phase IV, which is currently under implementation. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 3 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA 1.2 Cambodia’s Labor Market in private sector investors (World Bank 2017; Kelsall and Heng 2014). the Past Trade agreements and market access contributed Over the past two decades, Cambodia has been able to the entry of garment manufacturers into the to achieve political stability and economic progress. Cambodian economy. The United States granted Cambodia has succeeded in transforming itself from a war- Cambodia most favored nation (MFN) status in 1996, and the torn country to a peaceful one, and from a centrally planned following year the country received favorable access to the and ineffective economy to one that is regionally and US market through the Generalized System of Preferences globally linked and rapidly growing (World Bank 2017). The (GSP). In 2001, Cambodia’s received preferential access 1991 Paris Peace Agreement and the democratic elections to the European Union market through the enactment of of 1993 signaled a new phase of state-building. Against the European Union’s Everything-but-Arms (EBA) scheme. the backdrop of strong growth, poverty alleviation, and Cambodia joined the Association of Southeast Asian Nations structural transformation, some big shifts in the economy (ASEAN) in 1999 and became a member of the World Trade have defined the evolution of Cambodia’s labor market and Organization (WTO) in 2004. are reflected in its jobs picture today. Becoming strongly integrated into these global markets Cambodia grew from 1994-2004 as an open, was a key feature of Cambodia’s development and globally integrated economy jobs story during this period. Between 1994 and 2004, its exports grew from less than USD 1 billion to USD 5 billion Cambodia maintains a strong commitment to (Figure 1.4). Cambodia’s rapid economic development was economic openness. The government has adopted a also fueled by foreign inflows. Net Foreign Direct Investment series of reforms aimed at making Cambodia a highly open, (FDI) inflows remained steady during this period, averaging market-oriented economy and at fostering private sector at around USD 200 million per year. FDI to Cambodia came development. In 1994, it passed a law that allowed 100 almost exclusively from Asia, and increasingly primarily from percent foreign-owned investment and provided guarantees China. In response, labor rapidly flowed from agriculture against nationalization and price regulation. It has since to the emerging manufacturing and services sectors. created a one-stop service office for investors through the Approximately 80 percent of all jobs in Cambodia were in Council for the Development of Cambodia (CDC), which has agriculture in 1994, whereas this share had declined to 60 also helped to attract and channel foreign investment. This percent a decade later. The sector composition of FDI is was accompanied by the enactment of critical regulations changing rapidly, shifting from the tradeable sectors such related to land, banking, bankruptcy, and company law as agriculture and industry to tourism, construction and that made the economic framework more attractive to real estate. Figure 1.4: Exports of Goods and Services and Net FDI Inflows, 1991-2016 14 12 3 Exports (Billion USD) 10 FDI (Billion USD) 8 2 6 4 1 2 Exports 0 0 FDI 19 1 19 2 19 3 19 4 19 5 19 7 19 8 20 0 20 3 20 4 20 5 20 6 20 7 20 8 20 9 20 0 20 1 20 2 20 3 20 4 15 19 6 20 9 20 1 20 2 9 9 9 9 9 9 9 9 9 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 19 Source: Authors’ calculations using data from the World Bank’s WDI. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 4 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Agricultural employment rose during the crops (Figure 1.5). Declining prices coupled with unfavorable 2004-2011 commodity price boom weather events in 2013 and 2015 resulted in a slowdown and stagnation in agricultural GDP. The amount of cultivated During the global commodity price boom (2004- land also decreased after 2012. By 2014, agriculture’s share 2011), agriculture again became the main source of of GDP had fallen to pre-2004 levels (Figure 1.2). employment creation in Cambodia and made a strong contribution to poverty reduction. Rice and rubber The decline in agriculture led to a rural exodus. With prices peaked in 2008 and 2011, respectively, and have the collapse of the agricultural sector, many Cambodians only recently returned to their pre-2008 levels. Between left the countryside to migrate to urban areas or to other 2007 and 2011, more than 1 million net jobs were created countries. Estimates from the Cambodia Socio-Economic in Cambodia, with 533,000 of them being generated in Survey (CSES)4 indicate that domestic remittances sent agriculture as the sector continued to enjoy high commodity to recipient households increased significantly, reflected prices. The expansion in the amount of land under cultivation in the share of rural households who received domestic and high international food prices benefited both farmers remittances rising from 11 percent in 2004 to 33 percent and agriculture workers and lifted 3.3 million Cambodians in 2014. International migration also increased, with about out of poverty (World Bank 2017). 7 percent of rural households having a member who was an international immigrant in 2014 (World Bank 2017). During this period, agriculture’s declining significance International remittances increased when agriculture in Cambodia’s economy was temporarily reversed. By prices declined and vice versa. Migration and remittances 2004, agriculture’s share of GDP had fallen to 29 percent, contributed to the continued reduction in poverty during the but higher prices and increased output led this share to rise agriculture slowdown. to 35 percent by 2011 (Figure 1.2). FDI inflows picked up during the commodity price boom and have remained stable and high since 2011 Structural transformation has resumed from even after these prices fell (Figure 1.4). Despite not 2011 to the present being a commodity exporter, Cambodia has been among those countries with the highest rates of FDI, reaching an After 2011, commodity prices dropped, and agriculture average of 7.9 percent of GDP annually between 2005 growth slowed down again. In 2012, rice prices started and 2015. Large aid inflows (which amounted to to decline and rubber prices collapsed, having an especially 5 percent of GNI in 2014), along with tourism and export negative impact on farmers who had diversified into industrial receipts, have led to the dollarization of the economy. Figure 1.5: Trends in Rubber and Rice Prices, and Remittances 600 3.0% Remittances, Percent of GDP 500 2.5% Prices (2003=100) 400 2.0% 300 1.5% Rice Price 200 1.0% (Thai) 100 0.5% Rubber Price (Singapore) 0 0.0% Remittances 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 (%) Source: World Bank commodity prices database, WDI. 4 The CSES has been conducted annually since 2007 by the National Institute of Statistics (NIS) in the Ministry of Planning in Cambodia. The survey collects data on a wide array of indicators related to household demographics, consumption, education, and economic activities. This survey is the primary poverty monitoring and planning tool used by the Royal Government of Cambodia. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 5 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Cambodia’s continued progress in opening up to migration is through informal networks, with half of migrants global markets supported the resurgence of structural following earlier migrants from their own communities. transformation. Cambodia has introduced cross-border This can limit migrant job options and result in poor job trade facilitation improvements in recent years, which matching. While migrants, particularly women, sent home resulted in faster export and import procedures. Cambodia the equivalent of 3 percent of GDP in remittances in 2015 was recognized as having the fewest trade restrictions in (Roth and Tiberti 2017), it is not clear whether these funds services of all Southeast Asian countries (Ahsan et al. 2015). are used to invest in the local economy or whether they Cambodia now ranks in the top 5 percent of economies simply fund consumption. Furthermore, return migration in the world in terms of merchandise trade (imports plus rates and the employment trends of return migrants are exports) with 144 percent of GDP in 2015. largely unknown. As a result, employment growth in recent years has been led by a rise in manufacturing, notably in the 1.3 Cambodia’s Jobs Today garments and construction sectors. Manufacturing accounted for 11 percent of employment in 2009, a share Today, there are more than 8 million jobs in Cambodia. that rose to 17 percent by 2014 due to increases in the This is a high number of jobs given the size of Cambodia’s share of garment sector employment. Similarly, the share adult population. Around 80 percent of Cambodian adults of the construction sector in total employment increased aged 15 or higher are working.5 In comparison, the global from 3.5 percent to 6.8 percent during the same period, employment rate is 62.3 percent, employment rates in the while employment in service sectors increased by almost East Asia Pacific (EAP) region are 62.5 percent, and the 4 percent. Although the share of employment in tourism expected employment rate for a country with an income increased prior to 2009, it increased only by 0.6 percent level similar to that of Cambodia is 65 percent. In fact, the between 2009 and 2014. only EAP countries with similar levels of employment rates are Vietnam and Lao PDR (Figure 1.7). Though internal migration enabled this structural transformation, international emigration from In this report, we consider two aspects of Cambodian Cambodia is exacerbating the country’s skills jobs today: job quality and inclusiveness. Measuring shortage. Nearly 90 percent of internal migrants surveyed job quality can be challenging since different firm owners in 2011 had moved for jobs (MoP 2012). Nearly 1 million and workers value different aspects of a job, and survey Cambodian citizens are living abroad (Testaverde et al. instruments often do not measure many of these aspects. 2017). Both internal and international migrants tend to For example, does a worker prefer the exibility and freedom have completed more education than non-migrants. Recent of owning his own business or the stability of holding a wage estimates suggest that 15 percent of Cambodian tertiary job? Does an employer only value labor productivity or will graduates have migrated to OECD countries, contributing she place a premium on a stable labor force and good labor a net drain of Cambodia’s human capital (Figure 1.6). Most relations? We assumed that “better” jobs are characterized Figure 1.6: Net Effect of High-Skilled Emigration on the Share of High-skilled Workers in ASEAN Source Countries 1.5% Change in Share of High-Skilled Population Due to Emigration 1.0% 0.5% 0% -0.5% -1.0% -1.5% -2.0% Thailand Indonesia Philippines Malaysia Cambodia Vietnam Lao POR Source: Beine, Docquier, and Rapoport 2008. 5 Labor force participation (LFP) is the share of individuals age 15-64 who worked for at least one hour in the previous seven days or are unemployed. Since unemployment rates are so low in Cambodia, LFP rates and employment rates are nearly identical. See Annex 1.1 for the data sources and definitions of key employment categories. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 6 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA as having high labor productivity, high wages, and good Job Type: Most jobs are in traditional forms work conditions. We used four rough proxies for quality: of work • Type of job, in terms of ownership and nature of work, where we assumed being an employee in a registered Two of every three jobs in Cambodia are in traditional firm is a “better” job than owning one’s own micro- types of work. More than 3 million workers report their primary enterprise, which we referred to as a “traditional” job. job as being on a family farm, while another 1.4 million own a non-farm household enterprise. Non-wage farming is the • Occupation, where we used higher-skilled occupations primary job (Figure 1.8): 48 percent of rural working adults (such as managers, clerical workers, and professionals) compared to 7 percent of urban working adults. Household as a proxy for quality jobs as compared to low-skilled enterprise ownership is the primary job for 31 percent of urban occupations. working adults compared to 13 percent of rural working adults • Wages where a high-wage job is of a higher quality than (Figure 1.9(a)). While men and women have a similar incidence a low-wage job. of working on a family farm ( 37 and 40 percent of working adults), a higher percentage of women are household enterprise • Labor productivity where we assumed that higher labor owners than men (21 percent of all working adults compared to productivity is associated with better jobs. 14 percent for men). Another 7.5 percent of workers are wage Assessing the inclusiveness of jobs is also a challenge employees on farms, mostly in rural areas. It is notable that the given the lack of relevant available data. However, we were shares of farm and non-farm household enterprise jobs have been able to explore how job quality differs by gender—which is steadily declining since 2009, but that farm wage employment particularly fascinating given the structure of Cambodia’s has increased slightly during the same period (Figure 1.9(b)). economy—as well as by age. One in three jobs are wage-paying jobs in the private sector, which has expanded rapidly in recent years. In 2014, 2.6 million workers had private sector wage-paying 1.4 Job Quality: A Review of Job jobs, up from 1.1 million in 2009 (Figure 1.8 and Annex Types, Occupations, Wages, and Labor Table 1.2). This increase has occurred in rural areas (12 Productivity percent to 26 percent) as well as urban areas (20 to 38 percent), although private wage jobs are more common in We reviewed the quality of jobs in Cambodia as reflected in the urban areas than in rural areas. They are also more common four proxies: job type, occupation, wages, and labor productivity. among men than among women (Figure 1.9). Figure 1.7: Labor Force Participation Rates and GDP per capita, global, 2016 Cambodian employment rates are higher than those of comparator countries 100% Labor Force Participation Rate (age 15+) Cambodia Lao PDR 80% Vietnam China New Zealand Macao Thailand Myanmar Indonesia Australia Singapore Nauru Malaysia Bahrain 60% Kiribati Philippines South Korea Hong Kong Vanuatu Fiji Mongolia Japan Papua New Guinea 40% Marshall Islands All countries (2015) Samoa EAP Comparators Trend 20% Cambodia trajectory, 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 2007−2015 GDP per capita, PPP (log scale, base 10) Source: Author’s estimates using CSES 2007-2015 and the WDI online database. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 7 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Figure 1.8: Distribution of Job Types in Cambodia, 2014 Working 8,067k; 100.0% Owner/Self employment Wage Employment 4,481k; 55.5% 3,586k; 44.5% Farm HH Enterprise Farm Public Private Foreign 3,081k; 38.2% 1,400k; 17.4% 602k; 7.5% 398k; 4.9% 1,388k; 17.2% 1,199k; 14.9% Source: Author’s estimates using CSES 2014. The working age population estimate is from UNPOP. Note: See Annex 1.1 for definitions of the categories. Populations and percentages of jobs are presented next to the labels. The size of the marker represents the size of the population in that category. Nearly half of all wage-paying jobs in the private considered poor quality jobs (Figure 1.9). These jobs sector are in foreign firms. As of 2014, the domestic are mainly prevalent in rural areas (62 percent), while these private sector employed 1.4 million workers, while foreign- occupations constitute only a small share of jobs in urban owned firms employed 1.2 million workers (Figure 1.8). areas (15 percent). Women are slightly more engaged in The share of the working-age population involved in the agricultural occupations than men (39 percent of women domestic private sector increased from 8 percent in 2009 versus 36 percent of men), but men dominate in elementary to 14 percent in 2014.6 Wage employment in the foreign- occupations (11 percent of women versus 16 percent of owned private sector grew even more spectacularly from 5 men). Recent trends show that these occupations have percent of all jobs in 2009 to 15 percent in 2014. A higher been declining over time in all locations for both genders share of female workers than male workers hold wage- (Annex Figure 1.1). paying jobs in foreign-owned firms (19 percent versus 11 percent), while a higher share of male workers hold jobs The number of higher quality jobs (semi-skilled in domestically-owned firms (11 percent versus 23 percent). occupations)—such as machine operators, garment sector workers, craftsmen and tradesmen, and sales The quality of wage jobs, as measured by total monthly and service workers—has increased in recent years wage earnings, is higher in foreign-owned firms than and now constitute 40 percent of all Cambodian jobs. in domestically-owned firms. Only 4 percent of wage- These jobs are more prevalent in urban areas, with sales earners in foreign-owned firms earn less than the 2014 and services accounting for more than 30 percent of urban minimum wage level compared with 22 percent of wage- jobs. These jobs are also more prevalent among women earners in the private sector and 48 percent of farm wage- than men, with more women being employed in sales and earners. Earnings in both the domestic and foreign-owned services and in the garment sector. This is perhaps not sectors vary widely, with the middle 50 percent of domestic surprising given the highly feminized labor force generally wage-earners making between PPP$ 250 and PPP$ 470 employed in garment factories and in the restaurant compared to the PPP$ 300 to PPP$ 438 earned by the and hospitality sector globally. In fact, garment-related middle 50 percent of wage workers in foreign-owned firms. occupations in Cambodia have had the largest growth from about 5 percent in 2009 to 11 percent in 2014 (Annex Figure 1.1). Other crafts and related trades also grew from Occupation: Most jobs are low quality 7 percent to 9.6 percent over the same period as a result of (proxied by low-skilled occupations), but growth in construction sectors. some workers are moving into slightly better (semi-skilled) jobs Jobs in Cambodia are being upgraded in two ways. First, new jobs in semi-skilled occupations are being Half of Cambodia’s jobs are in agriculture or created as a result of new investments in various elementary occupations (51 percent), and can be sectors. When we analyzed finer occupation categories (up 6 There was not enough data to enable us to identify the number of private sector wage workers who are covered by labor legislations and those who are not. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 8 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Figure 1.9: Distribution of Jobs by Type (a) By Worker Type, 2014 A large share of the labor force is involved in traditional forms of work 100% Foreign wage 80% Percent of Jobs Domestic wage 60% Public wage 40% HH enterprise 20% Non−wage Farming 0% Wage Farming All Female Male Rural Urban (b) By Year, 2007-2015 Wage employment in private sector has boomed in recent years 100% Foreign wage 80% Percent of Jobs Domestic wage 60% Public wage 40% HH enterprise 20% Non−wage Farming 0% Wage Farming 2007 2008 2009 2010 2011 2012 2013 2014 2015 Source: Author’s estimates using CSES 2014. Note: Employment type refers to primary employment. The sample is restricted to the population aged 15-64 who are currently working. Foreign-owned firms were not identifiable in 2007 and 2008. to a three-digit level), we found that garments and related agriculture rising from 0.5 million in 2009 to 0.6 million in trades is the fastest growing occupation category (Table 2014. This shows some transition away from subsistence 1.1).7 The share of employment in this occupation category work and towards wage employment within agriculture increased by 1.1 percent per year between 2007 and 2015. (Annex Table 1.2 and Annex Table 1.3).8 Similarly, a move Other trades, such as building frames for construction, from sales in the streets to sales in shops shows a transition wood treating, and cabinet making have also increased in towards higher value-added salesmanship. numbers in recent years to cater to the growing construction sector. Service occupations such as shopkeepers, waiters, The best jobs—clerical, managerial, professional, and and bartenders that serve a growing urban population with technical occupations—are in short supply and are growing incomes are also expanding quickly, although some expanding only slowly. Only 9 percent of all jobs are in started at a very low base. these occupations, employing higher shares of men than women (11 percent versus 7 percent), and of urban dwellers Second, there is a shift towards more productive rather than rural dwellers (24 percent versus 5 percent). These variations of some of the lowest-quality subsistence occupations have expanded to some extent as they represented occupations. As seen in Table 1.1, although subsistence only 6 percent of jobs in 2009, but this rise was driven more farming is the fastest declining occupation, the employment by increases in clerical jobs (which contributed 2 percent of shares of laborers in agriculture, forestry, and fisheries this growth) than by increases in managerial, technical, and have risen during this period, with wage employment in professional occupations (contributing 1 percent of growth). 7 Because of sample size issues, we dropped any occupation category with a less than 0.5 percent employment share during the 2007-2015 period from our analysis. While our analysis using finer occupation categories tends to be noisy because of sample size issues, it is still insightful on the specific types of occupations that are growing and shrinking. Also, comparing these findings with changes in the broader categories (as in Annex Figure 1.1) confirms our overall results. 8 The shift is not completely driven by land ownership. Even within households that own farming land, wage work in farming is increasing. In fact, in rural areas, landless households were slightly less likely in recent years to be involved in wage work in farming than they were in the past. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 9 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Figure 1.10: Distribution of Jobs by Occupation, 2014 Agricultural and elementary occupations are predominant, particularly in rural areas 100% Agriculture Percent of Jobs by Occupation 80% Elementary Machine operators 60% Garments Craft and trades 40% Sales and services Clerical 20% Manager, professionals, 0% technical All Female Male Rural Urban Source: Author’s estimates using CSES 2014. Note: The sample was restricted to the working-age population aged 15-64 who had worked in the previous seven days before the survey. The occupations are those that the respondents indicated were their main job. Table 1.1: Fastest Growing and Declining Occupations, 2007-2015 Employment Share Change per year 2009 2014 Fastest growing occupations Garment and related trades workers 4.41% 10.18% +1.14% Building frame and related trades workers 0.56% 2.77% +0.69% General office clerks 0.85% 2.58% +0.29% Agricultural, forestry, and shery laborers 6.29% 6.93% +0.16% Shop salespeople 5.03% 6.05% +0.14% Waiters and bartenders 0.26% 0.83% +0.12% Motodup and tuktuk drivers 1.16% 1.50% +0.07% Transport and storage laborers 0.82% 0.82% +0.05% Wood treaters, cabinet-makers, and related trades workers 0.72% 0.93% +0.05% Forestry and related workers 0.37% 0.67% +0.05% Fastest declining occupations Subsistence livestock farmers 10.22% 5.38% -1.05% Subsistence crop farmers 25.69% 19.85% -1.01% Firewood and water collectors 4.67% 0.59% -0.88% Subsistence fishers, hunters, trappers, and gatherers 2.51% 1.91% -0.25% Street and market salespeople 4.87% 3.65% -0.16% Market gardeners and crop growers 7.55% 6.88% -0.11% Food processing and related trades workers 1.60% 1.39% -0.09% Fishery workers, hunters, and trappers 1.50% 1.44% -0.06% Animal producers 3.58% 3.47% -0.04% Street food snack sellers 0.77% 0.69% -0.04% Source: Author’s estimates using CSES 2009-2015. Note: Average growth was estimated from a regression of shares (for each year) of occupations that had an employment share of at least 0.5 percent between 2007 and 2015. Each occupation cell is a three-digit occupation code. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 10 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Wages have been increasing in recent years distribution (p=10 and p=25 in Figure 1.11) grew faster than median wages from 2009-2014. At the same time, Real wages paid to those in wage-earning jobs were on wages at the top of the wage distribution only grew modestly average above the international poverty line by 2014.9 (p=75 and p=90 in Figure 1.10). Similarly, the gap between Average real monthly wage earnings were PPP$ 393 in 2014, the (composition-adjusted) hourly real wages of workers in a 54 percent increase since 2009. By 2014, the share of the agricultural and non-agricultural sectors also narrowed jobs paying below the international extreme poverty line of since a modest increase in agricultural wages was greater USD 2.00 per day was 1.6 percent, and the share below the than the growth in non-agricultural wages. This was also the dependency adjusted income poverty line of USD 4.22 per case for jobs held by the least-skilled workers (Figure 1.12). day was 7.7 percent. This was down from 9 percent and 25 The wages for jobs held by workers with only a primary percent respectively in the year 2009 (Figure 1.11). education increased faster than those for jobs held by Wages in low-quality jobs increased the fastest in workers with a secondary school education. Nonetheless, the period 2009-2014. Wages at the bottom of the wage the education wage gap remains substantial. Figure 1.11: Evolution of Wages, by wage quintile, and Share of Wages Below the Poverty Line, 2007-2015 Inequality reduced growth in monthly real wage earnings between 2007 and 2015 25% 3.0 Percent Earning Below $4.22 Share Below Poverty Line Wage index, 2009=1 20% 2.5 p=10 p=25 15% 2.0 Median Mean 10% 1.5 p=75 5% 1.0 p=90 2007 2008 2009 2010 2011 2012 2013 2014 2015 Source: Author’s estimates using CSES 2007-2015. Note: The “p”s in the legend refer to the percentiles of the wage distribution for each year. Figure 1.12: Distribution of Wages in Agriculture and Non-Agricultural Jobs, by worker education The wage gap between agriculture and non-agriculture activities has narrowed in recent years (a) Primary Education (b) Secondary Education 8.0 8.0 Log Real Hourly Wage (Riel) Log Real Hourly Wage (Riel) 7.8 7.8 7.6 7.6 7.4 7.4 7.2 7.2 7.0 7.0 Ariculture 6.8 6.8 Non-Agriculture 2004 2006 2008 2010 2012 2014 2004 2006 2008 2010 2012 2014 Source: World Bank staff calculations using CSES 2004-2014. 9 Monthly wages are recorded in KHR in the CSES, so we converted them to 2014 using the consumer price index and to international PPP$ using the conversion rate in 2014 of 1597.30 KHR per PPP$. The USD 4.22 per day threshold that we used is USD 2 per day multiplied by the ratio of the total population to the share of individuals that are employers, employees, or own-account workers. We used the dependency ratio to convert the per-earner wage to the per capita poverty line. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 11 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA The increase in the wages for the lowest-paid wage distribution gained only 1 to 2 percent from the jobs was not likely due to the minimum wage. The minimum wage hike. Furthermore, the hours of low-paid minimum wage legislation was put into effect in 1997, workers were cut when the minimum wage was increased, mandating a wage floor specifically for the garment and which meant that their total monthly take-home pay stayed footwear industries. It was initially set at USD 40 per the same. month and was increased eight times between 2007 and 2016, reaching USD 153 by September 2016. Although Although the minimum wage is only mandated for the nominal minimum wage increased by over 380 percent the garment and footwear industries, it also affected over 19 years, the real value of the minimum wage was employment and wages in jobs in other sectors. below its 1997 real value until January 2015. While the Recent minimum wage hikes reduced wage earnings for 20 to 30 percent increase in the minimum wage raised those outside the garment sector by about 1.4 percent. At the average wages in the garment sector by 2.8 percent, household level, total earnings did not increase in response most of these gains were concentrated among higher-paid to the higher mandated minimum wage. Instead, workers workers, who experienced wage increases in excess of 4 adjusted their working hours and sectors of employment to percent whereas those in the bottom half of the garment maintain the same level of earnings.10 Figure 1.13: Impact of Minimum Wage Increase (a) Impact on Labor Participation (b) Impact on Hourly Wages All Activities Garment Sector Garment Sector .02 .01 .20 Impact, Percentage Point Impact, Percent of Wage .01 0 .15 0 -0.1 .10 -0.1 -0.2 .05 -0.2 -0.3 0 -0.3 -0.4 -.05 -8 -6 -4 -2 0 2 4 6 Self Employment Other Sectors .06 .05 Impact Percentage Point .04 Impact, Percent of Wage 0 .02 0 -.05 -0.2 -.10 -0.4 -0.6 -.15 -8 -6 -4 -2 0 2 4 6 -8 -6 -4 -2 0 2 4 6 Quarters from minimum wage hike Quarters from minimum wage hike Point estimates 95% Con dence Interval Source: Adapted from Shrestha (2018). Note: The plotted estimates are for each quarter with the quarter immediately preceding the change as the omitted category. The orange lines show the average for one to eight quarters before the change and for two to eight quarters after the change. The left plot shows the impact of the minimum wage increase on participation rates in all economic activities, on garment work, and on self-employment activities. The right plot shows its impact on hourly wages in the garment and non-garment wage sectors. 10 See Annex 2.1 for a detailed summary of the methodology underlying these findings and a deeper discussion of the findings. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 12 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA The increases in the minimum wage did not seem the bottom of the garment sector wage distribution, this to impact labor force participation rates in the long could reduce employment in the sector. This effect has run in either the garment or non-garment sectors. been found by cross-country studies comparing economies However, there have been some temporary patterns around with low minimum wages (relative to the market wage) the time of the minimum wage hikes. Employment in the and those with high minimum wages.11 Second, minimum garment sector increases in the quarters preceding the hike wage increases may outstrip labor productivity growth. but falls in the two quarters following the wage increase. The increased productivity that has been observed in the The availability of temporary workers is what likely allows Cambodian garment sector following minimum wage hikes firms to make these short-lived adjustments. After three is likely to have been the result of low-effort adjustments quarters have passed, employment rates go back to their that increasesd in productivity. Firm owners suggest that pre-change levels. The empirical strategy used for the study these productivity increases were the result of employers factors out all trends in employment patterns to focus just on setting steep performance targets for garment workers. the effects of minimum wage hikes. As such, such aspects Without significant changes to the production technology, as other sectors, general benefit level increases, or worker there is a limit to how many future increases in labor flow trends are unlikely to explain the results. However, it is productivity will be achievable, so any future minimum possible that because garment firms employ many migrant wage hikes could then lead to an increase in wages with no and temporary workers, it may be easy for them to make the increase in productivity. Such wage growth will lower output small adjustments in their labor demand at just about the and reduce labor in the garment sector and make Cambodia time of minimum wage increases. However, the data cannot less attractive to foreign firms. credibly distinguish between permanent and temporary workers, or between migrant and non-migrant workers, and it is therefore unable to test if the impacts are different Labor productivity increases are not along these dimensions. re ected in wages Future minimum wage increases may have a bigger Long-term labor productivity trends (output per impact than these trends from past years might worker) fell in the fastest-expanding occupations suggest. First, during the period of our analysis, the but increased in others, thus having a mixed effect minimum wage was lower than the market wage. Even on the overall quality of jobs (Figure 1.14). From though real per capita income increased drastically between the perspective of employers, jobs with higher labor 1997 and 2015 in Cambodia, the level of the real minimum productivity are better jobs, while from the perspective wage remained close to the 1997 level throughout that of workers, higher labor productivity can push wages up, whole period. Consequently, an overwhelming majority of thus improving the quality of the jobs. Labor productivity workers in the garment sector were already earning a wage in general in Cambodia has been following a downward that was higher than the new minimum wage. If the minimum trend since 2004, although this decline became more wage rises faster than wages (and productivity) increase at gradual in 2010 in the fastest-growing wage-paying Figure 1.14: Productivity Trends, by Sector (a) Agriculture and industry (b) Services 16 16 Trillion Riel (Year 2000 constant) Trillion Riel (Year 2000 constant) 14 14 12 12 10 10 Hotel and 8 8 Restaurant 6 Agriculture 6 Trade Garments 4 4 Read Estate and Construction Professional 2 Other 2 Services 0 Industry 0 Other Services 2004 2008 2010 2012 2014 2004 2008 2010 2012 2014 Source: World Bank staff calculations using CSES 2004-2014. 11 See Del Carpio and Pabón (2014) for a review of experiences in East Asia and Cunningham (2007) for a comparison of Latin American countries. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 13 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA jobs—garments, construction, and real estate. However, education, and only 3.8 percent has had a college or higher productivity has slightly increased in agriculture, which level of education. These educational attainment rates are is also slowly expanding, and in “other” industries and lower than those of other countries with comparable income services. The productivity gains in agriculture reflect, at levels and of other countries in the EAP region (Figure 1.16). least in part, higher yields due to the expansion of land Although schooling attainment for the younger cohorts has under cultivation, gains in crop yields due to mechanization, recently increased, the education level of the older cohorts higher-quality seeds, greater access to irrigation, and of the population remains low. The low level of education diversification into higher value-added crops including the appears to prevent individuals from accessing higher quality expansion of rubber, cashews, pepper, and other tree crops jobs. For instance, individuals with low education outcomes (Eliste and Zorya 2015). Nevertheless, there has been no do not tend to hold clerical, managerial, or professional improvement in job quality related to productivity overall. occupations in Cambodia. The combined effect of higher wages and limited Today’s education system is not providing students productivity growth indicates that job quality has with the fundamental skills that they will need to be declined as measured by the unit cost of labor. The competitive in tomorrow’s jobs. While Cambodia has unit cost of labor measures labor-related costs (mostly made significant progress in terms of increasing access to wages but also benefits) for each unit of output. From education, by 2014, only slightly more than 80 percent of an employer’s perspective, a high unit cost per worker is primary-school-aged children were attending school (World a drain on profits, and thus does not constitute a “good” Vision 2017). The percentage is even lower at the secondary job. Cambodia’s garment sector has the highest unit cost school level, where only Myanmar, Papua New Guinea, and of labor among competitor garment-producing countries Lao PDR have similar or lower levels of high school completion (Figure 1.15). rates than Cambodia, and only Papua New Guinea has lower college completion rates. However, attending school is not synonymous with acquiring fundamental skills: language, The low skill-level of the labor force may mathematics, and basic reasoning. Between 2010 and prevent Cambodia’s movement to higher 2012, 17 to 35 percent of second graders could identify only quality jobs a single letter, and 33 to 48 percent could not read a single word in a simple paragraph (USAID 2015).12 Consistent with Productivity is partly a function of the education this, in 2016, about one-third of sixth graders had “below level of the workforce, and, despite some progress basic” knowledge of the Khmer language and over half had in the past decades, the level of education of the “below basic” knowledge of mathematics (MoEYS 2017). Cambodian workforce is quite low. Only 13.5 percent of The quality of schooling is much worse in rural areas than the working age population has completed upper secondary in urban areas. Figure 1.15: Average Unit Labor Costs of Apparel Firms Across Comparator Countries 80% 70% 60% Percent of Value Added 50% 40% 30% 20% 10% 0% Cambodia Bangladesh Ethiopia India Myanmar Nicaragua Pakistan Philippines Turkey Vietnam Source: World Bank Enterprise Surveys. 12 International evidence has shown that the reading skills of poor readers deteriorate with age as opposed to improving (Stanovich 1986). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 14 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Figure 1.16: Global Secondary and Tertiary School Completion Rates and GDP Cambodia still lags behind comparator countries in educational attainment (a) Secondary School Completion Rate 70% 60% Percent who Completed Secondary 50% 40% 30% 20% 10% 0% Non−EAP 6 7 8 9 10 11 12 Log (GDP per capita, PPP) EAP Cambodia (b) Tertiary School Completion Rate Locally-linear relationship 30% Percent with College or Higher 20% 10% 0% 6 7 8 9 10 11 12 Log (GDP per capita, PPP) Source: Per capita GDP from the WDI online. Completed education data for 2010 from Barro and Lee (2013). Cambodia data from the 2009 CSES. Note: The size of the bubble denotes the population in 2010. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 15 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA 1.5 Inclusive Jobs: Women and Older Nearly 23 percent of Cambodian women worked in the manufacturing sector in 2014 compared with 21 percent in Workers Have Lower-Quality Jobs Thailand and 19 percent in Vietnam. In contrast, an equal or lower share of Cambodian men worked in manufacturing While there was a general lack of data regarding the compared with men in Thailand and Vietnam. Between inclusiveness of jobs in Cambodia, we were able to explore how job quality differs by gender and age. 2009 and 2014, the share of female employment in manufacturing increased by 7 percent, almost completely due to the growth in garment jobs. These jobs increased Gender: Women hold better jobs by some from 8.5 percent of female employment in 2009 to 17.3 measures, but fare worse than men on percent in 2014. The share of women working in the in the many others hospitality sector increased slightly by 0.4 percent, and in other service sectors their share increased by 2.6 percent Women hold a high share of Cambodia’s jobs. About 73 during the same period. These changes were slightly smaller percent of women were working in 2015 compared with 80 than those for men (0.9 and 3.5 percent), arguably due to percent of men. These rates are higher than women’s LFP rates the large increase in the share of women working in garment globally (52.6 percent), in the EAP region (54.5 percent), and manufacturing. However, growth in the construction sector in countries with a similar level of development (56.7 percent) has disproportionately benefitted men. (Figure 1.17). Vietnam and Lao PDR are the only EAP countries with similarly high levels of female LFP as Cambodia. Women have access to better jobs in the sectors where they dominate. Women fill 65 percent of jobs in The structure of Cambodia’s economy provides more the manufacturing sector—a sector where the majority of good job opportunities to women than in comparator wage-paying jobs are in garments. They also hold more countries. Cambodia has a disproportionately female than 65 percent of the managerial and professional jobs manufacturing sector that is driven by two engines of in this sector (Figure 1.18(a)). A similar situation exists in economic growth: garment production and tourism. the services and trade sectors where women are over- Figure 1.17: Women’s Labor Force Participation and Per Capita GDP in Cambodia and Comparator Countries 100% Cambodia Labor Force Participation Rate (age 15+) 80% Lao PDR Vietnam China New Zealand Macao Thailand Australia Singapore 60% Mongolia Malaysia Bahrain Vanuatu Myanmar South Korea Indonesia Hong Kong Kiribati Philippines Nauru Japan Papua New Guinea Fiji 40% Marshall Islands Samoa 20% All countries (2015) EAP Comparators Trend 0% Cambodia trajectory, 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 2007−2015 GDP per capita, PPP (log scale, base 10) Source: Author’s estimates using CSES 2007-2015 and the WDI online database. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 16 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA represented among professionals and among associate transport jobs), though they are over-represented among professional and semi-skilled jobs. However, the services clerical and sales and service jobs (Figure 1.18(b)). and trade sectors display a “glass ceiling” since, unlike in the garment sector, women are under-represented among The labor market rewards highly educated women the top jobs (managers). In contrast, only 10 percent more than equally well-educated men. Female of construction jobs are held by women, and women secondary-school graduates earn 20 percent more for an are clustered into the lowest quality jobs in elementary additional year of education than the workforce average of occupations (usually unskilled agriculture, construction, and 9 percent (Figure 1.19). Figure 1.18: Vertical Gender Segregation within Industries (a) Manufacturing (b) Construction Managers Professionals Professionals and Associates and Associates Clerical Clerical Services and Services and Sales Sales Crafts and Crafts and Trade Trade Machine Machine Operators Operators Agricultural Agricultural and Elementary and Elementary 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Female workers Male workers Industry Average Female Employment Rate Source: Guvalyugova (2018) based on the CSES 2016. Figure 1.19: Returns to Education by Gender Female Male 100 (relative to receiving no schooling) Returns to Year of Schooling 80 60 40 Point 20 estimate 0 95% Con dence -20 Interval 1 3 6 9 12 College+ 1 3 6 9 12 College+ Years of schooling Source: Author’s estimates using CSES 2014. Note: The sample was restricted to the working-age population aged 15-64 who had worked in wage employment for at least 20 hours in the previous seven days before the survey. The vertical dashed lines indicate completion of the primary, lower-secondary, upper secondary, and college levels. The dependent variable is the logarithm of hourly wages. The regressions control for individual covariates (gender, age, marital status, and household size and composition) and location (rural-urban x region). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 17 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Women are performing less well in the labor dollar of wages earned by a male worker (Cunningham and market according to other measures and are Pimhidzai 2018). The gap is due to a number of factors, over-represented in traditional forms of work. Non- including women’s lower education levels, 13 household- farm household enterprises constitute a larger share related gender roles, 14 their industries of employment, and of employment for women than for men, and this may their occupations (Figure 1.20). However, even equalizing be because they find the flexibility involved in running a across these factors still leaves an unexplained gap of household enterprise particularly attractive. However, 7 cents on the dollar. It is notable that gender wage gaps women and men are equally likely to own family farms have fallen faster in female dominated sectors than in (40 percent). male dominated sectors. While the gender wage gap was more or less the same in all sectors in 2011, with On average, female wage workers earns only women earning 86 cents for every dollar paid to a man, 86 cents for every dollar earned by male workers. by 2016 the gender wage disparity had improved to 97 The Cambodian average gender wage gap is similar to cents on the dollar in the manufacturing sector, but had that in other East Asian economies; for example, in 2014, deteriorated to 71 cents on the dollar in construction Vietnamese women earned, on average, 88 cents for every (Gavalyugova 2018). Figure 1.20: Gender Gap in Hourly Wages 25% 20% Gender Wage Gap 15% 10% 5% 0% 2007 2008 2009 2010 2011 2012 2013 2014 2015 Total pay gap Adjusted for demog. characteristics Adjusted for demographic characteristics + occupation & industry choices Source: World Bank staff estimates based on the CSES. 13 The 2016 round of the CSES showed that 46.1 percent of Cambodian women and 63.5 percent of men aged 25 and over had obtained at least a primary education. The available statistics show that, in this respect, Cambodia lags behind other countries in the region. However, more recent generations have higher educational attainment and less of a gender gap. Within the population of recent graduates from the general education system who are between 20 and 30 years old, 75.7 percent of women and 76.6 percent of men had completed at least primary schooling (Gavalyugova 2018). 14 Parenthood introduces gender gaps in the labor force. Men and women with no children have a similar likelihood of working in a family-owned household enterprise. However, after the birth of their first child, women’s probability of working in a household enterprise increases dramatically and remains 10 percent higher than that of men (which is the size of the overall gap in household enterprise employment). Similarly, while men with no children have higher wages than women with no children, the gap begins to widen when the first child is born and expands rapidly on the birth of the third child (Gavalyugova 2018). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 18 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Older adults, rather than young people, Figure 1.21: face the greatest challenges to accessing Source of Employment, by Age good jobs 100% Percent of Working Age Population While the school-to-work transition is a challenge for 80% young people, older workers fare worse in the labor market. Older workers are disproportionately employed 60% in low-productivity family farming or in household enterprises compared with new labor force entrants aged 15 to 24 40% years old (Figure 1.21). Young workers predominate in both 20% foreign and domestic private sector employment, which often provide higher wages than does owning a farm or 0% non-farm household enterprise, whereas the older population 15 20 25 30 35 40 45 50 55 60 65 has difficulty transitioning into the fastest growing areas Age of the economy.15 Employment in the foreign-owned garment sector is particularly popular among young women, Foreign wage Domestic wage Public wage HH Ent. while domestic wage employment remains attractive for Non−wage farming Wage farming Not working young men. Source: Authors’ estimates using CSES. Youth are particularly active in joining the fast- growing manufacturing and construction sectors skills and manual abilities that older workers have not (Figure 1.22). Nearly one-third of workers aged 20 to 29 acquired or can no longer do. Young men are most likely to work in these sectors compared to 10 percent of workers work in the construction sector, whereas young women tend aged 45 to 59. This may be because these sectors require to work in the garment sector. Figure 1.22: Occupation and Sectoral Composition of Employment by Age A younger labor force is engaged in rising sectors and occupations (a) Sector Distribution of Employment (b) Occupation Distribution of Employment 100% 100% Percent of Workers in Age Group Percent of Workers in Age Group 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 15 20 25 30 35 40 45 50 55 60 65 15 20 25 30 35 40 45 50 55 60 65 Age Age Agriculture Other services Trade Tourism Agriculture Elementary Sales and services Construction Other manufacture Garments Machine operators Craft and trades Garments Clerical Manager, professionals, technical Source: Author’s estimates using CSES 2014. Note: The sample was restricted to the working-age population aged 15 to 64 who had worked in the previous seven days before the survey. The sectors and occupations are defined as the sector of main employment. Point estimates for each age were estimated using a locally linear regression with a bandwidth of one year. 15 A similar situation exists in Vietnam according to panel employment data for 2012 and 2014. Transition matrices show that youths (aged 16 to 24) have a higher propensity than any other age group to transition into formal sector wage employment, especially into jobs in the FDI sector. In contrast, workers at the other end of the age spectrum did not have a significant rate of transition into these sectors (Cunningham and Pimhidzai 2018). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 19 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Workers aged 20 to 29 are more likely than workers Urbanization may also be re-shaping the preferences aged 45 to 59 to have mid-skilled occupations of Cambodian consumers. Twenty percent of Cambodians (46 percent versus 29 percent) as well as high- lived in urban areas in 2015, and this share is expected skilled occupations (12 percent versus 9 percent). to increase to 50 percent by 2040 (Baker et al. 2018). The difference is particularly stark for women, with only Urbanization means households will need to buy food that 4.5 percent of older female workers working in high- in the past they may have produced themselves. Urban skilled occupations compared to 11 percent of younger consumers spend more of their income on services such female workers. Over half of younger female workers as communications, transport, restaurants, and banking are involved in mid-skilled occupations, including those (rather than food or other basic necessities) than do non- in garment production, compared to only 28 percent urban consumers. However, perhaps the biggest opportunity of older female workers. The age differences are less for Cambodia is the potential growth in exports to China obvious in urban settings as 22 percent of older urban in response to China’s rapid urbanization and consumer workers work in high-skilled occupations compared to growth, particularly in services, aged care services, and 27 percent of younger female workers. The older urban education (McKinsey Institute 2016). cohort predominates in managerial positions whereas younger cohorts predominate in clerical occupations (Annex This has two implications for Cambodian jobs. First, Figure 1.3). This suggests that transitioning into better there will be growing domestic demand for more, and occupations is particularly difficult for women and for different, goods and services than in the past. Non-poor workers in rural areas. households purchase more expensive food baskets than poor households do—including non-rice cereals, fruits, and meat—and they expect these goods to meet solid standards 1.6 Cambodia’s Future Jobs of hygiene and food safety. They also buy more non-food products and services (Jamora and Labaste 2015). Second, Given this context, what will future jobs look like regional trade patterns will shift to accommodate wealthier in Cambodia? The future jobs picture will be driven by and more urban East Asian populations. Both of these trends changing mega-trends taking shape outside of Cambodia suggest that there will be an increase in the demand for jobs that will affect jobs inside Cambodia but also by policy in higher-value agricultural products, processes products, and reforms implemented today. service-oriented jobs, especially in urban zones in Cambodia and for regional export. The world is experiencing a number of Trade patterns and global value chains are being mega-trends… re-organized We identify four principal trends that are reshaping The re-organization of production globally—and the the global economy and the nature of production that rapid growth of global value chains (GVCs)—has will affect the shape of Cambodia’s job picture in the been an important driver of globalization in recent future: decades. GVCs are a production process in which separate activities are needed to take a product from an idea to the Global consumption patterns are changing due to purchaser, and these different processes can take place in wealth and urbanization several different countries. GVCs have enabled emerging economies to integrate into the global economy faster than The Asian and Cambodian consumer class is in the past, opening up new sources of exports, jobs, and expanding rapidly alongside higher urbanization, growth. However, the world economy is facing a number of which are shifting consumptions patterns. While Asia is structural shifts that may dramatically change the outlook home to some of the richest countries in the world, it is also for GVCs— and thus economic globalization—in the coming home to a growing number of middle-income countries. By years (OECD 2017). First, GVCs are maturing and losing 2030, more than 90 percent of developing Asian households momentum as the potential for the “unbundling of production” should have incomes high enough to have money available into its separate parts in different countries has largely to spend on excess consumption. In Cambodia, 15 percent been exhausted. Second, several factors are expected to of households were classified as being economically secure re-orient production within GVCs as follows: (i) production in 2004, and this proportion had doubled by 2014. costs have significantly increased in a number of emerging CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 20 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA economies; (ii) some new players in GVCs have vast and More jobs are being automated rapidly growing labor forces;16 (iii) digitization may reorient global production and trade back towards advanced While recent media headlines have villainized economies; and (iv) robotics, automation, computerized technology in the workplace, it has the potential to manufacturing, and artificial intelligence may reduce the create more, better, and more inclusive jobs. It is already advantages of siting production in emerging economies unlocking current bottlenecks to job creation. For example, with low labor costs. Production may become increasingly the expansion of mobile money systems in Cambodia in concentrated in regional or local hubs closer to end recent years has not only helped to create jobs, but has also markets, thereby strengthening regional (rather than global) facilitated the transfer of remittances, increased access to value chains. credit, and increased productivity and entrepreneurship. Technology has also improved the quality of a range of jobs by These trends will shape Cambodia’s future relegating manual tasks to machines and freeing up workers participation in GVCs, which are an important feature to carry out higher value-added tasks. For example, for those of the country’s export-oriented growth strategy. First, living in remote areas, those with household duties that tie while in part due to off-shoring and low-value production, them to their home, or those who are physically unable to China’s contribution to global manufacturing value-added leave home, technology can enable them to work from home. and employment is continuing to rise, while increased Technology will also make jobs more inclusive by giving wages within China are also opening opportunities for excluded groups the tools that they need to be integrated other developing countries, including Cambodia (Hallward- into the workforce. For example, household enterprises can Driemeier and Nayyar 2017). Outward FDI by Asian countries, use Facebook to interact with a larger client base than if all of particularly from China, into other developing countries has their interactions had to be face-to-face. also became a much more important share of global FDI and could be a source of new FDI to diversify Cambodia’s While automation is not an immediate threat to the exports. Second, shifts that re-orient production back to Cambodian economy, some jobs will disappear. advanced economies may not affect Cambodia, which sits For instance, the availability of online services for mobile on the doorstep of Asia’s rising consumer class. However, banking, travel bookings, and public services through Cambodia’s current export production is largely aimed at e-government negatively affects workers in medium-skilled, the United States and the EU, two countries that are rapidly routine-based occupations, such as tellers and clerical jobs. digitizing and may bring production home. If Cambodia Other actual and potential jobs may be lost as a result of does not find a way to diversify export markets, its future global firms reshoring (in other words, transferring their participation in global manufacturing industries may be at overseas business operations back to their home country). risk. This could be exacerbated as Cambodia climbs the Wage increases will speed up the pace of automation as income ladder and loses the tariff preferences currently firms displace workers with technology. Cambodian workers’ offered by the EU and the United States. skills will need to be upgraded to align with the factors that drive new jobs growth, or else their jobs may disappear. Box 1.2: Opportunities for Entrepreneurship and Self-Employment in the Global Digital Economy China’s State Information Center estimates that the recent boom in the country’s e-commerce sector has created 10 million jobs in online stores and related services, accounting for about 1.3 percent of the country’s employment. In Kenya, the M-Pesa digital payment system has created additional income for more than 80,000 agents. Samasource and Rural Shores link clients in the United States and the United Kingdom with workers in Ghana, Haiti, India, Kenya, and Uganda. Of all of the global online workers on the Upwork freelancing platform, 44 percent are women, many of whom do freelance work because they want to balance work and family life. Moreover, ICT jobs tend to pay well, and each high-tech job generates 4.9 additional jobs in other sectors in the United States. Source: World Bank (2016) 16 For example, Myanmar in the East Asia Pacific region is attracting new FDI in the garment sector, and Ethiopia is starting to play a larger role in manufacturing assembly activities in Africa. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 21 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA The knowledge economy will become increasingly up with current job creation and upgrading trends. However, important for jobs if policymakers implement some significant changes now, they can leverage these mega-trends to put Cambodia on Twenty-first century workers require a more complex a path to greater development and more, better, and more set of skills than in the past. There has been a global inclusive employment. shift away from manual, routine jobs and into non-routine, knowledge-intensive jobs (Acemoglu and Autor 2010; World To implement these radical changes, Cambodian Bank 2016). While basic cognitive skills (reading and writing) policymakers will need to make significant new and technical knowledge in a given field was the recipe for a investments to turn Cambodia into a modern and productive worker in the past, today’s employers are looking dynamic destination for new and higher-value FDI, for a range of skills and knowledge. This is being driven by and they will need to leverage this FDI to stimulate the automation as machines are taking over routine tasks or parts growth of the domestic private sector. This would mean of the production process and also by the increase in demand tackling key constraints at their root, including: (i) scrapping for high-value products and services driven by the expanding regulations that have no business or ethical rationale; (ii) consumer class and global value chains (Box 1.2). strengthening existing GVC-local economy links; (iii) heavily investing in electricity, logistics, banking, and other services Although Cambodia’s job growth has so far been that are currently constraints to the business climate; (iv) based on low-knowledge jobs, its occupational profile taking steps to regain exchange rate flexibility and monetary is showing signs of becoming more knowledge- independence to shield jobs from US dollar fluctuations; and intensive. Even in the short period between 2009 and (v) rapidly building a demand-driven skills training system that 2015, there has been a significant increase in knowledge- is flexible enough to respond to changing market needs while intensive jobs in Cambodia while growth in manual tasks at the same time reforming the education system to teach has been stagnant or negative. Of course, the number of students the broad range of skills needed in the 21st century knowledge-intensive jobs is still low, but the trends suggest workplace. The government has recently announced new that this is beginning to change. reforms targeting SMEs—including announcing short-term profit tax incentives, customs duty exemptions on select inputs, and the establishment of an SME bank and entrepreneurship …and Cambodia will have to implement fund—but future reforms will need to cut across the sectors of significant changes to respond to these the economy and strengthen their integration. trends and guarantee a better jobs future The rest of this report will explore and define these The challenge faced by policymakers in Cambodia significant reforms that can propel Cambodia toward is to develop policies that capitalize on these mega- a better jobs future. Each chapter explores a different trends to create more and better jobs and stimulate piece of the jobs puzzle—the macroenvironment, private economic growth. Simply maintaining the status quo sector jobs and household enterprises, trade, and skills—in may result in Cambodia falling behind other countries that order to identify current successes and the opportunities are adapting to the new jobs reality, and simply making that are in Cambodia’s grasp if it chooses to undertake changes on the margin will only allow Cambodia to keep these bold reforms. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 22 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA 1.7 References Acemoglu, Daron and David Autor. 2010. “Skills, Tasks, and Technologies: Implications for Employment and Earnings.” NBER Working Paper 16082. Cambridge: NBER. Ahsan, Ahmad, Aaditya Mattoo, Batshur Gootiiz, Juan Sebastian Saez, Martin Molineuvo, and Peter Walkenhorst. 2015. ASEAN Services Integration Report : A Joint Report by the ASEAN Secretariat and the World Bank. Washington, D.C.: World Bank. Baker, Judy, Sarah Xiyuan Lin, Huyen Thi Phuong Phen, Natsuko Kikutake, Erik Johnson, Frank Van Woerden, Bunlong Leng, Katie Mcwilliams, Makathy Tep, Chamroen Sok, Narya Ou. 2018. Cambodia - Achieving the potential of urbanization. Washington, D.C.: World Bank Cunningham, Wendy. 2007. Minimum Wages and Social Policy in Latin America. Washington D.C.: World Bank Group. Cunningham, Wendy and Obert Pimhidzai. 2018. Vietnam’s Future Jobs: Leveraging Megatrends for Greater Prosperity. Washington, DC: World Bank. Del Carpio, Ximena, and Laura Pabón. 2014. Minimum Wage Policy: Lessons with a Focus on the ASEAN Region. Washington D.C.: World Bank. Eliste, Paavo, and Sergiy Zorya. 2015. Cambodian Agriculture in Transition: Opportunities and Risks. Washington, D.C.: World Bank. Gavalyugova, Dimitria. 2018. “Gender Analysis of the Cambodian Labor Market.” Forthcoming. Goldin, Claudia. 1995. “The U-shaped female labor force function in economic development and economic history” In T.P. Schultz, ed. Investment in Women’s Human Capital and Economic Development Chicago: University of Chicago Press, 61-90. Hallward-Driemeier, Mary and Gaurav Nayyar. 2017. Trouble in the making? : the future of manufacturing-led development Washington, D.C.: World Bank. Jamora, Nelissa, and Patrick Labaste. 2015. “Overview of Food Demand Trends and Prospects in East Asia” Background paper prepared for the World Bank. Kelsall, Tim and Seiha Heng. 2014. “The Political Settlement and Economic Growth in Cambodia.” ESID Working Paper No 37. McKinsey Institute. 2016. Urban World: The Global Consumers to Watch. New York: McKinsey & Company MoEYS—Ministry of Education, Youth and Sport. 2017. Grade six national assessment findings in 2016. Flyer. Ministry of Education, Youth & Sport. Kingdom of Cambodia. MoP (Ministry of Planning). 2012. “Migration in Cambodia: Report of the Cambodian Rural Urban Migration Project (CRUMP).” OECD. 2017. “The Future of Global Value Chains: ‘Business as Usual’ or ‘A New Normal’?” Paris: Organisation for Economic Co-Operation and Development. Royal Government of Cambodia. 2015. “National Employment Policy: 2015-2025.” Shrestha, Maheshwor. 2018. “The impact of minimum wage hikes in participation and wages.” Washington, D.C.: World Bank. Stanovich, Keith E. 1986. “Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy” Reading Research Quarterly 21: 360-406. Testaverde, Mauro, Harry Moroz, Claire H. Hollweg, and Achim Schmellen. 2017. Migrating to Opportunity: Overcoming Barriers to Labor Mobility in Southeast Asia. Washington, D.C.: World Bank. USAID. 2015. Assessment of early grade reading in the education sector in Cambodia. World Bank. 2016. World Development Report 2016: Digital Dividends. Washington, D.C.: World Bank. World Bank. 2017. Cambodia - Sustaining strong growth for the benefit of all. Washington, D.C.: World Bank. World Vision. 2017. 2017 Policy Brief on Education in Cambodia. https://www.wvi.org/sites/default/files/WVI-C%20Education%20Policy%20Brief_2017_EN_Final.pdf CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 23 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA ANNEX 1.1: Data Sources and Definitions Data Sources This includes those who were working as wage workers as well as the self-employed and workers helping out on the This report uses data from the Cambodia Socio- family farm. Economic Survey (CSES) that has been conducted annually since 2007 by the National Institute of Public wage employment: The employer of those involved Statistics (NIS) in the Ministry of Planning in Cambodia. in non-farm wage work is either the government or a state- The CSES collects data on a wide array of indicators related owned enterprise. to household demographics, consumption, education, and economic activities, for example. This survey is the primary Private wage employment: All non-farm wage employees poverty monitoring and planning tool used by the Royal who are employed by a non-public entity are assumed to be Government of Cambodia. in the private sector. This may include being employed by a foreign enterprise or a non-profit institution as well by a Although the survey is conducted annually, the Cambodian private enterprise. sample size has differed from year to year. Usually, the CSES interviews about 3,500 to 4,000 households each Non-farm enterprise workers: This category consists of year, but every five years the NIS conducts a “big sample” workers who are neither in farming nor in wage employment. survey with a sample of about 12,000 households. These These workers are self-employed in non-farm sectors and big sample surveys were conducted in 2009 and 2014. include own-account workers, employers, and unpaid family workers. In this report, we use annual CSES data from 2007 through 2015 (the latest year for which data are Unpaid family workers: These are family members available). Although we used data from all survey years to who are making unpaid contributions to the household’s look at trends, for point estimates we usually only used data economic activities. From 2009 onwards, the manuals for from 2009 and 2014 because of their larger sample sizes the CSES have explicitly instructed enumerators to classify and the greater precision of the indicators. those who contribute to their own household activities and enterprises without payment or income as self-employed. Previously, these workers, who were mostly women, were Employment Categories classified as “unpaid family workers.” For conceptual clarity, this report will classify unpaid family workers as Unless otherwise specified, labor force participation farming or non-farm enterprise workers (self-employed) categories in this report are based on the primary as appropriate. economic activity performed by the respondent in the previous seven days before the survey. Sectors of Employment Not working: An individual is defined as not working if he/she had not spent even one hour in the previous seven The CSES uses four-digit industry codes derived from days (before the survey) doing any economic activities. ISIC-Rev 4, a United Nations industry classi cation However, they are considered to have been working if they system, to classify the industrial sectors of each job. had a regular job but were temporarily not working that In this chapter, we use the following aggregation: week because they were taking a period of time off from work. This also includes those who are traditionally defined Garments: This includes the categories of “Manufacture of as “unemployed,” or in other words, those who did not wearing apparel, except fur” and “Manufacture of footwear.” work but were actively looking for work. The share of such Over 92 percent of this category consists of manufacturing unemployed individuals is extremely low in Cambodia, with of wearing apparel. less than 0.2 percent of the working-age population falling into this category. Hence, we do not analyze this category Other manufacturing: This category includes all separately in this report. manufacturing sectors except garments and footwear. Farming: Those workers whose main activity falls under Construction: This category includes the construction of the “Agriculture, forestry, and fishing” categories in the buildings, civil engineering, and specialized construction- classifications of employment sectors used in the CSES. related activities. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 24 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Tourism: This category includes industries related to Craft and related trades: This category includes workers accommodation, food and beverage service activities, and who produce foodstuffs, textiles, and other articles by travel agency and related activities. applying their specific skills and knowledge. They typically work with their hands or with hand-powered tools and Trade: This category includes industries related to the understand all stages of the production process. Common wholesale and retail trade as well as to the repair of motor- occupations in this category in Cambodia are building vehicles and motorcycles. frame workers, food processing workers, wood treaters and cabinet workers, and those in related trades. Other services: This category includes all other non-farm related sectors. Garments and related trades: Although garment workers are officially coded as craft and trades work, jobs in the Farming: All agriculture, forestry, and fishing related garments industry in Cambodia tend to be different from industries fall in this category. the typical work done in other craft-related occupations. For instance, most garment manufacturing in Cambodia is Occupation Categories done in a production line, meaning that workers may not be familiar with the entire production process. The CSES uses three-digit occupation codes derived from ISCO-08, an occupational classification Service and sales: This category includes personal structure of the International Labour Organization services workers, sales workers, and personal care (ILO), to classify each occupation. In this chapter, we and protective services workers. In Cambodia, common use the following categories: occupations in this category are shop salespersons, street and market salespersons, waiters and bartenders, protective Agriculture: This category includes skilled agricultural, service workers, street food snack sellers, policemen, and forestry, and fishery workers. Typical occupations in this hairdressers and beauticians. category in Cambodia include subsistence crop and livestock farmers, market gardeners, and crop growers. Clerical: This category includes general and keyboard clerks, customer services clerks, and numerical and Elementary occupations: These are usually unskilled material recording clerks among others. In Cambodia, occupations in agriculture, manufacturing, and the service common clerical occupations include general office clerks, sector. Occupations typically in this category in Cambodia tellers, and client information workers. are laborers in agriculture, construction, and transport. Note that the way in which jobs are coded in Cambodia, it does Managers, professionals, and technicians: In Cambodia, not include workers in garment factories. common occupations in this category include armed forces, school teachers, legislators, finance professionals, and Machine Operators: This category includes plant and managers in various sectors. machine operators and assemblers, including drivers. Typical occupations are drivers of tutktuks, taxis, and other vehicles including heavy trucks, mobile plant operators, and machine operators in the food, rubber and plastic, and textile industries. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 25 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA ANNEX 1.2: Additional Figures and Tables Annex Figure 1.1: Garments and construction-related occupations boomed during 2007-2015 (a) Total Labor Force; 100% Percent of all Workers 80% 60% 40% 20% 0% 2007 2008 2009 2010 2011 2012 2013 2014 2015 (b) Female Labor Force; 100% Percent of Female Workers 80% 60% 40% 20% 0% 2007 2008 2009 2010 2011 2012 2013 2014 2015 (c) Urban Labor Force; 100% Percent of Urban Workers 80% 60% 40% 20% 0% 2007 2008 2009 2010 2011 2012 2013 2014 2015 (d) Rural Labor Force 100% Percent of Rural Workers 80% 60% 40% 20% 0% 2007 2008 2009 2010 2011 2012 2013 2014 2015 Agriculture Elementary Machine Operators Garments Craft & trades Sales and services Clerical Manager, professionals, technical Source: Author’s estimates using CSES 2007-2015. Note: The sample was restricted to the working-age population aged 15-64 who had worked in the previous seven days before the survey. The occupations were defined as the main occupation of employment. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 26 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Annex Figure 1.2: Sector Employment and Age, 2014 Younger labor force engaged in the rising sectors (a) Total Labor Force 100% Percent of Total Labor Force 80% 60% 40% 20% 0% 15 20 25 30 35 40 45 50 55 60 65 Age (b) Female Labor Force 100% Percent of Female Force 80% 60% 40% 20% 0% 15 20 25 30 35 40 45 50 55 60 65 Age (c) Urban Labor Force 100% Percent of Urban Labor Force 80% 60% 40% 20% 0% 15 20 25 30 35 40 45 50 55 60 65 Age (d) Rural Labor Force Agriculture 100% Percent of Rural Labor Force Other Services 80% Trade 60% Tourism 40% Construction 20% Other Manufacturing 0% Garments 15 20 25 30 35 40 45 50 55 60 65 Age Source: Author’s estimates using CSES 2014. Note: The sample was restricted to the working-age population aged 15-64 who had worked in the previous seven days before the survey. The sectors were defined as the main sector of employment. The point estimates for each age were estimated using a locally linear regression with a bandwidth of one year. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 27 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Annex Figure 1.3: Occupation and age, 2014 (a) Total Labor Force 100% Percent of Total Labor Force 80% 60% 40% 20% 0% 15 20 25 30 35 40 45 50 55 60 65 Age (b) Female Labor Force 100% Percent of Female Force 80% 60% 40% 20% 0% 15 20 25 30 35 40 45 50 55 60 65 Age (c) Urban Labor Force 100% Percent of Urban Labor Force 80% 60% 40% 20% 0% 15 20 25 30 35 40 45 50 55 60 65 Age Agriculture (d) Rural Labor Force Elementary 100% Sales and services Percent of Rural Labor Force 80% Machine operators 60% Craft and trades 40% Garments 20% Clerical Manager, professionals, technical 0% 15 20 25 30 35 40 45 50 55 60 65 Age Source: Author’s estimates using CSES 2014. Note: The sample was restricted to the working-age population aged 15-64 who had worked in the previous seven days before the survey. The sectors were defined as the main sector of employment. The point estimates for each age were estimated using a locally linear regression with bandwidth of one year. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 28 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Annex Table 1.1: Profile of the Cambodian Labor Force, 2014 Employed Labor Force Aged 15-64 Labor Force Employed Total Labor Force Male Female Rural Urban Female 0.514 0.482 0.483 0.480 (0.500) (0.500) (0.500) (0.500) Age: 15-24 0.336 0.290 0.298 0.282 0.311 0.221 (0.472) (0.454) (0.457) (0.450) (0.463) (0.415) 25-34 0.274 0.301 0.304 0.298 0.270 0.404 (0.446) (0.459) (0.460) (0.457) (0.444) (0.491) 35-44 0.159 0.175 0.176 0.175 0.174 0.178 (0.365) (0.380) (0.381) (0.380) (0.380) (0.382) 45-54 0.143 0.153 0.149 0.157 0.160 0.129 (0.350) (0.360) (0.356) (0.364) (0.367) (0.335) 55-64 0.089 0.081 0.073 0.089 0.085 0.068 (0.285) (0.273) (0.260) (0.285) (0.279) (0.251) Education: None 0.159 0.159 0.118 0.204 0.189 0.062 (0.366) (0.366) (0.322) (0.403) (0.391) (0.241) Incomplete Primary 0.287 0.307 0.288 0.328 0.346 0.180 (0.452) (0.461) (0.453) (0.470) (0.476) (0.384) Incomplete Secondary 0.420 0.408 0.441 0.373 0.399 0.440 (0.494) (0.492) (0.497) (0.484) (0.490) (0.496) HS graduate 0.097 0.082 0.100 0.063 0.056 0.168 (0.295) (0.275) (0.300) (0.243) (0.230) (0.374) College or more 0.038 0.043 0.053 0.032 0.010 0.150 (0.191) (0.202) (0.223) (0.176) (0.100) (0.358) Marital status: single 0.358 0.322 0.340 0.304 0.313 0.354 (0.480) (0.467) (0.474) (0.460) (0.464) (0.478) Ethnic minority 0.036 0.034 0.035 0.034 0.038 0.024 (0.186) (0.182) (0.184) (0.181) (0.190) (0.154) Household size 5.141 5.094 5.177 5.004 5.088 5.114 (1.889) (1.881) (1.856) (1.902) (1.862) (1.942) Rural 0.754 0.767 0.766 0.768 (0.430) (0.423) (0.423) (0.422) Region: East 0.048 0.048 0.048 0.047 0.054 0.029 (0.213) (0.213) (0.215) (0.212) (0.225) (0.167) South-East 0.207 0.209 0.204 0.216 0.252 0.069 (0.405) (0.407) (0.403) (0.411) (0.434) (0.254) South-West 0.175 0.176 0.176 0.176 0.198 0.102 (0.380) (0.381) (0.380) (0.381) (0.399) (0.303) North 0.148 0.151 0.153 0.149 0.171 0.086 (0.355) (0.358) (0.360) (0.356) (0.377) (0.281) North-West 0.159 0.161 0.166 0.155 0.166 0.144 (0.366) (0.367) (0.372) (0.362) (0.372) (0.351) Special Region 0.264 0.255 0.253 0.257 0.159 0.570 (0.441) (0.436) (0.435) (0.437) (0.366) (0.495) Source: Author’s estimates using CSES 2014. Note: Standard deviations in parentheses. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 29 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Annex Table 1.2: Employment of the Working-age Population Farming Non-farm Wage Work Non-farm HH Enterprise Not Working Year Wage Self + Family Private Public Self Family 2007 313 3696 1127 334 892 558 1429 2008 391 3497 1257 365 900 567 1597 2009 522 3816 1118 340 1218 407 1376 2010 532 3702 1386 386 1376 406 1218 2011 571 3923 1604 351 1301 271 1180 2012 604 3473 1834 393 1389 213 1499 2013 707 3207 2156 367 1414 103 1655 2014 602 3082 2588 398 1290 111 1729 2015 681 2761 2934 412 1348 98 1741 Source: Source: Author’s estimates using CSES 2007-2015. The working-age population estimates are from UNPOP. Note: See Annex 1.1 for definitions of the categories. Annex Table 1.3: Number of Wage Workers, in thousands Year Agriculture Trade Services Construction Manufacturing 2007 312.5 44.0 766.3 279.9 370.9 2008 391.2 36.8 824.5 261.3 499.7 2009 521.6 37.7 721.5 245.0 453.2 2010 531.5 66.8 881.8 282.2 541.2 2011 570.5 80.5 905.3 269.4 699.5 2012 604.0 72.0 1047.9 357.8 749.2 2013 707.1 86.6 1101.0 459.8 875.8 2014 602.2 110.8 1218.3 540.6 1116.2 2015 680.7 123.9 1383.4 636.2 1202.5 Source: Author’s estimates using CSES 2007-2015. The working-age population estimates are from UNPOP. Note: Trade includes retail as well as other trades. Services include all other sectors. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 30 TECHNICAL REPORT CHAPTER 1: AN INTRODUCTION TO JOBS IN CAMBODIA Annex Table 1.4: Largest Occupations in 2009 and 2014 2009 2014 Occupation Share Occupation Share Subsistence crop farmers 0.257% Subsistence crop farmers 0.198% Subsistence livestock farmers 0.102% Garment and related trades workers 0.102% Market gardeners and crop growers 0.076% Agricultural, forestry, fishery labourers 0.069% Agricultural, forestry, shery labourers 0.063% Market gardeners and crop growers 0.069% Shop salespersons 0.050% Shop salespersons 0.060% Street and market salespersons 0.049% Subsistence livestock farmers 0.054% Firewood and water collectors 0.047% Street and market salespersons 0.036% Garment and related trades workers 0.044% Animal producers 0.035% Animal producers 0.036% Mining and construction labourers 0.031% Subsistence fishers, hunters, gatherers 0.025% Building frame and related trades workers 0.028% Source: Author’s estimates using CSES 2009 and 2014. Note: Occupations are three-digit occupation codes. Primary occupations are used to categorize workers. Annex Table 1.5: Real Wage Earnings, 2007-2015 Monthly Wages, in 2014 PPP International $ Percent of Workers Earning Below Year p=25 p=50 Mean p=75 $2 per day $4.22 per day 2007 131.4 197.1 301.4 307.8 7.4% 23.9% 2008 120.1 187.7 261.8 285.3 9.1% 26.5% 2009 120.9 204.0 256.0 287.2 11.1% 25.4% 2010 145.3 218.0 272.3 303.0 7.0% 20.3% 2011 172.2 240.1 287.4 330.7 3.6% 15.4% 2012 167.3 247.7 293.9 328.0 3.5% 12.5% 2013 195.1 260.1 325.7 390.1 2.9% 9.0% 2014 250.4 325.5 393.7 425.7 1.6% 7.7% 2015 272.1 371.1 411.9 463.9 0.8% 6.6% Source: Author’s estimates using CSES 2007-2015. Note: Monthly wages were recorded in KHR and were converted to 2014 using the consumer price index and to international PPP$ using the conversion rate in 2014 of 1597.30 KHR per PPP$. The $4.22 per day threshold is $2 per day multiplied by the ratio of the total population to the share of individuals that are employers, employees, or own-account workers. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 31 TECHNICAL REPORT CHAPTER 2: Macroeconomic Policies Conducive for Creating Good Jobs Dilaka Lathapipat, Sodeth Ly, and Miguel Eduardo Sánchez 2.1 Introduction Over the past two decades, Cambodia has been one of the fastest growing countries in the world, but its labor productivity has not been as impressive. GDP growth benefitted from considerable additions to the workforce and an increase in working hours, but the contribution of capital and labor quality to growth and labor productivity has been smaller in Cambodia than in Vietnam or Thailand. Thus, while the number of jobs grew and contributed to economic growth, the quality of jobs was not as high as it could have been. In addition, while output per worker in agriculture and services has been rising in Cambodia, it has slightly declined in the garment and construction sectors in recent years. This chapter looks at the relationship between economic growth and jobs, and proposes a set of macroeconomic policies that will foster strong and stable growth and job creation going forward. We begin by describing the relationship between jobs and economic growth in recent years, focusing on areas of labor contribution to growth. Next, we discuss the need for investments in human and physical capital to improve job quality and increase labor contributions to growth—using a Solow-Swan growth accounting framework to predict how these investments might affect future growth, labor productivity, and employment. Finally, we explore how Cambodia’s monetary policy of high dollarization, while significantly contributing to past growth, undermines future job goals. 2.2 Structural Transformation Supported Job Creation in Cambodia Sustained economic growth has led to job creation across several sectors… Over the past two decades, Cambodia has enjoyed strong and sustained economic growth and macroeconomic stability. Cambodia ranked sixth in the world in economic growth between 1994 and 2015, with an average annual rate of 7.6 percent. This growth has also been resilient, with the economy expanding at around 4 percent even during the 1997-98 period of the East Asian financial crisis. Although growth was negatively impacted by the 2008-2009 global financial crisis, and decelerated to 0.1 percent in 2009, it quickly rebounded and remained strong at an average of 7.2 percent between 2010 and 2015 (Figure 2.1). Except during those two external crises, inflation has remained in the single digits. Per capita GNI grew at an average rate of 5.4 percent between 1996 and 2015, the year in which Cambodia became a lower-middle-income economy. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 33 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS Figure 2.1: Growth in GNI per capita and GDP, 1994-2016 Cambodia has been a champion in terms of growth and poverty reduction 14% 12% 10% 8% Annual Growth 6% 4% 2% 0% GNI per capita -2% GDP growth 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Source: Official sources and World Bank staff estimates. Table 2.1: Contributions to GDP Growth by Sector 2004-07 2007-11 2011-15 Agriculture 23.6% 26.5% 5.7% Industry Textile, Apparel, & Footwear 19.6% 26.1% 22.2% Construction 9.7% -3.7% 15.7% Other Industry 5.5% 9.8% 9.0% Services Hotel & Restaurants 6.2% 7.4% 6.4% Trade 7.2% 11.9% 10.2% Real Estate & Business 7.5% -4.1% 10.5% Other Services 20.7% 25.9% 20.3% Gross Domestic Product 100.0% 100.0% 100.0% Source: National Institute of Statistics. Cambodia’s economic growth has been driven by prices. However, construction and real estate grew rapidly, a few sectors: agriculture, apparel manufacturing, and, by 2011 to 2015, accounted for one-quarter of GDP tourism, and, more recently, construction (Table 2.1). growth as the sector was fueled by foreign direct investment Exports of goods and services—mainly of garments, rice, (FDI) and rapid credit growth. Textiles, apparel, and footwear and tourism—grew at an average rate of more than 19 accounted for 22.2 percent of total growth in 2011 to 2015, percent annually between 1994 and 2015. Agricultural and other industry, including other manufacturing and growth accounted for one-quarter of Cambodia’s real mining and quarrying, held steady at 9 percent. Supported growth between 2004 and 2011 but plummeted to only by sustained growth in tourism and retail and wholesale 5.7 percent between 2011 and 2015 because of adverse services, the services sector accounted for almost half of all weather conditions and depressed agricultural commodity of Cambodia’s GDP growth in the same period. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 34 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS Jobs growth has been volatile in recent years. During jobs. The share of agriculture in total employment dropped the commodity price boom of 2004-2011, more than from 58.3 percent in 2004 to 43.8 percent in 2015, while 967,000 agricultural jobs were created. However, these employment in garments, construction, and real estate were nearly wiped out when commodity prices fell, coupled each accounted for 11 percent of the total compared by a rise in the non-agricultural sectors, out-migration, and to less than 5 percent each a decade earlier. The share some difficult weather events. However, the boom in textiles held by other manufacturing and services, including retail and garments created more than 600,000 jobs between and trade, remained flat at around 30 percent of total 2007 and 2015. Hotels and restaurants also picked up post- employment. 2007, with more than 230,000 new jobs being created by 2015, while more than 300,000 new jobs were created in There are signs that fast-paced job growth may be construction in just four years (2011 to 2015). slowing down, even as GDP growth rates remain impressive. Between 2004 and 2015, more than 2 million Job growth was driven by different factors in each jobs were created while 1 million were destroyed, mostly in sector. Construction activity declined in the aftermath of agriculture. Job growth matched or exceeded the growth the global financial crisis, but the sector quickly recovered of the working age population (aged 15 to 64) from 2004 and has been generating a large amount of jobs in recent to 2011, but it fell far short of the growing working age years. In garments, the European Union’s decision in population between 2011 and 2015. During the latter period, 2010 to simplify the eligibility criteria for Everything But the number of adults aged 15 to 64 increased by 778,000, Arms (EBA) beneficiary countries 1 exporting duty-free while net number of jobs created was only 260,000 (Table merchandise (from requiring a double transformation—for 2.2). A growing share of young adults were choosing to example from yarn to textiles to apparel—to just requiring continue in school (as seen in the increase in secondary and a single transformation—from textiles to apparel). This tertiary education attainment) while some may have chosen resulted in a surge in Cambodia’s trade with the EU, and led to not enter market work in favor of homecare. Nevertheless, to significant job creation and a shift in the composition of a substantial share entered the labor force. Table 2.2: Net Employment Creation by Sector Agriculture was the main source of employment creation for a decade, but this trend has recently been reversed 2004-07 2007-11 2011-15 Agriculture 433,572 533,124 -945,646 Textile, leather, and footwear 16,477 307,737 335,156 Construction 95,047 - 1,190 358,734 Other industry 98,252 -24,162 1,218 Hotel & restaurant 2,710 130,892 132,612 Trade -3,514 121,663 -6,364 Real estate & professional services 137,944 82,003 109,978 Other services - 37,345 8,901 274,457 Total net employment creation 743,144 1,158,968 260,145 Source: World Bank staff estimates based in Cambodia Socioeconomic Surveys (CSES), several years, and ILO employment figures. 1 The EBA scheme gives the 47 least developed countries in the world, including Cambodia, duty-free access to the European Union for exports of all products, except arms and ammunition. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 35 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS …but economic growth has not contributed Labor quality, on the other hand, has made modest much to improved labor productivity contributions to GDP growth during the past two decades, with little volatility across years. Only 5 One-third of real economic growth in Cambodia percent of GDP growth can be attributed to labor quality over between 1993 and 2015 was attributable to increases any of the time periods presented in Figure 2.2. This is not in the average number of employed workers. When GDP surprising given the low education level of the labor force, as growth is decomposed, it becomes clear that employment discussed in Chapter 1, and partly underlies Cambodia’s low creation made a significant contribution to GDP growth total factor productivity (TFP) and labor productivity. (around 1.8 percentage points) from 1993 until 2011, at which point it strongly dropped off (Figure 2.2.). This was Capital has been responsible for more than half of likely due to the strong contraction in agriculture jobs, as Cambodia’s GDP growth since 1993. Capital accumulation discussed above. Hours worked made a much smaller but has been the largest contributor to growth in Cambodia, more consistent contribution to GDP growth, averaging 0.38 remaining so in spite of economic volatility. Nonetheless, percent over the 22-year period. the levels of capital formation in Cambodia have so far been below those of other East Asian countries. Figure 2.2: Cambodia GDP Growth Decomposition Employment and working hours contributed to one-third of GDP growth until recent years 12% 10% 8% Percent Growth 6% TFP 4% Capital 2% Employment 0% Avg. Hours -2% Labor Quality -4% 1993-2015 1993-1999 1999-2004 2004-2007 2007-2011 2011-2015 Source: World Bank staff calculations using national accounts and other official statistics. Notes: Average weekly work hours were computed using information from Cambodia Socioeconomic Surveys and the APO Productivity Database 2016. Figure 2.3: Labor Productivity Growth Decomposition Labor productivity growth has been slower in Cambodia than it was in Vietnam and Thailand because of low capital intensity 8% 7% 6% Productivity Growth (output per hour) 5% 4% 3% 2% Capital Intensity 1% Labor Quality 0% TFP -1% Cambodia Vietnam Thailand (1993-2015) (1993-2014) (1986-1996) Source: World Bank staff calculations using national accounts and other official statistics. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 36 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS In general, capital helps to improve the quality of jobs, since capital intensity to labor productivity growth has increased it often replaces more inefficient workers and thus increases significantly in recent years. productivity and improves job quality. However, in some cases, capital replaces workers, which reduces job quality. Average labor productivity in the various sectors has begun to converge. 3 The level of labor productivity TFP growth in Cambodia was especially high during in the industry and service sectors in the mid-1990s the 1990s and the 2000s as Cambodia started to was respectively around five and four times higher than integrate into the global economy. Changes in TFP are the level in agriculture. By 2015, agriculture productivity associated with the process of creative destruction, in which was catching up, and was now only three times lower less productive firms and outdated processes are replaced than in industry and services. This rising agricultural labor by new and more efficient ones and labor flows into more productivity stemmed from the expansion of smallholder productive sectors. TFP growth, usually faster during a land and gains in crop yields, partly due to better access to country’s early stages of economic development, has been irrigation and crop diversification (Eliste and Zorya 2015). volatile in Cambodia and has slowed down in recent years. Meanwhile, output per worker in garment manufacturing and construction has slipped below the levels of a decade Cambodia’s performance in terms of labor ago, from before the global financial crisis and the 2009 productivity looks less impressive.2 Between 1993 and slowdown in the Cambodian economy (Figure 2.4(a)). The 2015, Cambodia’s labor productivity grew by 3.6 percent other sectors that have grown significantly over the past compared to 4.7 percent in Vietnam during the same period decade are other industry, trade and retail, and other and to the 7.1 percent average growth rate in Thailand services such as travel services, education, or business during its boom years of 1986 to 1996 (Figure 2.3). Capital services (Figure 2.4). It is worth noting that these sectors deepening contributed only 2.1 percent to Cambodia’s started from a low level of productivity and are improving average productivity growth rate, a significantly lower rate as firms and processes are becoming more organized and than in Vietnam and Thailand, which suggests that Cambodia modern, but that they still have lower productivity levels has fallen behind in upgrading its production technology. It than the garment industry does. is nonetheless worth mentioning that the contribution of Figure 2.4: Productivity in Agriculture, Industry and Services Productivity in the garment and construction industries has stagnated in recent years (a) Agriculture and Industry (b) Services 16 16 14 14 Trillion riel (Year 2000 constant) Trillion riel (Year 2000 constant) 12 12 10 10 8 8 6 6 4 4 2 2 0 0 2004 2007 2008 2009 2010 2011 2012 2013 2014 2015 2004 2007 2008 2009 2010 2011 2012 2013 2014 2015 Agriculture Garments Hotel and Restaurant Trade Construction Other Industry Real Estate and Other Services Professional Services Source: Author’s estimates using CSES 2014. 2 This analysis is complementary to our GDP growth decomposition. Labor productivity is defined here as output per working hour, and the analysis focuses on identifying how much of it comes from mechanization and capital intensity, how much has to do with improvements in labor quality, and how much is related to other improvements in TFP such as better managerial practices or the use of new technologies. 3 This chapter discusses productivity from a macroeconomic point of view, drawing from the value added in the national accounts and dividing it by the number of workers in the sector. Other chapters discuss productivity using firm-level data, so results may differ. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 37 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS Labor productivity growth since 2004 has been mainly of potential workers is expected to expand by more than the result of improvements within sectors rather than 135,000 per year on average over the next 35 years. The of labor reallocation between sectors. Most of the labor increase in the potential workforce will not be uniform, and productivity growth between 1999 and 2004 was due to the the pool will start to contract around the year 2044. The total reallocation of labor from agriculture into the export sectors dependency ratio is projected to decrease gradually until (garment manufacturing and services) (Figure 2.5). From 2044 and then increase thereafter (Figure 2.6). Hence, the 2004-2007, as the economy boomed, significant aggregate gradual entry of Cambodia’s youth into the labor market will productivity gains occurred within all sectors primarily as a continue to be a natural engine for economic growth for the consequence of different improvements made within each foreseeable future. sector rather than of shifts of labor between sectors. There were also few shifts between sectors between 2007 and However, the contribution of employment to growth 2011 as employment was created in agriculture in the will not be as substantial as it has been over the context of a prolonged commodity price boom. Only in recent last two decades. From 1994 to 2015, a period in which years, as structural change resumed and agriculture started employment made a positive and significant contribution to to shed workers, have productivity gains again been driven GDP growth, the average growth rate in the working age by the reallocation of labor across sectors complemented by population was 3.15 percent. Over the coming decades, the continued within-sector productivity improvements. contribution of employment to growth will be more modest given the projected average growth rate of the working age population of 1.12 percent. Because labor quality does not 2.3 Investments are Needed for contribute greatly to labor productivity in Cambodia, in order for the demographic dividend to materialize, Cambodia will Cambodia to Capitalize on its have to invest more in its physical and human capital. Demographic Dividend Since economic growth is expected to slow in Over the next three decades, Cambodia will benefit the medium to long term with falling labor force from a demographic dividend as the dependency participation, investing in the accumulation of ratio keeps falling, which will give the country the physical and human capital will be increasingly opportunity to continue its strong growth trajectory. important for boosting productivity and, ultimately, UN population projections indicate that Cambodia’s supporting economic growth and job creation. Here, population will continue to grow at an average compounded investments in human capital can take the form of policies, rate of 1.06 percent per year between 2015 and 2050, while such as academic scholarships, to increase the average the working age population (aged between 15 to 64) is years of schooling, while investment in physical capital can projected to grow at a slightly higher rate of around 1.12 include policies to promote capital investments.4 How much percent. Taking population dynamics into account, the pool of an effect will such investments have? Departing from Figure 2.5: Real Output Share Weighted Average Components of Labor Productivity Growth Structural transformation has significantly accelerated in recent years 8% 7% (output per worker employed) Labor Productivity Growth 6% 5% 4% Structural Change 3% Within industry 2% 1% Output Price 0% -1% Interaction Terms -2% Aggregate 1993-1998 1998-2003 2003-2007 2007-2010 2010-2014 Source: World Bank staff calculation using national accounts and other official statistics. 4 More detailed discussions about the many types of policies that can be implemented to make strategic investments that contribute to national growth and job creation are made throughout this report, and are summarized in the final policy matrix. The focus of this section is on an analysis of how, broadly speaking, investments in these areas will contribute to national macroeconomic trends. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 38 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS the growth and labor productivity decomposition discussed decline to 3.2 percent by 2050 (compared to 15 percent above, our analysis in this section aims at offering a in 2015). Meanwhile, the percentage of people with an forward-looking view of Cambodia’s long-term growth and upper secondary education or above would increase to 57 job creation prospects. We used a Solow-Swan growth percent by 2050 (compared to just 19.5 percent in 2015) accounting framework to project future growth drivers that (Figure 2.7). Under a more aggressive education strategy depart from observed historical trends.5 in which we assume that the average number of years of schooling would increase by five years by the year 2050, To assess how an increase in education can boost TFP, the percentage of people with no formal education would we compare outcomes under the current “business decline to 2.9 percent by 2050, while 65.5 percent of the as usual” strategy with those under a modest reform population would have at least completed lower secondary scenario. Under the baseline “business as usual” scenario, education (Figure 2.7). This faster pace of educational the average education level of the workforce is expected to attainment may be feasible as policymakers have expand increase by four years by the year 2050. This implies that scholarship programs, increase awareness of the importance the percentage of people with no formal education would of educational attainment, and increase access to schools. Figure 2.6: Projected Population Growth and Dependency Ratios in Cambodia Cambodia’s demographic dividend is expected to last until 2044 (a) Projected Annual Change in the Working Age (b) Dependency Ratio (15-64 Years) Population Age 65, Relative to the Population Age 15-65 Share of Population Under Age 14 or Over 350,000 1.2 300,000 1.0 250,000 (number of people) Annual Change 200,000 0.8 150,000 0.6 100,000 50,000 0.4 0 0.2 -50,000 -100,000 0.0 1990 2010 2030 2050 1990 2010 2030 2050 Source: United Nations World Population Prospects: 2015 Revision. Figure 2.7: Estimated Percent of Population with At Least an Upper Secondary Education, Baseline and Reform Scenarios 70% Percent of Population Aged 15+ 60% 50% 40% 30% 20% 10% Baseline 0% Reform 07 09 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Source: World Bank staff calculations based on UN World Population Prospects and CSES. 5 In Annex 2.1, we explain the methodology and the assumptions upon which we based our projections under both scenarios. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 39 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS TFP growth in the reform scenario is projected to be more slowly than the returns to physical infrastructure 0.35 percentage points faster than in the baseline and construction. scenario. Using statistical methodologies that account for human capital externalities, we estimated that every one- Under the baseline scenario in which human and year increase in the average schooling of the population is physical capital remains at current levels, GDP associated with a positive effect on TFP growth. Specifically, growth can be expected to decline to 5.9 percent we estimated that a one-year increase in average schooling by 2040. Under the baseline scenario, there would be would have a 13.5 percent increase in TFP. a moderate increase in physical capital intensity, with the capital to hourly labor ratio increasing from around To assess how a more aggressive capital investment 1 in 2014 to 8 by 2050. Aggregate labor productivity, strategy might affect TFP, we again compared the defined as output (real GDP in constant 2005 US dollars) current scenario to a scenario with greater capital per working hour, would rise from USD 0.57 in 2015 to intensity. Under a baseline scenario mostly driven by FDI around USD 3 in 2050 (Figure 2.9(c)). Growth in both output in the context of low domestic savings, we estimated that per worker and GDP would slow in the long term (Figure the investment share would stabilize at 28.1 percent of 2.9(e)(f)). Under the baseline scenario, per capita GDP GDP by 2030 (Figure 2.8(a)). Under the reform scenario in income would reach USD 4,320 by 2050 (in constant 2005 which policymakers would encourage savings and develop US dollars), up from USD 1,230 in 2017. a domestic bond market, capital accumulation would reach 30.7 percent of GDP by 2030. Under the reform scenario, GDP growth would be sustained at around 7 percent until 2040. In this case, Under both scenarios, Cambodia’s high return the capital to labor hour ratio would increase from around to capital is expected in the long run to decline 1 in 2014 to more than 10 by 2050, a level similar to that to the level in other countries. Cambodia’s rate of enjoyed by China today. Aggregate labor productivity would return to capital is currently 29 percent, the highest rise exponentially, from USD 0.57 in 2015 to USD 4 in 2050 among 10 comparator economies in the region. Under (Figure 2.9(c)). Growth in output per worker would hover both scenarios, this would steadily decline to around 20 around 6 percent until 2040 (Figure 2.9(e)). Under this more percent by 2050 (Figure 2.8(b)). Returns to investment in optimistic scenario regarding educational attainment, per machinery, where Cambodia is still lagging significantly capita GDP income would reach USD 6,021 by 2050 (in behind the comparator countries, are likely to decline constant 2005 US dollars). Figure 2.8: Estimated Investment, Savings, and Rate of Return to Capital To foster investment, Cambodia will need to boost domestic savings (a) Investment and Savings (b) Rate of Return to Capital 35% 80% 30% 60% 25% Percent of GDP Percent of GDP 20% 40% 15% 10% 20% 5% 0% 0% 1990 2000 2010 2020 2030 2040 2050 70 80 90 00 10 20 30 40 50 19 19 19 20 20 20 20 20 20 Savings Investment (baseline) Cambodia (Reform) China Hong Kong Investment (reform) Japan Malaysia Singapore Korea Thailand Philippines Vietnam Source: Official sources and World Bank staff calculations. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 40 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS Figure 2.9: Projected Evolution of Key Economic Variables Under Baseline and Reform Scenarios (a) Capital/Labor Hour Ratio (b) Total Factor Productivity 20 0.9 18 0.8 Capital/Labor Hour Ratio Total Factor Productivity 16 14 0.7 12 10 0.6 8 6 0.5 4 0.4 2 0 0.3 1970 1990 2010 2030 2050 1990 2000 2010 2020 2030 2040 2050 Cambodia (baseline) Cambodia (reform) China Baseline Reform Malaysia Thailand Philippines Vietnam (c) Labor Productivity (d) Log of effective labor 4.5 25.5 4.0 25.0 3.5 2005 USD per Hour 3.0 24.5 2.5 Ln(N) 24.0 2.0 1.5 23.5 1.0 23.0 0.5 0 22.5 1990 2000 2010 2020 2030 2040 2050 1990 2000 2010 2020 2030 2040 2050 Baseline Reform Baseline Reform (e) Output per Worker Growth (f) GDP Growth 12% 14% 10% 12% 8% Output per Worker Growth 6% 10% GDP Growth 4% 8% 2% 6% 0% -2% 4% -4% 2% -6% -8% 0% 1990 2000 2010 2020 2030 2040 2050 1990 2000 2010 2020 2030 2040 2050 Baseline Reform Baseline Reform Source: World Bank staff calculations. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 41 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS These two scenarios highlight the need for Cambodian Nations-organized general election in the following year, and policymakers to pursue bold reforms to sustain strong the opening-up of Cambodia to foreign aid and investors growth and quality job creation going forward. Our brought in large inflows of US dollars, which displaced the analysis suggests that Cambodia will need to significantly Khmer riel (KHR) as the primary currency in just a few years. boost investment in physical and human capital to be able These days, the share of foreign currencies in broad money to sustain GDP growth levels at around 7 percent in the long hovers around 80 percent, and the share of US dollar deposits run. Thus, the country’s long-term development goals will to total deposits has remained above 90 percent for the past only be attained if Cambodia is able to increase its capital two decades. The Khmer riel is de facto softly pegged to the intensity and employ a large skilled workforce. According US dollar,6 which provides a nominal anchor for economic to population projections, in order to maintain the current agents and underpins stable prices. It also discourages employment rates, Cambodia will have to create 2.2 million the public sector from resorting to domestic financing and net jobs by 2030 and a total of 4.7 million jobs by 2050. printing excess money (World Bank 2015). Dollarization may Employing a more educated workforce and raising capital have shielded Cambodia from suffering a major currency that will increase the productivity of labor is expected to depreciation such as those experienced by Thailand and increase and sustain economic growth by improving labor Malaysia during the 1998 Asian financial crisis. quality and increasing TFP. While eliminating exchange rate risk has been highly beneficial for the Cambodian economy because of its high rate of exports to the United States, the 2.4 The Current Macroeconomic limitations of dollarization have become apparent now Situation Presents Challenges to Jobs that the European Union has become the country’s main trading partner. The recent appreciation of the US Cambodia has so far been able to maintain strong dollar vis-à-vis the euro and other currencies has negatively growth and macroeconomic stability under impacted Cambodia’s garment exports (Figure 2.10) and dollarization. In the late 1980s and early 1990s, Cambodia tourist trade because it has become more expensive for non- relied on domestic banks to finance its public sector deficits, American tourists to travel to Cambodia. This is negatively which resulted in high inflation and encouraged the use of affecting the pace of job creation in these sectors—which other currencies. The arrival of the United Nations Transitional are the two primary export sectors and sources of good Authority in Cambodia (UNTAC) in 1992 to oversee a United jobs—as well as the prospects of jobs being created around Figure 2.10: Export Growth to the EU and EUR/USD Exchange Rate US dollar appreciation often results in a decline in Cambodian garment exports to the EU 120% 10% 100% Change in EUR/USD Exchange Rate 5% Change in Exports to EU 80% 60% 0% 40% 20% -5% 0% -10% -20% -40% -15% Q2 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q4 Q1 Q2 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Exports to EU, yoy EUR/USD, quarterly change Poly. (Exports to EU, yoy) Poly. (EUR/USD, quarterly change) Source: Official sources and World Bank staff calculations. 6 The currency has been fluctuating between KHR 4,000 and 4,100 per US dollar since 2011. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 42 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS emerging manufacturing clusters. Unlike in other economies growth in construction but also stagnation in manufacturing in the region, this has also led to persistent current account employment. deficits in Cambodia, which tend to be driven by demand for imports of consumption goods, including investments in non- The stability of Cambodia’s macro economy and tradable goods such construction, land, and real estate. the continuation of employment growth are both threatened by this fast credit growth, the associated Dollarization has been fueling capital flows and supply-driven boom in the real estate and construction credit into the non-tradable sector, which is risky for sectors, and the absence of any effective monetary macroeconomic stability, especially in the absence policy tools (Ahmed et al. 2014; IMF 2017). If the of a monetary policy independent of the US dollar.7 construction and real estate bubble should burst, it would Following a sharp decline in construction activity in the not only reduce employment in those sectors but would aftermath of the 2009 global crisis, the amount of credit also, indirectly, put the sustainability of the financial sector extended to construction, real estate, and mortgages at risk—which would result in an economy-wide slowdown. increased from around 4 percent of GDP in 2010 to an These risks are aggravated by the fact that dollarization estimated 13 percent of GDP as of 2015, or around 20 would prevent the effective use of interest rate policy by the percent of total outstanding credit to the private sector. In government, restrain seigniorage gains from printing money contrast, loans to the manufacturing sector have flattened in local currency, and hamper the ability of the Central Bank since 2013, remaining at around 5 percent of GDP. Thus, this to be a lender of last resort in the event of a crisis may dim the prospects for expanding the manufacturing of (World Bank 2015). promising products such as electrical appliances and auto parts, which were expected to yield more stable and higher Gross capital formation, which has traditionally been value-added jobs. low in Cambodia, has also been increasingly directed towards construction. Gross capital formation averaged In 2016 and 2017, under US dollar appreciation, 18 percent of GDP annually between 1993 and 2017 and 20 Cambodia’s external competitiveness declined, which percent between 2011 and 2017. However, this is still below created disincentives to invest in tradable goods. As the average pace of capital accumulation in other fast-growing a result, nominal credit growth in construction, real estate, countries in the region.8 Moreover, part of the increase in and mortgages and in retail trade and personal lending has recent years has been in the construction sector, while the been above 30 percent, while growth in manufacturing has increases in infrastructure and in machinery acquisition have stayed at around 7 percent, which has driven continued job been more modest (Figure 2.11). Figure 2.11: Equipment and Construction in Total Investment in Cambodia Cambodia’s investment in equipment has been lower than in peer countries 30% 25% Percent of GDP 20% 15% 10% Durable Equipment 5% 0% Construction 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Source: World Bank staff estimates using data from the Cambodian authorities. 7 Because Cambodia’s economy is dollarized, it is affected by monetary policy of the United States, leaving Cambodia without all the tools available to conduct its own monetary policy. 8 Countries that have grown at high rates for many decades such as Thailand, Malaysia, Korea, Indonesia, and China have maintained high investment rates since the 1970s, and even as late as the period between 1980 and 2010, they averaged rates of 34, 31, 38, 28, and 40 percent of GDP respectively (Kim et al. 2014). The Growth Commission report (Deblock and Haji 2008) recommended that countries should maintain investment rates of at least 25 percent of GDP or more for sustained strong growth based on the example of successful economies. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 43 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS Cambodia continues to rely heavily on foreign Policy Area 1: Revamp Public Investment investors and donors to finance capital, which leaves Management and Review Tax Incentives it exposed to the risk of external shocks. The gap between national savings and capital formation has doubled, Cambodia has experienced sustained economic increasing from an average of 3.4 percent of GDP between growth that has led to job creation without signi cant 2000 and 2008 to 7 percent of GDP between 2011 and improvements to job quality in the form of labor 2016 (Figure 2.12). The country increasingly relies on FDI productivity. To compensate for slowing job creation, and to cover its current account deficit (around 10 percent of to capitalize on the demographic dividend the country can GDP) and continues to depend heavily on donors to nance expect to experience until 2044, Cambodia will need to make public infrastructure (5 to 6 percent of GDP). Government- substantial investments in its human and physical capital to financed investment has accounted for only a modest 2 increase labor contributions to sustained economic growth. percent of GDP over the past decade (World Bank 2018). Policies aimed at enhancing educational attainment and This means that there is a need for policymakers to skills will need to become a priority to sustain growth by adopt policies aimed at increasing domestic savings and improving the quality and increasing the productivity of labor. developing capital markets. In addition, there is a need to boost capital accumulation by fostering domestic savings and investment, improving public investment management, and streamlining existing tax 2.5 Macroeconomic Policies for incentives to encourage companies to acquire and upgrade machinery. To this end, we recommend two policies: Sustaining Strong Growth and Job Creation (1) Shift from tax holidays to cost-effective investment incentives to boost machinery acquisition and attract Implementing the right macroeconomic policies will higher value-added production processes. The current support strong growth and job creation in the medium incentives that are in place in Cambodia are not achieving to long run. Apart from investments in human capital, national objectives such as supporting innovation, linking which will be discussed in other chapters, Cambodia will the domestic private sector to global value chains, training need to boost public and private investments in physical workers, or focusing on higher value-added sectors. capital and to adopt and maintain monetary and exchange Policymakers should consider moving away from tax holidays rate policy stances that are conductive to growth. and adopting a more targeted incentive system that fosters Figure 2.12: Investment, Savings and Infrastructure in Cambodia Cambodia’s savings-investment gap has widened in recent years, and infrastructure gaps persist (a) Evolution of Investments and Savings (b) Global Competitiveness Report, infrastructure subindices 25% 6 20% 5 Percent of GDP 4 Index Score 15% 3 10% 2 5% 1 0% 0 Cambodia Vietnam Thailand 93 95 97 99 01 03 05 07 09 11 13 15 17 19 19 19 19 20 20 20 20 20 20 20 20 20 Transport Infrastructure Gross Capital Formation Gross Savings Electricity and Phone Infrastructure Source: World Development Indicators (a) and World Economic Forum (2017) (b). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 44 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS private investment in machinery and capacity building, Cambodian policymakers to keep the local currency pegged which would support economic diversification and higher in order to mitigate exchange rate risk. This has so far value-added jobs. The revisiting of the existing Investment resulted in remarkable macroeconomic stability, but during Law is already foreseen in the government’s Industrial periods of US dollar appreciation, it has also resulted in Development Policy. The new incentive might involve the declining competitiveness in Cambodia’s tradable sector following measures: (i) progressively phasing out corporate (garments and tourism), limiting output and the prospects income tax (CIT) holidays or at least limiting them to six for further job creation. Though dollarization has fueled a years; (ii) introducing investment tax credits as an effective credit boom in the construction sector, this has heightened way to lower the cost to firms of acquiring machinery and macro-financial risks in a sector prone to volatility, including equipment; and (iii) replacing VAT exemptions at customs in terms of job creation and destruction. We have two policy with import tax credits to plug loopholes. These policy recommendations in this area: actions can be expected to attract a new generation of FDI associated with higher quality jobs. (1) Re-gain control over exchange rate flexibility and independence of the Government to set monetary (2) Improve public investment and asset management policy to shield jobs from US dollar fluctuations. so that Cambodia’s public sector can take on more Cambodian policymakers might consider devising a market- responsibility for building the country’s public based de-dollarization strategy to progressively take back infrastructure and to depend less on donors, it will more autonomy in setting domestic monetary policy. In the first be necessary to improve public investment and short term, they could focus on gradually promoting the asset management. Cambodia currently has enough use of the local currency. For example they could publicly fiscal space to increase its budget for public infrastructure encourage private sector institutions to pay salaries in projects and to compensate for the recent decline in donor- Khmer riel, or make efforts to improve financial literacy. In funded capital expenditure (World Bank 2018). However, the medium term, policymakers might consider gradually Cambodia currently depends on its development partners increasing the reserve requirements for deposits in dollars to build infrastructure, with the government only funding and bringing down the riel benchmark interest rates to small-scale projects. A recent World Bank assessment of promote borrowing in Cambodia’s national currency. public investment management (Minh Le et al. 2018) found that there are no formal guidelines for feasibility studies and (2) Develop a domestic bond market to both project selection for government-funded projects. Moreover, encourage domestic savings and to progressively just a handful of ministries have the capacity to build needed increase exchange rate flexibility to support job infrastructure, and most of them do not seem to have creation going forward. Investment in the tradable sector minimum quality standards and implementation manuals also needs to be fostered as the combination of large in place. Strengthening public investment management capital inflows with limited investment options has led to (PIM) would not only ensure that such investments are private investment and credits being channeled into the made in high-return projects but also that these projects non-tradable sector, thus fueling the construction and real are implemented more effectively. This would require that estate boom. Intervention is justified by the fact that the the government pass a sub-decree on PIM, develop project construction and real estate sectors usually do not create selection and implementation manuals, and build capacity as much employment as the exporting sector does, and that within core and line ministries. it would also help maintain macro-financial sustainability and prevent bubbles. Developing a domestic debt market (both public and private) and starting to issue sovereign Policy Area 2: Increase the Government’s debt would promote domestic and foreign savings. This Monetary and Exchange Rate Policy would scale up investment and the accumulation of capital Autonomy and Financial Deepening stock, which, in turn, would help sustain Cambodia’s current high economic growth. The development of capital and bond More fundamentally, Cambodia must increase markets would also provide sources of long-term funding for its monetary and exchange rate flexibility, while the corporate sector and establish alternative investments preserving macroeconomic stability, to foster to construction and real estate, thus mitigating some of the balanced growth and employment creation going current macro-financial risks that may put existing jobs forward (Box 2.1). High levels of dollarization have led in jeopardy. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 45 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS Box 2.1: The Peru De-Dollarization Experience and its Policy Implications for Cambodia During its pursuit of de-dollarization since the early 2000s, Peru introduced macroprudential measures such as provisioning requirements for foreign currency loans, higher liquidity requirements for banking liabilities denominated in USD, or reductions in the maximum net open position9 in USD—though not all of them may have been equally effective.10 In recent years, the National Bank of Cambodia has already introduced differentiated reserve and liquidity requirements in USD and KHR; yet dollarization levels have not significantly decreased. One of the reasons Cambodia has not yet experienced a significant decrease in dollarization levels may have to do with the still limited trust in monetary authorities, a legacy of the instability experienced during the 1980s. One way of building institutional trust and further autonomy would be by articulating a medium- term macroeconomic policy framework and establishing a track record of delivering it. Peru, for instance, introduced an inflation targeting regime that helped reducing the exchange rate pass-through in the presence of dollarization (Armas and Grippa 2005; Leiderman et al. 2006). This was accompanied by a progressive increase in exchange rate exibility, which enhanced liquidity management and increased the appeal of holding assets and liabilities in local currency, since they were subject to less volatility than those in USD (Ize and Yeyati 2005). For this shift towards inflation targeting to be successful, significant institutional reforms aimed at significantly enhancing the autonomy, data collection ability, technical capacity, transparency, and accountability of the central bank were needed. Finally, macroprudential measures and a clearer monetary policy framework would need to be complemented by the development of capital markets in local currency. In particular, a domestic bond market with titles denominated in KHR would provide investors with a more diversified set of investment opportunities, increase financing options for both local firms and the government, facilitate hedging against exchange rate swings, and decrease the cost of financing in local currency. In Peru, credit dollarization decreased significantly as private sector bond issuances in local currency increased (Garcia-Escribano 2010). In sum, macroprudential measures per se may not be sufficient to de-dollarize, if not accompanied by a clear monetary policy framework (e.g. inflation targeting), institutional capacity and confidence building, and development of capital markets in local currency. 9 A company, financial institution, or individual have an open position in foreign currency when the liabilities in that currency exceed the assets, thus carrying foreign exchange risk. 10 See Garcia-Escribano (2010) for mote detail. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 46 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS 2.6 References Ahmed, Faisal, Jung Yeon Kim, and Nujin Suphaphiphat. 2014. “What can go wrong? Shifting vulnerabilities as Asia rebalances” in Olaf Unteroberdoerster, ed. Cambodia: Entering a New Phase of Growth. Washington, D.C.: International Monetary Fund. Armas, Adrian and Francesco Grippa. 2005. “Targeting inflation in a Dollarized economy: the Peruvian experience.” IDB Working Papers 1925, Washington, D.C.: Inter-American Development Bank. ASEAN Secretariat and World Bank. 2015. ASEAN Services Integration Report. Jakarta: ASEAN Secretariat; and Washington, D.C.: World Bank. Ciccone, Antonio, and Giovanni Peri. 2006. “Identifying Human-Capital Externalities: Theory with Applications.” Review of Economic Studies 73(2): 381-412. Deblock, Christian, and Rabii Haji. 2008. “Commission on Growth and Development. The Growth Report: Strategies for Sustained Growth and Inclusive Development.” Washington, D.C.: World Bank. Duma, Nombulelo. 2011. “Dollarization in Cambodia: Causes and Policy Implications.” IMF Working Paper No. WP/11/49. Washington, D.C.: International Monetary Fund. Eliste, Paavo and Sergiy Zorya. 2015. Cambodian Agriculture in Transition: Opportunities and Risks. Washington, D.C.: World Bank. Garcia-Escribano, Mercedes. Peru: Drivers of de-dollarization. No. 10-169. Washington, D.C.: International Monetary Fund. IMF—International Monetary Fund. 2017. 2017 Article IV Consultation. IMF Country Report No. 17/325. Washington, D.C.: International Monetary Fund. Ize, Alain, and Eduardo Levy Yeyati. 2005. “Financial De-Dollarization: Is it for Real?” IMF Working Paper No. 05/187. Washington, D.C.: International Monetary Fund. Ize, Alain, and Eduardo Levy Yeyati. 2006. “Financial de-dollarization: is it for real?” In Financial Dollarization London: Palgrave Macmillan. Jimenez, Emmanuel, and Harry Anthony Patrinos. 2008. “Can Cost-Benefit Analysis Guide Education Policy in Developing Countries?” Policy Research Working Paper No. 4568. Washington, D.C.: World Bank. Jorgenson, Dale W., Frank M. Gollop, and Barbara M. Fraumeni. 1987. Productivity and U.S. Economic Growth. Cambridge: Harvard University Press. Kim, Soyoung, Sunghyun H. Kim, and Yoonseok Choi. 2014. “Financing Investment in East Asia: Regional or Global Savings?” Japan and the World Economy 31: 1–7. Lay, Sok Heng, Makoto Kakinaka, and Koji Kotani. 2010. “Exchange Rate Movements in a Dollarized Economy: The Case of Cambodia.” Working Papers EMS_2010_18. Tokyo: Research Institute, International University of Japan. Leiderman, Leonardo, Rodolfo Maindo, and Eric Parrado. 2006. “Inflation Targeting in Dollarized Economies.” Foerder Institute for Economic Research Working Papers 275699. Tel-Aviv: Foeder Institute for Economic Research. Ly, Sodeth. 2016. “Improving Macroeconomic and Financial Resilience: Cambodia Economic Update, April 2016.” Washington, D.C.: World Bank. Minh Le, Tuan, Jay-Hyung Kim, Leah April, and Miguel Eduardo Sánchez-Martín. 2018. “Public Investment Management in Cambodia: Ready for the Next Stage of Development?” Cambodia Public Expenditure Review Discussion Paper. Washington: D.C.: World Bank. Siregar, Reza and Narith Chan. 2014. “Factors behind Foreign Currency Holding by Households in Cambodia.” CAMA Working Papers 2014-58. Canberra: Centre for Applied Macroeconomic Analysis, Crawford School of Public Policy, The Australian National University. World Bank. 2015. “Maintaining High Growth: Cambodia Economic Update, April 2015.” Washington, D.C.: World Bank. World Bank. 2017a. “Cambodia Calling: Maximizing tourism potential” in Cambodia economic update: Cambodia climbing up the manufacturing value chains. Washington, D.C.: World Bank. World Bank. 2017b. “Cambodia: Sustaining Strong Growth for the Benefit of All: Systematic Country Diagnostic.” Washington, D.C.: World Bank. World Economic Forum. 2017. The Global Competitiveness Report 2017-2018. Geneva: World Economic Forum. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 47 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS ANNEX 2.1: Methodology for Economic Growth Projections In this annex, we present a brief description of For simplicity, we assumed that there is no government the modelling approach that we used to reach our in the economy and hence aggregate savings and gross economic growth projections. Fuller details of the investment were equal to each other in every period. Hence, methodology are available upon request from the authors. we assumed that savings (or gross investment) evolved over Our projections in this chapter are derived from the Solow- time as a fraction of output: Swan modelling framework. We assumed that aggregate output is given by the constant returns to scale Cobb- Douglas technology in two production inputs, the physical We denoted the growth in total annual labor hours worked, capital stock and the (quality-adjusted or effective) labor and the change in human capital per unit of labor hour by hours as follows: and respectively. We then derived the law of motion of the economy, which shows how the stock of capital per unit of effective labor evolves over (1) the forecast horizon, as follows: where is total factor productivity (TFP), is the estimated level of human capital per unit of labor hour input, is the total hours worked in the economy in year and where and are the capital the parameter is the income share of capital. and output per unit of effective labor respectively, and We obtained the capital share of income from the APO We used this law of motion equation to forecast Productivity Database 2016 by averaging the shares the growth path of the stock of capital per unit of effective between 2004 and 2014. This turned out to be = 0.51. labor. Physical Capital Growth Path The solution to the above first order differential equation for a Solow-Swan economy under a Cobb-Douglas production We estimated the annual physical capital stock technology can be shown to be: using the perpetual inventory method as follows: where is the gross fixed capital formation (GFCF) in year and is the assumed capital depreciation rate of However, in our projection, the Cambodian economy will not 5 percent used in constructing the entire physical capital have constant TFP and savings (investment) rates or constant stock series. hours employed and human capital per unit of labor growth rates throughout the forecast horizon. Therefore, we needed To construct the initial capital stock, we used the following a formula for the capital path that would permit such changes formula: to occur. The formula is given by: where is the GFCF in year 0 and is the steady state It should be noted that, as long as and are growth rate of past investment estimated from the average not all constant, the steady state of the economy will also growth rate of GFCF in the first five years of the data. The keep changing. variables are all expressed in constant 2005 US dollars using the annual average official exchange rate and the capital deflator. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 48 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS Human Capital Growth Path Identifying Human Capital Externalities To obtain the predicted employment rates of workers The method that we used in this report to estimate aged 15 and over by age and education group, we the magnitude of human capital externalities estimated group-specific employment-to-population is developed by Ciccone and Peri (2006). In this ratios using household survey data and multiplied framework, workers are the only factor of production so them by the projected population of that group.11 We total output is equal to aggregate income. Consider the accounted for changes in labor skills by forecasting changes simplest possible case where there are two types of workers in human capital per unit of labor hour input. Specifically, – high skilled and low skilled . Normalizing total we used data on past educational attainment trends from employment to unity, we have highly-educated workers the Cambodian Socio-Economic Surveys (CSES) from 2007 to 2015 to produce educational attainment probability and less-educated workers. Total income is projections for the Cambodian population in the different thus given by: age groups.12 We computed the growth path of “quality-adjusted” or (2.1) “effective” labor input and the accompanying human where and are the wages of less-educated and capital per unit of labor . The level of human capital per better-educated workers respectively. If the employment of unit of labor or the human capital index that we used to better-educated workers produces positive externalities of compute the historical effective labor input strength on output (equivalent to the effects on from 1993 to 2014 was taken from the Penn World TFP), then their marginal social product exceeds Table, Version 9.0. The projection of Cambodia’s human the high-skill wage premium by : capital index from 2015 to 2050, on the other hand, was constructed in accordance with the approach of Jorgenson et al (1987) as explained in this section. (2.2) Specifically, for our five schooling groups and eleven age groups indexed by , we calculated the growth in effective Ciccone and Peri (2006) showed that externalities can be labor using this formula: expressed as the earnings-weighted average percentage change in wages in response to the change in the supply of highly educated workers. Taking the log change in the average wage and holding skills composition constant yields: where the weight is the average share of total wage bill for age-education group over years -1 and : (2.3) where the bars denote that the number of workers in the We estimated the age-education group wages using CSES two skill groups are held constant. 2014 data, and we assumed that the (relative) wages for Empirical Methodology for the Constant Composition all five schooling groups and eleven age groups between Approach in Identifying Human Capital Externalities 2015 and 2050 would remain at their 2014 level. Hence, the likely effects of technological change on relative wages The counterpart of the expression empirical (skill-biased technological change) are not accounted for in in equation (2.3) for each our growth projection, which means that it is likely that we province can be extended to encompass a wider range have under-estimated the growth in human capital per unit of schooling and experience groups. The change in the labor and effective labor input. 11 We assumed that the employment rate for each age-education group in 2016 (our first year of projection) would be equal to the average rate prevailing in 2013, 2014, and 2015. For the employment rates for the years 2017 and beyond, we used the average employment rates computed over the preceding three years as our predicted rates. We computed the total labor supply (in terms of annual working hours) of these employed workers by taking average weekly working hours (47.64 hours per week calculated from CSES 2015) and assumed it was the same for all age-education groups and remained constant throughout the forecast time period, and then multiplied it by 52 weeks per year and by total employment. 12 We categorized the CSES sample population into five schooling groups and eleven age groups. The schooling categories are: (1) no schooling (NS); (2) primary (PR); (3) lower secondary (LS); (4) upper secondary (US); and (5) post-secondary (PS). The eleven age groups are defined in five-year intervals. We used an ordered probit model to t the historical patterns of changes in educational attainment trends for every age cohort. We then used the fitted model to predict educational attainment probabilities for each age group up until the year 2050. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 49 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS log constant composition average (real) wage for province Empirical Evidence of Human Capital Externalities from over the the time period of interest, which effectively Cambodian Provinces, 2009-2014 captures the change in TFP in a province, is given by: Annex Table 2.1 presents the constant composition approach ordinary least squares (OLS) estimates of externalities from average years of schooling (columns (1) and (2)), average (3.1) years of secondary education (column (3)), and average years This variable serves as our dependent variable in the of secondary and post-secondary levels per worker estimated regression of the log change in TFP on the change in the simultaneously (columns (4) and (5)). All of the reported average schooling attainment of workers at the standard errors are clustered at the regional level. provincial level: The basic specification from equation (3.2) is used in columns (1) and (2), with column (1) including the change in average potential experience over the five-year period (3.2) and an intercept as controls, while column (2) further where the coefficient is the estimate of the strength controls for regional dummies.13 The estimates of the of externalities arising from a one-year increase in the strength of average schooling externalities (or the effect average schooling per worker in province on TFP) range from 0.127 in specification (1) to 0.142 in is a vector of provincial-level control variables, and is specification (2), with both being statistically significant at the error term. conventional levels. Annex Box 2.1: Assumptions of the Model i. Under the baseline scenario, the average number of years of schooling for the Cambodian working age population is projected to increase by 0.113 year per annum between 2016 and 2050. ii. Under the reform scenario, the average number of years of schooling for the Cambodian working age population is projected to increase by 0.139 years per annum between 2016 and 2050. iii. From CSES 2007-2015, we calculated that the actual average change in years of schooling per annum was 0.128 years. We refer to this rate as the reference educational attainment growth rate. iv. We estimated the historical average TFP growth rate at 1.34 percent (discarding the data from 2009 and 2010 because of the global economic crisis). We refer to this rate as the reference TFP growth rate. v. Using the strength of externalities estimate of 13.5 percent for each year’s increase in the average schooling of the population, we calculated that the TFP growth rate for the baseline scenario would average 0.205 percent less than the reference TFP growth rate. We estimated the difference in the average annual increase in years of schooling from the reference educational attainment growth rate (0.113-0.128) multiplied by the 13.5 percent strength of externalities, which equaled 0.205 percent. vi. A similar calculation yielded an average TFP growth rate for the reform scenario that is 0.149 percentage points higher than the reference TFP growth rate. Therefore, in the reform scenario, TFP growth is projected to be 0.354 percentage points faster than in the baseline scenario. 13 Region 1: Stung Treng, Ratanakiri, and Mondulkiri. Region 2: Kampong Cham, Prey Veng, and Svay Rieng. Region 3: Kampong Speu, Takéo, Kampot, Preah Sihanouk, Koh Kong, and Kep. Region 4: Siem Reap, Oddar Meanchey, Preah Vihear, and Kampong Thom. Region 5: Battambang, Pursat, Banteay Meanchey, and Pailin. Region 6: Kampong Chhnang, Kandal, and the greater municipality of Phnom Penh. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 50 TECHNICAL REPORT CHAPTER 2: MACROECONOMIC POLICIES CONDUCIVE FOR CREATING GOOD JOBS Annex Table 2.1 OLS Estimates of Human Capital Externalities from Average Years of Schooling, Years of Secondary Schooling, and Years of Post-Secondary Schooling at the Provincial Level, 2009-2014 (1) (2) (3) (4) (5) – ∆S 0.127* 0.142** (0.059) (0.050) – ∆S SR 0.118 0.117 0.116 (0.070) (0.073) (0.071) – ∆S PS 0.033 0.032 (0.131) (0.138) ∆ Experience -0.005 -0.005 0.002 0.003 (0.053) (0.062) (0.065) (0.066) 2.region -0.174* -0.233 -0.216 -0.212 (0.081) (0.135) (0.188) (0.152) 3.region 0.175** 0.167** 0.172* 0.174*** (0.055) (0.061) (0.076) (0.037) 4.region 0.034 -0.062 -0.046 -0.043 (0.086) (0.088) (0.130) (0.072) 5.region 0.043*** 0.018 0.027 0.027 (0.006) (0.022) (0.051) (0.047) 6.region 0.096* 0.051 0.057 0.059 (0.041) (0.065) (0.084) (0.062) Intercept 0.301*** 0.244*** 0.331*** 0.313** 0.313** (0.027) (0.026) (0.014) (0.081) (0.079) Observations 24 24 24 24 24 R-squared 0.220 0.361 0.280 0.282 0.282 Source: World Bank staff calculations. Note: Robust clustered standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Employing the same control variables as in column (2), from the preceding model. Interestingly, the TFP effect column (3) shows that the estimated effect on TFP of an from an additional year of post-secondary schooling per additional year of secondary education per worker is on the worker is 0.033 and very statistically insignificant (p-value order of 0.118. However, the estimate is not significantly of 0.814). This finding is consistent with the general view different from 0 at conventional statistical levels (p-value of that lower levels of education produce more externalities 0.151), although it is significant at the 10 percent level under than higher levels in less-developed economies (Jimenez a one-tail test. Column (4) reports estimates of the strength and Patrinos 2008). Column (5) eliminates the always of externalities for both secondary and post-secondary insignificant change in the average potential experience levels estimated simultaneously. The TFP effect of an extra variable. Nevertheless, the estimates of externalities from year of secondary education per worker is 0.117 (p-value of secondary and post-secondary education levels remain very 0.169), which is very similar to the estimate that we obtained similar to those shown in column (4). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 51 TECHNICAL REPORT CHAPTER 3: Creating More and Better Jobs through Private Sector Development Johanne Buba and Maheshwor Shrestha 3.1 Introduction Cambodian jobs are in transition. As shown by the previous chapters, a structural transformation is happening to employment. Jobs in agriculture contracted significantly between 2011 and 2015 while they increased substantially in both the apparel and construction sectors. The potential for accelerating this structural transformation lies in the ability of the private sector to generate more jobs and absorb the labor surplus coming from the agriculture sector. In this chapter, we look specifically at the contribution to job creation made by private sector firms, mostly from a micro perspective. The boundaries of the “private sector” will be broadly defined in this chapter as any kind of non-public firm and can range from household enterprises to formal and professionally run firms. The private sector is traditionally defined as the pool of formal and private firms operating within a given country that are formally registered with the authorities, are owned by an individual or a group of associates, and provide wage work. However, in a country like Cambodia, most firms are informal and working under the radar and can be described as informal small and medium-sized enterprises (SMEs). They function like a firm and are managed by dedicated business owners but are not formalized. In addition to these SMEs, many households produce and/or sell items to increase their revenues. We consider that households engaged in such income-generating activities differ from the SMEs in that they are not professionally run, referring to them as household enterprises (HHE). The continuum of business owners can be differentiated by their levels of professionalization, skills, and motivation. From the self-employed worker working from home to the CEO of a large multinational, all are “entrepreneurs.” While their contributions to the economy are different, they all contribute to employment. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 53 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT The objectives of the chapter are to increase our including high costs of doing business, limited access understanding of which firms (as broadly defined in to credit, access to only the small domestic market that this report) contribute to employment and to propose is already dominated by foreign products, and limited ways in which they can generate more, better, and entrepreneurial capacity. more inclusive jobs. First, we focus on the whole private sector, regardless of firm ownership (foreign or domestic), Finally, we turn to household enterprises, which make status (registered or not), or business-orientation (managed an important contribution to the Cambodian economy as a firm or managed as an extension of the household in several ways. Revenues from these enterprises as in household enterprises), and their employees. We represent a large share of the household’s total revenue, describe the contribution of these various firms to the they employ family members and, in some cases, outside creation of jobs, both in terms of their location and their workers, and they constitute a way to train young people. quality. Firms have significantly increased their contribution Household enterprises may be able to play a significant role to job creation, but conditions of employment in the private in absorbing workers shifting away from agriculture or new sector remain precarious. Very few firms provide their workers entering the labor market. employees with benefits (such as contributions to health care costs or pensions), salary levels vary widely, and We used four datasets in this chapter: many employees report having to work for more than 48 • A 2011 census of all enterprises in Cambodia. Some hours per week. The chapter organizes the discussion by key variables are missing such as data on skilled versus separately considering business-oriented firms and non- non-skilled workers, average salary per type of worker, farm household enterprises. investment, exports, and sourcing of materials. In total, the dataset contains information on 505,134 firms Second, we highlight the main challenges faced by (including self-employed workers). the Cambodian economy when it comes to foreign direct investment (FDI). The economy relies heavily for • A between-censuses economic survey (ICES) conducted growth and job creation on FDI firms in the apparel sector, in 2014. This used the same questionnaire as was used which have particularly spurred the development of local in the 2011 census with only minor changes. This survey economies around industrial parks, clusters, and special is a combination of a census (for firms with more than 50 economic zones (SEZs). However, the evidence also shows employees) and a representative survey (for firms with that these firms rarely buy inputs from Cambodian firms, and fewer than 50 employees). In total, the dataset contains that their products are not made for the domestic market 12,178 observations. but rather are directly exported to international markets. We discuss ways to build on this booming FDI apparel sector • World Bank Enterprise Surveys for various countries, to develop a domestic private sector. The FDI-dominated particularly data on the textile and apparel sector. garments industry is a labor-intensive sector with low level The surveys cover 100 countries with at least one of investment in machinery and equipment compared to firm operating in this sector with a total of 11,513 other sectors in Cambodia, and evidence suggests that observations. We also restricted some of our analysis labor productivity in these firms has not been on par with to countries that are key players in the textile and labor costs, negatively affecting the competitiveness of garment industry (large producers, large exporters, and the sector. Therefore, we also discuss how to ensure that Cambodia’s neighboring countries that are investing FDI firms remain in the country while increasing workers’ in the apparel industry), namely Bangladesh, Ethiopia, salaries and improving the quality of their jobs India, Myanmar, Pakistan, Turkey, and Vietnam. In total, there are 3,342 observations for these seven countries. Third, we explore domestic firms. Cambodia has a dual economy, with large FDI firms in export-oriented sectors on • The Cambodia Socio-Economic Survey (CSES) for the one hand and a myriad of micro domestic firms in more 2014 to analyze non-farm household enterprises. This traditional sectors like trade and services on the other. These nationally representative survey collects data from sectors have evolved in different business environments. 21,000 households, 3,800 of which have a non-farm While FDI firms are eligible for fiscal incentives and access enterprise. In addition, in February 2018, the report team to land (through SEZs), the business environment for micro, conducted numerous interviews with firms (including small and medium enterprises (MSMEs) in Cambodia is not household enterprises) of different sizes and operating conducive to growth. Domestic firms face many challenges in different sectors. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 54 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT 3.2 The Private Sector is an Important Wage-earning jobs are concentrated in a Contributor to Job Creation few regions and sectors The private sector represents a large share of jobs. As Most of Cambodia’s firms are located along two main shown in Figure 3.1, 40 percent of jobs in Cambodia are in geographical corridors. The first runs from Cambodia’s the private sector (either formal or informal)1 and the workers border with Thailand to Phnom Penh through the Great Lake, in these jobs consider them as their main occupation. If we and the second goes from Phnom Penh to Sihanoukville, consider non-farm household enterprises (in other words, with a high proportion of jobs located in or around the self-employed workers) as part of the private sector, then capital city (Figure 3.2 and Figure 3.3). Four main factors the proportion of jobs in the private sector goes up to explain this spatial distribution: (i) most industries are 57 percent. A large share of jobs are still in farming located in Phnom Penh or within a one-hour distance from (38 percent). the city; (ii) the temples of Angkor, located north of the Great Lake, attract hundreds of thousands of tourists every year Over the past decade, the number of private sector (5 million in 2016 according to the Tourism Statistics jobs has been a growing. According to various rounds Report); (iii) Battambang is the second largest city in the of the CSES, the number of wage jobs in the private sector country because of its strategic location close to the Great grew by an annual rate of 12 percent between 2007 and Lake, the rice plains, and the Thai border; and (iv) most 2015, while self-employment outside farming grew by imports and exports transit through the port of Sihanoukville 5 percent annually. This exceed the rate of growth of the (in the south). working age population, which increased by 2.2 percent on average over the same time period. Figure 3.1: Jobs Contributed by the Private Sector 9,000 8,000 7,000 6,000 Jobs (‘000) 5,000 4,000 3,000 Wage Earning Jobs 2,000 Non-Farm Self-Employment 1,000 0 Other Employment 2007 2008 2009 2010 2011 2012 2013 2014 2015 Source: Author’s estimates using CSES 2007-2015. The working age population estimate is from UNPOP. Note: These numbers are computed using current employment, defined as the worker’s main activity in the previous seven days. Other types of employment include: public wage work, self-employment in farming, and family members contributing to household enterprises but who are not paid. The wage-earning jobs include wage jobs in firms—regardless of whether they are formal and informal—as well as paid jobs in a household enterprise. 1 To be formally registered, firms are required to get several licenses and undergo many inspections. To be considered as a formal business, newly created enterprises are required to have received all necessary licenses from a wide range of government agencies. The number of licenses and inspections depends on business’s activities, size, and location. These licenses include, for example, a Certificate of Incorporation (Business Registration), a Tourism License, a Food Processing Operating License, Tax Registration for the Paten Tax and VAT Number, a Location Approval License, and Registration for Opening an Enterprise. Besides the high number of procedures, the entrepreneur needs to interact with minimum 4 administrations for the primary licenses and certificates (Ministry of Commerce, General Tax Department, Municipality and the relevant ministry for the business, e.g. Ministry of Tourism) and a minimum two administrations for the other licenses and certificates (Ministry of Labor and Vocational Training and General department of taxation). Depending on the sector in which the business is operating, inspections are required before it can start operating. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 55 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Figure 3.2: Figure 3.3: Number of Firms by District Total Jobs by District (quartile, by job density within the district) Source: Authors’ calculations based on the census and ICES. Note: Data are for 2011 and exclude self-employed workers. “Quartile” is defined by the number of jobs in a district. The “ rst quartile” are the 25 percent of districts that have the largest number of jobs. “Second quartile” are the 25 percent of districts that have the second-highest number of jobs, and so forth. The red lines indicate the two corridors along which most firms and jobs lie. Table 3.1: Wage Workers in Private Firms (Formal and Informal), by Sector (‘000) 2007 2008 2009 2010 2011 2012 2013 2014 2015 Farming 300.7 386.7 507.4 521.1 554.5 560.2 666.4 567.6 662.6 Apparel 250.5 321.1 339.9 421.1 559.0 568.9 683.8 883.2 942.4 Agro-processing 22.0 22.9 23.5 16.1 24.8 41.6 48.9 38.0 37.6 Tourism 36.1 26.5 63.6 92.7 84.3 130.7 149.8 169.5 197.6 Heavy industry 22.7 19.1 28.9 20.0 35.3 36.9 54.3 53.8 73.4 Medium and high tech 0.3 3.1 2.2 0.0 1.3 4.7 14.2 24.6 27.3 Other manufacturing 70.9 133.0 56.7 82.7 77.8 97.0 74.5 113.5 119.2 Commerce 43.8 36.8 36.6 66.8 80.0 72.0 86.6 110.2 123.9 Construction 278.9 260.2 241.9 281.1 268.0 357.8 459.5 539.2 636.2 Mining & utilities 13.8 16.7 19.5 13.5 25.6 60.4 47.1 44.8 42.7 Other services 392.9 421.0 316.8 396.5 463.5 506.7 577.9 639.2 752.1 Total 1,432.5 1,647.1 1,637.1 1,911.6 2,174.1 2,436.9 2,863.2 3,183.5 3,615.0 Source: Author’s estimates using CSES 2007-2015. Note: Agro-processing: manufacturing of food products and beverages. Textile/Apparel: textile, wearing apparel, and leather and related products. Tourism: accommodation and food service activities, travel agencies, tour operators, and creative, arts and entertainment. Heavy industry: coke and refined petroleum products, other non-metallic mineral products, basic metals, and fabricated metal products. Medium and high technology industry: chemicals, basic pharmaceuticals, rubber and plastics, computers, electronic and optical products, electrical equipment, machinery and equipment, motor vehicles, and other transport equipment. Other manufacturing: tobacco products, wood, paper, printing, and reproduction of recorded media, furniture, other manufacturing, and repair and installation of machinery and equipment. Commerce: wholesale and retail trade. Other services: all other services (excluded travel agencies and entertainment). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 56 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT As a result, wage-earning jobs are not distributed of wage jobs in privately owned, non-farm firms). equally across the country. Slightly over a third of wage According to the CSES 2015, their presence is concentrated jobs are located in the Special Region (Phnom Penh and in a few sectors, such as apparel (85 percent of employment surroundings). The Southern regions accounts for another in private firms in this sector), agro-processing (about one- third, while the rest are distributed across the Northern and quarter), and, very recently, medium and high technology Eastern regions, with the Eastern region accounting for only (about three-quarters, though the actual number of jobs 3 percent of wage jobs (CSES 2014).2 is very small). In terms of numbers of firms, data from the census and the ICES show that foreign-owned firms (mostly Wage-paying jobs are concentrated in four main Chinese, Vietnamese, and Korean) are present in the apparel, economic sectors, namely apparel, services, farming, tourism, wholesale and retail, and mining sectors. and construction. The largest sector is the apparel sector (textiles, apparel, and footwear), which employs nearly one- Cambodia has a lower employment-to-growth third of the wage workers, or about 1 million people (full time elasticity than other countries, at both the macro and temporary contracts), followed by services, farming, and and micro levels. As shown in Figure 3.4, which plots construction (Table 3.1). These four sectors account for 80 countries’ employment to GDP elasticity to their GDP levels, percent of all private wage-paying jobs. Cambodia stands below the fitted lines for all countries. Cambodia’s elasticity is also lower than that of more than In terms of the growth of jobs, the apparel sector half of the other countries in the East Paci c Asia region. has flourished since the global economic crisis and The sectoral composition of growth does not seem to be the continues to grow. Construction jobs grew over the main factor explaining this low elasticity since Cambodia five years between 2011 and 2015, mainly driven by real has grown in sectors that are traditionally known for being estate in Phnom Penh and Sihanoukville built in large part labor intensive, like textiles. The results also hold at the to accommodate foreigners who live or do business in micro level. Figure 3.5 plots the average employment the country or who want to invest. However, tourism and elasticity to sales growth against GDP level for all garment agro-processing, which seem promising sectors given the firms in each Enterprise Survey that contains a large share country’s natural endowments, represent only a small share of garment firms across a range of countries. Here again, of wage employment. A few other sectors have picked up Cambodia appears at the bottom of the distribution with recently, such as equipment and electronics (medium and a lower employment elasticity-to-sales growth than other high technology manufacturing). countries at a similar level of economic development. These findings suggest that Cambodian firms could be generating During the period 2010-2015, foreign-owned firms more jobs given the country’s level of development as well accounted for one-third of wage jobs (and 46 percent as the sectoral composition of growth. Figure 3.4: Employment Elasticity to GDP Across Countries 3 Employment−to−GDP Elasticity 2 1 Rest of the World East Asia 0 Cambodia 95% Con dence Interval -1 Fitted values 15 20 25 30 35 Log of GDP Source: Authors’ calculations using the World Bank’s World Development Indicators. 2 The Eastern region is less populous than the rest of the country but lags other regions in terms of the wage employment rate as well. Similarly, the Phnom Penh Special Region leads other regions in terms of wage employment rate. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 57 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Figure 3.5: Employment Elasticity to Sales Growth for Apparel Firms 8 Employment−to−Sales Growth Elasticity 6 4 2 95% Confidence Interval 0 Fitted values 0 5000 10000 50000 GDP per capital Source: World Bank Enterprise Surveys (WBES). Note: Only firms operating in the textile, garments, and leather sectors are included. The sample is comprised of countries where the World Bank Enterprise Survey has been fielded. Figure 3.6: Monthly Wages in Private Firms, 2014 2.0 Number of Jobs (million) 1.5 1.0 0.5 Apparel sector 0 minimum wage 0 1,000,000 2,000,000 3,000,000 Monthly Wages (KHR) Source: Author’s estimates using CSES 2007-2015. Figure 3.7: Number of Hours Worked in Main Wage Job, 2014 0.15 0.10 Density 0.05 0 0 20 40 60 80 100 Weekly Hours Source: Author’s estimates using CSES 2007-2015. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 58 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Not all jobs in the wage-earning private workers even though they still consider this job as their sector are of equal quality main occupation. Not all jobs offer similar wages. Figure 3.6 shows The wage differences can partly be explained by the the distribution of wages across all jobs in Cambodia. differences in skills between workers, as well as The orange dotted line indicates the minimum wage as by Cambodia’s segmented labor market. Figure 3.8 established in the apparel sector. Most of the jobs provide illustrates the distribution of monthly wage earnings of men salaries at or around this minimum wage. A quarter of all and women separately by sector and level of education. workers receive a monthly salary below or equal to this The wage distribution shifts slightly to the right for higher minimum wage. The upper tail of the distribution is much levels of education, particularly upon completing secondary more dispersed. education (see Chapter 5 for more details). However, the figure also indicates that wage earnings are very Nor do all jobs offer similar working conditions. The dispersed within each level of education and particularly labor law sets the number of working hours at eight hours for lower levels of education. For women with less than a a day for a maximum of six days or 48 hours a week. primary education, hourly wages are significantly higher in However, more than 60 percent of wage workers reported the garment and construction sectors than elsewhere. In working more than these legal working hours. Some jobs fact, hourly wages in the garments sector are no higher for are also more precarious than others. As many as 18 more educated women or for men at any education levels percent of wage workers have not been working for their and, in many cases, are statistically lower (see Chapter 5 company for a full year and are categorized as temporary for more detail). Figure 3.8: Distribution of Monthly Wages by Educational Achievement (a) Women Monthly Wage Distribution (b) Men Monthly Wage Distribution Garments Garments Agro-processing Agro-processing < Primary < Primary Tourism Tourism Manufacturing Manufacturing Construction Construction Other services Other services Garments Garments Agro-processing Agro-processing < Secondary < Secondary Tourism Tourism Manufacturing Manufacturing Construction Construction Other services Other services Secondary School Graduate Secondary School Graduate Garments Garments Agro-processing Agro-processing Tourism Tourism Manufacturing Manufacturing Construction Construction Other services Other services Garments Garments Agro-processing Agro-processing College + College + Tourism Tourism Manufacturing Manufacturing Construction Construction Other services Other services 11 12 13 14 15 11 12 13 14 15 Log (monthly wage, 2014 KHR) Log (monthly wage, 2014 KHR) Source: Author’s estimates using CSES 2014. Note: The sample is restricted to education-industry cells with more than 30 observations. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 59 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT There are a few possible reasons why wages are not income each month, up to eight times their monthly salary. equal across sectors for workers with the lowest level Third, a job may be considered less desirable because of education. First, certain occupations, for example those the employer is not reliable in paying wages. Fourth, the in construction or garment manufacturing, may require opportunity for personal growth within a firm can affect a speci c skills that take time and effort to acquire and that job desirability. For instance, the tourism industry seems may not be directly captured by education. Second, some to offer a career path for each low-skilled occupation occupations impose non-monetary costs, such as having whereas opportunities for advancement on a production to work away from home or not having flexibility in terms floor are more limited (for example, there is usually only of hours. Third, some occupations are perceived as being one supervisor for every 15 to 20 workers). Finally, a risky, precarious, or hazardous and thus can command a key factor is the working conditions associated with a compensating wage. Fourth, mobility barriers may prevent job, which seem to be quite difficult in the garment and workers from moving to places with a greater concentration construction sectors for instance. of higher-paying jobs. Because wages may not fully reflect the desirability Household Enterprises are the source of a and/or quality of a job, other dimensions need to be substantial share of households’ revenues considered. From the qualitative interviews with firms and employment undertaken for this study, we found that the salary for an entry-level job requiring zero or little education is the Household enterprises are ubiquitous, albeit very same across sectors: about USD 100 per month (except for heterogenous, in Cambodia. This business category trainees who are paid USD 25 to 45 per month). However, encompasses a young entrepreneur starting a high-tech other factors play a role. First, the job may provide other company in Phnom Penh, a tuk-tuk driver in a tourist zone, benefits. The garment sector provides workers with a tailor in a tiny shop with three sewing machines and two allowances for transportation and accommodation, and apprentices, and a woman operating a retail shop. Based other sectors, such as hotels, often provide two or three on the questionnaire provided in the CSES, we define meals a day plus free accommodation. Second, for workers a household enterprise as any “non-farm business or in the tourism sector, for example, service charges (such enterprise run by a household” regardless of whether they as tips) can account for a significant amount of their are entrepreneurs by opportunity or by necessity.3 Figure 3.9: Prevalence of Non-farm Household Enterprises in Cambodia Share of households with non-farm enterprise < 15% 15% - 20% 20% - 25% 25% - 35% 35%+ Source: Author’s estimates using CSES 2014. 3 The literature on the nature of entrepreneurship tends to discuss only two types of entrepreneurs (Djankov et al, 2006; Schoar, 2010; De Mel et al. 2010; Bruhn, 2013; and Calderon et al., 2016), but we prefer to talk about a continuum of entrepreneurs. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 60 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Household enterprises constitute a large part of storage. Women hardly ever own household enterprises in jobs picture, particularly in regions of Cambodia this sector. Manufacturing enterprises account for about 14 where there is a large formal private sector. Overall, percent of the enterprises, are slightly more common in rural 29 percent of the households in the country have a (non- areas than in urban areas, and are slightly more likely to have farm) household enterprise, with 51 percent being in urban a male owner than a female owner. The accommodation and areas and 23 percent in rural areas. However, among food sectors are the sectors where enterprises are most likely all household enterprises, over 60 percent are located to be female-owned. in rural areas where 80 percent of the population still lives. Households in the eastern provinces (Stung Treng, Non-farm household enterprises create jobs (both Ratanakiri, Kratie and Mondulkiri) are the least likely to paid and unpaid) through self-employment and operate a household business (19 percent) while those in the participation of family members. About half of the Special Region (Phnom Penh, Kandal, and Kampong them have at least one other household member involved Chhnang) are most likely to do so (40 percent). Households besides the owner, although this varies by sector. About 60 in the other regions are 20 to 30 percent likely to operate percent of household enterprises in the retail trade have a household enterprise. The eastern region contains more than one household member involved compared with almost no private sector activity, in terms of either firms or only 5 percent of enterprises in the transport and storage household enterprises. Most household enterprises seem to sector. Overall, female-owned enterprises are 12 percent be located not far from the two main economic corridors, more likely to have at least one other household member which suggests a link between those household enterprises working in the enterprise than male-owned enterprises. and economic density potentially led by the larger firms. However, in the retail, manufacturing and food sectors, men are more likely than women to have additional household Almost half (47 percent) of household enterprises are members working in the enterprise. Rural enterprises are in retail trade. Retail shops account for almost half of urban slightly more likely than urban enterprises to have other enterprises and for two-thirds of the enterprises operated family members involved, except in other trades and by women. About 14 percent of all household enterprises, accommodation and food sectors. Involving other household and about a fifth of urban enterprises, are in transport and members in the enterprise is a way to train younger family Figure 3.10: Number of Household Members Working in Household Enterprises 60% 60% 60% Male Owner Rural Female Owner Urban 50% 50% 50% 40% 40% 40% Percent of Household 30% 30% 30% 20% 20% 20% 10% 10% 10% 0% 0% 0% 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 Number of Working Household Members Source: Author’s estimates using CSES 2014. Note: These figures depict the number of household members employed in addition to the enterprise owner. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 61 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT members in conducting the household’s trade or the retail contribute to the development of other firms in the operation. During our qualitative interviews, many household country by buying products from local firms or sourcing enterprises also reported having trained other people from local products (such as trims, fabric, yarn, and leather). their villages. In some cases, these informal apprenticeships In Cambodia’s garment sector, only about 25 percent of were also a source of revenue for household enterprises as inputs and supplies are outsourced locally as compared families pay a small fee to them in return for training their to 60 percent in Vietnam, 62 percent in Bangladesh, and child in the business for a couple of months. nearly 100 percent in India. Apparel firms in Cambodia import products from neighboring countries, assemble the Household enterprises can also create jobs for garments in Cambodia, and reship the final products to the individuals outside of the family, but these numbers US and European markets. In addition, “locally sourced” are small. According to the CSES, only 6 percent of does not necessarily mean “sourced by Cambodian firms” household enterprises hire external workers.4 Enterprises but rather sourced by firms located in Cambodia, which can in the service sector, non-retail trades, and manufacturing include those that are internationally owned. This suggests sector hire the most external workers (over 10 percent). that the locally sourced trims and packaging used by Among household enterprises in retail trade (the most garment FDI firms actually came from foreign-owned firms common type of household enterprise), the probability of that operate in Cambodia. hiring an external worker is only 4 percent. These numbers are in line with other international experiences. Evidence suggests that there is an increase in the production of unregistered sub-contracting factories. According to an ILO bulletin published in May 2017 (ILO 3.3 Foreign Direct Investment Also 2017), there is a widening gap between the number of Contributes to Domestic Job Creation garment firms registered with the National Social Security Fund (NSSF), a database of all garment firms that employ While FDI firms have created jobs directly, they have workers, and the number of garment-exporting factories not had the same effect in terms of creating jobs registered with the Ministry of Commerce. The number indirectly because they have not built links with in the NSSF database exceeded that in the Ministry of local suppliers or purchasers. In the apparel sector Commerce database by 84 in 2014, 106 in 2015, and 244 where FDI firms predominate, there are very few backward in 2016. If we assume that the national market represents and forward links to the local economy, especially when a tiny share of garment sales, then this difference likely compared to other countries. In most cases (for example, reflects the increasing number of subcontracted garment Ethiopia, Morocco, and Turkey), garment multinationals enterprises selling to exporting garment factories. Figure 3.11: Percent of Garment Firms’ Inputs Sourced Locally 100% 80% Percent of Inputs 60% 40% 20% 0% Cambodia Bangladesh Ethiopia India Myanmar Nicaragua Pakistan Philippines Turkey Vietnam Source: World Bank Enterprise Surveys Note: Apparel includes garment, textiles, and leather. 4 Because the questionnaire used in the CSES does not ask how many workers are hired by the enterprise but only their labor costs, we assumed that the household enterprises had hired a worker when we observed a positive wage bill. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 62 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Firms can also have an indirect effect in terms were associated with 0.1 percent more households of promoting other local activities and household operating a non-farm enterprise. There also exists a enterprises. Using 2014 data, we tested the assumption correlation between the FDI sectors and type of household that the presence of FDI firms could generate spillover enterprises nearby. Higher garment employment was effects in the local economy, using agglomeration effects correlated with a higher incidence of household enterprises to proxy the creation of new household enterprises in the in the trade and transport sector, whereas higher employment same province as FDI firms. As we have already seen, many in domestic manufacturing seemed to encourage the household enterprises exist in the areas where large number creation of household enterprises in retail shops and of factories are located.5 We found a strong correlation manufacturing to provide the domestic manufacturing firms between wage employment in major sectors and the with inputs. Similarly, high rates of employment in tourism prevalence of household enterprises.6 For example, areas appeared to promote retail shops and enterprises in the with 1 percent more than average employment in garments hospitality sector.7 Figure 3.12: Correlation between Wage Employment Rates and the Number of Non-farm Enterprises Garments and Footwear Non-Garment Domestic Manufacture Non-Garment Foreign 12 12 12 10 10 10 8 8 8 Log (# HH enterprise) 6 6 6 0 1 2 3 4 5 0 1 2 3 0 1 2 3 4 Domestic Tourism Domestic Construction 12 12 10 10 8 8 Log (# enterprise) in Province x rural/urban Linear fit 6 6 0 1 2 3 4 0 1 2 3 4 Log (wage employment) in Province x rural/urban Source: Author’s estimates from CSES 2014. Note: The regression line is weighted by the population of the province x rural/urban cell. The size of the circle represents the population in the cell. 5 The same is likely true for SMEs, though we did not test for this relationship in this analysis 6 The regressions represented in the gure and the top panel of the table are: where is the rural/urban areas within each province, is the estimated wage employment, and is the estimated number of households with an enterprise. The relationship was estimated for year 2014. Full regression results are presented in Annex Table 3.3. 7 Of course, it is also plausible that the correlations are driven by localities with better investment opportunities or other attractive characteristics drawing both wage firms as well as household enterprises. To investigate this, we examined whether the growth in wage jobs in these locations was correlated with growth in household enterprises. The positive correlation remained and was statistically significant for the effect of wage jobs in domestic manufacturing, construction, and tourism. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 63 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT The cheap and abundant labor force is both availability as the main reason, with 42 and 39 percent citing a strength and weakness to Cambodia tax incentives and tariff incentives respectively as the second and third reasons. Cheap labor costs in Cambodia have attracted investments into labor-intensive sectors. China, Apparel firms in Cambodia seem to have fewer xed Vietnam, and other Asian countries have historically produced assets than other types of firms and so, in principle, low-quality and low sophistication products. However, could easily move. Figure 3.13 shows the level of fixed because of rising salaries in these countries, multinational assets (land, buildings, and equipment) in firms by sector and Asian companies have increasingly been looking for new based on data from the Census and the ICES on firms destinations for labor-intensive industries. Combined with a that reported having an income statement (i.e. formal and stable government and incentives for FDI, Cambodia’s cheap relatively large and organized firms). As can be seen in the labor has attracted many assembly firms, predominantly in figure, the garment sector has fewer fixed assets than other apparel and, more recently, electronics. The 2016 Investment sectors do. The potential downside for workers employed in a Motivation Survey in Cambodia asked investors for the top labor-intensive industry like apparel is that labor is the main reasons why they were investing in Cambodia, and all of production factor and is easily substitutable, whether through them mentioned the country’s low labor costs and labor mechanization or through relocation of the firm. Figure 3.13: Value of Fixed Assets per Full-time Permanent Worker, 2010 200,000 Value per Worker (USD) 150,000 100,000 50,000 0 log - m rin l Tra sale ilit g el on g Ot ing- es tu ria no ng Ut inin sin ar ris vic cti y fac ust r de ies g ch ri pp r he le es u Te factu tu M tru er To ho nu Ind s/A oc fac rS ns W pr tile nu nu he Co ro d Ma Ma Ot an x Ag Te Ma Source: Authors’ calculations based on the Census and the ICES. Note: Self-employed workers are excluded. Only firms reporting an income statement are included. Box 3.1: Minimum Wage in the Garment Sector Cambodia has a minimum wage scheme that is only applicable to the garment and footwear sector, the two largest manufacturing exports of the country. The country’s minimum wage legislation has been in effect since 1997 and has been revised multiple times since then (see ILO 2016 for a detailed review). The regular minimum wage was set at USD 40 per month in 1997 and had been revised eight times by January 2018 when the monthly minimum wage was set at USD 170. Although the nominal minimum wage has increased by over 380 percent in the past 19 years, the real value of the minimum wage was below its 1997 real value till the change in January 2015. Besides the minimum wage set by the government, employers in the garment and footwear sector are also required to pay mandatory allowances to their workers. These mandatory allowances include: (i) USD 7 per month for transportation and accommodation; (ii) attendance bonus of USD 10 every month; (iii) USD 0.50 per day for an overtime meal; and (iv) seniority bonus of USD 2 per month in their second and subsequent years of employment. See Chapter 1 and Shrestha (2018) for further discussion on the impacts of minimum wage hikes during 2008 and 2015. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 64 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT The relatively low productivity rates of Cambodian labor productivity (Box 3.1). This may result in a loss workers counteract the potential benefits of “cheap of competitiveness for FDI firms, which may result in labor.” Low labor costs alone do not attract foreign investors. bankruptcies and/or offshoring. It is the combination of low labor costs and high labor productivity that attracts them as investors need value for their The lack of competitiveness in the garment sector, money. Compared to other countries that have a large garment the most important export sector and job provider in sector, are exporters of apparel, or are entering international Cambodia, is alarming. In this mature sector, marginal markets, Cambodia offers comparable salaries to other key productivity gains are now costly and time consuming, and players, like Bangladesh or Ethiopia (Figure 3.14(a)). In firms may find it difficult to change fast enough to cope with other words, these salaries are quite low by international the increase in wages or to find ways of staying competitive standards. However, with regard to labor productivity in the production of low-quality and low-sophistication levels, Cambodia has the highest unit labor costs (as a products. Our interviews with firm managers in the private percentage of labor costs per value-added) (Figure 3.14(b)). sector indicated that, in order to adjust to rising costs, some This suggests that, although Cambodia has relatively low apparel firms have recently closed down. Some garment wages, it is the least competitive among the selected assembly firms may prefer to relocate to other countries comparator countries.8 where labor is cheaper. Firms that remain in Cambodia might try various coping mechanisms to compensate for higher In recent years, the minimum wage in the garment labor costs, such as mechanization, optimized industrial industry has been increased incrementally and processes, or providing more incentives for workers to frequently, which has led to an overall increase in be more productive. However, given the slow pace of wages in garment firms. Because of the workforce’s adjustments, it is likely that most firms will not adequately low productivity, FDI firms have not been able to fully upgrade to higher value-added products to survive in the compensate for this increase in labor costs by increasing international market. Figure 3.14: Average Wages and Unit Labor Costs in Apparel Firms (a) Average Wages (b) Unit Labor Costs 8,000 80% 70% 6,000 60% Percent of Value Added USD (Base 2010) 50% 4,000 40% 30% 2,000 20% 10% 0 0% ia ia ia Vie y ia y h ca r es am h r es am pia pia Pa a a Ph tan tan ma ma e ke gu gu es es od Ind od Ind rk pin pin hio hio tn tn r kis kis lad lad ra ra an an mb Tu mb Tu Vie ilip ilip ca Et Et Pa My My ng ng Ca Ca Ni Ni Ph Ba Ba Source: World Bank Enterprise Surveys Note: Apparel includes garments, textiles, and leather. 8 Other reasons that explain why investors come to Cambodia are the lack of restrictions on investment for FDIs, Cambodia’s trade agreements with the US and European markets, and low (or non-existent) import and export duties for exporting firms. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 65 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT 3.4 The Domestic Private Sector Figure 3.15 also indicates that Cambodian firms seem to be adding jobs more slowly than do domestic firms Could Create Many More Jobs, but is in other countries such as the US, India, Mexico or Struggling To Do So even Vietnam. For each age category of firms, Cambodian firms are smaller;9 most domestic firms in Cambodia are Domestic firms have the potential to generate more small and remain small. jobs and reduce inequalities across regions. While domestic firms outnumber FDI firms, they do not contribute Cambodia has a lot of young firms, a feature that as many jobs as foreign-owned firms do. Domestic firms is currently an untapped potential for job creation. average only eight employees compared with 124 in foreign- Compared with other developing countries, Cambodia has owned firms. After controlling for other factors like location, one of the highest shares of firms under five years old at sector, and year, foreign firms are still twice the size of nearly 60 percent of all firms (Figure 3.16). These firms are domestic firms. Nevertheless, domestic firms are more evenly already a significant source of jobs, but they find it difficult to distributed across provinces (except in the eastern part of grow (see the next section for a discussion of the constraints the country) while foreign-owned firms are mostly located in faced by young firms) or create even more jobs. Phnom Penh and the surrounding provinces. Figure 3.15: Lifecycle of Firms in Cambodia and Selected Comparator Countries 8 7 Average Employment US - manufacturing (Firm Age <5=1) 6 5 Mexico - manufacturing 4 India - manufacturing 3 Vietnam 2 Cambodia - manufacturing 1 Cambodia - all firms 0 Cambodia - rms above <5 5-9 10-14 15-19 20-24 25-29 30-34 35-39 >=40 10 employees firm age Source: Author’s calculations for Cambodia. Hsieh and Klenow (2014) for India, Mexico, and USA Note: The base of this calculation is firms that have been operating for fewer than five years. Figure 3.16: Share of Young Firms in the Economy 80% 60% 40% Share of Firms 20% that are Young Share of Employment 0% by Young Firms tan Vie la Ca am Za a ia rk da Mo e rra vo Ug e o te an ng e za sh bo a e Pa ru ut ay ca iqu i Ca ldov on Af fas Ba Ivoir rd go od mb Pe Sie oso So agu Mo ade fri Bu an Co nist tn kis Ve Le An mb mb hA ina K ji r l a d Ta gh Source: Various Censuses. 9 Based on the lifecycle of firms as described by Hsieh and Klenow (2014) we looked at the average firm’s size as the firm ages. We used comparable methodology using cross-sectional data Census and the ICES. This allowed us to compare cohorts’ growth between different types of firms and to benchmark Cambodia against other economies. We used three series, one with all firms, another for manufacturing (to compare Cambodia with the US, India, and Mexico), and the third being manufacturing firms with 10 or more employees. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 66 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT To generate more jobs, domestic firms need to grow in Cambodia by providing duty free imports and exports, their sales. The firm census data suggest that the higher access to land, and a one-stop-shop for processing the a firm’s sales, the greater the number of people it employs required documentation. According to the CDC, most firms (Figure 3.17).10 This section presents how the investment in these SEZs are foreign-owned. climate impedes sales growth by domestic firms. It then explores other barriers to sales growth that are faced by Cambodia’s convoluted and restrictive business domestic firms, ranging from access to inputs to worker regulations prevent domestic firms from thriving. On skills shortages. the World Bank’s Doing Business Index for 2018, Cambodia ranked 135th out of 190 countries in terms of the ease of doing business in the country. Also, it ranked 183rd and Domestic firms face an investment climate 179th respectively for ease of “starting a business” and not conducive to sales growth “dealing with construction permits.” This makes it very difficult for all domestic firms but particularly SMEs not only The main barrier to increased sales by domestic to grow but simply to function. firms—and thus potential job growth—is that the de facto business environment for FDI firms is much better than the environment in which domestic A range of other constraints impede sales firms have to operate. Cambodia offers attractive growth and job creation incentives such as large tax exemptions and investment guarantees for large investments (above USD 2 million), Subjective assessments by managers of domestic which are called Qualified Investment Projects (QIP). SMEs report that political instability and the skill- Most beneficiaries are in the garment, footwear, other level of the workforce are the key constraints to manufacturing, infrastructure, agri-business, mining, growing their sales and their workforce. The World and other services sectors. In the garment sector, 383 Bank Enterprise Survey interviews with private sector garment factories were given QIP status by the Council for managers in Cambodia showed that both SMEs and large the Development of Cambodia (CDC) between 2010 and domestic firms perceive their main constraints to sales 2014. According to data from the Cambodian Investment growth as being political instability and an inadequately Board database, the average fixed asset for each garment educated workforce. There are, however, some notable factory is USD 5.05 million while the average registered differences between firms of different sizes. Unlike in most capital is USD 1.6 million. This level of investment in other countries, small firms in Cambodia do not tend to be as setting up a business is much larger than is possible for concerned about access to finance as large and, to a certain a typical domestic SME. In addition, FDI firms have easy extent, medium-sized firms are. In contrast, large firms are access to SEZs, which ease the cost of doing business far more concerned about labor regulations (Figure 3.18). Figure 3.17: Domestic Firm Employment and Sales, 2010 8 6 Employment (log) 4 2 95 % Con dence interval 0 Real sales 0 5 10 15 Real Sales (log) Source: Authors’ calculations based on the Census and the ICES. 10 Output growth rates tend to lead to employment growth rates. Haltiwanger et al. (2016) tested the correlation between lagged variables and concluded that lagged output growth is associated with higher net employment growth in the current period but the reverse relationship (lagged employment growth associated with output growth in the current period) does not hold. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 67 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Objective measures suggest a broader set of considerably higher for large domestic firms (64.6 percent). constraints to sales and employment growth. To put Additionally, there is very little support available for managers the subjective perceptions reported in Figure 3.18 into of domestic businesses. What little help is available comes context, we use objective measures from the Enterprise from a handful of NGOs that operate at the local and micro Surveys to compare the severity of constraints to sales level. Unlike in other countries, there is a limited number of growth in Cambodia, in its comparator countries, and across consulting firms in Cambodia, and government institutions different groups of Cambodian firms.11 First, Cambodia fares do not play an active role in providing or funding business poorly compared to other countries in getting credit (Figure development services. Hence, in most parts of the country, 3.19(a)). Only 25 percent of Cambodian firms reported entrepreneurs are left to develop their business entirely on having an overdraft facility or a loan from a financial their own. However, the private sector is starting to fill the institution, a proportion far below all comparator countries gap. There is currently a surge in small-scale structures in except Myanmar. The problem seems to be accentuated Phnom Penh designed to foster entrepreneurship such as for firms that have been operating for less than five years co-working spaces, saving groups, and incubators. and for small and large firms. Surprisingly we did not find Third, the cost of registering a business, which opens a “missing middle” as in most countries, wherein medium- access to new markets and sources of financing, is sized firms are the underserved segment. However, fewer disproportionately high in Cambodia. Given the high than 40 percent of medium-sized firms had any access costs involved in formalizing, it is not surprising that only 6 to funding, which is still very low compared to in other percent of small firms have done so. The costs of registering comparator countries. The problem appears to be similar seem to be higher in Cambodia than in comparator countries, for both household enterprises and micro-firms according and small firms may not see the advantage of formalizing. to CSES data. This is the case even though micro- nance However, there is a threshold effect. Above certain levels institutions are well developed in Cambodia. In 2014, one- of sales and employment, it becomes difficult for firms to third of household enterprises had an outstanding loan but remain informal. As a result, 64 percent of medium-sized only 5 percent reported using this loan for their business. firms and 97 percent of large firms are registered while Hence, when it comes to borrowing for investment and hardly any micro firms are. business, the penetration of micro-loans is very limited. Fourth, the private and public dialogue is under- Second, firms seem to have limited entrepreneurial developed. Chambers of Commerce exist in many provinces skills. Business practices within firms are not very and cities and could support entrepreneurs in various ways, sophisticated. For example, most business owners do for example, by providing them with information about not separate their personal accounts from their business government regulations and laws, training programs, and account or track their expenses and revenues. The Census advocacy. However, beyond Phnom Penh and a handful of indicated that, in 2013, almost no micro and small domestic large cities (like Siem Reap), Chambers of Commerce are firms had a balance sheet or an income statement. The poorly structured and organized. They also usually exclude proportion of firms that had these key documents was slightly SMEs, which makes it difficult for these firms to raise issues higher for medium-sized domestic firms (3.9 percent) and with relevant government officials. Figure 3.18: Biggest Perceived Obstacles to Sales Growth Among Cambodian Firms, by Size Inadequately educated workforce Access to finance Political instability Labor regulations Access to finance Political instability Inadequately educated workforce Practices of informal firms Transport Large Inadequately educated workforce Political instability Medium Practices of informal firms Small 0% 5% 10% 15% 20% 25% 30% 35% Percent of Firm Size Group Source: WB Enterprise Surveys. 11 The Enterprise Surveys include subjective questions about managers’ perceptions of the investment climate. However, these variables may not be comparable across firms, regions, or countries. In addition, various biases have been shown to affect subjective questions in ways that might affect the reliability of regression coefficients. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 68 TECHNICAL REPORT Percent of Firms Percent of Firms Percent of Firms Percent of Firms 0% 10% 20% 30% 40% 0% 20% 40% 60% 80% 0% 20% 40% 60% 80% Ca Ca Ca Ca 0% 20% 40% 60% 80% 100% m m m m Ba bo Ba bo Ba bo Ba bo ng dia ng dia ng dia ng dia lad lad lad lad Figure 3.19: Eth esh Eth esh Eth esh Eth esh Gu iop Gu iop Gu iop Gu iop ate ia ate ia ate ia ate ia ma ma ma ma la la la la Financial Institution My India My India My India My India a a a a Ni nm Ni nm Ni nm Ni nm ca ar ca ar ca ar ca ar ra ra ra ra (c) Firms with a Website Pa gua Pa gua Pa gua Pa gua Ph kist Ph kist Ph kist Ph kist ilip an ilip an ilip an ilip an pin pin pin pin e e e e as a Major or Severe Constraint Tu s Tu s Tu s Tu s Source: World Bank Enterprise Surveys. Vie rkey Vie rkey Vie rkey Vie rkey tna tna tna tna m m m m S S S S Me mall Me mall Me mall Me mall diu diu diu diu m m m m La La La La 1− rge 1− rge 1− rge 1− rge 5 5 5 5 (g) Inputs and materials being sourced locally 6− yrs 6− yrs 6− yrs 6− yrs 9 9 9 9 >1 yrs >1 yrs >1 yrs >1 yrs 0 0 0 0 y y y y Fo rs Fo rs Fo rs Fo rs (a) Firms with an Overdraft Facility or a Loan from a rei rei rei r Do gn Do gn Do gn Do eign me me me me Objective Measures of Possible Barriers to Sales Growth sti sti sti sti c c c c (e) Firms Reporting an Inadequately Educated Workforce Percent of Firms Number of Years Percent of Firms 0 5 10 15 0% 20% 40% 60% 0% 05% 10% 15% 20% 25% Ca Ca Ca m m m Ba bo Ba bo Ba bo ng dia ng dia ng dia lad lad lad Eth esh Eth esh Et esh Gu iop Gu iop Gu hiop ate ia ate ia ate ia ma ma Severe Constraint ma la la la My India My India My India an an an Ni m Ni m Ni m ca ar ca ar ca ar ra ra ra Foreign-Owned Company Pa gua Pa gua Pa gua Ph kist Ph kist Ph kist ilip an ilip an ilip an pin pin pin e e e All countries Tu s Tu s Tu s Vie rkey Vie rkey Vie rkey tna tna tna m m m Sm Sm Sm Me all Me all Me all Cambodia − ownership diu diu diu m m m Cambodia − age of firms Cambodia − size of firms La La (d) Schooling Years of Production Workers La 1− rge 1− rge 1− rge 5 5 5 (b) Firms using a Technology Licensed from a 6− yrs 6− yrs 6− yrs 9 9 9 >1 yrs >1 yrs >1 yrs 0 0 0 y y y (f) Firms Reporting Labor Regulations as a Major or Fo rs Fo rs Fo rs r r rei Do eign Do eign Do gn me me me TECHNICAL REPORT CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW sti sti c c sti c CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT 69 CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Fifth, entrepreneurs have limited technical capacity small (Table 3.2). The Cambodia market is among the smallest to develop or adopt new technologies or to find these in the region with a small population, coupled with a low GDP technologies abroad. Two facts stand out. First, during per capita (at purchasing power parity). Interviews suggest our interviews with firm managers, many of them reported that this national market is also flooded with products having difficulty accessing the know-how on technologies from the three large neighboring countries: China, Vietnam or new business processes. With very few technologies and Thailand. developed in the country, firms need to import most of them. As shown in Figure 3.19(b), more than 15 percent Eighth, Cambodia’s workforce is low skilled. A quarter of firms reported using a technology licensed from a of all Cambodian firms have reported that the inadequately foreign-owned company. If this percentage seems low, it skilled Cambodian workforce is a major or severe constraint, is nevertheless higher than in most comparator countries. which is more than the percentages in most comparator Older firms in Cambodia seem to have easier access to these countries (Figure 3.19(e)). This issue seems to particularly technologies than young firms. This may hinder the ability of affect medium-sized and large firms and foreign firms. young firms to innovate and generate in-house knowledge and technologies. Second, Figure 3.19(c) compares the Domestic firms have also reported having difficulty proportions of firms with a website across comparator retaining workers, in particular in the northern provinces. countries and groups of firms. This is a good indicator of the Our interviews with the private sector indicated that way in which firms are reaching out to customers. Among domestic firms have to compete for workers with labor all of the countries, Cambodia has the lowest proportion of markets abroad. We did not have enough data to calculate firms with a website except for Myanmar. turnover in domestic firms, but the CSES data showed a large proportion of household members working abroad. For Sixth, domestic and smaller firms have little access example, 15 percent and 32 percent of wage workers living to inputs. Cambodia, Guatemala, and Nicaragua are at the in households in the north and north-west are employed bottom of the distribution of comparator countries when by a firm abroad, although this share is zero in Phnom it comes to using inputs that are locally sourced (Figure Penh and around 2 to 3 percent in the rest of the country. 3.19(g)). However, there are huge differences between The explanation is not that firms abroad offer better companies of different sizes and between domestic and contracts or living conditions, but that they offer higher foreign-owned firms. Foreign firms are not at all integrated wages (Figure 3.20). in the local economy and import the bulk of their inputs and intermediate materials from abroad. These numbers There are also some specific skills that, according must be interpreted with caution: foreign firms may directly to private sector managers, are rare in the country import from abroad, but it is likely that smaller firms also and thereby limit the growth of the private sector. For use imported inputs but bought from local retailers and instance, manufacturing firms reported that they need more wholesalers. engineers and skilled and managerial staff. The tourism sector reported a lack of Chinese speakers in Cambodia, Seventh, firms also struggle to have access to a market meaning that hotels are poorly equipped to welcome Chinese and hence to be profitable. The national market is very tourists who are coming to Cambodia in large numbers. Figure 3.20: Wages in Cambodia by Type of Firm, 2014 800,000 Average Wage (KHR) 600,000 400,000 200,000 0 Domestic Foreign abroad Foreign Source: Author’s estimates from CSES 2014. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 70 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Table 3.2: Size of Regional Markets Population (million) GDP/Capita (PPP) Cambodia 15.8 $2,347 Vietnam 92.7 $4,051 Malaysia 31.2 $20,426 Lao 6.8 $3,945 Thailand 68.9 $12,573 Philippines 103.3 $5,392 Source: WDI (2017). 3.5 Household Enterprises are An Household Enterprises offer good economic Attractive Alternative to Many opportunities for some groups As discussed earlier, household enterprises are an important On average, there is a slight financial advantage source of employment in Cambodia and benefit from to operating a household enterprise over being increased employment in other sectors of the economy. employed in Cambodia. As illustrated in Figure 3.21, Households may start an enterprise as a flexible means monthly earnings for the average household enterprise to complement household income from other sources or, (defined as profits for the main person operating the for a limited few, as a way towards expanding this into a household enterprise) are 11 percent higher than are the larger business. In the rest of this section, we examine the average wages for an employee. These differences, which characteristics of these enterprises and explore determinants are statistically significant, persist even after controlling for of household enterprise formation and productivity, as well gender and locality, and the results hold in both rural and as the determinants of them hiring an external worker—a urban areas. Is this enough to conclude that there are more sign of business expansion. opportunities available in operating a household enterprise than in working for a wage? Figure 3.21: Earnings for Household Enterprise Workers and Wage Workers (a) Monthly Earnings (b) Hourly Earnings 1 1 Smoothened density Smoothened density .8 .8 .6 .6 .4 .4 .2 .2 0 0 10 11 12 13 14 15 16 3 4 5 6 7 8 Log (monthly earnings) in all areas Log (hourly productivity) in all areas Wage worker Household enterprise worker Source: Author’s estimates using CSES 2014. Note: In (a), monthly earnings for household enterprise workers consist of the total profit made by the household enterprise that accrues to the person mainly responsible for the enterprise. In (b), productivity for wage workers consists of monthly wage earnings per hours worked. Productivity for household enterprise workers consists of the share of profits accruing to the worker for every hour worked in the enterprise. Only those whose main occupations are wage work or household enterprise work were used in the estimation. The size of the total sample of household enterprise workers is 3,800 and that of wage earners is 13,200. The top and bottom 0.5 percent of the distribution have been removed to avoid outliers. Dotted lines represent average earnings. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 71 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT The benefits of working in a household enterprise are operator is earning much less than the lowest earning not just monetary. Although average hourly earnings are wage worker is. For instance, over 37 percent of household lower in household enterprises than in wage work, there enterprise operators make profits that are lower than the are other benefits that come with operating a household garment sector’s monthly minimum wage level, whereas enterprise. For instance, our interviews with household only 30 percent of the private sector wage employees do enterprise operators in Cambodia indicated that running so. This may reflect the greater exibility that individuals a household enterprise gives the operator more flexibility have in operating their household enterprises but may also than a wage job, which is particularly appealing to women reflect the higher risks and uncertainty involved in operating who have disproportionately more household chores and an enterprise compared with having a wage job. child care responsibilities. We found that individuals with an infant child in a household, particularly women and those in The profits from a household enterprise are rarely the rural areas, were more likely than those without a child to only source of revenue for the household, particularly operate a household enterprise. Presumably, this gave them in rural areas. Although 29 percent of Cambodian the opportunity to schedule their own working hours to fit households have at least one household enterprise, this them in around their other household responsibilities. is rarely the household’s only source of employment. Only a third of them (10 percent of all households) have no However, operating a household enterprise involves adult members who are employed outside the household working more hours than a wage job while facing enterprise. The share of households relying solely on uncertain success. We found that hourly earnings are 11 household enterprise income is much higher among urban percent lower for household business owners than for wage households (20 percent) compared to rural households workers, which suggests that the owners are compensating (8 percent) (Figure 3.22). In urban areas, households for their lower hourly productivity by working longer hours. typically mix non-farm household enterprises with private Also, there is a wider range of monthly and hourly earnings wage employment, whereas in rural areas, households mix among household enterprises than among wage workers non-farm enterprise employment with agriculture or wage (Figure 3.21). The lowest earning household enterprises employment in agriculture. Figure 3.22: Composition of Household Employment in Cambodia 100% 80% Percent of Households 60% Only Non-Farm Enterprise (NFE) NFE + Ag. SE NFE + Private Wage 40% NFE + Other Only Agriculture Self-Employed (Ag. SE) 20% Ag. SE + Private Wage Ag. SE + Other Only Private Wage Private Wage + Public 0% All Female Head Male Head Rural Urban Source: Author’s estimates from CSES 2014. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 72 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT A household enterprise’s sector of business, Education, gender, and location are the three main education of the operator, and size of the market are drivers of job creation in household enterprises. With the strongest determinants of their earnings. With regard to education, as can be seen in Figure 3.23, if the regard to the sector of business, we found that household operator of a household enterprise has a higher education, enterprises in the retail trade had higher productivity than the probability of their employing a wage worker in the those in other sectors. With regard to the education of household enterprise increases. One year of additional the operator, each additional year of schooling increases education increases this probability by 0.5 percentage points hourly earnings in non-farm enterprises by 3 percent. In (which is about 7 percent of the average probability of hiring rural areas, however, education increases the chance of a wage worker) after controlling for relevant covariates.12 having a household business but does not significantly With regard to gender, women are about 4 percent less increase returns. Another strong determinant is the likely than men to hire an outside worker in both rural and amount of community wealth in local markets. Operating urban areas. This gender difference in rural areas is driven by an enterprise in richer communities translates into higher different choices of industries, but in urban areas, it persists productivity, suggesting that local demand for the goods even after controlling for industries. With regard to location, and services produced by these household enterprises urban enterprises are more likely to hire workers than rural increases with the wealth of the community in which the enterprises. This holds across all sectors, except transport enterprise is located. and storage. Enterprises in wealthier communities tend to have a higher probability of hiring an outside worker than those located in poorer communities, particularly if they are operated by a female owner. Figure 3.23: The Role Played by Education in the Decision to Hire an External Worker in a Household Enterprise All Male Female .15 .15 .15 .10 .10 .10 .05 .05 .05 Probability of Hiring a Wage-worker 0 0 0 0 2 4 6 8 10 12 0 2 4 6 8 10 12 0 2 4 6 8 10 12 Rural Urban .15 .15 .10 .10 Mean at each percentile .05 .05 Local−linear regression 95% Con dence Interval 0 0 0 2 4 6 8 10 12 0 2 4 6 8 10 12 Household Enterprise Owner's Years of Completed Schooling Source: Author’s estimates from CSES 2014. Note: An enterprise was considered to have hired a worker if they had a positive wage bill in the previous 12 months. The plots do not control for other covariates. 12 The full regression estimate is shown in Annex Table 2. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 73 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Table 3.3: Probability of Hiring a Wage Employee Gender of Owner Locality All Male Female Urban Rural Sectors: Manufacturing 10.4% 13.7% 6.4% 19.6% 8.5% Retail trade 3.7% 7.1% 2.5% 5.6% 2.4% Other trade 13.7% 14.8% 10.5% 16.7% 12.4% Transport & storage 3.7% 3.7% 0.0% 2.1% 5.5% Accommodation & food 3.7% 9.2% 2.4% 5.8% 2.1% Service 14.2% 9.9% 18.3% 13.2% 14.9% Others 11.1% 12.1% 9.0% 10.2% 11.7% All 6.3% 8.8% 4.1% 7.2% 5.7% Source: Author’s estimates using CSES 2014. Note: An enterprise was considered to have hired a worker if they had a positive wage bill in the previous 12 months. Household enterprises provide an attractive income- services than would be the case in poorer communities. generating option for those who are unable to Physical remoteness is negatively correlated with households participate in the wider labor market. Being a woman, operating enterprises. Having access to credit (for example, married, a minority, or having a child in the family increases having a bank in the village) and having internet access the chance of operating a household enterprise. Women are both correlated with a higher probability of starting and and ethnic minorities have historically found it difficult to expanding household enterprises, so encouraging the spread access a wide range of economic opportunities, while being of credit services and internet access throughout the country married and having children often limits the flexibility with is likely to foster the formation of more household enterprises which anyone can engage productively in the economy. in Cambodia. Household enterprises provide these individuals with an option to generate income. In rural areas where economic Evidence from impact evaluations in other countries opportunities are scarce, and in the case of women with suggests that any intervention by governments, household responsibilities, household enterprises provide donors, or NGOs designed to support household an opportunity for individuals with higher levels of education enterprises tends to result in increasing revenues for to make use of their skills and ability. In urban areas, where the operators (i.e. better jobs for them) rather than in economic opportunities abound, more educated individuals the creation of additional jobs. Self-employed workers tend to choose to move away from household enterprises or microenterprises usually face a wide range of challenges, and towards wage jobs with steadier, and potentially higher, but interventions tend to focus only on one or two of these. incomes.13 Our interviews with household enterprise operators in Cambodia suggest that they face the following barriers to Improving the business and economic environment growth: (i) a lack of technical skills; (ii) limited access to and increasing access to markets may stimulate information and technologies; (iii) a lack of business skills; the creation of household enterprises. Conditional on (iv) a lack of motivation; and (v) limited access to markets individual and community characteristics, those living in for their products because of their inability to compete with wealthier communities have a higher probability of operating Thai or Vietnamese products. Therefore, it will be important a household enterprise (see Annex Table 3.3 for the estimation to ensure that any future interventions are based on solid details). This could be a result of these communities having evidence of what is needed to help these household a higher local demand for locally produced goods and enterprises not only to grow but also to create job. 13 Based on the Heckman Selection model regressions reported in Annex 3.1. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 74 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT 3.6 Policies for More and Better which employ almost 1 million people, have contributed substantially to employment. These industries have also Private Sector Job Creation driven the development of a local economy (proxied by the increasing number of household enterprises in these In this chapter, we have explored the contribution that the areas). In recent years, there has also been a surge in sub- various parts of the private sector make to the employment contracting factories, which constitutes a growing domestic of Cambodia’s workers and have concluded that FDI firms, private sector in garments, though this is mostly informal. domestic SMEs, and household enterprises will all be critical to creating more, better, and more inclusive jobs in However, the downside of this FDI industry is that Cambodia in the future. its comparative advantage in low-value segments of the international garment market lies in labor costs, Our analysis has shown that Cambodia not only which are rising. The minimum wage has been increased relies heavily on private sector FDI firms for current incrementally and frequently which has led to an overall employment but will also continue to rely on them to increase in wages in garment firms. Firms have not been create jobs in the long run. While these FDI firms benefit able to fully compensate for this increase in labor costs with from a conducive business environment in Cambodia, they an increase in labor productivity. This may result in a loss face increasing challenges when it comes to matching their of competitiveness and consequently in bankruptcies and labor costs with labor productivity. offshoring. Meanwhile, domestic firms are hampered in their Therefore, we make the following recommendations efforts to increase sales, and thus create jobs, by a to continue attracting FDI firms to participate in the range of factors. These include an investment climate Cambodian economy: that is not favorable to domestic producers, a high cost of doing business, limited entrepreneurial capacity, limited (1) Transition to higher value-added segments of access to and information on technology, limited access to GVCs. See Chapter 4 for speci c recommendations markets (both for imports and exports), and an unskilled in this area. This transition will need to be carefully and and unstable workforce. Policymakers will need to address appropriately supported. Existing garment firms will need these constraints to sales growth before it will be possible to have time to adjust by mechanizing, optimizing their for most domestic firms to create more jobs. industrial processes, or developing higher value-added products. The non-farm household enterprise sector is often overlooked, but collectively these enterprises have (2) Increase the skills of the workforce: See Chapter the potential to make a substantial contribution to 5. A continuous effort needs to be made to attract more creating employment and increasing the revenue of labor-intensive and foreign investment to Cambodia, such Cambodia’s households. as the production of higher value-added garment or leather products or assembly for electronics. However, this will In this section, we focus on what policies can require policymakers to make additional investments in the support the growth and job creation of these three human capital of the Cambodian population to build the different segments of the private sector in the future. necessary pool of semi-skilled and skilled workers. Specifically, we make three recommendations: (i) continue efforts to attract and accommodate job-friendly FDIs; (ii) (3) Encourage domestic firms to enter the apparel create a more conducive business environment for domestic sector. It will be necessary to build a stronger domestic firms; and (iii) development of the necessary ecosystem for sector in the most labor-intensive segments (apparel, non-farm household enterprises. footwear, and electronics) while FDI firms transition into higher value-added products. For example, policymakers Policy Area 1: Continue Efforts to Attract could consider giving further support to the small but Job-Friendly FDI Firms significant existing pool of sub-contracting apparel factories in the form of technical assistance, financial FDI firms have been attracted to Cambodia by its support, or tax incentives to help them to grow and cheap and abundant workforce. This workforce has become more sophisticated. In time, these firms may primarily attracted labor intensive industries in low-value see the benefits in formalizing their status (as, currently, segments of the garment industry and with low levels most of them are informal). This support will enable them of fixed assets. Assembly factories in these industries, to create not only more and better jobs (since these firms CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 75 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT do not currently come under the aegis of the Better (1) Reduce the costs of doing business for domestic Factories).14 They could also help keep the low value-added firms. To achieve this aim, we recommend: (i) reducing segment of the apparel GVC in Cambodia. the fees involved in starting a business, considering implementing a flat fee that reflects the cost of the service, and introducing a single platform for business registration Policy Area 2: Create a More Conducive by integrating the current registration processes at the Business Environment for Domestic Firms Ministry of Commerce, the General Department of Taxation, and the Ministry of Labor; (ii) finalizing the issuing of There are many young and small domestic firms land titles in the country and developing an integrated in Cambodia, but they do not contribute as many online registry system for land that includes ownership jobs as do foreign-owned firms. Since they are much and cadaster information; (iii) reviewing the existing case smaller and less dynamic, they do not grow in terms management system in the courts for resolving commercial of revenues, profits, and jobs as much as FDI firms or disputes and considering developing an electronic case equivalent firms in other countries. However, with the management system for judges and lawyers; and (iv) right support and a conducive business environment, they establishing an insolvency administration profession and have the potential to generate more jobs nationwide, as increasing the capacity of the public and private sector recognized in Cambodia’s Industrial Development Policy related to insolvency matters. 15 for 2015-2025. (2) Develop an ecosystem for domestic firms to Domestic firms in Cambodia suffer from limited access to finance, poor entrepreneurial skills, a increase their business skills and knowledge. At the lack of information about up-to-date equipment and moment, Cambodia’s nascent ecosystem for supporting potential suppliers, and an absence of supporting business growth only exists in the Phnom Penh, and it needs structures such as business associations, consulting to be deepened and extended to the rest of the country firms, and support programs. They also face immense (Box 3.2). First, more structured institutions are needed pressure from foreign-owned competitors that have lower to facilitate a public-private dialogue where private sector production costs and larger economies of scale. As a result, companies can raise their concerns and discuss possible they cannot compete in the national or, consequentially, the solutions. There have been some successful examples of international market. this in other countries. For example, in Vietnam, the Private Sector Forum has helped to eliminate a dual pricing system Since the main comparative advantage of Cambodia that was discriminating against foreign-owned companies is its cheap and abundant workforce, domestic firms in favor of domestic firms. Second, the government needs to could potentially invest in labor-intensive sectors provide more support for entrepreneurship and innovation, of the economy. However, even there, domestic firms for example through research centers and incubators, struggle to hire and retain workers because they cannot which are usually privately run and provide co-working afford to offer the same salaries as do the FDI apparel spaces and marketing and other services to entrepreneurs. firms in Phnom Penh or in Thailand. Furthermore, during The public funding provided to these structures could be our qualitative interviews with FDI firm managers, it became made conditional upon their results, for example, based clear that some FDI firms may provide additional non- on the number of entrepreneurs ready for the next stage financial incentives to workers, such as a medical center of investment in the case of incubators, or the number of within the facility. new technologies being developed by research institutes. Policymakers could foster the development of privately- Therefore, we make the following recommendations owned consulting firms to provide support to domestic for creating a more conducive business environment enterprises.16 The government could also foster the for domestic firms in the Cambodian economy: development of a consulting sector by starting a public- 14 Better Factories was created in 2001 as a partnership between the International Labour Organization (ILO) and the International Finance Corporation (IFC). The program aims at improving working conditions in the garment industry in Cambodia. betterwork.org/where-we-work/cambodia/ 15 See the World Bank (2018) for details. 16 Impact evaluations carried out in Mexico (Bruhn 2013) and India (Bloom et al. 2013) showed that providing one-to-one and personalized support to firms of a certain size can contribute to job creation. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 76 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT funded project that would aim to link domestic firms with to link domestic suppliers with existing and nascent FDI- consulting firms. The project could first provide technical dominated value chains, such as apparel and electronics, assistance to consulting firms to make sure their service is particularly existing sub-contracting apparel companies. of a suf ciently high quality. It could then partially subsidize Second, incubators could play a role in nurturing early- the cost of providing the services to SMEs while advertising stage startups and in identifying startups that need further the benefits of using a consulting firm around the country. financing, which could be provided through the Job Fund. Finally, the government should build a website to provide Third, in a more long-term approach, the government online resources for entrepreneurs, including some massive could consider lowering the level of investment needed open online courses (MOOC), some key documents such for domestic firms to be eligible for fiscal incentives. As of as a template for a business plan or a spreadsheet for now, very few domestic firms in Cambodia are able to meet preparing financial projections, and lists of consulting firms, the requirements to be eligible for being a QIP. The fees information centers, and incubators. involved are also quite high and the application process (3) Introduce a grant program and fiscal incentives to quite burdensome. According to the Investor Motivation increase the access of domestic firms to financing. Survey carried out by the IFC in 2015, 72 percent of the First, the government should consider setting up a Job QIP firms surveyed reported that they had hired a broker Fund to provide domestic firms that have high job creation or another third party to process their QIP application. potential with technical assistance, legal support, and It is estimated that it costs about USD 12,000 for firms matching grants (see Box 3.3). One aim could be to help in the garment and footwear industry and approximately several small firms (for example, in agro-processing) to USD 20,000 for those in the other manufacturing sector consolidate to more effectively compete on the national to process their investment incentives package. This is a (and possibly international) market. Another aim might be clear barrier to the entry of SMEs. Box 3.2: The Importance of Good Business Practices for Sales Growth and Employment The productivity and performance of any firm is strongly correlated with the sophistication of its business practices, i.e. financial management, marketing and communication strategies, human resources management, inventory management, etc. Bloom and Van Reenen (2007) collected data on the management practices of 732 medium-sized firms in four high-income countries and showed that good managerial practices are strongly correlated with firm-level productivity and profitability. In a subsequent study (Bloom and Van Reenen 2010), the authors increased their geographical coverage to developing countries. They surveyed 6,000 firms in 17 developing countries, including China, India, and Brazil, and came up with similar results as their findings for developed countries. In a recent paper, McKenzie and Woodruff (2017) investigated whether this conclusion held for micro and small firms operating in developing countries. Looking at seven developing countries, they found that variations in business practices explained as much as the variation in the outcomes of microenterprises as in those of larger enterprises. They also found that good business practices predict higher firm survival rates and faster sales growth. Some firms are better managed than others, but the question is why. Bloom and Reenen (2010) observed that two factors explain most of the differences between national average management scores: the level of competition in the country and who owns the firm. More intense competition in the segment where the firm operates is positively correlated with the use of better business practices. Moreover, firms that are family-owned and managed by the eldest adult child have worse than average business practice scores. Bloom et al. (2016) complemented these conclusions by analyzing data from over 30,000 US plants and found four main drivers of management practices: product market competition, the country’s business environment, learning spillovers, and human capital. These drivers explained 30 to 50 percent of the total variation among firms in their management practices. Other studies have explored the role played by the entrepreneur’s characteristics in explaining variations in management practices. Bertrand and Schoar (2003) studied top executives in 600 firms (including CEOs, CFOs, and Presidents) who had managed at least two firms for three years each during their 30-year sample period (1969 to 1999). The authors found that the individual manager fixed effects (particularly for CEOs) had significant explanatory power over firms’ returns on their assets. Adding a fixed effect for the entrepreneur raised the adjusted R-squared from 0.72 to 0.77. These results reflected performance differences that could be explained by the attributes of the entrepreneurs themselves. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 77 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT (4) Provide more incentives to young people to become Policy Area 3: Develop the Necessary entrepreneurs. Being an entrepreneur is a risky activity, Ecosystem for Non-farm Household especially in Cambodia where the business environment Enterprises is quite challenging and the chances of domestic firms succeeding are limited. Because entrepreneurs everywhere Household enterprises represent a large share of need extra support to reduce the risks involved in launching economic activity in Cambodia. Twenty-nine percent of their own businesses, some countries have provided the households in the country have a non-farm household them with some fiscal (and non-fiscal) incentives. For enterprise with 1.4 million Cambodians considering their example, France and Germany both offer a continuation of enterprise as their main activity or a complementary activity unemployment benefits for a certain period of time while even when they work in the wage sector. The enterprises individuals try out their new business ideas. In France, the discussed in this subsection exist at the subsistence end of enactment of this policy resulted in a surge of new firms the enterprise spectrum. Those with high potential for growth being created and many unemployed workers going into were included in the previous section. self-employment (Hombert et al. 2014). Policymakers should consider the best ways to provide such incentives Most household enterprises do not generate wage to entrepreneurs in Cambodia, with one possibility being to jobs. Nevertheless, they generate revenues for the business provide them with a stipend to cover their basic needs for owners and their families who would otherwise be out of the 18 months. labor market and, hence, they help to reduce poverty. The objective of our recommendations, therefore, is to increase (5) Make it easier for domestic firms to hire and retain the revenues of these non-farm household enterprises. workers. This might involve developing and strengthening job matching platforms through, among others, the National Many initiatives (Box 3.4) and impact evaluations Employment Agency (see Chapter 5 for recommendations) and around the developing world have shown that subsidizing some of the basic benefits (such as health care interventions focusing on those micro enterprises coverage and a retirement pension) that are usually provided can enhance the living conditions for the households to workers by their employers. This could be accompanied that operate them (Table 3.4). by an information campaign informing domestic firms about the importance of providing these benefits to workers and informing workers about their legal rights. Box 3.3: Strengthening the Entrepreneurial Ecosystem—The Jobs Fund in South Africa Launched in 2011, the Jobs Fund is designed to support job creation in South Africa. One of its windows (enterprise development) aims at fostering jobs through small businesses while building a sustainable ecosystem for firm support. Rather than allocating grants to small businesses, the Fund works supplies funding to existing structures, NGOs or private firms to leverage their existing capacity and scale up their services to small businesses. One example of an incubator project funded through the Job Fund is the Awethu Project, whose objective was to support 500 entrepreneurs and to create an additional 1,000 jobs (awethuproject.co.za). The target group of this incubator is high-potential entrepreneurs from vulnerable groups. This incubator provides a face-to-face program for 6 months after screening of the business idea. It also developed a mobile application for virtual incubation with training in basic business concepts, business tools and networking. Other examples of Fund recipients include: the Cape Craft Design Institute that provides support to 24 growth orientated companies in craft (improvement of production processes, certification and quality assurance, research and development into materials etc.); Hot Dog Café, that aims to support young unemployed black South Africans in starting a Hot Dog Café franchise; and Shanduka Black Umbrellas, an incubator for businesses with the potential to grow and generate at least four jobs per incubatee. The selection of the projects is based on the number of potential jobs being created; the grants are disbursed conditional upon jobs being created. So far, the Fund has supported about 71 projects under this window with an average grant per project of about 2 million USD. As per an independent evaluation, this window created 84 percent of the total jobs generated by the Fund. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 78 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Box 3.4: YouWin! In Nigeria: Generating Jobs through a Business Plan Competition that Identifies a Pool of High-Growth Firms The Youth Enterprise with Innovation in Nigeria (YouWin!) program is a business plan competition for young entrepreneurs launched in 2011 by the Nigerian Ministry of Finance. To be eligible for the program, applicants had to be Nigerian citizens aged 40 or younger and propose, via a concept note, the creation of a new or expansion of an existing business venture within Nigeria. In 2012, the top 6,000 out of 23,888 applicants were selected for a four-day business plan training course. 4,510 business plan applications were received and scored, and the 1,200 most innovative winners were selected to receive prizes averaging USD 50,000 each. The grant was given in four tranches conditional upon a close monitoring. McKenzie (2017) indicates that the program was able to identify high-growth firms, address their constraint on finance through a grant, and as a result, accelerate the growth of the winners. The author found that conditional on reaching the semi-finalist stage of the business plan competition, there seems to be no good predictor for high-growth firms. The total business plan score indeed does not appear as a significant predictor for any of the success outcomes for new and existing businesses. The selection process was more important than the business plan itself. Indeed, the selection process was quite intense: candidates were asked to fill out a concept note on Excel and register online; complete an in-class training program and submit a very detailed business plan online. As a result, applicants were older and more educated on average. Three follow-up surveys were done in 2012, 2013 and 2014 which enabled the tracking of results over time. Three years after applying, new firm applicant winners were 37 percentage points more likely than the control group to be operating a business and 23 percentage points more likely to have a firm with 10 or more workers (relative to a control mean of 11 percent), while existing firm winners were 20 percentage points more likely to have survived, and 21 percentage points more likely to have a firm with 10 or more workers (relative to a control mean of 17 percent). The winners are also innovating more and are earning higher sales and profits. The latest follow-up survey conducted in 2016 finds that the program continues to have significant impacts three years after the last tranche was received, despite the economic crisis in Nigeria. 17 Based on a rigorous impact evaluation, it was possible to attribute jobs to the program: 2,500 in 2012, 6,800 in 2013, 7,000 in 2014 and 4,200 in 2016. Given the program budget, the cost per job is about 2,300 dollars over 5 years. 17 In 2016, Nigeria suffered its worst economic performance in thirty years, driven by a contraction in the oil sector which is the main export and accounts for 70 percent of government revenues. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 79 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Table 3.4: Interventions Targeting Self-employed Workers and Micro-enterprises Possible Barriers Support Example Capacity/ Cognitive ability - insufficient technical Technical skills training Blattman et al. (2014) in Ghana Information skills Insufficient non-cognitive skills Life skills training Graduation programs: Banerjee et al. (2015) Lack of entrepreneurial mindset/ Personal initiative training Campos et al. (2017) in Togo. Lafortune et motivation Alumni talks al (2017) in Chile Consulting in the choice of the market Pre-creation support None that has been rigorously evaluated Basic business practices Basic business training Drexler et al. (2014) in the Dominican (financial literacy, business Republic; Lafortune et al. (2017) in Chile; plan), peer-to-peer and/or Valdivia (2013) in Peru; and Brooks et al. mentoring (2016) in Kenya Access to Limited access to finance due to See above - nance lack of capacity Limited access to finance due to Grants and micro-credit McKenzie and Woodruff (2008) in Mexico; friction on the financial markets De Mel et al. (2008) in Sri Lanka; and Fafchamps et al. (2011) in Ghana Access to Coordination failure – asymmetry Information provided (for None specific to firms market of information example, on prices) Coordination failure – matching Intermediary to train and Atkin et al. (2017) in Egypt frictions carry out quality control Source: Buba (2018). Therefore, we make the following recommendations may lack motivation and self-confidence, these websites for developing the necessary ecosystem for nonfarm could also include online training programs designed to household enterprises: develop an entrepreneurial mindset. Other ways to develop this outlook in household business owners have been tried (1) Generate more knowledge about household successfully in other countries, including personal initiative enterprises. The available data provides a general training in Togo, Mexico, and Ghana. description of who these households are but not enough detail to guide policymaking. The data lacks information on (2) Link household enterprises to markets. Many employees in household enterprises (their wages, skills, household enterprises and small firms rely on their personal and relationship to the owner), their work conditions, or networks to supply them with inputs. Often, this means the number of apprentices working for the enterprise. contacts in cities (Phnom Penh), and this distance can be Furthermore, it provides little information into the a significant barrier to opening or expanding enterprises, aspirations, motivations, and skills of these households, particularly in rural and remote areas. For example, small thus yielding little insight into the areas where policy could shop-owners in the villages find it costly to travel to the city improve the situation. Therefore, there is a need to generate to buy the limited amount of goods that they need for their more information about Cambodia’s household enterprises individual shops, while large wholesalers in cities find it hard through more in-depth studies. Despite the data limitations, to know what to sock to supply these rural retail shops. it is clear that household enterprises lack the basic minimum Therefore, policymakers should investigate ways (including of information and skills needed for doing business, including IT solutions) to help household enterprises to aggregate their basic financial management information, basic knowledge demand to make it worthwhile for large suppliers to engage on marketing, and technical skills. Therefore, providing them productively with them (Box 3.6). with simple business training should help them to increase their revenues. Another option is to promote innovative IT (3) Reduce the cost of credit in rural areas. Households solutions (such as cell-phone applications and websites) in villages with a bank are more likely to open a non-farm that deliver basic training on these skills and provide a household enterprise than those located in a community toolkit for better management practices, which could be a without a bank. This suggests that increasing access to cost-effective ways of improving the productivity of these credit and reducing the cost of credit can facilitate the enterprises (Box 3.5). Since household business owners creation of household enterprises to meet local demand. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 80 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Box 3.5: Using Mobile Phones to Deliver Simplified Financial Advice in the Dominican Republic, India, and the Philippines Managing the finances of an enterprise, even a household operated micro-enterprise, is a complex task. Several studies around the world have shown that micro- and small-entrepreneurs do not have good financial management practices. Identifying this gap, several training programs around the world have attempted to teach better business practices to these micro-entrepreneurs. Few examples of such programs include, but are not limited to, the International Labor Organization’s Start and Improve Your Business program and the International Finance Corporation’s Business Edge program. However, only a few evaluations of traditional business training programs and financial education courses have found impacts on better financial behaviors, and even fewer have found impacts on profits and productivity. Even if they did succeed, these programs are too expensive to scale. Behavioral science provides some insights on why traditional training programs have failed. Micro- entrepreneurs, particularly the poorer and low-skilled ones, are juggling with multiple tasks in their daily lives. This constrains their cognitive bandwidth and makes it difficult for them to grasp complex concepts, and even more difficult to translate them into practice. Furthermore, because of the strain attending in-person training programs places on their time (and resources), they are unlikely to regularly attend the training making the problem even worse. Recognizing these problems, ideas42, an organization that uses behavioral science to develop solutions for social issues, along with researchers from MIT and Harvard have designed a training program that delivers management best practices in an easy-to-adopt “rules-of-thumb” (heuristics). They engaged with micro-entrepreneurs in the Dominican Republic for months to distill important financial management practices into lessons that were easy to learn and implement. A rigorous impact evaluation conducted over 2006-2008 found that the adopters of the “rules-of-thumb” were 10 percent more likely to engage in good business practices, even more so than those who received the traditional training program. Better business practices also led to higher revenues, with the effects much larger for entrepreneurs with low-levels of human capital. Over the past 10 years, the researchers and ideas42 have continued to work on this to improve the training program as well as its delivery. In India and Philippines, they adapted the program to deliver the training through mobile phones, completely replacing the need for expensive classroom-based trainings. The simplified “rules-of-thumb” are delivered through short automated calls in form of a narrative storytelling style using professional voice actors. One important feature of this is that the entrepreneurs could receive these calls at a time convenient for them, making it easier and more convenient for them to absorb the lessons and translate them in practice. Rigorous evaluation of this scalable solution found that, in both India and Philippines, the training translated into meaningful changes in financial behaviors of the micro-entrepreneurs. For example, about 8 percent more people separated their household and business cash—a key first step in determining business profit. In India, where airtime costs were cheaper, the mobile based training was 67 percent cheaper than a comparable classroom-based financial education program. Source: ideas42 (2018) CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 81 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Box 3.6: Key Characteristics of Successful Programs for Integrating Household Enterprises into the Economy There are many initiatives around the world that aim to help household enterprises integrate into the wider economy. A recent study of these existing initiatives concluded that the essential elements of the successful programs have been: (i) using a participatory methodology; (ii) providing one-on-one coaching and training to the owners of the enterprises; (iii) carefully selecting target households; and (iv) tailoring program outcomes to the specific needs of the target population in order for initiatives to succeed in integrating household enterprises. Being participatory means involving local leaders and community members in the program’s design to increase its effectiveness while also using the existing talents and skills of those in the target group and putting them to effective use. This participatory approach has been used by the NGO Aid to Artisans in its work in developing artisanal value chains. Programs can also become more successful by motivating community members to participate. In Cambodia, the NGO Mlup Baitong launched an ecotourism initiative to support households in selected rural communities to create and operate sustainable tourism activities. The NGO further empowered the community members to use part of the income generated by their ecotourism activities to implement community development projects such as a community water supply system to provide more access to safe drinking water. Elsewhere in the world, the NGO Fonkoze partners strategically with organizations like Zanmi Lasante in Haiti to offer free health care, while Village Enterprise has worked with Marie Stopes and Mercy Corp in Uganda to provide health training and services. Providing household enterprises with training and mentoring is at least as essential as providing them with productive assets. In its interventions, the NGO TechnoServe puts a higher value on installing managerial capacity than inserting capital into a business that is operating inefficiently. Furthermore, many programs exist to help household enterprises to manage their finances better by providing them with training in how to generate income and savings, how to use their savings to reinvest in their enterprises, and where they can minimize losses and inefficiencies. These programs often, though not always, also provide small loans or seed capital. One important factor in the success or failure of these interventions is the selection of the beneficiary household enterprises. It is important to choose business owners who have the right motivations and attitude rather than on the basis of their business plan or financial profile alone. The applicants should be participating in the program for the right reasons; in other words, their aspirations should align with the outcome goals of the program in order to avoid high dropout rates and unsuccessful business ventures. Finally, the effectiveness of these programs is also contingent on tailoring the support to the specific needs of the target population, coordinating with existing interventions, and integrating this program into the country’s existing social protection framework. Another important consideration is to identify and develop skills that already exist in the target households and their communities rather than introducing a completely new crop or handicraft that will require more investment and training. The replicability and relevance of these successful outcomes in other countries can be maximized, according to TechnoServe and CIAT, an agricultural NGO, by dividing target household enterprises into groups—based on commodity, size, or entrepreneurial characteristics—and customizing initiatives for each segment. What is clear is that these interventions have the potential to help household enterprises to overcome market failures and structural gaps and integrate into the economy. Source: Buba (2018) CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 82 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT 3.7 References Atkin, David, Amit K. Khandelwal, and Adam Osman. 2017. “Exporting and firm performance: Evidence from a randomized experiment.” The Quarterly Journal of Economics 132(2): 551-615. Banerjee, Abhijit, Dean Karlan, and Jonathan Zinman. 2015. “Six randomized evaluations of microcredit: Introduction and further steps.” American Economic Journal: Applied Economics 7(1):1-21. Bender, Stefan, Nicholas Bloom, David Card, John Van Reenen, and Stefanie Wolter. 2016. “Management Practices, Workforce Selection and Productivity.” Journal of Labor Economics 36(S1). Bertrand, Marianne, and Antoinette Schoar. 2003. “Managing with style: The effect of managers on firm policies.” The Quarterly Journal of Economics 118(4): 1169-1208 Blattman, Christopher, Nathan Fiala, and Sebastian Martinez. 2014. “Generating skilled self-employment in developing countries: Experimental evidence from Uganda.” The Quarterly Journal of Economics 129(2): 697-752. Bloom, Nicholas, and John Van Reenen. 2007. “Measuring and Explaining Management Practices Across Firms and Countries.” The Quarterly Journal of Economics 122(4):1351-1408. Bloom, Nicholas, and John Van Reenen. 2010. “Why do management practices differ across firms and countries?” The Journal of Economic Perspectives 24(1): 203-224 Bloom, Nicholas, Benn Eifert, Aprajit Mahajan, David McKenzie, and John Roberts. 2013. “Does management matter? Evidence from India.” The Quarterly Journal of Economics 128(1): 1-51. Brooks, Wyatt, Kevin Michael Donovan, and Terence Johnson. 2016. The Dynamics of Inter-Firm Skill Transmission among Kenyan Microenterprises. Helen Kellogg Institute for International Studies. Bruhn, Miriam. 2013. “A tale of two species: Revisiting the effect of registration reform on informal business owners in Mexico.” Journal of Development Economics 103: 275-283 Buba, Johanne. 2018. “Firm-level interventions for jobs,” Washington, D.C.: World Bank. Calderon, Gabriela, Jesse M. Cunha, and Giacomo De Giorgi. 2013. “Business literacy and development: Evidence from a randomized controlled trial in rural Mexico.” National Bureau of Economic Research. No. w197410. Campos, Francisco, Michael Frese, Markus Goldstein, Leonardo Iacovone, Hillary C. Johnson, David McKenzie, and Mona Mensmann. 2017. “Teaching personal initiative beats traditional training in boosting small business in West Africa.” Science 357(6357): 1287-1290. Cooper, Russell, John Haltiwanger, and Jonathan L. Willis. 2007. “Search frictions: Matching aggregate and establishment observations.” Journal of Monetary Economics 54: 56-78. De Mel, Suresh, David McKenzie, Christopher Woodruff. 2008. “Returns to capital in microenterprises: evidence from a field experiment.” The Quarterly Journal of Economics 123(4): 1329-1372. De Mel, Suresh, David McKenzie, and Christopher Woodruff. 2010. “Who are the microenterprise owners? Evidence from Sri Lanka on Tokman versus De Soto.” in Josh Lerner and Antoinette Schoar, eds. International Differences in Entrepreneurship, Chicago: University of Chicago Press, 63-87. Djankov, Simeon, Yingyi Qian, Gerard Roland, and Ekaterina Zhuravskaya. 2006. “Entrepreneurship in Brazil, China, and Russia.” Center for Economic and Financial Research (CEFIR) Working Paper w0066, Moscow. Drexler, Alejandro, Greg Fischer, and Antoinette Schoar. 2014. “Keeping it simple: Financial literacy and rules of thumb.” American Economic Journal: Applied Economics 6(2): 1-31. Fafchamps, Marcel, David McKenzie, Simon R. Quinn, and Christopher Woodruff. 2011. “When is capital enough to get female microenterprises growing? Evidence from a randomized experiment in Ghana.” Cambridge: National Bureau of Economic Research. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 83 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Haltiwanger, John, Ron S. Jarmin, Robert B. Kulick, and Javier Miranda. 2016. “High growth young firms: contribution to job, output, and productivity growth” in John Haltiwanger, Erik Hurst, Javier Miranda, and Antoinette Shoar eds. Measuring Entrepreneurial Businesses: Current Knowledge and Challenges Chicago: University of Chicago Press. Hombert, Johan, Antoinette Schoar, David Sraer, and David Thesmar. 2014. “Can unemployment insurance spur entrepreneurial activity?” No. w20717. National Bureau of Economic Research, Cambridge. Hsieh, Chang-Tai and Benjamin A. Olken. 2014. “The missing ‘missing middle’.” The Journal of Economic Perspectives 28(3): 89-108. Hsieh, Chang-Tai and Peter Klenow. 2014. “The Life Cycle of Plants in India and Mexico.” The Quarterly Journal of Economics 129(3):1035-1084. ideas42. 2018. “The Financial Heuristics Training: A proven mobile-based tool for teaching business lessons that stick.” www. ideas42.org/wp-content/uploads/2018/08/Ideas42_ProjectBrief_FinHeuristics_3.pdf ILO—International Labour Organization. 2016. Cambodia Garment and Footwear Sector Bulletin. Geneva: International Labour Organization. ILO. 2017. “What explains strong export and weak employment figures in the Cambodian garment sector?” Cambodian Garment and Footwear Sector Bulletin 6(May). Lafortune, Jeanne, Marcela Perticará, and José Tessada. 2013. “Are (Random) Friends Good for Business? Peer Effects in Training and Entrepreneurship Courses.” McKenzie, David and Christopher Woodruff. 2008. “Experimental evidence on returns to capital and access to finance in Mexico. The World Bank Economic Review 22(3): 457-482. McKenzie, David, and Christopher Woodruff. 2017. “Business practices in small firms in developing countries” Management Science 63(9). Schoar, Antoinette. 2010. “The divide between subsistence and transformational entrepreneurship” Innovation Policy and the Economy 10(1): 57-81. Shrestha, Maheshwor. 2018. “The impact of minimum wage hikes in employment and wages in Cambodia.” Working Paper. Valdivia, Martin. 2013. “Business training plus for female entrepreneurship? Evidence from a field experiment in Peru.” Unpublished manuscript GRADE, Lima. World Bank. 2018. Cambodia Investment Climate Assessment. Washington, D.C.: World Bank. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 84 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT ANNEX 3.1: Household Enterprises (Regressions) Annex Table 3.1: Determinants of Productivity and Selection into Household Enterprises All Male Female Rural Urban Prod Sel Prod Sel Prod Sel Prod Sel Prod Sel Completed education 0.031*** 0.012*** 0.033*** 0.001 0.035*** 0.024*** 0.016 0.042*** 0.051*** -0.027*** (0.008) (0.004) (0.012) (0.006) (0.009) (0.006) (0.012) (0.006) (0.014) (0.005) Age 0.002 0.102*** -0.022 0.092*** 0.032 0.112*** 0.009 0.104*** -0.036 0.110*** (0.029) (0.009) (0.046) (0.013) (0.024) (0.012) (0.023) (0.012) (0.044) (0.013) Age squared -0.000 -0.001*** 0.000 -0.001*** -0.000* -0.001*** -0.000 -0.001*** 0.000 -0.001*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Female -0.049 0.210*** -0.076 0.284*** -0.025 0.108** (0.058) (0.032) (0.062) (0.044) (0.064) (0.049) Household size 0.032* -0.041*** 0.030 -0.019 0.030* -0.064*** 0.035* -0.065*** 0.038* -0.001 (0.018) (0.014) (0.029) (0.021) (0.017) (0.019) (0.019) (0.019) (0.023) (0.020) Never married 0.106 -0.641*** 0.419 -0.687*** -0.128 -0.617*** -0.086 -0.738*** 0.421** -0.489*** (0.204) (0.049) (0.301) (0.073) (0.163) (0.064) (0.185) (0.071) (0.201) (0.069) Ethnic minority -0.089 0.302*** -0.120 0.370** -0.041 0.268** -0.020 0.264** -0.261 0.341** (0.156) (0.092) (0.252) (0.148) (0.141) (0.120) (0.159) (0.118) (0.266) (0.142) Community wealth index 0.205*** 0.178*** 0.117 0.179*** 0.269*** 0.190*** 0.158** 0.191*** 0.242*** 0.147*** (0.065) (0.036) (0.105) (0.046) (0.055) (0.049) (0.066) (0.053) (0.083) (0.056) Financial penetration 1.083*** 0.990*** 1.155*** 1.270*** 0.861 (0.266) (0.373) (0.339) (0.312) (0.566) squared -1.082*** -0.935 -1.175** -1.238*** -1.614 (0.406) (0.599) (0.492) (0.452) (1.169) Log (dist to district HQ) -0.033 -0.039 -0.020 -0.034 -0.040 (0.024) (0.031) (0.033) (0.029) (0.039) Village has a bank 0.163*** 0.129* 0.192** 0.169* 0.121* (0.054) (0.072) (0.075) (0.088) (0.068) Village has internet cafe 0.114** 0.116 0.096 -0.032 0.150*** (0.054) (0.073) (0.074) (0.114) (0.056) # aged 15-64 in HH -0.045*** -0.073*** -0.018 -0.029 -0.078*** (0.017) (0.024) (0.022) (0.023) (0.023) # aged 0-5 in HH 0.077*** 0.027 0.139*** 0.097*** 0.042 (0.026) (0.031) (0.036) (0.032) (0.036) Parameters artanh (rho) -0.429 -0.646 -0.168 -0.456*** -0.663* (0.303) (0.415) (0.212) (0.164) (0.397) log (sigma) 0.142 0.216 0.082** 0.139** 0.218 (0.087) (0.161) (0.034) (0.057) (0.144) Rural x Region Yes Yes Yes Yes Yes Observation 16197 8159 8038 9269 6928 Source: Author’s estimates using CSES 2014. Note: The Heckman selection model was estimated for each subgroup with the productivity and selection equations presented in alternate columns with parameter estimates in the bottom panel. The sample was restricted to the working age population involved primarily in household enterprises and in private wage employment. The productivity of an enterprise worker is the profit per hours worked in the enterprise with the profit split proportional to the hours that the worker put into the enterprise. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 85 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Annex Table 3.2: Determinants of Hiring a Wage Employee All Male Female Rural Urban Completed education 0.005*** 0.005** 0.005*** 0.005*** 0.004** (0.001) (0.002) (0.002) (0.002) (0.002) Age 0.000 0.003 -0.002 0.001 -0.001 (0.002) (0.004) (0.003) (0.003) (0.004) Age squared 0.000 -0.000 0.000 -0.000 0.000 (0.000) (0.000) (0.000) (0.000) (0.000) Female -0.042*** -0.044*** -0.039*** (0.009) (0.012) (0.013) Household size 0.000 -0.006 0.006 0.002 -0.004 (0.004) (0.005) (0.005) (0.005) (0.006) Never married -0.015 -0.028 -0.001 -0.022 -0.004 (0.018) (0.033) (0.022) (0.026) (0.025) Ethnic minority -0.002 -0.014 0.008 -0.007 0.001 (0.019) (0.030) (0.017) (0.024) (0.033) Community wealth index 0.013 0.009 0.016** 0.019 0.023 (0.009) (0.015) (0.008) (0.012) (0.014) Financial penetration -0.066 0.043 -0.153 -0.111 0.078 (0.088) (0.140) (0.095) (0.104) (0.175) squared 0.215 0.058 0.336* 0.253 0.106 (0.151) (0.221) (0.184) (0.169) (0.369) Log(dist to district HQ) -0.001 0.000 -0.000 -0.001 -0.022* (0.006) (0.011) (0.006) (0.007) (0.012) Village has a bank 0.015 -0.007 0.033* 0.002 0.006 (0.016) (0.026) (0.018) (0.023) (0.023) Village has an internet café 0.008 0.003 0.014 -0.040* 0.023 (0.014) (0.022) (0.014) (0.021) (0.015) # aged 15-64 in HH 0.000 0.004 -0.003 0.001 0.001 (0.004) (0.007) (0.006) (0.006) (0.008) Rural x Region Yes Yes Yes Yes Yes Observation 4150 1923 2227 2048 2102 R-squared 0.025 0.013 0.024 0.024 0.029 Source: Author’s estimates using CSES 2014. Note: A firm was considered to have hired worker if they had a positive wage bill in the previous 12 months. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 86 TECHNICAL REPORT CHAPTER 3: CREATING MORE AND BETTER JOBS THROUGH PRIVATE SECTOR DEVELOPMENT Annex Table 3.3: Employment in the Wage Sector and the Number of Households with Enterprises Sector of HH enterprise : All Trade Hospitality log (2014 employment) in: Garments & footwear 0.115*** 0.108*** 0.142 (0.033) (0.030) (0.110) Non-garment domestic manufacturing 0.422*** 0.405*** 0.178 (0.093) (0.099) (0.167) Domestic construction 0.106 -0.005 0.326 (0.125) (0.120) (0.328) Domestic tourism 0.329*** 0.404*** 0.542*** (0.083) (0.076) (0.170) Non-garment foreign owned 0.181* 0.237** 0.118 (0.104) (0.099) (0.232) Observation 45 45 45 R-squared 0.900 0.903 0.533 5 year change of Log (employment) in: Garments & footwear 0.012 0.049 -0.469 (0.077) (0.153) (0.435) Non-garment domestic manufacturing 0.279*** 0.3818** 0.147 (0.090) (0.153) (0.438) Domestic construction 0.239** 0.294 0.049 (0.106) (0.188) (0.365) Domestic tourism 0.226** 0.242 0.947*** (0.097) (0.153) (0.250) Non-garment foreign owned -0.006 0.018 -0.078 (0.049) (0.077) (0.216) Year FE Y Y Y Observation 89 89 89 R-squared 0.275 0.312 0.227 R-squared 0.907 0.868 0.621 Source: Author’s estimates using CSES 2014, 2009, and 2004. Note: The top panel is a cross-sectional regression for year 2014, which regresses the logarithm of household enterprises on the logarithms of wage employment categories. The bottom panel regresses a five-year change in the logarithm of household enterprises on five-year changes on the logarithm of wage employment categories and survey year fixed effects. The changes are between 2004 and 2009 and between 2009 and 2014. For both panels, each observation is a province x rural/urban cell for each year. Regressions are weighted by the population in 2014 in the top panel and the population in 2004 in the bottom panel. Robust standard errors reported in parentheses. *: p<.10. **: p<.05. ***: p<0.01. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 87 TECHNICAL REPORT CHAPTER 4: Jobs and the Export Sector Claire H. Hollweg and Anne Ong Lopez 4.1 Introduction The Royal Government of Cambodia (RGC) has a vision to transform and modernize the country’s industrial structure from labor-intensive to skill-driven, which, in large part, will happen though its export sector. As identified in the government’s Industrial Development Policy for 2015-2025, the factors that will be crucial to realizing this vision will be: (i) connecting to regional and global value chains; (ii) becoming integrated into regional production networks; and (iii) developing interconnected production clusters to diversify exports. Labor-intensive export sectors—particularly garments and tourism—have played an important role in Cambodia’s growth and wage employment but have remained stagnant in low-value-added activities. Cambodia’s export growth—among the fastest in the world—has been driven mostly by garment exports and tourism receipts. However, the country’s industrial sector remains weak and narrowly based as reflected by the simple structure of manufacturing and its low level of export sophistication. For example, Cambodia has had limited success in moving up the value chain in garments, with most of its activities in the sector being limited to low-end, low-wage assembly work with little added value. Therefore, there is a long way to go to achieve the RGC’s vision. The factors behind Cambodia’s impressive job creation, poverty reduction, and economic growth over the past decade—exports of goods and services in garments, tourism, and agriculture through labor-intensive low-skilled jobs—are unlikely to be the factors that will drive the growth of good jobs in the future. The labor-intensive garment industry is facing rising competition from other low-wage countries in the region, particularly Myanmar, and real wages seem to be rising significantly faster than productivity (see Chapter 3). Garment (and tourism) exports are also under pressure in the context of the recent appreciation of the US dollar given Cambodia’s high dollarization and pegged exchange rate (see Chapter 2), raising the prospect of Cambodia losing special access for its exports to important markets such as the European Union. This would further threaten key exporting industries, as well as deterring FDI, and would have immediate and severe negative implications for Cambodia’s jobs that are dependent on exports. Other mega-trends—such as the rise of regional competitors vying for higher-valued added spots in regional and global value chains, slowing international trade and maturing global value chains, and the growth of an Asian consumer class—will also reshape Cambodia’s job picture (see Chapter 1). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 89 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Cambodia’s export diversification strategy going 4.2 How Do Cambodian Exports forward must strive to create more, better, and more inclusive jobs at the same time finding ways Support Wages and Jobs? to move the economy into higher-value-added Export growth can be a powerful way to increase export activities. In this chapter, we look closely at employment and earnings, both directly within how Cambodia’s export sector has supported structural exporting firms and indirectly through these firms’ transformation and job creation to discover what export demand for goods and services from the domestic diversification and upgrading as envisioned in the economy. The extent to which exports support domestic government’s Industrial Development Plan could mean in labor depends on several factors, including the labor terms of creating more, better, and more inclusive jobs. intensity of export sectors and the extent to which exporting The need for the economy to make this transition is more firms purchase inputs from domestic firms. In this section, immediate given recent EU statements that Cambodia we measure the labor sophistication of Cambodia’s will temporarily lose its preferential trade status until it merchandise exports (how the goods that Cambodia exports makes “clear and demonstrable improvements” in its are associated with different labor market outcomes) and the human rights.1 We will: (i) profile Cambodia’s export labor content of Cambodia’s exports (how many wages are structure as well as the factors that underpin its current supported by exports) and relate these to other indicators of export competitiveness; (ii) measure the labor content of Cambodia’s export performance (such as firms’ participation Cambodia’s exports, in terms of the jobs and wages that in value chains and the extent of their diversification). Cambodia’s exports currently support and their evolution over time; (iii) assess whether the profile of workers in current export activities and in other light manufacturing The labor content of Cambodian exports sectors into which Cambodia is looking to diversify, as well is high… as any other macroeconomic factors, are constraining the economy from diversifying into modern, higher-quality jobs The export sector is responsible for a substantial within global and regional value chains; (iv) explore the proportion of the wages paid in Cambodia. In 2014, importance of services exports (particularly in tourism) for exports paid USD 2.6 billion in wages, representing about creating jobs; and (v) recommend policies to support export 15 percent of GDP, both directly to workers within the export competitiveness in job-friendly sectors. sector and indirectly to workers employed by domestic input Box 4.1: Concepts and Measures The definitions of various concepts and measures used throughout this chapter to describe the relationship between exports and labor outcomes are as follows: Labor sophistication of exports: This is an indirect measure of the trade-weighted average (labor) outcomes of the exports that appear in Cambodia’s export basket, also known as EXPY. The labor outcomes include: value added, wages, skills, physical capital, and human capital. See Annex 4.1 for more detail. The logic behind these measures is that information on labor markets and other characteristics of the economies that intensively export a given product is useful for determining the potential for specializing in certain products exported to certain markets. Based on these observations, it is possible to draw inferences about how export growth and product and regional trade patterns will affect future employment, wages, and the demand for skills in Cambodia. Labor content of exports: This is a direct measure of the contribution of labor to the value-added of a given sector’s exports and the wages paid by the companies within it. The latter can be broken down into direct and indirect wages. The direct wages are the sector’s domestic labor value-added as embodied in its own exports. The indirect wages are the sector’s domestic labor value-added as represented by inputs to and from other sectors’ exports. See Annex 4.2 for more detail. Labor intensity of exports: This constitutes total wages paid per unit of a sector’s exports. It is measured as the labor content of exports of a given sector divided by the sector’s gross exports. 1 On February 11, 2019 the European Commission launched a procedure to temporarily withdraw tariff preferences granted under the Everything But Arms (EBA) arrangement over concerns related to human and labor rights violations. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 90 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR suppliers (Figure 4.1). This was more than in Nepal, Lao in sectors providing inputs to the exporting sector. In PDR, and Malaysia (2 percent, 8 percent, and 11 percent comparison, in Vietnam, exports paid USD 19 billion in of GDP respectively), but less than in Thailand (19 percent). direct wages and USD 11 billion in indirect wages. Exports in Vietnam contributed a similar amount to GDP as they did in Cambodia (16 percent). Cambodia’s exports have grown strongly since 1997 (Figure 4.2). In 1991 (the first year of available data), exports As in its comparator countries, more wages were of goods and services totaled only USD 0.3 billion, but this paid directly for export production than indirectly in had increased to over USD 13 billion by 2016, equivalent to Cambodia. In 2014, exports paid USD 1.8 billion in direct 61 percent of GDP. Net inflows of foreign direct investment wages to workers employed in the exporting sector, and (FDI) also grew strongly after 2003, reaching USD 2.3 billion USD 0.8 billion in indirect wages to workers employed in 2016 or 11.4 percent of GDP.2 Figure 4.1: Labor Content of Cambodian Exports, 2004-2014 2,500 2,000 USD Millions 1,500 1,000 Total 500 Direct 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Year Source: Authors’ calculations using the LACEX database. Note: Values in constant 2014 prices. Figure 4.2: Exports of Goods and Services and Net FDI Inflows, 1991-2016 14 12 10 USD Billions 8 6 4 2 Exports 0 FDI 19 1 19 2 19 3 19 4 19 5 19 6 19 7 19 8 20 9 20 0 20 1 20 2 20 3 20 4 20 5 20 6 07 20 8 20 9 20 0 20 1 20 2 20 3 20 4 15 9 9 9 9 9 9 9 9 9 0 0 0 0 0 0 0 0 0 1 1 1 1 1 19 20 Year Source: Authors’ calculations using data from World Bank World Development Indicators. 2 Foreign direct investment (FDI) consists of direct investment equity flows to a given economy. It is the sum of equity capital, reinvestment of earnings, and other capital. FDI net inflows are the net value of inward direct investment made by non-resident investors in the economy. These data are from the World Bank’s World Development Indicators. The data on equity flows are based on balance of payments data from the International Monetary Fund (IMF). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 91 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Consistent with Cambodia’s export performance, However, Cambodia’s labor content of exports remains the labor content of exports has also grown rapidly extremely concentrated and has changed little since since 2004. The compound annual growth rate (CAGR) of 2004. The labor content of exports is highly concentrated in the labor content of exports was 9.2 percent between 2004 the apparel sector in Cambodia. Total wages in exports in the and 2014, and this was similar to the CAGR of exports (8.2 apparel sector, for example, were over USD 1.67 billion in percent) and the CAGR of the labor content of domestic 2014 compared with around USD 481 million in 2004 (in real demand (7.4 percent) over the same time period (Figure 2014 values) (Figure 4.4). Apparel and textiles accounted 4.3). The growth in the labor content of Cambodia’s exports for 64 percent of the wages supported by exports in 2014, was greater than that in Malaysia, Thailand, Nepal, and compared to 61 percent in 2004. The increase was driven Vietnam, but lagged behind that of Lao PDR. primarily by apparel, as Cambodia exports very few textiles Figure 4.3: CAGR of Exports and Labor Content of Exports and Output, 2004-2014 (a) Exports and Labor Content of Exports (b) Labor Content of Exports and Labor Content of Output Cambodia Cambodia Lao PDR Lao PDR Malaysia Malaysia Nepal Nepal Thailand Thailand Vietnam Vietnam -5% 0% 5% 10% 15% -5% 0% 5% 10% 15% CAGR between 2004 and 2014 CAGR between 2004 and 2014 Labor Content of Exports Exports Labor Content of Exports Labor Content of Output Source: Authors’ calculations using the LACEX database. Note: Values in constant 2014 prices. Figure 4.4: Total Labor Value-Added of Cambodian Exports (backward linkages) (a) Labor Content of Exports, 2004 (b) Labor Content of Exports, 2014 Wearing Apparel Wearing Apparel Textiles Leather Products Other Private Services Other Private Services Trade and Transport Services Trade and Transport Services Leather Products Processed Foods Processed Foods Agr, Forestry, Fisheries Chemical, Rubber, Plastic Products Chemical, Rubber, Plastic Products PubAdmin/Defence/Health/Educat PubAdmin/Defence/Health/Educat Agr, Forestry, Fisheries Machinery and Equipment nec Wood Products Wood Products Manufactures nec Textiles Transport Equipment Minerals nec Machinery and Equipment nec Transport Equipment Construction Manufactures nec Beverages and Tobacco Products Metals nec Mineral Products nec Beverages and Tobacco Products Paper Products, Publishing Construction Ferrous Metals Paper Products, Publishing Electricity, Gas, Water Mineral Products nec Metal Products Metal Products Metals nec Electricity, Gas, Water Minerals nec Ferrous Metals Energy Extraction Energy Extraction Dwellings Dwellings 0 500 1,000 1,500 2,000 0 500 1,000 1,500 2,000 USD Million USD Million Direct LVAX Indirect LVAX Source: Authors’ calculations using the LACEX database. Note: Values in constant 2014 prices. "nec" stands for "not elsewhere classified." CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 92 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR (see Annex Figure 4.3). The majority of the labor content in plastic products; and textiles. By 2014, the four most this sector was concentrated in direct export production. important sectors for export wages were electronic Leather products such as footwear, handbags, and travel equipment; chemical, rubber, and plastic products, trade goods accounted for slightly more labor content in 2014 and transport services; and business services. than they did in 2004. In Cambodia, most inputs for these leather products are imported from abroad due to a lack of The fact that only a few export sectors support labor in raw materials in Cambodia itself and then are assembled Cambodia reflects the country’s high concentration of for export. Other private services and trade and transport exports. Goods exports are concentrated in labor-intensive services remained top export sectors in terms of wages export activities, namely, apparel and, to a lesser extent, in Cambodia. footwear. Between 2000 and 2016, Cambodia’s export basket to the world was highly concentrated in apparel, In Vietnam, in contrast, exports diversified into although its share in total exports was gradually decreasing, higher-value-added sectors over the same period. In from 83 percent in 2000 to 61 percent in 2016. The shares 2004, the four most important sectors for export wages in of footwear in total exports has remained relatively stagnant, Vietnam were oil; leather products; chemical, rubber and from 8 percent in 2000 to 10 percent in 2016 (Figure 4.5). Figure 4.5: Share and Value of Cambodia’s Total Exports to the World, 2000-2016 (a) Sector Share of Total Exports (b) Sector Export Value 80% 20 Percent of Total Exports 60% 15 USD Millions 40% 10 20% 5 0% 0 2000 2002 2004 2006 2008 2010 2012 2014 2016 2000 2002 2004 2006 2008 2010 2012 2014 2016 Year Year Vegetable Products Textiles Footwear Precious Metals Base Metals Transport Equipment Apparel (c) Sector Share of Total Exports (d) Sector Value of Total Exports 10% 3 Percent of Total Exports 2 USD Millions 5% 1 0% 0 2000 2002 2004 2006 2008 2010 2012 2014 2016 2000 2002 2004 2006 2008 2010 2012 2014 2016 Year Year Foodstuffs Chemicals Footwear Machinery Transport Equipment Source: Authors’ calculations using UNCOMTRADE data from World Integrated Trade Solutions (WITS). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 93 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR There is some evidence of recent export diversification. coupled with the high concentration of exports, confirms The declining share of apparel in total exports reflects that Cambodia’s recent export performance continues to the Cambodian government’s commitment to its strategy support labor in only a few sectors. to diversify the country’s export base. The government’s Industrial Development Policy for 2015-2025 aims to Labor intensity is low in Cambodia compared to reduce the export share of footwear and garments to 55 most of its peer countries. The labor intensity of exports percent of total exports by 2020 and to 50 percent by 2025. measures the wages paid per 1 US dollar of exports or, The recent decline in the Herfendehal Hershman Index (HHI), when multiplied by 100, the percentage of export value a measure of export product concentration, is evidence of paid as wages. In Cambodia, the labor intensity of exports progress towards this goal (Figure 4.6). There is also some is less than 20 percent, which is only a little higher than evidence that exports have been growing in electronics, that of Thailand (Figure 4.7). Labor intensity declined in machinery, and transport equipment since 2011. many sectors between 2004 and 2011 but recovered between 2011 and 2014 (Annex Figure 4.7). There was Nevertheless, Cambodia has grown on the intensive an increase in labor intensity in the apparel sector during rather than the extensive margin. Export growth has both of these periods (Annex Figure 4.11), particularly after come from an increase of existing products in established 2011, whereas there was a steady decline in labor intensity markets (the intensive margin) rather than from either new other manufactures over time. This increase in the apparel products to old markets or old products to new markets or sector coincided with increases in the minimum wage both (the extensive margin). An example of export growth after 2013. on the intensive margin is increasing the sales of tee- shirts to the United States. Export growth on the extensive Cambodia’s labor intensity in apparel, its most margin has been much weaker and has primarily come from important sector, is low for the country’s level of product diversification in established markets rather than development as measured by per capita GDP. Both in introduction of new products in new or established markets 2004 and 2014, Cambodia’s labor intensity was lower than (Annex Figure 4.5). One example of Cambodia’s export growth in other countries at similar levels of development (Figure on the extensive margin through product diversification in 4.8). Nepal has a higher labor intensity in the apparel sector established markets is that it has started selling shoes to the than Cambodia though both countries have similar per US that had previously only been exported to the EU. This, capita income levels. Figure 4.6: Herfendehal Hershman Index (HHI) of Product Concentration .20 .15 Product Concentration Index .10 Cambodia Lao PDR Malaysia .05 Nepal Thailand 0 Vietnam 2000 2005 2010 2015 Year Source: Authors’ calculations using UNCOMTRADE data from WITS. Note: The Herfindahl Hershman Index measures levels of product concentration. A higher number reflects greater concentration, a lower number greater diversification. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 94 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Figure 4.7: Labor Intensity of Exports, 2014 Cambodia Lao PDR Malaysia Nepal Thailand Direct Vietnam Indirect 0% 10% 20% 30% Total Worker Compensation as Percent of Export Value Source: Authors' calculations using the LACEX database. Figure 4.8: Labor Intensity in Apparel Exports Across Countries (a) Apparel Labor Intensity, 2004 70% 60% Total Worker Compensation as Percent of Export Value 50% 40% Nepal 30% Lao PDR Thailand 20% Malaysia Vietnam Cambodia 10% 0% 4 6 8 10 12 Log GDP per capita (USD) (b) Apparel Labor Intensity, 2014 70% 60% Total Worker Compensation as Percent of Export Value 50% 40% 30% Nepal Lao PDR Thailand 20% Cambodia Malaysia 10% Vietnam 0% 4 6 8 10 12 Log GDP per capita (USD) Source: Authors’ calculations using the LACEX database and World Development Indicators. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 95 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Cambodia’s labor intensity is also low in other in some of the services sector (such as electricity, gas, sectors compared to its peers in the region. Its total water, trade and transport, and public services) (Figure 4.9). labor intensity was lower than the level in other comparator Furthermore, while apparel was the largest source of wages countries in the agriculture, forestry, and fisheries sector, among all export sectors in Cambodia in 2004, 2007, 2011, in some manufacturing sectors (such as beverages and and 2014, labor intensity in this sector was lower than in all tobacco, metal, paper, textiles, and wood products) and other peer countries except Vietnam (Box 4.2). Figure 4.9: Labor Intensity of Exports across Sectors and Countries, 2014 Agr, Forestry, Fisheries Beverages and Tobacco Products Chemical, Rubber, Plastic Products Construction Dwellings Electricity, Gas, Water Energy Extraction Ferrous Metals Leather Products Machinery and Equipment Manufactures Metal Products Metals Cambodia Mineral Products Minerals Lao PDR Other Private Services Paper Products, Publishing Malaysia Processed Foods PubAdmin/Defence/Health/Educat Nepal Textiles Trade and Transport Services Transport Equipment Thailand Wearing Apparel Wood Products Vietnam 0% 20% 40% 60% Total Worker Compensation as Percent of Export Value Source: Authors’ calculations using the LACEX database. Note: "nec" stands for "not elsewhere classified." Box 4.2: Labor Intensity in Cambodia’s Garments Sector Labor content has increased in the apparel sector, and in terms of wages paid for exports in this sector, Cambodia performs better than its peers. In nominal terms, labor content of exports increased in wearing apparel between 2004 and 2014 by a factor of 6.5. In real terms, it increased by a factor of four. Also, in terms of wages paid in this sector, Cambodia exceeds Thailand, Malaysia, Lao PDR, and Nepal and only lags behind Vietnam (Box Figure 1). However, labor intensity is weaker in this sector in Cambodia than in these comparator countries. In 2014, labor intensity (the labor content of exports as a percentage of total exports) in wearing apparel was second lowest among the comparator countries (with the lowest being in Vietnam) (Box Figure 2). This is indicative of the fact that this sector in Cambodia is characterized by low-value-added activities. Box Figure 1: Labor Content of Apparel Exports Box Figure 2: Labor Intensity of Apparel Exports 2,000 35% Total Apparel Wages (Million USD) Total Worker Compensation as Percent of Export Value 30% 1,500 25% 1,000 20% 500 15% 0 10% 2004 2006 2008 2010 2012 2014 2004 2006 2008 2010 2012 2014 Cambodia Lao PDR Malaysia Nepal Thailand Vietnam Source: Authors’ calculations using the LACEX database. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 96 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR What are job-friendly and export-friendly sectors? There Overall, Cambodian exports have tended to have a is a clear trade-off between the export intensity of a sector low level of sophistication.3 Comparing trends in the (measured as the percentage of output that is exported globally) labor sophistication of exports over time, Cambodia had the and its labor intensity (measured as the wages paid per unit of weakest performance across many measures between 2000 export) (Figure 4.10). While a country’s export diversification and 2016 compared to its peers. Cambodia’s export basket strategy should be based on many factors including the contains products that pay lower wages, have lower value- country’s comparative advantage in an export sector, the jobs added, and employ fewer skilled workers than the products intensity of that export sector can be one consideration. Sectors exported by its peer countries (Figure 4.11). Cambodia’s like oil and gas that have a high export intensity also have a exports to the world also have the lowest median value- low labor intensity, meaning that few wages are paid per unit of added and lower median wages paid, and neither of these exports. On the other hand, other sectors like trade and transport, labor sophistication (EXPY) measures has changed much construction, and other business services have a higher labor over the past decades. While Cambodia’s output per intensity but tend not to have a high export intensity. employee (productivity) has increased, it still performed worse than all of its peers except Vietnam in recent years. Thailand and Malaysia’s exports have more sophistication …while the labor sophistication of than Cambodia’s across all labor sophistication measures Cambodian exports is low (Figure 4.11). Figure 4.10: Export Intensity and Labor Intensity Across Sectors, 2014 200% (Wage Value as Percent of Export Value) 150% Export Labor Intensity 100% 50% 0% 0% 20% 40% 60% Export Intensity (Percent of Sector Output Exported) Source: Authors’ calculations using the LACEX database. Note: atp = air transport. b_t = beverages and tobacco products. c_b = sugar cane and sugar beet. cmn = communication. cmt = meat: cattle, sheep, goats, and horses. cns = construction. coa = coal. crp = chemical, rubber, and plastic products. ctl = cattle, sheep, goats, and horses. ele = electronic equipment. ely = electricity. fmp = metal products. frs = forestry. fsh = fishing. gas = gas. gdt = gas manufacture and distribution. gro = cereal grains not else classified (nec). i_s = ferrous metals. isr = insurance. lea = leather products. lum = wood products. mil = dairy products. mvh = motor vehicles and parts. nfm = metals nec. nmm = mineral products nec. oap = animal products nec. obs = business services nec. ocr = crops nec. ofd = food products nec. ofi = financial services nec. oil = oil. ome = machinery and equipment nec. omf = manufactures nec. omn = minerals nec. omt = meat products nec. osd = oilseed. osg = public administration, defense, health, and education. otn = transport equipment nec. otp = transport nec. p_c = petroleum and coal products. pcr = processed rice. pdr = paddy rice. pfb = plant-based fibers. ppp = paper products and publishing. rmk = raw milk. ros = recreation and other services. sgr = sugar. tex = textiles. trd = trade. v_f = vegetables, fruit, and nuts. vol = vegetable oils and fats. wap = wearing apparel. wht = wheat. wol = wool and silk-worm cocoons. wtp = sea transport. wtr = water. "nec" stands for "not elsewhere classified." 3 Labor sophistication of exports is measured according to methodology devised by Hausmann et al. (2007) and Shirotori et al. (2010) in which six aspects of weighted average values of export labor market outcomes are measured (EXPYs), namely, the median wage, the median value-added per worker, the ratio of skilled to total workers, output per employee, years of schooling, and capital stock per worker. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 97 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Figure 4.11: Labor Sophistication of Cambodia’s Exports Relative to Its Peers (a) Output per Employee (b) Human Capital 5 10 4 9 Years of School PPP Millions 3 8 2 7 1 6 0 5 2000 2002 2004 2006 2008 2010 2012 2014 2016 2000 2002 2004 2006 2008 2010 2012 2014 2016 Year Year (c) Physical Capital per Worker (d) Median Value Added 120 100 100 80 80 PPP Thousands USD Thousands 60 60 40 40 20 20 0 0 2000 2002 2004 2006 2008 2010 2012 2014 2016 2000 2002 2004 2006 2008 2010 2012 2014 2016 Year Year (e) Median Wage (f) Share of Skilled Workers 25 80% 20 75% Percent of Workforce PPP Thousands 70% 15 65% 10 60% 5 55% 0 50% 2000 2002 2004 2006 2008 2010 2012 2014 2016 2000 2002 2004 2006 2008 2010 2012 2014 2016 Year Year Cambodia Lao PDR Malaysia Myanmar Nepal Thailand Vietnam Source: Author’s calculations based on data from UNCOMTRADE, the World Bank’s Business Environment and Enterprise Performance Surveys, and Shirotori et al. (2010). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 98 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Figure 4.12: Labor Sophistication of Cambodia’s Exports by Destination Market (a) Output per Employee (b) Human Capital 12 25 20 8 Years of School PPP Millions 15 10 4 5 0 0 2000 2002 2004 2006 2008 2010 2012 2014 2016 2000 2002 2004 2006 2008 2010 2012 2014 2016 Year Year (c) Physical Capital per Worker (d) Median Value Added 120 200 150 80 PPP Thousands USD Thousands 100 40 50 0 0 2000 2002 2004 2006 2008 2010 2012 2014 2016 2000 2002 2004 2006 2008 2010 2012 2014 2016 Year Year (e) Median Wage (f) Share of Skilled Workers 30 80% 25 60% Percent of Workforce 20 PPP Thousands 15 40% 10 20% 5 0 0% 2000 2002 2004 2006 2008 2010 2012 2014 2016 2000 2002 2004 2006 2008 2010 2012 2014 2016 Year Year World EAP EU27 USA Source: Author’s calculations based on data from UN COMTRADE, World Bank Business Environment and Enterprise Performance Surveys, and Shirotori et al. (2010). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 99 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR The labor sophistication of Cambodian exports to the How Cambodia participates in GVCs is key to how its exports world, the US, and the EU27 has remained stagnant. support labor today and in the future in terms of strengthening Interestingly, Cambodian exports destined for the East Asia the links between domestic firms and GVCs and creating and Pacific region (EAP) are more labor-sophisticated than opportunities to upgrade into higher value added and more its exports to the EU and the US, in terms of having higher sophisticated products or tasks within the value chain. wages, more value-added, and more human capital (Figure 4.12). This is because Cambodia’s export basket to the EAP Cambodia is located “downstream” in GVCs (in other is more diversified than its export basket to the US and the words close to the final consumer) and is primarily EU27 (Annex Figure 4.3) and includes goods that have a higher a buyer of foreign inputs. A country can participate in factor content (such as machinery) at higher percentages GVCs in two ways. The first is as a buyer of foreign inputs than the US and EU27 baskets. Cambodia’s most important that it then uses to create its own exports. This type of machinery exports to the EAP in 2016 included insulated GVC participation can be measured as foreign value-added wire and cable, telephone sets, and electrical transformers embodied in Cambodia’s exports. The second way is as a and converters. This more diversified and labor sophisticated seller of inputs for further downstream processing. This type export basket to the EAP can also explain Cambodia’s of GVC participation can be measured as domestic value- improving trends in the “skilled workers” measure (in which added embodied in third countries’ exports. In Cambodia, Cambodia’s exports were weaker than those of their peers). domestic value-added in the exports of third countries Conversely, textiles dominated Cambodia’s manufacturing (DVAX) increased from 8.4 percent in 2004 to 12 percent export basket to US and EU27 between 2000 and 2016, in 2011 (Figure 4.13(a)). However, its forward participation which helps to explain why the factor content of exports to remains low compared that of its peers, such as Malaysia (20 percent) and Vietnam (16 percent), meaning that these countries remained constant. Cambodia supplies few inputs for third country’s exports. This reflects Cambodia’s “downstream” position in GVCs. Cambodia participates in key global value The foreign value-added embodied in Cambodia’s gross chains exports (FVAX) is much higher, about 40 percent, but has remained stagnant since 2000 (Figure 4.13(b)). The share Cambodia’s main export sectors—garments and of Cambodia’s FVAX is almost on a par with or even higher footwear—are in important global value chains (GVCs). than that of the ASEAN average. Figure 4.13: Cambodia’s Integration in GVCs (a) Domestic Value Added in Third Country Exports (b) Foreign Value Added of Gross Exports 25% 60% Percent of Third Country Export Value 50% 20% Percent of Exports Value 40% 15% 30% 10% 20% 5% 10% 0% 0% 00 01 02 03 04 05 06 07 08 09 10 11 00 01 02 03 04 05 06 07 08 09 10 11 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Cambodia Malaysia Cambodia Malaysia Thailand Thailand Vietnam Vietnam ASEAN Source: Author’s calculations based on data from OECD-WTO TiVA database. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 100 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR World Bank Enterprise Survey data show that The Comprehensive and Progressive Agreement on imported inputs are used by many Cambodian firms Trans-Pacific Partnership is not expected to yield any for their exports. In 2016, 46 percent of Cambodian significant benefits for Cambodia. The Comprehensive manufacturing firms used imported inputs compared to and Progressive Agreement for Trans-Pacific Partnership the average 41 percent of the EAP region. Imported inputs (CPTPP), also known as TPP11, is a signed but not-yet-ratified represented 33.8 percent of the country’s total inputs on trade agreement between Australia, Brunei, Canada, Chile, average compared with 25.8 percent of total inputs in the Japan, Malaysia, Mexico, New Zealand, Peru, Singapore, and EAP as a whole. This is consistent with Cambodia’s typical Vietnam. These 11 countries represent 13.4 percent of global participation in the value chain, which is primarily in low- GDP (USD 13.5 trillion), making it one of the world’s largest value-added assembly functions. trade agreements. The CPTPP will impact both production and trade structures. In the case of Vietnam, which will be Trade agreements are key to Cambodia’s participation the biggest beneficiary of the agreement, the sectors that in GVCs and to continued job creation through the will benefit most are food, beverages, and tobacco; apparel export sector. Cambodia is a party to many regional free and leather; and textiles. Most services sectors are also trade agreement agreements including the ASEAN Free expected to expand faster as a result of the CPTPP. Increases Trade Area and ASEAN agreements with Australia, New in output will mostly be driven by higher exports, some of Zealand, India, Japan, China, and Korea. Cambodia also is which are expected to redirect trade towards members of the participating in the 16-member Regional Comprehensive partnership and away from non-members, such as Cambodia, Economic Partnership negotiations. Cambodian exports have who are expected to lose out. Cambodia’s projected income preferential access to the US under the Generalized System gains under the CPTPP are in the range of -0.42 percent of of Preferences (GSP) and to the EU under the Everything income (Maliszewska et al. 2018). But Arms (EBA) scheme. However, these countries grant this preferential access based on a country’s level of economic development, so, as Cambodia continues to climb the Indirect links with exporting firms are nascent development ladder, this access is likely to be withdrawn. In Cambodian exports generally support labor directly addition, the European Commission has launched a procedure through employment in exporting firms rather than to temporarily withdraw tariff preferences granted under the indirectly through input purchases from domestic EBA arrangement over concerns related to serious human firms. In 2014, the share of direct labor value-added in and labor rights violations. The US-Cambodia Trade and total labor value-added in Cambodia was 69 percent, Investment Framework Agreement (TIFA), signed in 2006, whereas indirect labor value-added accounted for only 31 provides a forum for both countries to address any bilateral percent (Figure 4.7 and Figure 4.14(a)). The share of direct trade issues and to coordinate on regional and multilateral labor value-added is higher in Cambodia than in any of its issues, but it is not a free trade agreement. Given that the peer countries. In Vietnam, for example, the direct share is US and the EU are two of the most important destinations for 63 percent, while in Malaysia and Thailand it is 51 percent Cambodia’s apparel exports, it will be crucial to ensure the and 55 percent respectively. country’s continued preferential access to these markets, for example through bilateral free trade agreements (FTAs). Between 2004 and 2014, direct wages paid to support This will be necessary not only for the garment sector, exports grew slightly faster than indirect wages currently Cambodia’s main export sector, but also to give to support exports. In that decade, total labor intensity firms the opportunity to diversify into other GVC sectors. coincided with indirect labor intensity, which declined Also, the EU recently informed the Cambodian government between 2007 and 2011 before increasing thereafter (Figure that it will lose preferential trade status unless it makes 4.14(b)). Indirect labor intensity was around 6 percent “clear and demonstrable improvements” to its human in 2007, 4.6 percent in 2011, and 5.6 percent in 2014. rights records. The loss of this status, which exempts Direct labor intensity conversely increased steadily over Cambodia’s exports from tariffs, would have an immediate time. Direct labor intensity was most substantial in apparel, negative impact on Cambodia’s competitiveness in the EU which supports jobs more directly than indirectly. However, market, which accounts for 40 percent of the country’s other sectors had stronger backward links with the domestic merchandise exports. economy, which underscores the importance for Cambodia of diversifying exports to support wages and jobs in a wider variety of sectors. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 101 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR There are surprisingly few forward links across added of agriculture, forestry, and fisheries inputs went sectors, meaning that exports are not supporting to processed foods, 20.7 percent to leather, and 16.2 jobs and wages in domestic input-providing percent to apparel. sectors. However, there are some exceptions such as exporting companies in the apparel, trade and transport, Cambodia’s lack of indirect links in the services and agriculture, forestry, and fisheries sectors. In 2014, sector is a challenge because research has shown 51.1 percent of the labor value-added of apparel inputs that, for an economy to be integrated into GVCs, it went to leather, while 18.1 percent went to trade and needs considerable services inputs Between 2004 and transport services, and 9.1 percent went to other private 2014, Cambodian services had relatively stable but low services. In the same year, 64.1 percent of the labor indirect labor value-added (around 5 percent) (see Annex value-added of trade and transport inputs went to apparel, Figure 4.6). In general, Cambodia’s services sectors tend 13.7 percent to leather, and 4.5 percent to machinery and to support labor more in direct export production than in other equipment. Also, 31.5 percent of the labor value- indirect purchase of inputs (Figure 4.15). Figure 4.14: Direct, Indirect, and Total Labor Content of Cambodian Exports (a) Labor Content of Exports (b) Labor Intensity of Exports 2,500 20% Total Worker Compensation as USD Millions (2014 prices) Percent of Export Value 2,000 15% 1,500 10% 1,000 500 5% 04 05 06 07 08 09 10 11 12 13 14 04 05 06 07 08 09 10 11 12 13 14 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Direct Indirect Total Source: Authors’ calculations using the LACEX database. Note: Values in 2014 prices. Figure 4.15: Labor Content of Cambodian Exports, 2014 (Forward Linkages) Wearing Apparel Trade and Transport Services Agr, Forestry, Fisheries Other Private Services Leather Products Textiles PubAdmin/Defence/Health/Educat Chemical, Rubber, Plastic Products Processed Foods Machinery and Equipment nec Minerals nec Transport Equipment Manufactures nec Wood Products Construction Beverages and Tobacco Products Electricity, Gas, Water Mineral Products nec Metals nec Paper Products, Publishing Direct Energy Extraction Metal Products Ferrous Metals Indirect Dwellings 0 500 1,000 1,500 Million USD Source: Authors’ calculations using the LACEX database. Note: "nec" stands for "not elsewhere classified." CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 102 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Although the percentage of firms exporting indirectly Firm-level data provide insights on the composition in Cambodia is slightly above the EAP average, the of indirect/domestic linkages. Data on Cambodia from percentage of inputs sourced domestically is far the World Bank Enterprise Survey (WBES) and the Business below that in peer countries. 4 Firm-level data from the Environment and Enterprise Performance Survey (BEEPS) Word Bank’s Enterprise Surveys show that, in 2016, 4.4 show that, in 2016, indirect exporters were mostly small percent of Cambodian firms exported indirectly by providing businesses,5 even though most workers in Cambodia inputs to exporters compared to an EAP average of 3.7 were employed in large firms (Table 4.1).6 Fifty-six percent percent. This rate was higher than in all of Cambodia’s of indirectly exporting firms were in “other services,” peer countries except Malaysia (Figure 4.16). However, followed by 40 percent in manufacturing, and the rest the percentage of total inputs and supplies sourced being in construction, wholesale, and transport. Within the domestically was only 20 percent compared to Vietnam, manufacturing sector, most indirect exporters were in the for example, where this share was closer to 60 percent garments and food sectors. The surveys also found that (Figure 4.17). indirect exporters were generally domestic-owned, located in the capital (Phnom Penh), and had been in operation for an average of about 14 years. Figure 4.16: Firms Indirectly Exporting At Least 10 Percent of Their Sales 9% 8% 7% Percent of Firms 6% 5% 4% 3% 2% 1% 0% Cambodia Lao PDR Malaysia Nepal Thailand Vietnam East Asia (2016) (2016) (2015) (2013) (2016) (2015) & Pacific Source: World Bank Enterprise Surveys, various years. Figure 4.17: Share of Inputs Sourced Domestically Percent of Inputs Sourced Domestically 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Bangladesh Cambodia Ethiopia India Myanmar Pakistan Turkey Vietnam (2013) (2016) (2015) (2014) (2016) (2013) (2013) (2015) Source: World Bank Enterprise Surveys, various years. 4 Among manufacturing firms, 6.9 percent of firms export indirectly. 5 The fact that there were so few medium-sized firms that were indirectly exporting may be due to constraints preventing small-scale firms from becoming medium-sized firms. Because of the small sample sizes in the two surveys, it is difficult to tell for sure whether this was the case. 6 This is similar to the case among directly exporting firms, where large firms provide the most jobs. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 103 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Cambodia’s services exports are important countries except for Nepal, where services exports accounted for jobs for 65 percent of the country’s total exports. In Vietnam, for example, the share is 6 percent and in Thailand, 24 percent. Tourism accounts for the majority of Cambodia’s services Travel services represent 76 percent of Cambodia’s services exports, which are an important part of Cambodia’s exports, and transport services (which includes passenger overall export basket. In 2016, services exports accounted transport) constituted another 10 percent. Other business for 31 percent of Cambodia’s export basket, totaling USD 4 services, such as lawyers and accountants, were more billion. This is a higher share than those in Cambodia’s peer important in Malaysia, Nepal, and Thailand (Figure 4.18). Table 4.1: Composition of Indirect Exporters Distribution of Firms Size Industry Small (0-19) 69.72 % Food 10.74 % Medium (20-99) 80.00 % Garments 23.40 % Large (100+) 28.48 % Leather 2.72 % Sectors Fabricated metal products 3.23 % Manufacturing 40.09 % Construction services 23.97 % Retail 3.67 % Wholesale 18.69 % Other services 56.24 % Retail 3.67 % Hotel & restaurant 1.39 % Type of ownership Transport 12.19 % Domestic 78.02 % Employment share Foreign 21.98 % Small (0-19) 2.00 % Region Medium (20-99) 0.20 % Phnom Penh 70.59 % Large (100+) 97.90 % Plains 11.59 % Share of full-time permanent workers who Mountain 0.19 % 42.60 % completed secondary school Coastal 17.63 % Average age 13.7 years Source: World Bank Enterprise Survey, Cambodia (2016). Note: Sample size equals 25. Figure 4.18: Composition of Services Exports, 2016 100% 90% 80% Goods-related Transport Percent of Services Exports 70% Travel 60% Other 50% Construction 40% Insurance Financial 30% Intellectual Property 20% ICT 10% Other Business 0% Personal, Cultural, Cambodia Lao PDR Malaysia Nepal Thailand Vietnam Recreational Source: Authors’ calculation using data from UNCTADstat. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 104 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Exports of tourism are also important for relatively constant in these peer countries, it has grown Cambodia’s economy, and have strong backward significantly in Cambodia in the last two decades (Figure links. According to the World Travel and Tourism Council, 4.19). This suggests that the sector is a major generator of travel and tourism to Cambodia’s GDP directly contributed 14 employment and that sector-specific targeted policies may percent of GDP in 2016—a higher rate than in comparator be needed. countries. That year, this rate was less than 5 percent in Malaysia, Nepal, and Lao, and was 6 percent in Thailand. Cambodia has significant resources for tourism Beyond this direct economic impact, the travel and tourism that are diverse and well placed to be a pillar for industry has a significant indirect impact as, in 2016, its socioeconomic development as identi ed by the total contribution across all sectors of the economy was 32 government. However, Cambodia was ranked low (101st) percent of GDP—compared with only 20 percent in Thailand on the competitiveness list. Although tourist arrivals exhibit and less than 15 percent in all other comparator countries.7 a stable pattern of travel through regional hubs in Thailand, Thus, tourism in Cambodia has significant backward links. Vietnam, and China, there are some worrying trends that could threaten the chances for developing tourism as a The impact of tourism’s backward links on jobs in central pillar of an inclusive economy. Value captured per Cambodia is much higher than in its peer countries. tourist has not changed much in the last decade (from USD In 2014, travel and tourism directly supported 13 percent 585 in 2005 to USD 655 in 2016), low-end businesses have of total employment, but this increases to 29 percent after mushroomed, stays remain short with limited repeat visits, accounting for its indirect impact across all other sectors and there are signs of overcrowding and degradation of the of the economy, and for its contribution to growth (and key assets at the Angkor temples (World Bank 2017). The thus jobs). This was double the employment contribution average expenditure per visitor in 2016 in Cambodia (USD made by tourism in Cambodia’s peer countries. While the 655) was much lower than it was in Thailand (USD 1,489) total contribution of tourism to employment has stayed or Vietnam (USD 925). Figure 4.19: Total Contribution of Tourism to Employment, 1995-2016 35% 30% Percent of Total Employment 25% 20% 15% 10% 5% 0% 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Year Cambodia Lao PDR Malaysia Nepal Thailand Vietnam Source: World Travel and Tourism Council. 7 See www.wttc.org/economic-impact/country-analysis/methodology/for the methodology that the World Travel and Tourism Council applied consistently across all countries. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 105 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Constraints to jobs exist in the broader information, or limited management capability) that may be business environment preventing small indirect export firms from becoming more productive as they grow. It also highlights the importance The labor productivity of Cambodia’s indirect of focusing on increasing productivity at the firm level to exporters of all sizes is flat, though larger direct increase exports. exporting firms tend to become more productive (Figure 4.20). This trend of flat productivity across firm Data from the World Bank Enterprise Survey for sizes for indirect exporters goes against the stylized fact Cambodia show that direct exporters are more likely to presented in the literature that says larger firms tend to be be large and medium-sized firms than non-exporters more productive. This trend may indicate that economies of (Figure 4.21(a)). Workers in the exporting economy also scale are unattainable, which would also explain why small- tend to be employed in large firms (Figure 4.21(b)). In the scale firms are not growing bigger (see Chapter 3). Given manufacturing sector, firms that export directly (on average) this situation, it will be crucial to identify and dismantle any sell more diverse types of products and import more inputs obstacles (for example, regulatory constraints, a lack of than non-exporting firms. Figure 4.20: Productivity and Firm Size of Exporters versus Non-exporters (a) All Firms (b) Direct Exporters 25 22 Log Labor Productivity (output per worker) Log Labor Productivity (output per worker) 20 20 18 15 16 10 14 2 4 6 8 10 2 4 6 8 Log Firm Size Log Firm Size (c) Indirect Exporters (d) Non−Exporters Log Labor Productivity (output per worker) Log Labor Productivity (output per worker) 22 25 20 20 18 15 16 14 10 2 4 6 8 2 4 6 8 10 Log Firm Size Log Firm Size Source: World Bank Enterprise Survey, Cambodia (2016). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 106 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Figure 4.21: Characteristics of Exporters versus Non-exporters (a) Firm Distribution (b) Employment Share 100% 100% 90% 90% 80% 80% 70% 70% Percent of Employees Percent of Firms 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Non-exporters Indirect Direct Total Non-exporters Indirect Direct Total Exporter Exporter Exporter Exporter Small (0-19) Medium (20-99) Large (100+) Small (0-19) Medium (20-99) Large (100+) Source: World Bank Enterprise Survey, Cambodia (2016). Table 4.2: Characteristics of Exporters verses Non-Exporters, Manufacturing Sector Mean Median Non- Non- Exporters Exporters Exporters Exporters Share of foreign ownership 47.4%*** 11.9% 50% 0% Share of firms with at least 50% foreign ownership 52.6%*** 11.6% 1% 0% Share of firms using imported inputs 78.9%** 52.7% 1% 1% Share of goods that are imported 67.8%*** 37.7% 100% 15% Share of firms importing directly 38.9%* 20.0% 0% 0% Share of firms for whom labor regulations are a major or severe obstacle 57.9%*** 10.8% 1% 0% Share of firms for whom workforce is a major or severe obstacle 42.1%** 20.5% 0% 0% Share of firms for whom labor regulations is a top business obstacle 26.3%*** 6.3% 0% 0% Source: World Bank Enterprise Survey, Cambodia (2016). Notes: Exporters are defined as those firms that export directly. *Difference in means statistically signi cant at the 10 percent confidence level. **Difference in means statistically signi cant at the 5 percent con dence level. ***Difference in means statistically signi cant at the 1 percent con dence level. The characteristics that are not statistically different between exporters and non-exporters (not shown) are: employees, job growth, whether the establishment is part of a larger firm, multiproduct (which identifies if the share of the most important product in total sales is less than 100 percent), product innovation, process innovation, share of skilled workers, training, and the workforce as a top business obstacle. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 107 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Exporters tend to cite an inadequately educated Exporters and foreign firms provide more training workforce as a top business obstacle more often to their workers than do non-exporters or domestic than non-exporters (Figure 4.22). What factors constraint firms. A 2012 Employer Skills Needs Survey undertaken export competitiveness in Cambodia? According to the World in Cambodia by the ILO and National Employment Agency Bank’s Enterprise Survey for 2016, exporters reported that (NEA) gathered information on training by firm ownership the top constraints that they face are: (i) an inadequately (foreign versus Cambodian) and by their main market educated workforce; (ii) having to compete against (international, national, or local). Nearly three-quarters of unregistered and informal firms whose working practices foreign firms and exporting firms reported providing training are different from those of formal enterprises;8 and (iii) tax to their workers compared to 57 percent of domestic firms rates. Although non-exporters also face these constraints, and 61 percent of firms that service the national market. the lack of an educated workforce is more problematic for This suggests that domestic firms and those that serve the exporters than for non-exporters (with the difference being domestic market have the potential to enhance the skills of statistically significant). This concern on the part of direct their workers by providing on-the-job training. exporters had increased significantly in 2016 compared to 2007 and 2013.9 Figure 4.22: Biggest Obstacles Faced by Exporters and Non-exporters 100% 90% 80% 70% 60% Percent of Firms 50% 40% 30% 20% 10% 0% Direct Exporter Non-exporter Direct Exporter Non-exporter Direct Exporter Non-exporter 2007 2013 2016 Access to Finance Access to Land Business Licensing and Permits Corruption Courts Crime, Theft and Disorder Customs and Trade Regulations Electricity Inadequately Educated Workforce Labor Regulations Political Instability Practices of the Informal Sector Tax Administration Tax Rates Transportation Source: World Bank Enterprise Survey, Cambodia (2016). 8 Given that this is a constraint, it will be necessary to understand precisely what this means in order to develop a tailored solution. 9 Results for 2016 biggest obstacles are similar when looking at perception of top business constraints for exporting and non-exporting firms that are only available in the panel dataset. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 108 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR 4.3 Policies for Fostering a domestic producers, will be vital for creating more export jobs. Since not all FDI has spillovers, it will be Job-friendly Export Agenda for important to attracting the “right” foreign investors. Cambodia Therefore, it will be necessary to avoid those that concentrate on low-value-added activities, that retain There is scope for Cambodia to leverage its existing external control of sourcing, and that employ expatriates export sector to create more and better jobs. The in manager and technical positions (Staritz and Frederick export sector can continue to be a source of jobs that 2014). A lack of spillovers can be aggravated by weak provide better pay and work conditions than most jobs in domestic absorptive capacity (including a low-skilled the current labor market (see, for example, Bambrilla et al. labor force, for example), a poor domestic business 2012). In this chapter, we have identified those parts of the environment, a lack of adequate support for SMEs, and export sector that are more conducive to job growth and the absence of a local entrepreneurial support for firms. have argued that adopting policies designed to target these Instead, the government should consider adopting and characteristics may lead to higher quantity and quality of implementing an investment strategy aimed at attracting jobs in the sector than if the market alone were to drive the FDI into new productive sectors and activities with a transformation. In this section, we identify a variety of policy high potential to generate spillovers. As a first step, the interventions that would support the creation of more, better, government should review and potentially reform its and more inclusive jobs in the competitive export sector, both existing investment incentives (World Bank 2018b), for directly and indirectly, while also taking into consideration example by progressively phasing out tax holidays and the emerging megatrends that were discussed in Chapter introducing investment tax credits that foster private 1. These policy interventions involve: (i) encouraging firms investment in machinery and worker training. Another part to move into higher value-added activities in their current of the strategy should involve streamlining the approval value-chain sectors and to move into entirely new value- process for receiving these incentives and implementing chain sectors; (ii) encouraging foreign-owned firms to buy clear, pre-defined, and objective eligibility criteria and more inputs and services from domestic small and medium- establishing a standard duration for the credits. sized enterprises (SMEs); and (iii) encouraging firms to diversify their exports. (2) Reduce rms’ trade and transportation costs and improve logistics to attract investment—and These three policy priorities are closely related. All create jobs—in GVC sectors. Governments that facilitate three can be pursued in parallel in terms of both short-term the movement of imports and exports are likely to attract and medium-term actions. Also, all three have a common investment and help domestic private firms to participate underpinning, which is discussed in other chapters of this and compete in the international trading system. Therefore, report. For example, skills development and improving the for Cambodia to attract investment in GVC sectors, it will be business environment underpin export competitiveness and necessary to speed up lead-time in supply chains, modernize thus also underpin export diversification, backward links, and trade procedures, and develop a logistics infrastructure. the participation of exporters in GVCs. Important first steps have been taken, including the establishment of a National Logistics Council and a National Policy Area 1: Encouraging Firms to Move Up Single Window. Although Cambodia has reduced the Existing Value Chains number of documents required for export and import, it is still a long, costly, and complicated process to comply with The first general policy area is to encourage firms all regulatory, documentary, and border requirements to to move into higher value-added activities within import to and export from the country (World Bank 2018a). existing value chains. This will help to create more Alleviating these constraints will support not only GVC and better export-based jobs. Given the fact that wages sectors but also domestic exporting firms. The business are rising in Cambodia, this will be necessary in order to community should have easy access to and be aware of all maintain a competitive edge. Specific policy areas under regulatory requirements and the different procedural steps this heading might involve: (i) attracting foreign investment required for exporting and importing. Currently, although the in new GVC segments and/or sectors; (ii) reducing firms’ requirements are stated in laws and regulations, they often trade and transportation costs; and (iii) supporting training are not well understood by entrepreneurs. Local authorities in the skills needed in the country’s export sectors. can help by publishing a user-friendly guide for the most frequently used procedures (World Bank 2018a). Increasing (1) Attract foreign investment in new GVC segments transparency in this way would also reduce the amount of or sectors, especially those with spillovers to discretion that of cials are able to exercise. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 109 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR A legal framework for electronic government services Policy Area 2: Encourage Exporting Firms to needs to be put in place. Providing these services for Create Links with Small and Medium-sized businesses can increase transparency and support the implementation of the new, streamlined procedures, but the Domestic Businesses legal framework governing electronic transactions has not A second key policy area relates to encouraging yet been developed in Cambodia (World Bank 2018a). A few exporting firms (including those in tourism) to buy processes can be completed online, but in most cases— inputs and services from domestic supply firms. including the registration of property transactions, construction These links can be an important way for exports to create permits, and approvals to import and export—electronic jobs indirectly, and the government can foster these procedures have not yet replaced the use of paper documents. links in three main ways: (i) by supporting local supplier However, these processes will need to be simplified before development programs; (ii) by providing value chain these electronic government services are developed. financing to suppliers, distributors, and retailers; and (iii) Various other issues will need to be addressed to helping indirect exporters, particularly SMEs, to increase facilitate trade in the future. The prevalence of informal their productivity and capacity to enable them to participate payments in the export sector currently increases trade in global value chains. costs in Cambodia (World Bank 2018a). If these payments are fair compensation for the service provided, they should (1) Local supplier development programs can help be formalized. With regard to the practicalities of trade, domestic enterprises to become qualified suppliers encouraging the appropriate use of technology such as GPS or subcontractors for FDI firms. However, there is a systems would make shipments more secure and cargo debate about whether these types of programs would be transit more efficient (World Bank 2018a). Automating effective in Cambodia. Cambodia does not have a lot of the process of sanitary and phytosanitary certification domestic companies that are active in the FDI-dominant would ensure the quality of goods being exported, while manufacturing sectors, which makes it challenging to streamlining customs requirements and clearance for realize any supply chain spillovers. The garment value chain export and import licenses would reduce the time and costs is relatively short as the sector needs only a small number involved in trading across Cambodia’s borders. of inputs that are produced globally in large quantities and at low cost. A recent World Bank survey has shown that FDI (3) Upgrade the skills of Cambodian workers to firms in the garments sector in Cambodia have little interest enable them to participate in GVCs that require in increasing local sourcing and have little decision-making workers to perform more sophisticated tasks. Skills power about purchasing inputs, where the decisions are mismatches are currently a significant constraint on firms made by the lead firms (World Bank 2018a). Therefore, the operating in Cambodia, and in order to increase productivity best chances of realizing spillovers in the garment sector in and innovation, the skills of the Cambodian labor force will Cambodia would be for domestic firms to provide services need to be upgraded. This will mean improving their basic rather than inputs to FDI firms or for new domestic firms to education outcomes as well as increasing their specific enter the garment sector. and practical skills (for example, in ICT, craftsmanship, and language competency). With the right policies, the Nevertheless, FDI firms operating in Cambodia are government will be able to attract more private investment the most obvious market for domestic firms besides in training institutions, encourage firms to provide their the limited domestic market and a viable path to workers with on-the-job training, facilitate training for transfer technology. A recent World Bank survey found workers employed by small firms, or foster partnerships that Cambodia’s SEZs operators have reported that local among large enterprises to train their workers. Introducing suppliers are desired by some FDI tenants (World Bank tax incentives for R&D and staff training can also give firms 2018a). This suggests that policymakers should consider an incentive to provide training. The policies needed to creative approaches to developing the capacity of local firms support skills development will be discussed in Chapter 5. and finding them opportunities to supply FDI firms. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 110 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Providing domestic entrepreneurs and SMEs with to purchase or source from local SMEs would be to simplify information and other support can help them to build and improve the administration of VAT reimbursement so links with firms in FDI sectors. There are various ways that firms can get their refunds in a timely manner. As for for policymakers to do this, including compiling a guidebook supporting the development of new entrepreneurs, the or database of suppliers with adequate technology and forthcoming Entrepreneurship Promotion Fund can be used capacity to provide inputs and services for FDI firms and/or to help workers who have left employment in foreign firms providing matching services to coordinate between suppliers to start a business to gain access to capital, finance, labor, and buyers (see Box 4.3). Another way to persuade FDI firms and land. Box 4.3: International Experience in Backward Linkages Programs Inter-sectoral linkages between FDI/lead firms and domestic firms can be the conduit for transferring knowledge, new technology, management skills, marketing techniques that can lead to increased firm productivity. However, establishing linkages is not an automatic process, and different countries have used varying approaches. Costa Rica launched a three-year pilot project, Provee (CR Provee), in 2001 to strengthen SME linkages with FDI firms. This was co-funded by The Inter-American Development Bank (which provided 60 percent support), Costa Rica’s investment promotion agency (CINDE) and the country’s export promotion agency (PROCOMER). After the three-year pilot, CR Provee was formally integrated into PROCOMER in 2004. To date, CR Provee focuses on key sectoral components namely: ICT, Electrical Electronics, Metal Mechanics sectors; Medical, Chemical, Pharmaceutical sectors; Agribusiness; and Textile sectors. Among over things, CR Provee addresses market failures related to lack of information and coordination failures, though various support services. These include (1) an Integrated Information System that manages buyer and supplier data, (2) mapping potential suppliers to multinational corporation (MNC)- buyers’ value chain after identifying MNCs’ input and raw material requirements in a value chain, and (3) the identification of business opportunities that SMEs and MNCs can engage in through ‘small’ projects. In the Czech Republic, CzechInvest was created in 1992 to be the country’s investment and business development agency. In 2000-2003 during the country’s entry to the EU the agency began focusing on creating a strong local supplier network that can meet investor demand for inputs in key sectors. Among other services, CzechInvest attempts to address coordination failures by organized matchmaking services and related support services, such as (i) supplier forums with/for local suppliers and MNCs; (ii) Meet-the-Buyer events between local suppliers and MNCs, where MNCs can communicate their specific input needs and suppliers can share their product portfolio; and (iii) submission of suitable supplier proposals to potential MNCs buyers. As part of Malaysia’s industrialization strategy, government linkages programs started in 1988 through the Vendor Development Program (VDP), which focused initially in automotive before expanding into electronics by 1992. The program aimed to support local firms to be suppliers to large domestic firms and/or MNCs. There are complementary SME financial support programs to support SMEs attempting to develop linkages that address lack of access to finance. For example, the Small Loans Guarantee Scheme by the Credit Guarantee Corporation Malaysia Berhad (CGC) was created in 1973 to guarantee loans for small businesses which cannot afford collateral. Also, the SME Credit Bureau was established in 2008 and serves as a repository of SME information to facilitate loan access from financial institutions such as banks. Another example is the SME Bank, which works similarly as a commercial bank but is targeted for SMEs. SME Bank also has complementary services, such as business advisory services to support loan approval processes. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 111 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Joint ventures between FDI firms and local firms can (3) Increasing the productivity of indirect exporters, help to develop the capacity, technology, and skills in particular SMEs, will make it easier for them to be of the local firms. Box 4.4 gives an example of how Sri integrated into GVCs. FDI firms often have high quality Lanka used joint ventures between foreign and local firms in and productivity standards for their local suppliers, and, as the apparel sector to enable the local firms to upgrade their we have shown in this chapter, the productivity of indirect processes to the level needed for them to provide inputs in exporters in Cambodia is particularly low. Specific policies GVCs. that would help to increase the productivity and capacity of these direct exporters and of SMEs more generally as (2) Consider providing value chain financing to well as removing barriers to growth and formalization domestic firms to help them grow and become more were discussed in Chapter 3. Other policies that would competitive. This financing would enable suppliers to increase the productivity of domestic firms might include obtain financing for their working capital operations at lower providing incentives to firms to encourage local work costs or distributors and retailers to obtain financing to buy force development, R&D, local sourcing, or activities to help larger orders from local suppliers. Through the recently their suppliers to improve their processes and increase their established SME Bank, policymakers could target value capacity (see Box 4.5). chain financing to SMEs who wish to become suppliers with no requirement that they should already have a relationship with a buyer. Box 4.4: The Role Played by Joint Ventures in Upgrading the Sri Lankan Apparel Sector Joint ventures (JVs) have played a key role in establishing and developing the apparel industry in Sri Lanka. These JVs have taken several different forms. Some have been formed between local manufacturers and the global buyers of their products or, in the case of textiles and sundries, between a global supplier of the textile product and a local firm that supplied them with inputs. Other JVs also included end buyers. These direct relationships, which were a requirement for foreign investors in Sri Lanka, have enabled local firms to increase their capacity and improve their processes and products. This has particularly been evident in the case of intimate apparel products such as lingerie and swimwear, of which Sri Lanka has become a leading global supplier with the support of foreign investors who enabled domestic firms to invest in equipment and human resources. JVs have helped domestic rms not only to adopt modern technology but also to upgrade their technical and managerial skills. For example, Textured Jersey of the UK sent 10 of its managers to its affiliate Textured Jersey Lanka, to train their local counterparts over three to five years. Furthermore, when Textured Jersey’s shares were bought by Pacific Textiles of Hong Kong SAR, China, that company based full-time seconded personnel in Sri Lanka to transfer best practices and technical expertise to Sri Lankan workers. The result was a significant reduction in foreign (mainly Chinese) technical personnel in favor of qualified Sri Lankans. Not only have JVs between suppliers and buyers in apparel encouraged Sri Lankan suppliers to upgrade their production and skills, but buyers have also encouraged their other textile and sundry suppliers to move to Sri Lanka. For instance, because of the close relationship that MAS Holdings has with Victoria’s Secret, Victoria’s Secret asked some of their input providers to relocate to Sri Lanka, including the suppliers of various components that go into making bras and panties like lace and pads and also warp knit fabric. Source: Staritz and Frederick (2014). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 112 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Box 4.5: Export Promotion in Jamaica Despite its small size, Jamaica has managed to outperform other small countries in attracting foreign investment. Jamaica’s economy relies on commodity exports (aluminum), tourism, and remittances. It has a narrow production base and weak, underdeveloped infrastructure. However, in 2014, Jamaica attracted almost USD 700 million in FDI, 7 percent more than in 2013 and the largest amount since 2008, and equivalent to 4.28 percent of GDP. The ICT sector has been increasingly important, with foreign investors setting up call centers and, more recently, software development facilities. Jamaica has also diversified its export base, with 73 percent of exports now being in services. JAMPRO (Jamaica Promotions Corporation, formerly known as Jamaica Trade & Invest) is in charge of both export and investment promotion. In terms of export promotion activities, JAMPRO provides local businesses, particularly SMEs, with a range of technical and advisory support for export readiness and competitiveness. In 2011, it launched the “Export Max: Enterprise Development for Export Growth” program, aimed at providing focused capacity building and market penetration support to a group of Jamaican exporters and export-ready firms. Under the program, a number of services are offered, including company diagnostics and needs assessment, business matchmaking, quality standards development, and mentoring activities. JAMPRO relies on a targeted strategy to increase Jamaican exports, focusing on agro-processing and agriculture, creative industries, manufacturing, energy and mining, ICT, and trade in services. In 2012, JAMPRO received the award for best TPO from a Small Island Developing State, the third time the agency has achieved international recognition since 2004. JAMPRO is also in charge of investment promotion, offering investors a range of services from information about opportunities in Jamaica, to support in the early stages of establishing a presence within Jamaica, to aftercare services on their investments are operational. Working in partnership with other government agencies, JAMPRO guides investors through the necessary processes and regulatory requirements. Taking advantage of its position as both and TPO and IPA, JAMPRO has a “Business Linkages Program” seeking to promote links between foreign investors and local suppliers. The program has resulted in over 500 linkages contracts. It also helps in identify joint ventures of financing partners for exporting companies seeking to expand or retool. Finally, JAMPRO also assists SMEs in understanding how they might fit into global value chains. Indeed, the export development and investment promotion divisions of the organization work closely together to identify inter-departmental opportunities. A Global Business Connect Strategy seeking to increase the value of the country ‘s global trade and investment relations was launched in 2014. The program seeks to identify Jamaicans and members of diaspora that can act as potential investors and distributors. In 2014, JAMPRO was selected as the winner of UNCTAD’s award for excellence in promoting export-oriented FDI. Source: Gomez-Mera (2016) Policy Area 3: Encouraging Firms to (1) In order to support the emergence of new export Diversify Their Exports products, it will be necessary for policymakers to identify the sector-speci c barriers preventing Export diversification will be key to improving the domestic firms from expanding into new products. jobs and wages available in Cambodia. Policymakers The next step will then be to provide them with the need to find ways to encourage this diversification with necessary inputs to enable them to move into new activities a particular focus on jobs-friendly export sectors. There and to operate their existing activities more productively. are four policy areas that have the potential to do this: (i) Establishing a public-private dialogue, for example in supporting the identification and creation of new export a business dialogue forum, could help the government products; (ii) increasing Cambodia’s integration with its to identify the nature of challenges faced by firms. In regional neighbors and other international trading partners; the medium run, the government should ensure that the (iii) building the human capital of service providers; and country’s export promotion efforts are aligned with good (iv) reducing the regulatory burden on SMEs to make them international practices (World Bank 2018a). more competitive. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 113 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR (2) There is scope for exploiting Cambodia’s that are increasingly in demand are those that require international links to diversify its exports. Export higher-value skills (rather than those needed for low-skilled diversification and better labor outcomes will come by assembly for export), including design, R&D, marketing, diversifying in terms not only of products but also of and after-sale services as well as supporting professions destinations. For example, the analysis above shows that such as accounting and legal services. Unless Cambodia Cambodia’s exports to EAP region are more sophisticated has workers with the necessary skills, its ability to provide than are its exports to the US or the EU. Therefore, regional these services will be greatly undermined. While Cambodia trade agreements will be key to increasing Cambodia’s has one of the most open policies in services in the region market access to these countries as well as enabling the as a result of the country’s regulatory reform, it still has movement of capital and labor. There also appears to be some restrictions on professionals from other countries untapped potential for Cambodia to increase services filling jobs in Cambodia. Concluding mutual recognition exports with specific countries in the region (World Bank agreements on the free movement of professionals between 2015) in such sectors as tourism (World Bank 2017). In countries would help to attract foreign professionals with the the short run, signing, finalizing, and enforcing international necessary skills to work for Cambodia’s service providers in trade and investment agreements with Cambodia’s key the near term. However, in the longer term, Cambodia will trade partners will help to reinforce its existing markets need to nurture its own service professionals with skills up (World Bank 2018a). These include agreements with the to international standards in logistics, marketing skills, and, EU and the US to ensure preferential trade treatment once most importantly, management. Cambodia graduates from LDC status, as well as supporting the deepening of the ASEAN Economic Community and the (4) Reducing the regulatory burden on SMEs will Regional Comprehensive Economic Partnership (RCEP) FTA reduce their operating costs and help them grow and between ASEAN and Asia-Pacific countries. become more competitive. Increased competitiveness is essential not only for export competitiveness but also to (3) It will also be important to build the human capital enable the development of new export products and markets. of workers in firms that provide professional services. Some of the ways in which policies can mitigate the burden The emerging knowledge economy and the rise of an East on domestic SMEs would be to reduce their electricity Asian middle class means that services have the potential costs, increase the ease of doing business by abolishing to become a larger share of Cambodia’s export earnings, unnecessary procedures and informal fees including those both directly and indirectly, as inputs into manufacturing for opening a business (see Chapter 3), and reducing trade and agricultural production. However, the kind of services transaction costs (see above). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 114 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR 4.4 References Brambilla, Irene, Daniel Lederman, and Guido Porto. 2012. “Exports, Export Destinations, and Skills.” American Economic Review 102(7): 3406-38. Cadot, Olivier, Céline Carrère, and Vanessa Strauss-Kahn. 2011. “Export Diversification: What’s behind the Hump?” The Review of Economics and Statistics 93(2): 590-605. Calì, Massimiliano, Joseph Francois, Claire H. Hollweg, Miriam Manchin, Doris Anita Oberdabernig, Hugo Rojas-Romagosa, Stela Rubinova, and Patrick Tomberger. 2016. “The Labor Content of Exports Database,” World Bank Policy Research Paper No. 7615, Washington, D.C.: World Bank. Francois, Joseph, Miriam Manchin, and Patrick Tomberger. 2013. Services Linkages and the Value Added Content of Trade. Policy Reach Working Paper No. 6432, Washington, D.C.: World Bank. Gomez-Mera, Laura. 2016. “Designing Institutions to Promote Trade and Investment in Nepal” in Guillermo Arenas ed. From Evidence to Policy: Supporting Nepal’s Trade Integration Strategy. Washington, D.C.: World Bank. Hausmann, Ricardo, Jason Hwang, and Dani Rodrik. 2007. “What you export matters.” Journal of Economic Growth 12(1): 1-25. Hertel, Thomas W. 2013. “Global Applied General Equilibrium Analysis using the Global Trade Analysis Project Framework.” In Peter Dixon and Dale Jorgenson, eds. Handbook of Computable General Equilibrium Modeling, North Holland: Elsevier. Maliszewska, Maryla, Zoryana Olekseyuk, and Israel Osorio-Rodarte. 2018. Economic and Distributional Impacts of Comprehensive and Progressive Agreement for Trans-pacific Partnership: The Case of Vietnam. Washington, D.C.: World Bank. McDougall, Robert, ed. 2001. The GTAP Database: Version 5 West Lafayette: Center for Global Trade Analysis, Purdue University. Mera, Gomez. 2016. “Policy Note 5: Designing Institutions to Promote Trade and Investment in Nepal” in Varela, Gonzalo, ed. From Evidence to Policy: Supporting Nepal’s Trade Integration Strategy Washington, D.C.: World Bank. Shirotori, Miho, Bolormaa Tumurchudur, and Olivier Cadot. 2010. “Reveal Factor Intensive Index at the Product Level.” Policy Issues in International Trade and Commodities Study Series #44, New York: United Nations Conference on Trade and Development. Staritz, Cornelia, and Stacey Frederick. 2014. “Sector Case Study: Apparel” in Thomas Farole and Deborah Winkler, eds. Making Foreign Direct Investment Work in Sub-Saharan Africa: Local Spillovers and Competitiveness in Global Value Chains Washington, DC: World Bank. World Bank and ASEAN Secretariat. 2015. ASEAN Services Integration Report. Washington D.C.: World Bank. World Bank, 2017. Cambodia Economic Update: Cambodia Climbing Up the Manufacturing Value Chain. Washington D.C.: World Bank. World Bank. 2018a. Cambodia Investment Climate Assessment Washington, D.C.: World Bank. World Bank. 2018b. Improving the Effectiveness of Public Finance: Cambodia PER. Washington, D.C.: World Bank. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 115 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR ANNEX 4.1: Labor Sophistication of Exports We examined trends in export sophistication, by export destination, to infer potential implications for labor market outcomes. We constructed six indices or “EXPYs” that measured the following aspects of labor sophistication: (i) median wage; (ii) median value-added per worker; (iii) the ratio of skilled to total workers; (iv) output per employee; (v) years of schooling or human capital; and (vi) capital stock per worker. We benchmarked each measure—over time, across destination markets, and relative to Cambodia’s peer countries—to infer the potential implications of export growth on labor market outcomes. These EXPYs reflect the trade-weighted average labor market outcome of the exports that appear in Cambodia’s export basket. The logic behind these measures is that labor market and other characteristics of the economies that intensively export a given product provide information about the implications of specializing in certain products exported to certain markets and determine how export growth will contribute to job creation, wages, and skill mix. Based on these observations, it was possible for us to draw inferences about how product and regional trade patterns can impact employment, wages, and skill demand in Cambodia. We calculated EXPYs for each of the six labor market outcomes (wages, value-added, skill ratio, output per employee, human capital, and physical capital) in two steps following Hausmann et al (2007). In the first step, we calculated a PRODY value for each product, in other words, the global trade-weighted average labor market outcome covering all of the countries in the world that export the product. Therefore, we calculated a wage PRODY reflecting the average international wage of countries that export the product, a skill PRODY reflecting the average share of skilled workers in countries that export the product, and so on. In the second step, we then weighted the PRODYs in Afghanistan’s export basket by the share of each product in Afghanistan’s total exports, denoted as EXPY. The formulas are: where Xi,j is exports from country i in sector j, Xi is total exports of i, and Yi is the labor-market outcome of i. These resulting EXPYs reflect the global average labor market outcome of goods that appear in Afghanistan’s export basket. In calculating the PRODYs and EXPYs, we used mirror data from UN Comtrade, the World Bank’s Business Environment and Enterprise Performance Surveys, and Shirotori et al (2010). For each of these labor-related EXPYs for Cambodia, we: (i) looked at how its level had evolved over time since 2000; (ii) considered variations in labor sophistication across different destination markets, comparing the average for all Afghan exports (in other words, destined for the “world”) with those destined for the EU28, the US, and South Asia; (iii) made bilateral comparisons of labor sophistication with a range of regional exporters; and (iv) repeated each of these analyses across the entire distribution to see what products were driving the observed differences. Two caveats are warranted. We measured the PRODYs as the average across all countries across multiple years (in other words, holding them constant). This ensures that changes in the EXPYs reflect changes in the composition of countries’ export basket rather than changes in the global average of the labor-related variables. Because the EXPYs are an indirect measure, they may over- or under-estimate the actual factor content of a country’s export basket. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 116 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR ANNEX 4.2: Labor Content of Exports The World Bank’s Labor Content of Exports (LACEX) database encompasses information on the contribution of labor to a given country’s domestic production and exports measured as employees’ compensation (wages) or the number of jobs. The LACEX database was computed by Calì et al. (2016) on the basis of a panel of global input-output tables and other aggregate data spanning intermittent years from 1995 to 2011 from The Global Trade Analysis Project (GTAP) and employment data from the ILO.10 The input-output tables in the GTAP data set make it possible to exploit a form of social accounting data, a variation on the social accounting matrix (SAM) where incomes are shown in the rows of the SAM and expenditures are shown in the columns (see Hertel, 2013 and McDougal, 2001). The structure of the underlying social accounting data provides a comprehensive and consistent record of national income accounting relationships between different sectors, including intermediate and final demand links. These are then used to construct country-specific measures of the contribution of labor to the value-added contained in a given country’s domestic production and exports.11 In this annex, we illustrate how Calì et al (2016) generated the (direct and indirect) value-added associated with each sector- country using the panel of global input-output tables from GTAP. The methodology follows Francois et al (2013), which formed the basis for the computation of the compensation of employees. Calì et al. (2016) began by denoting a representation of intermediate and final demands as follows: Y = Z – AZ (1) In equation (1), the term Y denotes a final demand vector, Z denotes a gross output vector, and A denotes a matrix of intermediate use coefficients, which can be calculated using data from input-output tables. Equation (1) therefore defines final output with respect to intermediate input requirements. The first matrix that Cali calculated was the widely used Leontief inverse matrix, also known as the multiplier matrix M. With some manipulation, Calì et al. (2016) arrived at the Leontief inverse matrix: Z = (I – A)-1Y = MY (2) The multiplier matrix M measures the inputs contained in a unit of final output. This M matrix contains both direct and indirect inputs. In particular, if assigning the sector indexes i,j to the M matrix, then a representative element of the M matrix, Mij, gives the direct and indirect inputs (and thus receipts) of sector i linked to each unit (for example each dollar) of sector j receipts in the data. This implies real production activities measured by value of output. For our purposes, it provided a way to trace through these income flows, the flow of gross activity (and eventually labor value-added) from intermediate to final goods and services, ostensibly across borders as well as sectors. Because links will vary by industry, each industry will be characterized by different multipliers. Next, Calì et al (2016) needed to calculate a diagonal matrix that measured the share of labor value-added (compensation of employees) in total gross output, which can be calculated using data from SAMs. To focus on labor value-added rather than gross activity, Calì et al (2016) noted first that, in terms of gross output values Z, some share of this involves labor value- added within each sector. Calì et al (2016) defined B ˆ as the diagonal matrix indexed over i,j with diagonal elements equal to the compensation of employees’ shares of output Z. These shares can be split between types of workers (for example, skilled versus unskilled). 10 GTAP represents a massive combined effort of international institutions and universities. Over time, the data set has grown to include more countries and more sectors. To maintain backward compatibility, Calì et al. (2016) started with the 1997 structure of regions and sectors and have carried this forward in the more recent iterations of the data set. The GTAP website provides extensive documentation on the underlying data structure, its sources, and the GTAP model structure for each release (www.gtap.org). 11 In the remainder of the paper, we refer interchangeably to total labor value-added content, wages, and compensation to employees. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 117 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR ˆ to provide a breakdown of the flow of labor value-added across activities in the form of Calì et al (2016) then used M and B the matrix V: ˆM V=B (3) Similar to the Leontief inverse matrix itself, the V matrix identifies the inputs of labor value-added in each sector related to a unit of final demand. Using the V matrix as a multiplier for exports and final outputs, one can obtain the total labor value-added of exports and final outputs respectively. Calì et al (2016) multiplied V by the diagonal matrix Ŷ whose non-zero elements are the vector of final demand. The resulting matrix yielded a breakdown of economy-wide labor value-added (both direct and indirect), which is the primary component of gross national product. Similarly, by multiplying V by the diagonal matrix X ̂ whose non-zero elements are the national export vector, Calì et al (2016) recovered the labor value-added content of exports X (both direct and indirect): G = VŶ (4) H = VX̂ (5) The G matrix and the H matrix thus gave the set of links, both direct and indirect, between value-added across sectors for total domestic demand and export demand respectively. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 118 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR ANNEX 4.3: Additional Figures and Tables When we examined the cumulative distribution of labor measures across the entire range of products that Cambodia exports, we found quality improvements in many export product lines. We plotted cumulative distribution functions across export products rather than just comparing weighted average export labor values. The distributions for 2000 and 2016 showed labor sophistication/quality improvements in both the lower and upper ends of the distribution for median wage, median value-added, capital per worker, and human capital. This implies that most of Cambodia’s exports are in products associated with higher wages, value-added, physical capital, and human capital. However, output per employee had already increased in 2000 for products with higher output per employee, but by 2016, output per employee seems to have slightly declined for those products with lower output per employee in 2000. Nonetheless, product lines that employed fewer skilled workers in 2000 seemed to have become slightly more sophisticated by 2016, increasing their exports of goods produced by more skilled workers. Overall, this shows that Cambodia’s export product mix is heterogenous and that there are product lines, depending on their location on the labor sophistication distribution, where there is still scope to improve labor quality and increase labor demand. Across almost all product exports in the distribution, Cambodia’s exports to EAP have higher wages, value added, physical capital, and human capital content than their exports to the EU27 and the US. This supports the findings of previous research and adds further evidence of the need for export diversification to products that are more sophisticated in terms of labor content. Overall, there has been export diversification on some margins driven by markets. Market concentration has increased much faster than product concentration across sectors, and these increases have been driven by the textiles and clothing and footwear sectors. Currently, market concentration levels in Cambodia are similar to the levels in Malaysia, Thailand, and Vietnam. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 119 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Annex Figure 4.1: Cumulative Distributions of Cambodia’s Export Basket to the World, 2000-2016 (a) Median Wage (b) Median Value Added 100% 100% 80% 80% Cumulative Distribution Cumulative Distribution 60% 60% 40% 40% 20% 20% 0% 0% 8 9 10 11 12 13 8 9 10 11 12 13 PRODY Median Wage (PPP Thousands) PRODY Median Value Added (PPP Thousands) (c) Physical Capital per Worker (d) Output per Employee 100% 100% 80% 80% Cumulative Distribution Cumulative Distribution 60% 60% 40% 40% 20% 20% 0% 0% 8 9 10 11 12 13 11 13 15 17 PRODY Physical Capital per Worker (USD Thousands) PRODY Output per Employee (PPP Millions) (e) Skilled Workers (f) Human Capital 100% 100% 80% 80% Cumulative Distribution Cumulative Distribution 60% 60% 40% 40% 20% 20% 0% 0% 0 .2 .4 .6 .8 1 2 4 6 8 10 12 PRODY Skilled Workers (Share of Workforce) PRODY Human Capital (Years of School) 2000 2016 Source: Author’s calculations based on data from UNCOMTRADE, the World Bank’s Business Environment and Enterprise Performance Surveys, and Shirotori et al (2010). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 120 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Annex Figure 4.2: Cumulative Distribution of Cambodia’s Export Basket to the EU28, EAP, and the US, 2016 (a) Median Wage (b) Median Value Added 100% 100% 80% 80% Cumulative Distribution Cumulative Distribution 60% 60% 40% 40% 20% 20% 0% 0% 8 9 10 11 12 13 8 9 10 11 12 13 PRODY Median Wage (PPP Thousands) PRODY Median Value Added (PPP Thousands) (c) Physical Capital per Worker (d) Output per Employee 100% 100% 80% 80% Cumulative Distribution Cumulative Distribution 60% 60% 40% 40% 20% 20% 0% 0% 8 9 10 11 12 13 11 13 15 17 PRODY physical capital per worker (USD Thousands) PRODY Output per Employee (PPP Millions) (e) Skilled Workers (f) Human Capital 100% 100% 80% 80% Cumulative Distribution Cumulative Distribution 60% 60% 40% 40% 20% 20% 0% 0% 0 .2 .4 .6 .8 1 2 4 6 8 10 12 PRODY Skilled Workers (Share of Workforce) PRODY Human Capital (Years of School) EU27 USA EAP Source: Author’s calculations based on data from UNCOMTRADE, the World Bank’s Business Environment and Enterprise Performance Surveys, and Shirotori et al (2010). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 121 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Annex Figure 4.3: Share and Value of Cambodia’s Total Exports to EU27, China, the US, and EAP (a) Share of Exports to EU27 (b) Value of Exports to EU27 100% 5 Percent of Exports to EU27 80% 4 USD Millions 60% 3 40% 2 20% 1 0% 0 2000 2002 2004 2006 2008 2010 2012 2014 2016 2000 2002 2004 2006 2008 2010 2012 2014 2016 Year Year (c) Share of Exports to China (d) Value of Exports to China 100% 0.20 Percent of Exports to China 80% 0.15 USD Millions 60% 0.10 40% 0.05 20% 0% 0 2000 2002 2004 2006 2008 2010 2012 2014 2016 2000 2002 2004 2006 2008 2010 2012 2014 2016 Year Year (e) Share of Exports to the United States (f) Value of Exports to the United States 100% 3 Percent of Exports to USA 80% 2 USD Millions 60% 40% 1 20% 0% 0 2000 2002 2004 2006 2008 2010 2012 2014 2016 2000 2002 2004 2006 2008 2010 2012 2014 2016 Year Year Vegetable Products Foodstuffs Textiles Footwear Transport Equipment Other Apparel CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 122 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR (g) Share of Exports to the East Asia Pacific (h) Value of Exports to the East Asia Pacific 50% 2.0 Percent of Exports to EAP* 40% 1.5 USD Millions 30% 1.0 20% 0.5 10% 0% 0 2000 2002 2004 2006 2008 2010 2012 2014 2016 2000 2002 2004 2006 2008 2010 2012 2014 2016 Year Year Vegetable Products Foodstuffs Textiles Footwear Transport Equipment Other Apparel Source: Authors’ calculations using UNCOMTRADE data from WITS. Annex Figure 4.4: Composition of Cambodia’s Export Basket (a) 2000 (b) 2016 100% 100% 80% 80% Percent of Exports Percent of Exports 60% 60% 40% 40% 20% 20% 0% 0% World USA EU27 EAP World USA EU27 EAP Animal Products Vegetable Products Fats & Oils Foodstuffs Mineral Products Chemicals Plastics & Rubber Leather & Skins Wood & Pulp Textiles Footwear Precious Metals Base Metals Machinery Transport Equipment Other Apparel Source: Authors’ calculations using UNCOMTRADE data from WITS. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 123 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Annex Figure 4.5: Intensive and Extensive Margins (a) All Sectors Increase of existing products** Decrease in existing products** Export extinction of products** Intro of new products* Intro of new products** Intro of existing products* Product diversification** 0 20 40 60 80 Intensive and Extensive Margins between 2000 and 2016 New markets (*), Established markets (**) (b) Textiles and Clothing Sector Increase of existing products** Decrease in existing products** Export extinction of products** Intro of new products* Intro of new products** Intro of existing products* Product diversification** 0 20 40 60 80 100 Intensive and Extensive Margins in Textiles & Clothing New markets (*), Established markets (**) Source: Authors’ calculations using UNCOMTRADE data from WITS. Note: The intensive margin measures the selling of the same product to the same market. The extensive margin measures doing something new—by selling a new product to an existing market, an old product to a new market, or both. Annex Figure 4.6: Labor Intensity by Sector Manufacturing 2004 Services Primary Manufacturing 2007 Services Primary Manufacturing 2011 Services Primary Manufacturing 2014 Services Direct Primary Indirect 0% 10% 20% 30% 40% LVAX Source: Authors’ calculations using the LACEX database. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 124 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Annex Figure 4.7: Labor Intensity of Exports across Sectors and Years PubAdmin/Defence/Health/Educat Agr, Forestry, Fisheries Processed Foods Other Private Services Chemical, Rubber, Plastic Products Wearing Apparel Energy Extraction Beverages and Tobacco Products Leather Products Textiles Construction Electricity, Gas, Water Trade and Transport Services Wood Products Minerals Manufactures Mineral Products 2004 Paper Products, Publishing Metals 2007 Machinery and Equipment Dwellings Transport Equipment 2011 Metal Products Ferrous Metals 2014 0% 10% 20% 30% 40% 50% Total Worker Compensation as Percent of Export Value Source: Authors’ calculations using the LACEX database. Annex Figure 4.8: Labor Intensity of Exports across Sectors and Countries (a) Direct Labor Intensity Agr, Forestry, Fisheries Beverages and Tobacco Products Chemical, Rubber, Plastic Products Construction Dwellings Electricity, Gas, Water Energy Extraction Ferrous Metals Leather Products Machinery and Equipment Manufactures Metal Products Metals Mineral Products Minerals Other Private Services Paper Products, Publishing Processed Foods PubAdmin/Defence/Health/Educat Textiles Trade and Transport Services Transport Equipment Wearing Apparel Cambodia Wood Products 0% 20% 40% 60% Lao PDR Direct Worker Compensation as Percent of Export Value Malaysia (b) Indirect Labor Intensity Agr, Forestry, Fisheries Nepal Beverages and Tobacco Products Chemical, Rubber, Plastic Products Thailand Construction Dwellings Vietnam Electricity, Gas, Water Energy Extraction Ferrous Metals Leather Products Machinery and Equipment Manufactures Metal Products Metals Mineral Products Minerals Other Private Services Paper Products, Publishing Processed Foods PubAdmin/Defence/Health/Educat Textiles Trade and Transport Services Transport Equipment Wearing Apparel Wood Products 0% 10% 20% 30% 40% Indirect Worker Compensation as Percent of Export Value Source: Authors’ calculations using the LACEX database. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 125 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Annex Figure 4.9: Herfindahl Hershman Market Concentration Index in All Sectors Market Concentration Index in All Sectors 0.4 0.3 Cambodia Lao PDR 0.2 Malaysia Nepal 0.1 Thailand 0 Vietnam 2000 2005 2010 2015 Year Source: Authors’ calculations using UNCOMTRADE data from WITS. Annex Figure 4.10: Her ndahl Hershman Market Concentration Index in Selected Sectors (a) Textiles and Clothing 0.5 Market Concentration Index 0.4 0.3 0.2 0.1 Cambodia 0 2000 2005 2010 2015 Lao PDR Malaysia Year Nepal (b) Footwear Thailand Vietnam 0.6 Market Concentration Index 0.4 0.2 0 2000 2005 2010 2015 Year Source: Authors’ calculations using UNCOMTRADE data from WITS. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 126 TECHNICAL REPORT CHAPTER 4: JOBS AND THE EXPORT SECTOR Annex Figure 4.11: Labor Intensity of Wearing Apparel Exports, 2004-2014 2004 2007 Year 2011 2014 Direct Indirect 0% 5% 10% 15% 20% 25% Total Worker Compensation as Percent of Export Value Source: Authors’ calculations using the LACEX database. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 127 TECHNICAL REPORT CHAPTER 5: Workforce Skills for More, Better, and More Inclusive Jobs Maheshwor Shrestha and Une Lee 5.1 Introduction Cambodia’s success in creating more, better, and more inclusive jobs in the future will depend heavily on its ability to increase the skills of its workforce. While FDI and SME growth may yield new job opportunities, they can only be realized if enough adequately skilled workers are available to fill them. Only if a job is filled by someone with the appropriate skills will it yield the highest possible labor productivity, profits, and (potentially) earnings. As noted in previous chapters, labor productivity is not as high in Cambodia as it is in comparator countries, even in low-skilled jobs. As global and regional mega-trends produce jobs that are increasingly knowledge-intensive and technology-based, it is becoming even more important to enhance the skills of the working population. While the government has made commendable progress in recent decades in increasing educational attainment, Cambodia still has a large stock of low-educated and low-skilled individuals in its workforce. Only the workers in the younger, more educated age cohort have taken advantage of the increased opportunities in the economy’s booming sectors because they are more likely than older cohorts to have the skills that are demanded by employers. The government’s focus on improving the educational quality and increasing attainment is critical and should be continued, but it will have little effect on today’s workers. It is the low skill level of the current workforce that prevents them from getting better jobs, prevents foreign higher-value firms and investors from entering the Cambodian market, and keeps Cambodia trapped in an unskilled, low-level economy. Perhaps even more worrying is the absence of a robust, market-oriented skills development system to address these shortcomings. In this chapter, we investigate these issues in depth. We start by describing the educational profile of the Cambodian labor force and investigating how the labor market rewards education, the first indicator of skills. We then analyze the skills that employers value in their workers, as well as the formal and informal training they provide to them. Next, we explore the skill and task contents of Cambodia’s main occupations, with a particular emphasis on those valued by employers, to discover whether or not the current skill profile of Cambodian workers is aligned with the future of work as shaped by technological change. We go on to analyze discrepancies in labor market returns by gender, decomposing the differences in the wages paid to men and women with different characteristics as well as those paid to men and women with the same characteristics. We conclude by recommending a set of policy actions regarding skills development that will help Cambodia achieve its goal of becoming an upper-middle-income economy in the near future. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 129 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS 5.2 Most Cambodian Workers See Low Despite these advances, educational attainment in Cambodia remains low. Only 13.5 percent of the working Economic Benefits from Education age population has 12 or more years of schooling (high school completion), and only 3.8 percent have a college Low education attainment characterizes the degree. The educational attainment status of Cambodia is labor force, despite recent progress below the level that could be expected given its income level, particularly for high school completion (Figure 5.2). Cambodia has made good progress in the past decade Only Myanmar, Papua New Guinea, and Lao PDR have in educating its labor force. In 2004, a quarter of its similar or lower levels of high school completion rates in working age population had no schooling, and 62 percent the region, and only Papua New Guinea has lower college had not completed primary school (6 years) (Figure 5.1). completion rates. In 2010, Cambodia’s high school and By 2014, these shares were down to 16 percent and 45 college completion rates were in the bottom one-fifth in a percent respectively. Average years of schooling increased ranking of all countries globally. from 4.4 years to 6.3 years, a rise of 0.18 years of education per year. The annual increase was much higher (0.26 years) for the 20 to 29 age-group. Figure 5.1: Educational Attainment of the Working Age Population, 2004 and 2014 There has been a remarkable shift in Cambodia’s educational profile over a decade 25% Percent of Labour Force 20% 15% 10% 5% 2004 0% 2014 0 1 2 3 4 5 6 7 8 9 10 11 High College+ Years of Schooling School+ Source: Author’s estimates using CSES 2004 and 2014. Figure 5.2: Secondary and Tertiary Completion Rates and GDP Cambodia still lags behind comparator countries in educational attainment (a) Secondary School (High School) Completion Rates (b) Tertiary School (College+) Completion Rates 70% 30% 60% 50% Completion Rate Completion Rate 20% 40% 30% 10% 20% 10% 0% 0% 6 7 8 9 10 11 12 6 7 8 9 10 11 12 Log (GDP per capita, PPP) Log (GDP per capita, PPP) Non−EAP EAP Cambodia Locally-linear relationship Source: Per capita GDP from the WDI online. Completed education data for 2010 from Barro and Lee (2013). Cambodia data from the 2009 CSES. Note: The size of the bubble denotes the population in 2010. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 130 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS The continued low education level of the workforce, Vocational Education and Training (TVET).3 The share barely in spite of recent improvements in educational increases to 0.75 percent after including those who have attainment, results from the large stock of poorly taken non-formal training classes or participated in post- educated workers and a low flow of slightly more literacy programs. This suggests that the Cambodian TVET educated (young) individuals into the workforce. system has had little impact in terms of improving the skills Even among the relatively younger cohorts aged 20 to 29, of those who have left formal mainstream education—in the high school completion rate is only 23 percent, and other words, the bulk of the workforce. the college completion rate is only 7 percent. The older cohorts have much lower formal educational attainment Those with higher levels of education also have better rates (Figure 5.3). In order to improve the skills of the jobs. The increase in opportunities available in the growing Cambodian labor force, it will be necessary to upgrade the economy has opened up a wider range of occupations to skills not only of the current working age labor force but those with higher levels of education. The share of individuals also of students leaving school and entering the labor force. involved in farming and other elementary occupations declines If the current high school completion rates of the younger steadily with years of education and drops drastically for cohorts (aged 15 to 24) stay at the same level as in 2014, those who completed high school (Figure 5.4). Similarly, the only 19 percent of the working age population will have shares of machine operator jobs or those in craft and trades completed high school by 2030. 1 Even if the high school gradually increase with education but then drop for those completion rates of the younger age groups increased at with a high school or higher education. This is consistent the same rate during the 2004 to 2014 period, only 26 with the fact that these occupations are most suitable for percent of the working age population will have completed mid-skilled individuals. On the other hand, occupations such high school by 2030.2 At this rate, it will take until 2050 as clerks, managers, professionals, and technicians are for half of the working age population of Cambodia to have more common among those with a high school or higher completed high school. level of education. About 47 percent of college graduates work as clerks, and 31 percent as managers, professionals, The public provision of formal and non-formal or technicians. This drastic change in employment pattern training is limited in scope. Only 0.6 percent of the upon completion of high school and college holds across working age population has completed formal Technical and genders and locations (Annex Figure 5.2). Figure 5.3: Age and Education Profile of Cambodian Workforce Older cohorts are still largely low-skilled and low-educated 100% 80% Percent of Labor Force 60% No schooling 40% < Primary < Secondary 20% Upper Secondary 1% College+ 15 20 25 30 35 40 45 50 55 60 65 Age Source: Author’s estimates using CSES 2014. Note: School category coded is based on completed years of schooling. Primary schooling is 6 years, Higher secondary is 12 years, and College is 16 years of schooling. 1 These estimates are based on population projections from UN DESA for each age group and on the current educational attainment levels of the population. This projection assumes that the educational attainment of the younger age groups in the future will be the same as that of the younger age group in 2014. 2 This projection assumes that the high school completion rate of the 15 to 19 age group increases every year by 0.57 percent and that of the 20 to 24 age group increases every year by 1.73 percent. These rates were the observed rates for these age groups in the 2004 to 2014 period. 3 This does not include those who may have completed TVET education and transitioned back into the mainstream education system (to upper secondary schools or universities), but the share of such transitions is likely to be very low. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 131 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Figure 5.4: Occupation Choices by Years of Schooling Occupation choices change drastically upon completing high school 100% Percent of Labor Force 80% Agriculture and elementary 60% Machine operators 40% Craft and trades Service and sales 20% Manager, professionals, technical 0% Clerical 0 1 2 3 4 5 6 7 8 9 10 11 HS+ Col+ Years of Schooling Source: Author’s estimates using CSES 2014. Note: Sample restricted to the working age population aged 15 to 64, who had worked in the previous seven days before the survey. The occupation categories are defined for main employment. Figure 5.5: Returns to Education by Years of Schooling The returns to education are relatively at across primary and secondary school All Female Male 100% 100% 100% 80% 80% 80% 60% 60% 60% 40% 40% 40% 20% 20% 20% 0% 0% 0% Returns Relative to No Schooling -20% -20% -20% 1 3 6 9 12 16 1 3 6 9 12 16 1 3 6 9 12 16 Primary Higher College Primary Higher College Primary Higher College School Secondary School Secondary School Secondary Rural Urban 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% Point estimate 0% 0% 95% Con dence Interval -20% -20% 1 3 6 9 12 16 1 3 6 9 12 16 Primary Higher College Primary Higher College School Secondary School Secondary Years of Schooling Source: Author’s estimates using CSES 2014. Note: Sample restricted to the working age population aged 15 to 64 who had worked in wage employment for at least 20 hours the previous seven days before the survey. The vertical dashed lines indicate completion of primary, lower secondary, upper secondary, and college levels. The dependent variable is the logarithm of hourly wages. The regressions control for individual covariates (gender, age, marital status, and household size and composition) and location (rural-urban and regions). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 132 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Returns to education in the labor market increasing returns, at 2.9 percent per year at the primary are low, but increase at higher levels of level and 3.4 percent per year for the first 11 years of schooling. Although the urban returns are much higher and education more linear than are rural returns, urban returns for the first Low levels of education have a very low pay-off in 11 years are still lower than are international benchmarks. the labor market. Although Cambodia has made progress However, completing the upper secondary schooling in the recent decade in terms of educating the labor force, and college has larger payoffs. Individuals who have the labor market does not put much value on the skills that completed upper secondary schooling earn wages that are are acquired in the initial years of schooling. For instance, 9 percent higher than those earned by workers with 11 an individual with six years of schooling (a complete primary years of schooling (Figure 5.5). The returns to completing education) only earns 8 percent more than those without upper secondary (rather than just completing grade 11) is any schooling at all, suggesting a return of only 1.4 percent significantly higher for women (21 percent) than it is for for each of these years of schooling (Figure 5.5). Globally, men (3.4 percent) and is higher for those in rural areas (11 the average return for each year of primary schooling is percent) than those in urban areas (4.6 percent). 10.6 percent, which is more than 7.5 times higher than the returns for Cambodia (Montenegro and Patrinos 2014). For those who have completed upper secondary What is even worse is that the subsequent three years of schooling, completing college (or even higher) schooling (lower secondary) do not increase wages at all.4 provides a return of 14 percent per year of schooling. The first 11 years of schooling (before completing upper These returns are slightly lower in rural areas (11 percent) secondary) have an average return of only 1.3 percent than in urban areas (14 percent), but are much higher than for every year of schooling, which is much lower than the returns at lower levels of schooling. Internationally, the average returns of 10.6 percent per year for primary average returns to a year of tertiary schooling are similar to and 7.2 percent per year for secondary in other countries those in Cambodia, at around 15 percent.5 As seen in Figure (Montenegro and Patrinos 2014). 5.5, returns are highly convex in terms of years of schooling. Men experience worse returns from schooling than Overall, throughout the years of formal schooling in do women with lower levels of education. For men, Cambodia, each additional year provides an average primary schooling provides a return of 0 to 0.6 percent return of 3.6 percent per year, which is much lower per year, which is much lower than 1.8 percent per year than the global average of 10.1 percent. Cambodia’s provided to women. Returns from the first 11 years of returns, as estimated here, are in the same range as schooling are only 1.0 percent per year for men compared the five countries with the lowest returns among 139 with 1.4 percent per year for women. These differences countries around the world, as estimated by Montenegro help to narrow the gender gap in hourly wages from 30 and Patrinos (2014). As might be expected, the Cambodian percent for the uneducated workforce (with 0 to 10 years of yearly returns are much lower in rural areas (2 percent schooling) to 10 percent for the workforce with incomplete per year) than they are in urban areas (6 percent), and secondary schooling (with 11 years of schooling). Although are lower for men (3.2 percent) than they are for women the difference in returns in favor of women has been (4.2 percent). consistently observed in other countries as well, the levels of the returns in Cambodia are much lower than elsewhere in the world. The global averages for primary returns are Several factors are responsible for low rates 10 percent per year for men and 10.9 percent for women, of return to education which are several times higher than those observed in Cambodia (Montenegro and Patrinos 2014). Low quality of schooling could be one of the reasons for low returns. In 2016, about one-third of 6th graders As seen in Figure 5.5, urban and rural areas have had “below basic” knowledge of the Khmer language, and, vastly different returns to lower levels of education. even worse, over half of the sixth-graders had “below basic” In rural areas, returns are low but positive for the first years knowledge of mathematics (MoEYS 2017). The quality of and start to decline through grade 11. Consequently, the education in Cambodia is on a par with lower- and middle- returns to primary schooling are 1.2 percent per year, income countries in the region, but significantly lags behind whereas the returns to the first 11 years of schooling is more developed countries in East Asia (Figure 5.6). Perhaps only 0.5 percent per year. Urban areas experience steadily more importantly, the gains that Cambodia has made in 4 They have a return of statistical zero. 5 Montenegro and Patrinos (2014). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 133 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS educational attainment are being undercut by the low quality (oral communications, teamwork, and customer service), of education. While four-year-old Cambodian children today problem-solving skills (leadership, performing complex are expected to complete 9.5 years of school, their learning tasks, and planning), and technical skills (use of equipment, will be equivalent to only 6.9 years of school because of the languages, and IT) are all in high demand yet short supply poor quality of the schooling that they are likely to receive.6 (Figure 5.7).7 In terms of quality-adjusted years of education, Cambodia is at a significant disadvantage relative to middle-income The higher returns for those who have completed countries in East Asia. The quality of schooling is much upper secondary and beyond could be because of worse in rural areas than in urban areas. The poor quality of the scarcity of highly educated workers. As seen in basic education may be leading to a situation where only the Figure 5.4, the highly educated are more likely to become brightest of students make it to higher levels of schooling. clerks and to perform managerial and technical tasks. If the Those with low levels of education leave the schooling relative scarcity of workers with requisite skills for these system with very low levels of skills. tasks is the main reason for their experiencing higher returns, then those returns can be expected to dwindle as The skills valued by employers may not be those a larger share of the workforce becomes better educated taught in schools, so acquiring more education does (assuming that the demand for the highly educated does not necessarily increase an employee’s value (and not grow faster than the supply). Indeed, the returns to thus pay). For example, the rapid growth of the garment, upper secondary and college education have fallen slightly footwear, and apparel industry requires skills such as in recent years (Annex Figure 5.3). It may be necessary manual dexterity, the ability to use a sewing machine, to make further quality improvements to higher levels of punctuality, teamwork, and problem-solving, but it has little education to make them more productive and to ensure use for the curriculum taught in primary and secondary that they continue to yield high returns even when there schools. Indeed, employers in various sectors and is no longer a scarcity of well-educated workers. If the particularly in the fastest growing occupations, have stated demand for highly skilled workers increases in Cambodia in recent surveys conducted by the National Employment at a pace faster than its supply, then the returns at the top Agency (NEA) that a combination of inter-personal skills may continue to rise. Figure 5.6: Quality-Adjusted Years of Education 14 600 12 550 10 500 Test Score 8 Years 450 6 400 4 2 350 0 300 p. DR es r ia ia d sia a ina am ng re ma oli an od es Re po pin Ko lay oP Ch tn ng ail an on mb ga Vie a, ilip Ma ng Th Mo La My Ind Sin re Ca Ph Ho Ko Expected Years of Education (today's cohort) Learning-Adjusted Years of School Harmonized Test Score Source: Derived from the World Bank Human Capital Index. 6 Quality adjustments are completed based on harmonized test-scores across countries. For more details, see the World Bank Human Capital Index, www.worldbank.org/ en/publication/human-capital. 7 The National Employment Agency (NEA) surveys build and expand on previous ILO-NEA employer skills surveys from 2012/13. The NEA’s Employer Skills Needs Survey of 2014 surveyed more than 595 registered establishments (with 10 or more employees) in 10 sectors across Cambodia: food and beverages; garments, apparel, and footwear; rubber and plastics; construction; finance and insurance; accommodation; transportation, warehouse, and logistics; human health; education; and ICT. The surveyed firms represented 15 percent of all firms in Cambodia. The survey covered topics such as the characteristics of the selected sectors, their employment structures, the work readiness of first-time jobseekers, recruitment difficulties, skills shortages, skills gaps, workforce training, and business strategies. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 134 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Figure 5.7: Skill Shortages in the Cambodian Labor Market By sector Garment, footwear Accommodation Construction Food and beverages and apparel Taking initiative (31.7%) Team work (44.9%) Taking initiative (50%) Oral communications (27.5%) Manual dexterity (27%) Oral Communications (38.5%) Team work (33.3%) Team work (25%) Adapting to new equipment/ Customer handling (35.9%) IT literacy/using IT (27.8%) Customer handling (22.5%) materials (22.2%) Advanced IT application/ Oral communications (19%) Foreign languages (33.3%) development (27.8%) Manual dexterity (20%) Management responsibilities/ Public speaking/instructing/ Management responsibilities/ Learning new ideas/methods/ taking lead (14.3%) training (20.5%) taking lead (27%) concepts (15%) Rubber and plastics ICT Logistics, warehousing Finance and insurance and transportation Public speaking/instructing/ Customer handling (66.7%) Oral communications (66.7%) Customer handling (48.1%) training (25%) Foreign languages (48.9%) Customer handling (52.4%) Taking initiative (30.8%) Foreign languages (25%) Foreign languages (52.4%) Team work (38.1%) Oral communication (26.9%) Manual dexterity (25%) Solving complex tasks/ Taking initiative (38.1%) problems (47.6%) Planning and organizing (25%) IT literacy/using IT (27.8%) Team work (12.5%) Management responsibilities/ Management responsibilities/ Advanced IT application/ taking lead (28.6%) taking lead (28.6%) development (19.2%) Cognitive/soft skills (+50%, 30-50%, <30%) Manual/technical skills (+50%, 30-50%, <30%) In booming occupations Plant and machine operators, Craft and related Elementary occupations Technical and associated and assemblers trades workers professionals Oral communications/ Oral communication (36.6%) Taking initiative (39.5%) Oral communication (28.8%) customer handling (53.2%) Manual dexterity (26.8%) Team work (31.6%) Team work (25%) Solving complex tasks problems (40.3%) Adapting to new equipment/ Manual dexterity (23.7%) materials (24.4%) Manual dexterity (15.4%) Taking initiative (37.1%) Oral communication (21.1%) Taking initiative (15.4%) Taking initiative (17.1%) Team work (25.8%) Team work (14.6%) Adapting to new equipment/ materials (24.4%) Foreign languages (24.2%) Cognitive/soft skills (+50%, 30-50%, <30%) Manual/technical skills (+50%, 30-50%, <30%) Source: NEA (2014 and 2015). The meager returns to lower levels of education can of schooling. If this is the case, then the children that do not reduce the demand for education. The convex nature of receive even low levels of schooling will not be able to benefit returns lowers the incentives for families to invest in schooling. from any future job opportunities that might materialize from, For instance, if a family cannot afford to educate its children for example, future government policies and programs—thus beyond the upper secondary level, then they might choose not reducing the benefits to the economy as a whole from the to spend any money on their child’s schooling at all since they creation of these new job opportunities. Therefore, improving expect to see few benefits from the investment. This suggest the quality of the future Cambodian workforce will require that returns to education might also be low in part because incentives guaranteeing that students continue to attend there are few job opportunities for workers with lower levels school, in addition to quality improvements of the school system of education, rather than simply because of the low quality as a whole. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 135 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS 5.3 Workers are poorly equipped to dexterity (Figure 5.8).8 The exact hard skills that are needed tend to be speci c to each occupation, but include technical supply the skills that employers demand skills, such as foreign languages, or physical skills, such as manual dexterity. Soft interpersonal and social skills, such as One implication of the poor quality of education is teamwork, some types of problem-solving, and managerial that the workers are ill-prepared for the working skills, are widely demanded across many different types life ahead of them. Students end their schooling without of jobs. having gained the appropriate skills to prepare them to participate in the labor market. Figure 5.7 showed the range The scarcity of soft skills affects a wide array of of skills that employers require but for which they cannot firms across sectors and occupations. For instance, find enough workers who possess them. These skills, both many firms in the construction, finance and insurance, and “hard” and “soft,” are crucial for both workers and employers garment sectors find that “taking initiative” is a skill that is in to reap the rewards that the labor market can offer. short supply (Figure 5.7). Similarly, firms in accommodation, However, beyond being just a shortcoming of the construction, food and beverages, and ICT sectors put a formal educational and training institutions, this glut high value on teamwork skills, but find workers with these of skilled workers also suggests that workers have skills hard to come by. Meanwhile, workers with oral not been able to acquire those skills in the course communication skills and the ability to deal with customers of doing their jobs, either through formal training or are in short supply in most major sectors. Four of the top through on-the-job learning. Indeed, employer surveys five skills in demand by the ICT sector are soft skills, such suggest that very few employers have formal training as being able to handle customers, work in teams, take arrangements in place for their workers. initiative, and perform managerial tasks. The same holds true for firms hiring workers for elementary occupations. Employers value certain skills, many of However, there is also a great demand for workers which are scarce among Cambodian workers with occupation-specific technical skills. Problem- solving is a scarce skill for firms wishing to hire professional Employers demand a mix of “hard” and “soft” skills. workers, while manual dexterity is hard to find among According to a 2013 survey of 222 large enterprises workers in elementary occupations, machine operators, and with at least 100 employees (ADB-HRINC-DBLink 2013), crafts and related workers. Similarly, firms in construction, foreign language skills are in highest demand, followed by finance and insurance, and rubber and plastics find their teamwork, problem-solving, managerial skills, and manual workers particularly lacking in IT skills. Figure 5.8: Opinions of Large Employers about Which Key Skills Their Workers Lack Foreign Language Barrier 55% Team Work 39% Problem Solving 38% Management Skills/Supervision/Leadership 36% Manual Dexterity 25% Basic Technical Skills/Practical IT 25% Planning and Organising 18% Reading 14% Numeracy 12% Customer Handling/Sales/Marketing 12% Writing/Written Communication 12% 0% 10% 20% 30% 40% 50% 60% Percent of Large Employers Source: ADB-HRINC-DBLink (2013). 8 ADB-HRINC-DBLink (2013) conducted a survey of 222 large enterprises (with at least 100 employees) in six regions (Phnom Penh, Siem Reap, Kampot, Battambang, Svay Rieng, and Takeo) and the following sectors: garments and footwear (61 percent); hospitality (22 percent); other manufacturing (4 percent); and other professional services (13 percent). The survey focused on the needs of industry from the perspective of Cambodia’s largest employers and identified skills gaps, challenges, and opportunities for TVET education. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 136 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS If workers are working with a sub-optimal level of either on-the-job opportunities or through informal training skills, it reduces their productivity, thus damaging (Acemoglu and Pischke 1998). Anecdotal evidence in the firm and even the broader economy. A large share Cambodia also suggests that firms tend to provide training of firms in the NEA survey recognized that a lack of skills to their professional and semi-professional staff who have could lead to lower worker productivity and hurt business as had a longer tenure at the firm rather than to lower-level a result of delays in developing new products and services workers whose turnover rates they assume will be high (49 percent), an increased workload for other workers (38 (ADB-HRINC-DBLink 2013). percent), difficulties in meeting customer service objectives (34 percent), lost business to competitors (31 percent), and difficulties meeting quality standards (28 percent) (NEA 2015). 5.4 The Tasks Required in Common The poor quality of Cambodia’s workers may also be making Cambodian Occupations are Changing the country less competitive for foreign-owned employers. Any occupation consists of multiple tasks that require Employers rarely provide training to workers different skills. Therefore, it is instructive to examine the tasks that are carried out in any given occupation and then While workers can learn on the job, few formal training identify the skills needed to carry out those tasks, as this opportunities are provided by employers in Cambodia. can then demonstrate the range of skills that Cambodia’s Only 22 percent of manufacturing firms offer formal training labor force needs to develop. However, most household and according to the World Bank’s 2016 Enterprise Survey for labor force surveys, including those available for Cambodia, Cambodia. This is low compared to the EAP average of 57 do not collect data on the nature of tasks performed and the percent, and the lower middle-income country average of skills used in specific jobs. Instead, it has become common 25 percent. Even within those firms that offer training, it is practice when studying these issues to use not only provided to only 44 percent of the workers. Again, this is occupation data from a country’s labor force or household lower than the 85 percent for EAP countries and 60 percent survey, but also data from the O*Net—an occupation-task- for lower middle-income countries. Medium-sized firms, with skills database based on the US labor market—to match 20-99 employees, are much more likely to offer any training tasks and skills to country-specific occupations.9 Therefore, (61 percent) than either small firms (13 percent) or large we also chose to use this methodology to explore the task firms (31 percent). These firms cite difficulties in organizing content of Cambodian jobs. training or in finding good trainers as the key reasons for not providing more formal training for their employees (NEA Since there are hundreds of tasks and skills, it 2014, 2015). is common to summarize the tasks within each occupation using only two general categories of tasks The provision of informal training is slightly more and three skill types. The two task-types are routine common. Over 50 percent of firms provided some form and non-routine tasks.10 Routine (repetitive) tasks are job of training, either formal or informal, to their workers activities that involve a series of well-defined steps that could, (NEA 2014), with 40 percent of the small firms providing in theory, be programmed to be performed by a machine. training. Although this is lower than the provision by Non-routine tasks require judgment, which can only be medium-sized firms (61 percent) and large firms (58 provided by a human. The skill types are cognitive, manual, percent), the discrepancy is not as stark as with formal or interpersonal. Cognitive skills are mental abilities and training provision. knowledge such as mathematics or problem-solving. Manual skills include physical dexterity and strength. Interpersonal This pattern is consistent with the literature findings skills cover such qualities as teamwork, leadership, and regarding the under-supply of general training by initiative. Each occupation is a combination of a task type firms. Anticipating that a worker who receives formal and a skill type. For the purposes of this chapter, we used training in general skills may wish to leave the firm to six common task-skill categories as follows: repetitive mental obtain higher wages elsewhere, employers do not have an tasks, repetitive physical tasks, (non-routine) analytical tasks, incentive to invest in increasing their productivity. However, non-repetitive physical tasks, managerial interpersonal because firm-specific training is often not useful in other tasks, and social tasks (see Table 5.1 for a brief description firms, they do not hesitate to provide such training through of the categories). 9 National Center for O*Net Development, www.onetonline.org. 10 The terminologies used here are adapted from Autor et al. (2003) and have been widely adopted since. Please see Autor (2013) for a review of the methodology and literature related to the task-based approach. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 137 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Table 5.1: Definitions of Skill-Task Categories Task type Repetitive Tasks Non-Repetitive Tasks Repetitive Mental Tasks Analytical Tasks Definition: Require individuals to perform Definition: Require individuals to analyze and repetitive mental tasks with a high degree process abstract information and actively use it of accuracy. They are very structured and to solve complex problems. do not allow workers to determine their own tasks and priorities. Task examples: Identifying underlying principles, Cognitive reasons, or facts by breaking down information (analytical) Task examples: Typing, data-entry, or data into separate parts. Thinking creatively book-keeping, or checking ledgers. to develop, design, or create new applications, ideas, systems, or products. Interpreting and Occupation examples: Clerks, keyboard using information. operators, and tellers. Occupation examples: Engineers, doctors, lawyers, and scientists. Repetitive Physical Tasks Non-Repetitive Physical Tasks Definition: Require individuals to spend a Definition: Require judgment and physical signi cant amount of time making repetitive dexterity to perform certain tasks. physical motions. Task examples: Maneuvering or navigating tasks Task examples: Operating machines or by driving and operating mechanized equipment Manual controlling processes by using direct physical such as forklifts or other vehicles. Using hands activity or mechanical mechanisms, not and arms to grasp, manipulate, or assemble including operating vehicles or computers. objects or control mechanized equipment. Occupation examples: Textile workers, metal Occupation examples: Drivers, crews on boats processers, and plant and machines operators or ships, mining and construction laborers Skill-Type Managerial Interpersonal Tasks Definition: Require the ability to manage people and relationships. Task examples: Develop and maintain constructive and cooperative working relationships with others. Guiding and directing subordinates by setting performance standards and monitoring their performance. Coaching or helping others to improve their performance. Occupation examples: Managers, managing directors, legislators, and business administrators Interpersonal Social Tasks Definition: Require being aware of others’ reactions and understanding why they react as they do. Task examples: Using persuasion and negotiation skills. Adapting to and coordinating with others’ actions. Occupation examples: In addition to managerial occupations, occupations such as primary schoolteachers, policemen, sales-people, and public relations Note: Skill categories adapted from Acemoglu and Autor (2011). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 138 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS We created similar task measures in the context of The three largest occupations—subsistence farming, Cambodia (see Box 5.1 for methodological details) garment manufacturing, and agricultural labor—rank and used them to analyze the task content and very low in terms of their skills requirements. With evolution of Cambodian jobs. regard to basic skills, the farming occupations use some degree of science in their work (presumably related to High-order skills are not required for most agriculture), and garment workers use some mathematics occupations, but are needed in the fastest (presumably related to measurements and understanding growing ones design patterns). Otherwise, these occupations lag behind others in basic comprehension and communication skills, The most common occupations in Cambodia do not critical thinking and active learning skills, social skills, feature tasks that use skills that are in high demand and problem-solving skills. Furthermore, they do not have by employers and do not need higher-order analytical much need for communications and team-work in their and interpersonal skills. Employers in Cambodia have workplaces. Unsurprisingly, given the task content of these consistently reported a shortage of technical and problem- occupations, they need only a negligible amount of higher- solving skills among the workforce as well as soft skills order skills (analytical, managerial, and interpersonal) and such as communication, client engagement, and teamwork social skills. These occupations tend to consist mainly of skills (Figure 5.7, NEA 2016). However, many of the largest physical tasks and repetitive cognitive (mental) tasks. occupations do not tend to use these skills (Figure 5.9). Box 5.1: Methodology Used to Create the Task Classification for Cambodian Occupations Cambodia has no O*NET-like database specifically built using tasks from each of the country’s occupations. Therefore, we used the O*NET classification, which assigns scores for each occupation code along a wide array of tasks and skills and match them to the most recent Cambodia Socio-Economic Survey (CSES). We did this conversion in five steps: 1. Since the CSES uses occupation codes (NIS OCC codes), which are a slight variation of the ISCO 08 codes, we converted the three-digit Cambodia NIS OCC codes into ISCO 08 four-digit codes. This was done using the mapping provided by the National Institute of Statistics (http://www.stat.go.jp/info/meetings/cambodia/pdf/c8_occup.pdf). 2. We then converted the four-digit ISCO 08 codes into six-digit SOC 2010 codes used by the O*NET database. The methodology for making this conversion came from the US Bureau of Labor Statistics (https://www.bls.gov/soc/ ISCO_SOC_Crosswalk.xls). 3. For each of the SOC codes, O*NET provides occupation scores along several dimensions. Acemoglu and Autor (2011) used these dimensions to compute aggregate task scores for each of the occupations. We followed the same algorithm to create the task scores. (The code for this classification is available from the authors’ website: https://economics.mit.edu/files/7974). 4. In cases where the same Cambodian NIS OCC codes mapped to multiple SOC codes, the scores for the occupation were set as a simple average of the SOC codes that mapped into them. 5. The scores were then standardized for each of the Cambodian occupations using the composition of non-subsistence occupations in Cambodia for the year 2009, our base year. In other words, in terms of standard deviations, the scores measure the task intensity of an occupation in 2014/2015 compared to its mean score in 2009. There are valid concerns about using the O*NET classifications for Cambodia since the same occupations might involve different tasks in Cambodia than in the United States. One major difference between the countries is that the same occupation involves more higher-order tasks in the US than in Cambodia. However, since the scores are standardized within Cambodia, the overall level of difference between the two countries would not matter. What could affect our analysis is that the difference in the task content of occupations between two countries differs drastically by occupations. Even with such differences, the aggregate level of the analysis in this report will minimize the biases due to such errors. In fact, a similar methodology using the US-based O*NET classification has been used to study the labor markets in a wide array of countries, including East Asian economies such as China and Vietnam. Note: Skill categories adapted from Acemoglu and Autor (2011). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 139 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Figure 5.9: Skill-Task Intensity of the Ten Largest Occupations Agricultural, forestry and fishery labourers Building frame and related trades workers Garment and related trades workers Market gardeners and crop growers Mining and construction labourers Street and market salespersons Subsistence livestock farmers Subsistence crop farmers Very high intensity Shop salespersons High intensity Animal producers Low intensity Very low intensity Reading Comprehension Active Listening Writing Basic Skills Speaking Mathematics Science Critical Thinking Process Active Learning Monitoring Social Perceptiveness Coordination Social Skills Persuasion Negotiation Service Orientation Complex Problem Solving Operations Analysis Technology Design Techical and Equipment Installation and Maintenance Problem Solving Programming Operation Monitoring Operation and Control Troubleshooting Repairing Judgment and Decision Making Systems Analysis System and Time Management Resource Management Management of Financial Resources Management of Personnel Resources Communication at workplace Salesmanship Communication Communication with clients and Teamwork Team−building Work with team Deal With External Customers Analytical Managerial and interpersonal Social Aggregate Non−repetitive physical Repetitive mental Repetitive physical Source: Author’s estimates using CSES 2014 and the O*NET database of the skills associated with occupations. Note: Aggregate task intensity measures follow the classification used in Acemoglu and Autor (2011). Task intensity scores are standardized by the mean and standard distribution of the scores in non-subsistence occupations weighted by the labor share of those occupations in 2009. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 140 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Figure 5.10: Skill-Task Intensity of the Ten Fastest-Growing Occupations Building nishers and related trades workers Hotel and office cleaners and helpers; Build Tellers, money collectors and related clerks Agricultural, forestry and shery labourers Building frame and related trades workers Garment and related trades workers Motodup and tuktuk drivers Waiters and bartenders Very high intensity General office clerks Shop salespersons High intensity Low intensity Very low intensity Reading Comprehension Active Listening Writing Basic Skills Speaking Mathematics Science Critical Thinking Process Active Learning Monitoring Social Perceptiveness Coordination Social Skills Persuasion Negotiation Service Orientation Complex Problem Solving Operations Analysis Technology Design Techical and Equipment Installation and Maintenance Problem Solving Programming Operation Monitoring Operation and Control Troubleshooting Repairing Judgment and Decision Making Systems Analysis System and Time Management Resource Management Management of Financial Resources Management of Personnel Resources Communication at workplace Salesmanship Communication Communication with clients and Teamwork Team−building Work with team Deal With External Customers Analytical Managerial and interpersonal Social Aggregate Non−repetitive physical Repetitive mental Repetitive physical Source: Author’s estimates using CSES 2014 and the O*NET database of the skills associated with occupations. Note: Aggregate task intensity measures follow the classification used in Acemoglu and Autor (2011). Task intensity scores are standardized by the mean and standard distribution of the scores in non-subsistence occupations weighted by the labor share of those occupations in 2009. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 141 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Mining and construction laborers also tend to perform learning skills. Bank tellers also tend to perform repetitive mainly physical or repetitive tasks, but they also mental tasks as well as some higher-order managerial and perform a higher than average amount of analytical interpersonal tasks stemming from the need to engage and tasks. For example, construction laborers have to follow build relationships with clients. the designs of the buildings on which they are working. They do not need to use many social skills and do not have to Service sector occupations, such as waiters, interact with others apart from their direct work colleagues bartenders, and drivers, have to engage with clients and, therefore, have little need of managerial interpersonal and tend to require social perceptiveness in their tasks or social perceptiveness. jobs. Other social skills such as coordination and service orientation are also extensively used in these occupations. Shopkeepers, the fth largest occupation, however, have many tasks requiring higher-order skills in Having completed more education, particularly addition to basic skills, including process monitoring, high school and college, helps individuals perform social skills, problem-solving, resource management, and occupations with a higher intensity of analytical and communication skills. managerial interpersonal tasks. The manual task content of occupations, both repetitive physical and non-repetitive Some of the fastest growing occupations are more physical, decrease with the level of education of the workers likely to use higher-order skills. The exceptions, as in those occupations (Figure 5.11, Annex Table 5.6). As with discussed above, are the fastest growing occupations the returns to schooling, the repetitive physical and non- (garment workers) and agricultural labor. However, the repetitive physical task content of occupations falls drastically second fastest growing occupation, frame builders, involves upon high-school completion and college graduation (Annex tasks that require a high intensity of higher-order analytical Table 5.6). The opposite happens with analytical, managerial and interpersonal skills (Figure 5.10). Analytical skills are interpersonal, and socially perceptive skills, with these needed to interpret or create designs for the frames, while tasks increasing steadily with education and then jumping the interpersonal skills are needed to collaborate with up drastically with high-school and college completion. The or supervise the teams creating the final product. This repetitive mental task content jumps up only for college occupation also involves many non-repetitive physical tasks graduates, while completing lower levels of education causes as it requires great deal of manual dexterity. This pattern less dramatic rises. In addition to yielding higher earnings also holds for building finishers. (as discussed in an earlier section), education, particularly completing high school and college, enables individuals to be The third fastest growing occupation, general clerks, selected for occupations that use higher-order skills such as uses a mix of higher-order analytical and repetitive analytical, managerial, and interpersonal tasks. It suggests mental tasks. Basic reading, writing, and numeracy skills that education helps individuals take on occupations that are integral parts of their occupation, and they also use a require analytical and interpersonal tasks rather than manual fair amount of problem solving, critical thinking, and active tasks, even though the wage returns to schooling are very low. Figure 5.11: Task Intensity, by education level There is a strong correlation between educational attainment and task content 1 0.5 Analytical Managerial Task Intensity 0 interpersonal Repetitive mental -0.5 Repetitive physical -1 Non−repetitive physical -15 Social No schooling < Primary < Secondary HS graduate College + Source: Author’s estimates using CSES 2014 and the O*NET database of the skills associated with occupations. Note: Task intensity measures follow the classification used in Acemoglu and Autor (2011). Task intensity scores are standardized by the mean and standard distribution of the scores in non-subsistence occupations weighted by the labor share of those occupations in 2009. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 142 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Female-dominated occupations are more intense different from those assigned to younger ones (Figure 5.12 in social tasks and less intense in all other task and Annex Table 5.6). For instance, the intensity of physical categories. As seen in the previous section, there is a large tasks (both repetitive and non-repetitive) fell with age, which gender difference in terms of broad occupation categories. could suggest a lower level of ability or a preference against This difference also seeps down to the tasks performed by having to carry out physically demanding tasks as a worker women within those occupations (Annex Table 5.6). Women ages. At the same time, the intensities of both managerial in Cambodia tend to be employed in occupations that use interpersonal and socially perceptive tasks increased with more interpersonal skills such as social perceptiveness, age, suggesting that older workers tend to go into jobs that persuasion, and coordination but fewer analytical, managerial, involve managerial as well as social skills. However, the and physical tasks than their equally educated male peers. intensities of repetitive mental and analytical tasks also fell with age, particularly for the oldest segments of the labor Rural occupations are more intense in physical force. This could suggest that, even conditional on education tasks and less intense in mental, analytical, and level, the older cohorts of workers do not have the same interpersonal tasks. The difference arises partly because level of analytical skills as their younger peers and sort into of the prominence of subsistence farming occupations occupations that use these skills less intensely as they age. in rural areas. These occupations do not require many skills beyond performing physical tasks. However, these Both analytical and repetitive mental tasks have subsistence occupations do not completely explain the become increasingly common in Cambodia’s difference between rural and urban areas. Agriculture still occupational mix in recent years. Compared to the occupies a sizable share of non-subsistence occupations distribution of occupations in 2009, occupations in 2014 in rural areas, and opportunities in manufacturing sectors have become more intense in all types of aggregate outside garments, which uses only a limited set of skills, task measures except non-repetitive physical tasks. are not common. In a similar vein, the northern provinces, This is because of the decline of activities associated where farming is widespread, also have similar task content with farming (even non-subsistence occupations) and as rural areas (Annex Table 5.6). the rise of manufacturing, particularly in the garment industry. The intensity of analytical and repetitive mental Conditional on education, older workers still sort into tasks in Cambodian occupations dramatically increased occupations that are less intense in physical and by over 13 to 18 percent of a standard deviation just in analytical tasks, and into occupations that are more the five years between 2009 and 2014. The intensity intense in interpersonal tasks. One of the issues in the of repetitive physical tasks increased by 7 percent of a Cambodian labor market is the low education level of the standard deviation, and the intensity of managerial and older population. However, conditional on education levels, interpersonal tasks also increased slightly by 2 to 4 we found that the tasks assigned to older people were percent of a standard deviation (Figure 5.13). Figure 5.12: Task Intensity and Age of Workers Older workers work in occupations that involve few physical and analytical tasks 0.4 Analytical 0.2 Managerial Task Intensity interpersonal 0 Repetitive mental Repetitive physical -0.2 Non−repetitive physical -0.4 Social 15 20 25 30 35 40 45 50 55 60 Age (Residualized) Source: Author’s estimates using CSES 2014 and the O*NET database of the skills associated with occupations. Note: Task intensity measures follow the classification used in Acemoglu and Autor (2011). Task intensity scores are standardized by the mean and standard distribution of the scores in non-subsistence occupations weighted by the labor share of those occupations in 2009. The sample is restricted to those who had worked for at least 20 hours in the previous week before the survey. Both age and the task intensity measures are residualized with the same set of controls as in Table 1. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 143 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Figure 5.13: Evolution of Task Intensity, 2009-2015 0.2 Analytical Managerial Task Intensity 0.1 interpersonal Repetitive mental 0 Repetitive physical Non−repetitive physical -0.1 Social 2009 2010 2011 2012 2013 2014 2015 Year Source: Author’s estimates using CSES 2009-2015 and the O*NET database of the skills associated with occupations. Note: Aggregate task intensity measures follow the classification used in Acemoglu and Autor (2011). Task intensity scores are standardized by the mean and standard distribution of the scores in non-subsistence occupations weighted by the labor share of those occupations in 2009. The sample is restricted to non-subsistence occupations in all years. Automation of routine tasks could restrict Other booming employment sectors in non-textile job growth in fast-growing sectors manufacturing and services focus on non-routine tasks as well. Non-garment manufacturing jobs in As seen above, the fastest growing occupations Cambodia involve non-routine analytical and non-routine in Cambodia are mostly in the garment sector and physical tasks intensively in addition to more routine tasks. involve manufacturing either in or for foreign-own Jobs in construction require not only technical skills, but enterprises. One of the reasons that the number of these also social skills, which can be developed by working in jobs has exploded in Cambodia is the low cost of the labor teams or by managing such teams. Similarly, jobs in trade needed to perform the repetitive physical and the repetitive and tourism require interpersonal and social skills as they mental tasks associated with the garment production line. involve interacting with clients and customers. Skills gained The routine nature of these tasks means that in theory they from manufacturing jobs may be limited to the specific trade could be done by computers or machines. Rising labor costs in question, but the skills developed from the interpersonal in Cambodia and the falling costs of automation may lead tasks, specifically socio-emotional and communication foreign firms to consider automation as an increasingly skills, can be adapted to other jobs as well. These skills are attractive option. Until recently, non-repetitive physical tasks unlikely to be adversely affected by technological progress were dif cult to automate, but that is slowly changing with in the workplace. advances in machine learning and artificial intelligence. Moreover, recent advances in artificial intelligence On the other hand, interpersonal and social tasks have dramatically extended the range of tasks that have proven to be much harder to automate (Autor can be automated (Brynjolfsson and McAfee 2014). 2015). With technological advances changing the nature Even a few years ago, tasks requiring manual dexterity, of modern work across the board, skills that prepare such as driving a car or sewing a tee-shirt, were thought individuals to conduct higher-order analytical, managerial, to be immune to automation. However, these tasks are and social tasks are, and will be, increasingly valued by the already within the reach of the most modern technology.11 labor market. Indeed, studies in the US have found that the Although technology has not yet managed to fully automate labor market increasingly rewards social skills, which are in the production of more complex textiles, with the rapid increasing demand (Deming 2017). Employers all over the expansion in the field of artificial intelligence, this could world, including those in Cambodia, report that they need happen in the not-so-distant future. workers with social and other soft skills but find it difficult to find many (Figure 5.7 and Figure 5.8). 11 For instance, SoftWear, a US-based company, intends to produce tee-shirts using its automated robots and sell them at prices on a par with even the lowest wage countries within the next year (Emont 2018). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 144 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Such technological advances are likely to take place employment in the garment sector itself in in the developed economies and not necessarily in Cambodia. While the garment sector has provided high countries like Cambodia where front-line production wages and incomes for a large segment of the society, it occurs. Consistent with this trend, some advanced has imparted only a limited set of skills to the workforce. countries are already “re-shoring” tasks that were previously It would be difficult to integrate these workers into other outsourced to other countries. This has several implications parts of the economy as their skills would not easily be for Cambodia: transferrable to other large sectors in the Cambodian economy.12 • Cambodia cannot solely rely on the low cost of labor to attract jobs from abroad in the garment sector. Although the entire production process in all 5.5 A Gender Wage Gap Exists in the sorts of garments produced in Cambodia may not be fully automated soon, the growth in garment-related Cambodian Labor Market jobs in unlikely to increase at the same pace as it has If labor markets are not inclusive, then having done in the recent past. This means that new entrants the right set of skills may not be sufficient for the to the Cambodian labor market will need to find work excluded workers to reap the rewards from the labor outside the garment industry, which is likely to require a market. One such dimension of exclusion is gender. In different set of skills. this section, we examine the gender gap in the wage labor market to assess the extent of the pay gap and the role of • In general, the range of manufacturing jobs that played by skills and other characteristics in filling the gap. could be located in Cambodia is shrinking as a result of a restructuring of many global value Existing gender disparities in wages suggest that chains. Indeed, as automation begins eliminating women may not be able to take full advantage of repetitive and manual tasks and related jobs, countries available economic opportunities. On average, a female higher up the value chain may stop outsourcing these wage worker in Cambodia earns only 86 cents for every tasks to countries that are just entering the sector dollar of wages made by male workers (Table 5.2).13 This (Maloney and Molina 2016). Automation could also difference might arise from the fact that female wage make it less likely for unskilled workers in agriculture workers have different attributes than do male workers. to find higher paying jobs in low-skilled manufacturing. Alternatively, these differences might be the result of the Rodrik (2016) argues that the gains that countries can labor market treating male and female wage workers with make from increased industrialization in the future will the same attributes differently or even of factors that the be more limited than has been the case historically. For data just cannot measure.14 instance, Cambodians are unlikely to be able to benefit from future outsourced jobs in the automotive and Background characteristics explain only 27.9 percent electronics industries as was the case for Korea, Taiwan, of the observed gender wage gap. If women had the and China, because these two industries are now same level of education as their male counterparts, the heavily robotized. Cambodia will need to attract higher- wage gap would narrow by 2.6 percentage points so that end products (for example, designer garments) or move women would earn 89.4 cents for every dollar earned by higher up in the value chain where most innovations a man (Table 5.2). If women had the same attributes as happen. Both of these changes will require the labor male workers in terms of demographics and location in force to have a different skill profile than it currently has. addition to education, then the wage gap would fall by only 4.3 percentage points. In other words, women with • In the worst-case scenario, where advances in the same education and demographic background as male technology restructure global value chains in workers still make only 89 cents for every dollar earned by garment production, there could be declining these men. 12 Although some of the skills needed for garment manufacturing could be partially transferred to other manufacturing sectors, those sectors are small. The entire non- garment manufacturing sector employs only 4 percent of female workers, whereas the garment sector alone employs 17 percent of female workers. 13 The total wage gap of 0.154 log points translates to a female-to-male wage ratio of exp(-0.154) = 0.857. 14 In this analysis, we used the Blinder-Oaxaca decomposition to split the gender wage gap into components that can be explained by the differences in attributes and components that are not explained by the attributes alone. This decomposition separates the average difference in (log) wages into two components as follows: Δy ̅ = βmΔX ̅ + ΔβX ̅ f where the rst component represents the difference due to differences in covariates (or endowments), and the second component is due to differences in how the market values the same characteristics differently for men and for women, also capturing factors that are not enumerated in the data (such as actual years of employment and differential preferences for work). The second component is often referred to as the unexplained gender gap. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 145 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Barriers also exist that prevent women from accessing (second column, Table 5.2). This is indicative of the existence certain occupations and industries, explaining a large of barriers preventing women from working in higher paying share of the remaining gender pay gap. Women are sectors and occupations, even when they have the same more likely than men to be employed as wage workers in the level of education and other background characteristics garments and trade and tourism sectors and are less likely as men. to be employed as wage workers in all other sectors (except farming), including the public sector. Similarly, women are Furthermore, even within the broad occupation more likely to be in craft and trades occupations and less categories, women are more likely to perform tasks likely to be in the “other occupations” category (except in that require fewer cognitive skills and are less likely sales and service occupations). In other words, women to perform tasks that are physically more intensive. are often involved in occupations and sectors that are low Adjusting for the differences in tasks required in their jobs paying. When comparing wages of men and women within as well as for occupation and industry categories further the same industries and occupations, the pay gap lowers lowers the gender pay gap by 4.6 percentage points (third by 3.4 percentage points to 89.4 cents to every male dollar column, Table 5.2). Table 5.2: Decomposition of the Gender Gap in Wage Work, 2014 Basic Speci cation Full Specifcation Full + Tasks Percent Percent of Percent of Gap Gap Gap of gap gap gap (log points) (log points) (log points) explained explained explained Education Variables 0.026*** 16.9 0.022*** 14.3 0.023*** 14.9 (0.004) (0.003) (0.003) Demographic Variables 0.014*** 9.1 0.013*** 8.4 0.013*** 8.4 (0.002) (0.002) (0.002) Location 0.003 1.9 0.003 1.9 0.003 1.9 (0.002) (0.002) (0.002) Industry + Occupation 0.034*** 22.1 0.017 11.0 Categories (0.008) (0.013) Task content variables 0.029** 18.8 (0.012) Aggregate summary: Total Explained 0.043*** 27.9 0.073*** 47.4 0.086*** 55.8 (0.005) (0.010) (0.010) Unexplained gap 0.111*** 72.1 0.081*** 52.6 0.068*** 44.2 (0.010) (0.013) (0.013) Total gap 0.154*** 100.0 0.154*** 100.0 0.154*** 100.0 (0.011) (0.011) (0.011) Source: Author’s estimates using CSES 2014. Notes: Oaxaca-Blinder decomposition of the wage gap. Education variables include years of schooling as well as completed levels of education. Demographic variables include age, age-squared, household size, marital status, and ethnic minority. Location variables are provinces and rural/urban location. Occupation and industry include broad categories, and task content variables are measures of various skills required by the occupation. The sample is restricted to those with wage employment as main activity and those who worked for at least 20 hours in the previous week before the survey. The wage gap is measured in terms of the logarithm of hourly wages. Standard errors are reported in parentheses. *: p<0.1. **: p<.05. ***: p<0.01. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 146 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Nevertheless, differences in worker characteristics rose from 74 cents to 80 cents in urban areas. However, and occupation choices cannot explain half of the the unexplained gap, a measure of factors that could not be gender pay gap. In other words, if female wage workers measured by our data but do affect how men and women had the same education, demographics, location, industry, are remunerated, did not change dramatically between 2009 and occupation as male wage workers and performed and 2014. Such factors include differential preferences for similar tasks within their occupations, they would still earn types of remuneration (for example, wages versus social only 93 cents for every dollar of male wages. This is an benefits, as discussed in Chowdhury et al. 2018), gender improvement over the raw female to male wage ratio of 86 norms that affect job types, women’s intermittent work due cents per dollar, but still only cuts the gap in half. The rest to childbearing, and employers’ preferences for male or of the gap remains unexplained. female workers. The unexplained wage gap, after accounting for covariates and occupation choices, barely fell from 0.09 The gender gap is sensitive to economic fluctuations. log points to 0.08 log points. Since the total wage gap fell The raw gender gap rose steadily from 2008 to 2011 following faster than the unexplained wage gap, the share of the the global financial crisis and the movement of labor towards total gap that remains unexplained has in fact increased agriculture (earnings are not reported in the wage data) and in recent years. This suggests that, during economically started to decline after 2012 (Figure 5.14). At its worst, in difficult times, women disproportionately bear the brunt in 2011, the raw gender gap was equivalent to a female wage the form of lower wages. Not only does the labor market of only 78 cents for every dollar of the male wage. Between favor male characteristics (as measured by the explained 2009 and 2014, the raw gap in wages fell slightly, driven difference), but during such times, the market value of men’s by urban areas. Women’s earnings for every male wage characteristics compared to that of women’s characteristics dollar went up from 82 cents to 86 cents nationwide, and (as measured by the unexplained difference) also rises. Figure 5.14: Decomposition of the Gender Wage Gap, 2007-2015 The gender wage gap widens during difficult economic periods. All 0.25 0.20 Total pay gap (percent) 0.15 Adjusted for demographic characteristics 0.10 Adjusted for demographic characteristics Unexplained Gender Gap, in Log Points 0.05 + occupation & industry choices 0 2007 2008 2009 2010 2011 2012 2013 2014 2015 Year Rural Urban 0.20 0.4 0.15 0.3 0.10 0.2 0.05 0.1 0 -0.05 0 -0.10 -0.1 2007 2008 2009 2010 2011 2012 2013 2014 2015 2007 2008 2009 2010 2011 2012 2013 2014 2015 Year Year Source: Author’s estimates using CSES 2007-2015. Note: Oaxaca-Blinder decomposition of the wage gap. Adjustments for basic covariates include the variables in the first column of Annex Table 5.6. Adjustments for occupation and industry choices include the same set of variables as in the second column of Annex Table 5.6. The sample is restricted to those with wage employment as their main activity and those who worked for at least 20 hours in the previous week before the survey. The wage gap is measured in terms of the logarithm of hourly wages. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 147 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS 5.6 Policies to Improve the Skills of laws, the policies presented in this section are particularly relevant to the current and changing nature of jobs. Cambodian Workers Cambodia needs to dramatically increase the skill Policy Area 1: Modernize What and How level of both its current and future workforce. This People Learn from Pre-school Through implies a need for policymakers to adopt a two-pronged Retirement strategy. The rst prong would be to make broad and deep investments in providing today’s children and youth with (1) Develop and reform pedagogical systems to high quality basic and secondary education, which will yield provide the range of skills valued by the labor market, long-term returns in the form of an educated and skilled including cognitive, socio-emotional, and digital workforce. The second prong would be to invest in practical, literacy skills. Jobs will increasingly require a mix of applied retraining for today’s adult workers. higher-level skills, which the current school system does not provide.15 While it is crucial to enhance the relevance of Most of the jobs of today and tomorrow will require the knowledge-based curriculum and improve the learning workers to have both technical or vocational skills of such material16—as is well documented in the Education and soft skills. As has been discussed in this chapter, Strategic Plan 2014-2018—there is a growing consensus Cambodian jobs are becoming more skill-intensive and that the classroom can also be used to teach a range of non- are increasingly requiring a mix of soft and hard skills. academic skills. Emerging pedagogical methods that rely on In particular, the fastest growing jobs are moving toward student-based, experiential learning allow students to use a requiring semi-skilled workers. The skills appropriate for these jobs are not acquired in university. Instead, the training range of socio-emotional and cognitive skills while learning provided by market-oriented technical and vocational the knowledge that schools typically focus on. The methods institutions along with intensive, on-the-job learning for teaching these skills will differ by the age of the student opportunities are the best way to train workers to qualify for and context. For example, play is an effective technique for and be productive in these jobs. Clearly, university is, and young children in pre-school settings, pedagogical methods will continue to be, important for training the professional that require teamwork and other socio-emotional skills are and managerial classes, but even in high-income countries, effective for school-aged children, while TVET institutions semi-skilled jobs are prevalent. teach socio-behavioral skills that are needed on the job 17 (see table 5.3). Mega-trends are constantly and rapidly changing the nature of jobs and worker skills need to change (2) Use technology to deliver individualized levels of with them. This points to the need to enable Cambodian instruction to improve student learning.18 Computer- workers to develop a broad range of skills, to learn to assisted learning interventions that use technology to provide learn, and to have access to skill upgrading opportunities individual-level instruction have proven to be extremely across a 40 (plus) year worklife. The old model of learning effective at improving learning outcomes at the primary and while young and then working needs to be replaced with post-primary levels.19 An effective computer-aided learning a model of building a strong foundation in youth, then (CAL) system can tailor the level of instruction to each individual constantly upgrading throughout one's work life. student and proceed at the pace at which each student learns best. This kind of technological change in the classrooms will These findings underscore the need for the not displace teachers but, on the contrary, will vastly increase modernization of Cambodia’s skill development their productivity by enabling them to target their teaching system. The system will need to do more for more people to the level and pace of the individual students. The role for a longer period of time than a traditional school-based of the teacher will change from being an active instructor system does. While Cambodia is well on its way to improving to an enabler of student learning. Therefore, policymakers education quality through multiple strategies, plans, and should incentivize and/or facilitate the development 15 The National Employment Policy 2015-2025 of Cambodia has also made the expansion and enhancement of soft skills provision as one of its key objectives. 16 For a detailed discussion of strategies to improve teaching, see Crawford et al. (2018). For a detailed analysis of Cambodia’s education system and its results, see National Education for All Committee (2015). 17 See Cunningham et al. (2016) and Guerra et al. (2014) for examples and evidence of the impact of different interventions throughout the lifecycle. 18 Rigorous evaluations around the world have shown that interventions that only provide hardware support (such as computers or laptops) at home or at school do not improve student learning, and those that allow students to review grade-appropriate content have only a modest impact. See Bulman and Fairlie (2016) for a recent review. 19 See Kremer et al. (2013) for a review and Muralidharan et al. (2017), Barrow et al. (2009), and Banerjee et al. (2007) for specific examples and discussions. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 148 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS of CAL systems suitable for the Cambodian context their policies are effective. There are already national and find ways to make them widely available in schools. assessment systems in place in Cambodia for students in Virtual learning can also substantially reduce the cost grades 9 and 12 in mathematics and language, and the and increase the effectiveness of TVET where part of the Ministry of Education, Youth, and Sport (MoEYS) is preparing curriculum is learning to manipulate new technologies. Also, to implement the Programme for International Student 3-D technologies can virtually replicate modern machines Assessment (PISA) examinations. Testing Cambodian used in various industries, thus providing students with students from an earlier age than at present will make it a virtual “hands-on” experience. Such technologies are possible to identify any problems with student learning still rare but are more cost-effective than purchasing from early on. Furthermore, the government is currently actual machinery. developing and implementing a skills assessment and certification system in Cambodia, which will yield information Technology can also open opportunities for formal or on the technical skills possessed by students and workers. informal self-directed learning. Massive Online Open Courses (MOOCs) provide instruction on a range of courses (4) Integrate and coordinate between a larger number given by a range of instructors from the best universities. of actors in the skills development process. The days Many of these courses are low-cost or cost-free, only when students stopped learning as soon as they left school requiring the user to have good internet connection, are over. Instead, people continue to learn throughout their language skills (usually English), and a willingness to learn. lifetimes in a number of ways: (i) through formal institutions, such as schools and well-designed programs (such as the Similarly, technology can be used to improve the Skills Bridging Program); (ii) less formal programs such as quality of teachers and instruction at scale. A vast amount of high-quality resources exists online, but teachers learning on the job, in an apprenticeship, or through short may not know of the existence of this content. Helping programs; and (ii) very informally through the in uences teachers to find such high-quality content and adapt it for around them (Table 5.3). In the future, it will be important their own teaching needs can be a cost-effective means for learning at all of these levels to be recognized and of improving teacher quality and student learning (Jackson coordinated. This highlights the need to expand two policy and Makarin 2018). reforms that are already underway. First, it will be necessary to expand the certification system so that people who have (3) Develop assessment mechanisms to monitor learned skills through any means can acquire certification. student learning at various levels and in various Second, the Cambodia Qualifications Framework will need subjects. To improve the quality of education, policymakers to be adapted to enable, for example, TVET learners to will need more information about the academic performance transition into higher education or for learning through of students at various education levels to measure whether MOOCs to be credited course-work for qualifications. Table 5.3: A Broad Set of Actors for Lifelong Learning Type Certificate Learning “provider” by stage of lifecycle Formal Certi cate ECD/Primary school Secondary school Higher education (qualifications Second chance education recognition) (Skills Bridging Program) Vocational, technical, professional training Apprenticeships Non-formal Non-formal Youth and adult literacy On-the-job training, professional development certi cate programs Up-skilling and re-skilling short courses Informal No award Self-directed, family-directed Learning in the workplace, family, local community Incidental learning by reading, listening to the radio, etc. Source: Adapted from World Bank (2017b). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 149 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Policy Area 2: Ensure Workers Have Access of a website for both workers and firms containing information to Skills Development Throughout their about short-term training courses including information and feedback on the quality of the short courses. Working Lives (1) Provide additional incentives to students to Policy Area 3: Prioritize Employer encourage them to continue acquiring skills. Since Engagement and the Results Focus of TVET the returns to the first 10 years of school are very low, it might be a rational response under the current economic Reforms environment for students to drop out of school at earlier (1) Enhance the quality, access, governance, and levels. However, as jobs become more sophisticated, an public-private partnerships of the TVET system. The under-educated workforce will constraint the creation of better jobs. Thus, policymakers could consider expanding National Policy on Technical and Vocational Education the cash transfers that are conditional on secondary school and Training provides a comprehensive list of policy completion as this is likely to be a high-return investment of interventions to improve the current TVET system. Given the public resources. low skill level of Cambodia’s current workers and the long time that it takes for the education system to upgrade the Policymakers should also consider providing greater labor force, the TVET sector will need to play an increasingly support to the Skills Bridging Program in order to important role in the future. However, as reported in the open TVET and higher education to the many people National Policy, the current system has many shortcomings, who did not complete secondary school. Many and the most fundamental elements of an effective TVET countries are currently experimenting with incentives, such system are missing. A significant number of policy actions as individual learning accounts, to encourage early school are needed to enable the system to upgrade the skills of leavers to upgrade their skills. For example, Singapore’s the current workforce to meet the needs of the jobs market SkillsFutureCredit aims to encourage skills development at of the near future. The immediate challenge is structuring different stages in life by providing an opening credit of SGD the reforms so that short-term changes can be made while 500 to all Singaporeans aged 25 years old and older (World longer-term systematic reforms are being undertaken to Bank 2017b).This credit has no expiration date and can only produce a results-based, market-responsive TVET system. be used for accredited skills courses. Since women tend to have lower educational attainment than men and because (2) Carry out a systems assessment of the current they have fewer opportunities to enter higher paying jobs, TVET system. While a systems assessment of the TVET additional incentives could be offered to women. system was carried out in 2008 and formed the basis of the National Policy, it did not identify priority areas where reform (2) Tailor adult learning to the working population by was most needed or international best practices to consider expanding and co-financing technical short courses. when addressing these issues, and it did not quantify the Working adults face time constraints (because they need depth of the problems to help set benchmarks against which to work) and have immediate earnings needs (they need to to monitor the effectiveness of reforms. Therefore, in order provide for their families) that leave little time for learning to prioritize and sequence the reforms, we propose that the new skills. However, they will be in great need of upgrading Ministry of Labor and Vocational Training (MLVT) carry out a their skills as the nature of jobs continues to change. To new systems assessment using the World Bank’s Systems serve the adult working population, the skills sector will Approach for Better Education Results (SABER) tool. SABER need to offer more short, practical, and modular skills allows for a systematic assessment and benchmarking of a development courses. For example, an older worker might country’s policy intent (policies, laws, regulations, and the enroll in a short course to learn Excel to help her to manage her small business better or in a short course on electronic practices of ministries and relevant government agencies in repair to become qualified to take on a supervisor position technical and vocational training) and the implementation in the electronic assembly plant where she works. of policies by public and private TVET providers (their characteristics, actions, values, and outcomes). The tool Policymakers can play several roles in expanding and has been used in more than 35 developed and developing improving the system. First, they can identify the market- countries to identify the shortcomings, both system-wide demanded skills for which the private sector is not providing and within institutions, in order to inform policy (see Figure training and then incentivize the private and public sectors to 5.15 for an example of SABER results). This assessment provide such courses. Second, they can provide incentives to can also be used to assess institutions under the purview of the private sector and learning institutions for the development other ministries and private sector training programs. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 150 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Figure 5.15: Sample Workforce Development SABER Results 1.Direction 4 Strategic 9. Accountability 3 2. Demand-led Framework 2 System 1 8. Relevance 3. Coordnation Oversight 0 Malaysia 2013 7. Excellence 4. Funding Vietnam 2012 China 2014 6. Pathways 5. Standards Sri Lanka 2014 Service Delivery (3) Develop policies to encourage enterprises to play labor market entrants. The vouchers successfully opened up a bigger role in skills development. Strategy 3 of the on-the-job learning opportunities to the young women even National Strategy for TVET defines policy actions to foster though they did not lead to most of them being employed public-private partnerships. The challenge will be to set up long term (Groh et al. 2012). incentives and processes to ensure that these partnerships can play a meaningful role. There are four ways in which (b) Engage with the education and training institutions. private firms can become more engaged in the provision The public sector can help to engage employers and training of skills development and training with the support of the institutions in a common conversation to find solutions. public sector. For example, Australia’s publicly financed Skills Service Organisations, works with industry associations and the (a) Provide workers with skills development training sector to identify what skills are in demands and opportunities. Firms already provide the majority of non- how they can be learned. In Malaysia, industry associations formal skills development through on-the-job learning, are also involved in identifying skill needs through sectoral but they tend to provide less formal training than their training committees. Korea’s SME training consortiums are workers need. Various policies have been adopted in other groups of SMEs in the same sector and region who come countries to increase the amount of training provided by together to identify training needs and available providers. firms. First, several countries have imposed training levies Policymakers provide incentives to the SMEs in the form of on firms with the money going into a central public fund tax rebates to allow their workers to participate in training that is used to finance the provision of training. In the case as well as subsidizing training providers. of Malaysia, firms now seem more willing to send their workers to training courses in order to reap the benefits (c) Finance their own and public/private skills of the mandatory payment. Second, in some cases, firms development services. Firms can finance the training may not realize that their workers need training. Several of their workers in two different ways. First, they can pay OECD countries have financed firm or industry-wide training institutions to provide training to their workers. For example, needs assessments to identify the skills that are lacking enterprise and training institutions in Cambodia have recently in the workforce that the firms need to fill (World Bank come together to jointly design and implement the pilot 2017a). Third, firms may not know how to structure learning Cambodia Skills Development Fund with a heavy subsidy opportunities such as apprenticeships. Several countries from the public sector. Careful monitoring of the pilot will be have provided technical support to firms to help them to important since this program design has had mixed results in structure and implement their apprenticeship programs to other countries. Second, firms can provide in-kind support to lower their costs and increase the quality of the experience training institutions by donating old machinery for students for learners. Firms may need to receive additional incentives to practice on, allowing their staff to give guest lectures in to train hard-to-employ workers. For example, a program training institutions, allowing training instructors access in Jordan gave “employment vouchers” to young women to the factory floor to enable them to upgrade their own who had graduated from college as a way to subsidize their technical skills, and allowing their staff to take courses during wages and benefits in firms that temporarily hired these new their working day. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 151 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS (d) Advocate for broader policies. Firms can also use their (2) Provide incentives to private sector job services platform to lobby policymakers for greater transparency and to extend their reach to include employers of low- information about skills providers, for financial assistance skilled workers. A vibrant private sector has emerged to for students, and social policies to enable workers to take create job boards, but few of them undertake the labor- time off work for skills development or to alleviate the costs intensive task of engaging with employers of low-skilled to workers of changing jobs. workers. Initiatives in other countries have shown that it is feasible to persuade employers of low-skilled labor to use Lay the groundwork for the introduction of results- online job matching services. For example, Kenya’s Duma based financing for TVET. The National Policy for TVET Works (www.dumaworks.com) is a mobile social network contains various policy statements about results-based that enables informal workers to find jobs through friends financing (RBF), but these will need to be brought together of friends. Souktel (www.souktel.com), an international into a cohesive policy. The key elements of the system tech company specializing in emerging markets, uses text should include: (i) an information system that brings together messaging to reach a wider clientele than just internet outcome information (job placements, post-graduation users. Informal employers can post vacancies via text salaries, and employer satisfaction for each field of study) for without having to go through an extensive registration each institution as collected by tracer studies and employer process. In July 2011, Souktel surveyed their clients in surveys, processed using simple statistical software Palestine and found that 84 percent of job seekers reported packages, and shared through communications materials a 92 percent reduction in the time that they spent on job and websites; (ii) clear and transparent requirements for hunting compared to using traditional job search methods. institutional accreditation to make it easier for institutions Meanwhile, Palestinian employers reported a 50 percent to comply with basic standards of safety and competence; or greater reduction in the costs and time spent on hiring and (iii) contracting and procurement processes that allow workers. TaskRabbit (www.taskrabbit.com) takes many for public financing of private institutions. forms in different countries, but the basic premise is that it posts requests from people who need workers to carry Policy Area 4: Develop Labor Market out short-term tasks such as moving and packing or home repairs and other manual tasks, and the workers receive a Intelligence ranking by anyone who has employed them. This may be a (1) Facilitate the job search process by providing model for a centralized vacancy information service for low- labor market information to students, workers, firms, skilled workers driven by the job seekers themselves rather TVET institutions, and policymakers. Poor job search than by employers’ needs. methods lead to an allocative inefficiency and lower quality (3) Support the National Employment Agency to jobs. The analysis in this chapter did not cover the topic become the knowledge hub for jobs and job searches of job search because of data limitations. However, the in Cambodia. The NEA is already developing tools to be a job search situation in Cambodia is likely to be similar to knowledge hub for jobs in Cambodia by disseminating labor that in other countries in the region where job searches market information for jobseekers. The NEA’s “Cambodia are largely carried out through informal networks, leading Jobs Outlook” (financed by the Swedish Aid Agency SIDA) to an inefficient allocation of workers to jobs and thus to publishes information on the jobs available in today’s labor lower labor productivity and poorer quality jobs than if market, based on data collected by the NEA. Cambodia matches were more efficient. Students, workers, firms, could learn lessons from Korea, which has developed job TVET institutions, and policymakers are not able to easily forecasting models and a national website for sharing and access the kind of information that they need to make analyzing information on job openings and labor market informed decisions about jobs and skills development. trends. One vision for Cambodia is to build a knowledge Learning throughout one's working life requires workers to hub similar to interactive labor market intelligence websites take control of their own skill development process. While that guide users toward jobs and skill development paths, the process of transitioning from school to higher education such as Canada’s system. For the NEA to play the key role is well understood, planning a career path that involves in the development of this information hub, it will need to constant skills upgrading is much more difficult to map. The scale up from small pilots funded by donors to having a National Institute of Statistics’ annual report on the Socio- secure budget that will allow for it to expand its programs, Economic Survey is rich in statistics but is not the most engage new technologies and technical experts to improve accessible way to present this information to Cambodian its data collection methods, and maintain an interactive students or workers (who may not have an adequate and updated job search website that is compatible with the understanding of statistics) who need to make decisions private sector’s job vacancy and search tools. about their skills development CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 152 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS 5.7 References Acemoglu, Daron, and David Autor. 2011. “Skills, tasks, and technologies: Implications for employment and earnings” in Card, David and Orley Ashenfelter, eds. Handbook of Labor Economics Vol. 4, North Holland: Elsevier, 1043-1171. Acemoglu, David, and Jorn-Steffen Pischke. 1998. “Why do firms train? Theory and evidence” The Quarterly Journal of Economics 113(1): 79-119. ADB-HRINC-DBLink. 2013. “The supply and demand of TVET skills in Cambodia: A study of Cambodia’s large enterprises and TVET Institutions.” Strengthening Technical and Vocational Education and Training Project II: Technical Assistance Consultant’s Report. https://www.adb.org/projects/documents/supply-and-demand-tvet-skills-cambodia-tacr Andrabi, Tahir, Jishnu Das, and Asim Ijaz Khwaja. 2013. “Students today, teachers tomorrow: Identifying constraints on the provision of education” Journal of Public Economics 100: 1-14. Autor, David, Frank Levy, and Richard Murnane. 2003. “The skill content of recent technological change: an empirical exploration.” The Quarterly Journal of Economics 118(4):1279-1333. Autor, David. 2013. “The ‘Task Approach’ to Labor Markets: An Overview” Journal for Labour Market Research 46(3): 185-199. Autor, David. 2015. “Why are there still so many jobs? The history and future of workplace automation” Journal of Economic Perspectives 29(3): 3-30. Banerjee, Abhijit V., Shawn Cole, Esther Duflo, and Leigh Linden. 2007. “Remedying education: Evidence from two randomized experiments in India” The Quarterly Journal of Economics 122(3): 1235-1264. Barro, Robert and Jong-Wha Lee, 2013, "A New Data Set of Educational Attainment in the World, 1950-2010." Journal of Development Economics, vol 104, pp.184-198. Barrow, Lisa, Lisa Markman, and Cecilia Elena Rouse. 2009. “Technology’s edge: The educational benefits of computer-aided instruction” American Economic Journal: Economic Policy 1(1): 52-74. Bulman, George, and Robert W. Fairlie. 2016. “Technology and education: Computers, software, and the internet” in Eric Hanushek, Stephen Machin, and Ludger Woessmann, eds. Handbook of the Economics of Education Vol. 5, North Holland: Elsevier: 239-280 Burde, Dana, and Leigh L. Linden. 2013. “Bringing education to Afghan girls: A randomized controlled trial of village-based schools” American Economic Journal: Applied Economics 5(3): 27-40. Brynjolfsson, Erik, and Andrew McAfee. 2014. The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies New York: WW Norton & Company. Chowdhury, Iffat, Elizaveta, Perova, Aneesh Mannava, and Hillary Johnson. 2018. “Gender Gap in Earnings in Vietnam: Why Do Vietnamese Women Work in Lower Paid Occupations?” World Bank Policy Research Working Paper No. 8433. Washington, D.C.: World Bank. Crawford, Michael, Amer Hasan, and Raja Bentaouet Kattan. 2018. Growing Smarter : Learning and Equitable Development in East Asia and Paci c. World Bank East Asia and Pacific Regional Reports. Washington, D.C.: World Bank. Cunningham, Wendy, Pablo Acosta and Noël Muller. 2016. Minds and Behaviors at Work: Boosting Socioemotional Skills for Latin America’s Workforce. Washington, D.C.: The World Bank. Deming, David J. 2017. “The growing importance of social skills in the labor market.” The Quarterly Journal of Economics 132(4): 1593-1640. Duflo, Esther. 2001. “Schooling and labor market consequences of school construction in Indonesia: Evidence from an unusual policy experiment” American Economic Review 91(4): 795-813. Emont, Jon. 2018. “The robots are coming for garment workers: That’s good for the U.S., bad for poor countries.” Wall Street Journal. February 16, 2018. https://www.wsj.com/articles/the-robots-are-coming-for-garment-workers-thats-good-for- the-u-s-bad-for-poor-countries-1518797631 CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 153 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Groh, Matthew, Nandini Krishnan, David McKenzie, and Tara Vishwanath. 2012. “Soft skills or hard cash? the impact of training and wage subsidy programs on female youth employment in Jordan.” Policy Research Working Paper No. 6141. Washington, D.C.: World Bank. Guerra, Nancy, Kathryn Modecki, and Wendy Cunningham. 2014. “Developing social-emotional skills for the labor market : the PRACTICE model.” Policy Research Working Paper No.7123. Washington, DC: World Bank.. Jackson, Kirabo, and Alexey Makarin. 2018. “Can online off-the-shelf lessons improve student outcomes? Evidence from a field experiment” American Economic Journal: Economic Policy 10(3): 226-254. Kremer, Michael, Conner Brannen, and Rachel Glennerster. 2013. “The challenge of education and learning in the developing world” Science 340(6130): 297-300. Maloney, William, and Carlos Molina. 2016. “Are automation and trade polarizing developing country labor markets too?” World Bank Policy Research Working Paper. No. 7922. Washington, D.C.: World Bank. Montenegro, Claudio E. and Harry Anthony Patrinos. 2014. “Comparable estimates of returns to schooling around the world.” World Bank Policy Research Working Paper. No 7020. Washington, D.C.: World Bank. MoEYS—Ministry of Education, Youth, and Sports. 2014. Education Strategic Plan 2014-2018. Royal Government of Cambodia. http://www.moeys.gov.kh/images/moeys/policies-and-strategies/559-en.pdf MoEYS. 2017. “Grade six national assessment findings in 2016.” Flyer. Royal Government of Cambodia. Muralidharan, Karthik, Abhijeet Singh, and Alejandro J. Ganimian. 2017. “Disrupting education? Experimental evidence on technology-aided instruction in India.” NBER Working Paper. No 22923. National Center for O*Net Development. U.S. Department of Labor. www.ontonline.org National Education for All Committee. 2015. Cambodia National Education for All 2015 National Review. Royal Government of Cambodia. http://unesdoc.unesco.org/images/0022/002297/229713e.pdf NEA—National Employment Agency. 2014. Skills Shortages and the Skills Gap in the Cambodian Labor Market: Evidence from the Employer Skills Needs Survey. Ministry of Labor and Vocational Training, Royal Government of Cambodia. NEA. 2015. Skills Shortages and Skills Gap in the Cambodian Labor Market: Evidence from the Employer Skills Needs Survey. Ministry of Labor and Vocational Training, Royal Government of Cambodia. NEA. 2016. Skills Shortages and Skills Gap in the Cambodian Labor Market: Evidence from the Employer Skills Needs Survey. Ministry of Labor and Vocational Training, Royal Government of Cambodia. Rodrik, Dani. 2016. “Premature deindustrialization” Journal of Economic Growth 21(1): 1-33. Stanovich, Keith E. 1986. “Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy.” Reading Research Quarterly 21: 360-406. USAID. 2015. Assessment of early grade reading in the education sector in Cambodia. Washington D.C: United States Agency for International Development. World Bank. 2017a. Cambodia: Sustaining strong growth for the benefit of all. Washington D.C: World Bank. World Bank. 2017b. HRDF National Workforce Human Capital Development Blueprint 2018-2020 & 2021-2025. Washington D.C: World Bank. World Bank Human Capital Index. http://www.worldbank.org/en/publication/human-capital CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 154 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS ANNEX 5.1: Additional Figures and Tables Annex Figure 5.1: Occupation and Age of Cambodia’s Labor Force, 2014 (a) All Labor Force (b) Female Labor Force 100% 100% Percent of Female Labor Force 80% 80% Percent of Labor Force 60% 60% 40% 40% 20% 20% 0% 0% 15 20 25 30 35 40 45 50 55 60 65 15 20 25 30 35 40 45 50 55 60 65 Age Age c) Rural Labor Force (d) Urban Labor Force 100% 100% 80% 80% Percent of Urban Labor Force Percent of Rural Labor Force 60% 60% 40% 40% 20% 20% 0% 0% 15 20 25 30 35 40 45 50 55 60 65 15 20 25 30 35 40 45 50 55 60 65 Age Age Agriculture and elementary Sales, services Machine operators Craft and related trades Clerical Manager, professionals, technical Source: Author’s estimates using CSES 2014 data. Note: Sample restricted to the working age population aged 15-64 who had worked in the previous seven days before the survey. The sectors are defined for sector of main employment. Point estimates for each age estimated using locally linear regression with bandwidth of 1 year. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 155 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Annex Figure 5.2: Occupation and Education by Sub-group, 2014 (a) Male Workers 100% Percent of Male Labour Force 80% 60% 40% 20% 0% 0 1 2 3 4 5 6 7 8 9 10 11 HS+ Col+ Years of School (b) Female Workers 100% Percent of Female Labour Force 80% 60% 40% 20% 0% 0 1 2 3 4 5 6 7 8 9 10 11 HS+ Col+ Years of School (c) Rural Workers 100% Percent of Rural Labour Force 80% 60% 40% 20% 0% 0 1 2 3 4 5 6 7 8 9 10 11 HS+ Col+ Years of School (d) Urban Workers 100% Percent of Urban Labour Force 80% Agriculture and elementary Machine operators 60% Craft and trades 40% Service and sales Clerical 20% Manager, professionals, 0% technical 0 1 2 3 4 5 6 7 8 9 10 11 HS+ Col+ Years of School Source: Author’s estimates using CSES 2014 data. Note: Sample restricted to the working age population aged 15-64 who had worked in the previous seven days before the survey. The occupation categories are defined for main employment. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 156 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Annex Figure 5.3: Returns to Education in Wage Jobs, 2007-2015 All 120% 100% Returns Relative to Primary 80% 60% 40% Incomplete Secondary 20% HS graduate 0% College graduate 2007 2010 2012 2014 Female Male 150% 100% Returns Relative to Primary 80% 100% 60% 40% 50% 20% 0% 0% 2007 2010 2012 2014 2007 2010 2012 2014 Rural Urban 120% 150% 100% Returns Relative to Primary 100% 80% 50% 60% 40% 0% 20% -50% 0% 2007 2010 2012 2014 2007 2010 2012 2014 Source: Author’s estimates using CSES 2007-2015. Note: Sample restricted to the working age population, aged 15-64 who had worked in wage employment for at least 20 hours in the previous seven days before the survey. The vertical dashed lines indicate completion of primary, lower secondary, upper secondary, and college levels. The dependent variable is the logarithm of hourly wages. The regressions control for individual covariates (gender, age, marital status, and household size and composition) and location (rural-urban and region). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 157 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Annex Table 5.1: Profile of the Cambodian Labor Force, 2014 Labor Force aged 15-64 and working Full labor working labor force force Male Female Rural Urban Female 0.514 0.482 0.483 0.480 (0.500) (0.500) (0.500) (0.500) Age: 15-24 0.336 0.290 0.298 0.282 0.311 0.221 (0.472) (0.454) (0.457) (0.450) (0.463) (0.415) 25-34 0.274 0.301 0.304 0.298 0.270 0.404 (0.446) (0.459) (0.460) (0.457) (0.444) (0.491) 35-44 0.159 0.175 0.176 0.175 0.174 0.178 (0.365) (0.380) (0.381) (0.380) (0.380) (0.382) 45-54 0.143 0.153 0.149 0.157 0.160 0.129 (0.350) (0.360) (0.356) (0.364) (0.367) (0.335) 55-64 0.089 0.081 0.073 0.089 0.085 0.068 (0.285) (0.273) (0.260) (0.285) (0.279) (0.251) Education: None 0.159 0.159 0.118 0.204 0.189 0.062 (0.366) (0.366) (0.322) (0.403) (0.391) (0.241) Incomplete Primary 0.287 0.307 0.288 0.328 0.346 0.180 (0.452) (0.461) (0.453) (0.470) (0.476) (0.384) Incomplete Secondary 0.420 0.408 0.441 0.373 0.399 0.440 (0.494) (0.492) (0.497) (0.484) (0.490) (0.496) HS graduate 0.097 0.082 0.100 0.063 0.056 0.168 (0.295) (0.275) (0.300) (0.243) (0.230) (0.374) College or more 0.038 0.043 0.053 0.032 0.010 0.150 (0.191) (0.202) (0.223) (0.176) (0.100) (0.358) Marital status: single 0.358 0.322 0.340 0.304 0.313 0.354 (0.480) (0.467) (0.474) (0.460) (0.464) (0.478) Ethnic minority 0.036 0.034 0.035 0.034 0.038 0.024 (0.186) (0.182) (0.184) (0.181) (0.190) (0.154) Household size 5.141 5.094 5.177 5.004 5.088 5.114 (1.889) (1.881) (1.856) (1.902) (1.862) (1.942) Rural 0.754 0.767 0.766 0.768 (0.430) (0.423) (0.423) (0.422) Region: East 0.048 0.048 0.048 0.047 0.054 0.029 (0.213) (0.213) (0.215) (0.212) (0.225) (0.167) South-East 0.207 0.209 0.204 0.216 0.252 0.069 (0.405) (0.407) (0.403) (0.411) (0.434) (0.254) South-West 0.175 0.176 0.176 0.176 0.198 0.102 (0.380) (0.381) (0.380) (0.381) (0.399) (0.303) North 0.148 0.151 0.153 0.149 0.171 0.086 (0.355) (0.358) (0.360) (0.356) (0.377) (0.281) North-West 0.159 0.161 0.166 0.155 0.166 0.144 (0.366) (0.367) (0.372) (0.362) (0.372) (0.351) Special Region 0.264 0.255 0.253 0.257 0.159 0.570 (0.441) (0.436) (0.435) (0.437) (0.366) (0.495) Source: Author’s estimates using CSES 2014. Note: Standard deviations in parentheses. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 158 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Annex Table 5.2: Employment of the Working Age Population, thousands Wage Not Family Year Private Public Farming Non Farm Ent Family help working Farming 2007 1127 334 2024 889 2542 1433 4566 2008 1257 365 2061 896 2394 1601 4455 2009 1118 340 2981 1202 1754 1400 4736 2010 1386 386 3148 1368 1488 1231 4635 2011 1604 351 3555 1300 1209 1181 4764 2012 1834 393 3578 1386 712 1502 4290 2013 2156 367 3646 1412 371 1656 4017 2014 2588 398 3336 1281 455 1742 3791 2015 2934 412 3238 1340 302 1749 3540 Source: Author’s estimates using CSES 2007-2015. Working age population estimate from UNPOP. Annex Table 5.3: Number of Wage Workers, thousands Year Agriculture Trade Services Construction Manufacturing 2007 312.5 44.0 766.3 279.9 370.9 2008 391.2 36.8 824.5 261.3 499.7 2009 521.6 37.7 721.5 245.0 453.2 2010 531.5 66.8 881.8 282.2 541.2 2011 570.5 80.5 905.3 269.4 699.5 2012 604.0 72.0 1047.9 357.8 749.2 2013 707.1 86.6 1101.0 459.8 875.8 2014 602.2 110.8 1218.3 540.6 1116.2 2015 680.7 123.9 1383.4 636.2 1202.5 Source: Author’s estimates using CSES 2007-2015. Working age population estimate from UNPOP. Note: Trade includes retail as well as other trades. Services include all other sectors. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 159 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Annex Table 5.4: Largest Occupations, 2009 and 2014 2009 2014 Occupation Share Occupation Share Subsistence crop farmers 0.257% Subsistence crop farmers 0.198% Subsistence livestock farmers 0.102% Garment and related trades workers 0.102% Market gardeners and crop growers 0.076% Agricultural, forestry, fishery labourers 0.069% Agricultural, forestry, shery labourers 0.063% Market gardeners and crop growers 0.069% Shop salespersons 0.050% Shop salespersons 0.060% Street and market salespersons 0.049% Subsistence livestock farmers 0.054% Firewood and water collectors 0.047% Street and market salespersons 0.036% Garment and related trades workers 0.044% Animal producers 0.035% Animal producers 0.036% Mining and construction labourers 0.031% Subsistence fishers, hunters, gatherers 0.025% Building frame and related trades workers 0.028% Source: Author’s estimates using CSES 2009 and 2014. Note: Occupations are 3-digit occupation codes. Primary occupation is used to categorize workers. Annex Table 5.5: Real Wage Earnings, 2007-2015 Monthly Wages, in 2014 PPP International $ Percent of Workers who Earn Below Year p=25 p=50 Mean p=75 $2 per day $4.22 per day 2007 112.6 187.7 283.2 300.3 0.121% 0.284% 2008 112.6 180.2 252.1 277.8 0.118% 0.291% 2009 113.4 188.9 245.4 272.1 0.143% 0.285% 2010 145.3 218.0 267.0 294.3 0.089% 0.224% 2011 155.0 226.0 277.9 322.4 0.060% 0.183% 2012 167.3 241.0 289.6 321.3 0.043% 0.138% 2013 195.1 260.1 322.2 390.1 0.032% 0.098% 2014 250.4 325.5 389.6 425.7 0.018% 0.085% 2015 266.0 371.1 411.2 463.9 0.009% 0.070% Source: Author’s estimates using CSES 2007-2015. Note: Monthly wages were recorded in KHR, were converted to 2014 prices using consumer price index, and were converted to international PPP $ using the conversion rate in 2014 of 1597.30 KHR per PPP$. The $4.22 per day threshold is $2 per day multiplied by the ratio of the total population to the share of individuals that are employers, employees, or own-account workers. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 160 TECHNICAL REPORT CHAPTER 5: WORKFORCE SKILLS FOR MORE, BETTER, AND MORE INCLUSIVE JOBS Annex Table 5.6: Determinants of Task Content of Jobs, 2014 Managerial Repetitive Repetitive Non-rep. Analytical interpersonal mental physical physical Social (1) (2) (3) (4) (5) (6) Schooling: No-schooling omitted Some schooling 0.112*** 0.108*** 0.029* -0.099*** -0.146*** 0.101*** (0.030) (0.024) (0.016) (0.016) (0.016) (0.018) Primary complete 0.271*** 0.241*** 0.066*** -0.200*** -0.275*** 0.214*** (0.035) (0.029) (0.019) (0.019) (0.019) (0.023) L. Sec. complete 0.422*** 0.384*** 0.084*** -0.329*** -0.404*** 0.359*** (0.040) (0.035) (0.025) (0.025) (0.024) (0.029) U. Sec. complete 0.803*** 0.813*** 0.047 -0.798*** -0.899*** 0.805*** (0.050) (0.049) (0.044) (0.039) (0.036) (0.043) College complete 1.049*** 0.718*** 0.462*** -1.130*** -1.472*** 0.887*** (0.057) (0.053) (0.039) (0.036) (0.039) (0.040) Female -0.185*** -0.089*** -0.146*** -0.187*** -0.396*** 0.058*** (0.016) (0.016) (0.012) (0.011) (0.012) (0.013) Age -0.514*** -0.507*** -0.084*** 0.356*** 0.372*** -0.397*** (0.037) (0.038) (0.027) (0.032) (0.030) (0.034) Age-squared (/100) 0.010** 0.013*** -0.003 -0.016*** -0.008*** 0.011*** (0.005) (0.004) (0.003) (0.003) (0.003) (0.003) Rural -0.016*** -0.010* -0.006* 0.008** 0.002 -0.001 (0.006) (0.005) (0.003) (0.003) (0.003) (0.004) Region, Phnom Penh Special zone omitted East -0.218** 0.027 -0.197*** 0.083* 0.209*** 0.083 (0.104) (0.085) (0.042) (0.050) (0.054) (0.057) South-East -0.136** -0.027 -0.105*** 0.080** 0.133*** -0.003 (0.055) (0.049) (0.029) (0.035) (0.034) (0.039) South-West -0.158*** -0.074 -0.076** 0.005 0.014 -0.045 (0.059) (0.054) (0.035) (0.039) (0.038) (0.043) North -0.187*** 0.027 -0.202*** -0.040 0.094** 0.080* (0.067) (0.057) (0.033) (0.038) (0.037) (0.044) North-West -0.157*** 0.084 -0.168*** 0.040 0.163*** 0.068* (0.058) (0.051) (0.032) (0.039) (0.039) (0.041) Source: Author’s estimates using CSES 2014 and O*NET database on skills associated with occupation. Notes: The dependent variables are the standardized scores of the tasks indicated in the column heading. The task intensity measures follow the classification used in Acemoglu and Autor (2011). Task intensity scores are standardized by the mean and standard distribution of the scores in non- subsistence occupations weighted by the labor share of those occupations in 2009. The sample is restricted to those who worked for at least 20 hours in the previous week before the survey. Omitted controls include marital status, household size, and ethnicity. Standard errors, clustered at the PSU level, are reported in parentheses. * p<0.1. ** p<.05. *** p<0.01. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 161 TECHNICAL REPORT CHAPTER 6: Policy for Future Jobs Consistent with Cambodia’s Vision 2050 Wendy Cunningham and Claire H. Hollweg 6.1 Introduction Although the current growth model in Cambodia is still delivering new and good jobs, it may be slowing down. Several factors are reducing the returns from Stage I global integration (entering global production networks), the stage in which Cambodia is located (see Box 6.1). These factors include weak employment growth, partly driven by low productivity and high unit labor costs as well as a low-skilled labor force, weak links between the FDI sector and the rest of the economy (where most of the jobs are), and the concentration of exports in sectors of low sophistication. For Cambodia, this implies that, to improve the quality and inclusiveness of its jobs, it will need to move up to Stage II of global integration: attracting higher value-added FDI and involving more local firms in the production network. Box 6.1: The Stages of Global Integration for Development and Jobs Global value chains (GVCs) and the flows of foreign direct investment associated with them can provide developing countries with an opportunity for economic development and higher quality and inclusive jobs. However, becoming integrating into GVCs is not a one-off event but rather a cumulative, multi-stage, deliberate process. Stage I: Entering global production networks. This involves leveraging the country’s comparative advantage to attract foreign investors. This usually means offering: (i) abundant, low-skilled labor; (ii) a friendly business climate for tangible and intangible assets; (iii) few regulatory or restrictive measures on trade and investments; and (iv) a stable regulatory and political environment to attract multinational corporations that will provide low value-added jobs in GVCs. Stage II: Attracting higher value-added FDI and increasing domestic densification. Countries can expand and strengthen their participation in GVCs by becoming competitive in higher-value-added products, tasks, and sectors. Densification—engaging more local actors in the production network—is also desirable at this stage. Not only will this help domestic private firms benefit from doing business with multinational corporations, but it will also give them an opportunity to learn and to increase their productivity-led growth. In order for an economy to move into this stage, it needs a labor force with relevant skills, absorptive capacity, and strong links between domestic firms and GVCs. Stage III: Sustainability. Once an economy is at this stage, the domestic sector is sufficiently robust that it does not depend on the presence of FDI firms. In fact, as China has shown, domestic firms may have become the lead companies in GVCs. However, in order to sustain the GVC-centric approach to development, policymakers need to pay keen attention to the macroeconomic, social, and environmental sustainability of the economy’s global integration. Source: Taglioni and Winkler (2016) CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 163 TECHNICAL REPORT CHAPTER 6: POLICY FOR FUTURE JOBS CONSISTENT WITH CAMBODIA’S VISION 2050 While current trends predict a somewhat benign jobs challenges. These policies were selected based on existing picture, emerging mega-trends present Cambodia analytical work within Cambodia as well as other countries’ with new and better job opportunities provided that experiences in overcoming similar challenge, and are shown policies are put in place to take advantage of those in Figure 6.1. trends. This chapter proposes policies aligned along four strategic areas for leveraging mega-trends to create We also chose these policies because each meets the better jobs: (i) getting more benefits out of FDI, (ii) aligning criteria of being impactful, feasible, sustainable, and domestic policies with the goal of creating more, better, and measurable. Impactful policies are those that have a direct more inclusive jobs; (iii) strengthening linkages between the and significant effect on the challenge that they are designed export and domestic sectors; and (iv) building the skills and to alleviate. Policies should be feasible both in terms of competitiveness of the Cambodian labor force. These two the country having enough capacity to implement the new goals are inter-related. The key to upgrading Cambodia’s policies but also in terms of there being enough fiscal space jobs is to move to Stage II of global integration, which will and political support to implement them. Sustainable policies involve attracting higher-value FDI, supporting links between are those that not only support sustainable development but exporting firms and input-supplying firms, and ensuring a also can be sustainably implemented by the government with transition for workers. Doing so will require policymakers no outside support and with no threat of being reversed. to adopt policies designed to attract higher-quality foreign Finally, policies should be measurable to ensure that their investment, to build the competitiveness of domestically impact can be monitored over time. owned businesses, and to upgrade and expand the skills of Cambodia’s workers. These policies overlap with all four rectangles in Phase IV of the Government of Cambodia’s “The Seven sets of policies can help firms, workers, and Rectangular Strategy for Growth, Employment, Equity, foreign investors overcome the constraints that are and Efficiency: Building the Foundation Toward currently preventing the creation of higher-value Realizing the Cambodia Vision 2050.” as well as with FDI jobs and FDI spillovers into jobs in the domestic the “environment” and the “core” policies. The discussion economy. In the previous chapters, we identified the key in this chapter maps the policy options in this document challenges to upgrading Cambodia’s jobs along with areas to those in Phase IV of the Rectangular Strategy and also where policy intervention would be helpful. In this chapter, identifies the government agencies that could take the lead we set out the policies that we identified to overcome these in each policy area. Figure 6.1: Seven Policies for More and Better Jobs Domestic Economy Export Sector Enhance Skills Development Boost HHE Productivity Diversify Into Higher Strengthen Linkages Between Exporters and Domestic Firms Value Added Exports Support SME Expansion Ensure Efficient Labor Mobility Macroeconomic Stability Source: Authors. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 164 TECHNICAL REPORT CHAPTER 6: POLICY FOR FUTURE JOBS CONSISTENT WITH CAMBODIA’S VISION 2050 6.2 Seven Sets of Policies for More, In the medium term, policymakers should focus on: Better, and More Inclusive Jobs • Introducing super tax deductions for R&D and staff training to incentivize firms to undertake these 1. Diversify Exports and FDI into Higher activities. These investments will not only yield returns Value-added Value Chains (or Segments of for investors, but they are also expected to generate spillovers to the domestic economy through knowledge Value Chains) and technological transfers. Diversifying Cambodia’s export basket to include more • Automating the process of sanitary and phytosanitary higher-value activities could attract more productive certification while fostering more use of IT. By meeting jobs that pay higher wages. In order to encourage the global certification standards, Cambodia will not only creation of better jobs, policymakers will need to focus on ensure the high quality of agricultural goods but will also policies to attract jobs-friendly FDI (from global or emerging support agricultural exports to feed expanding regional regional markets) and to move the economy into higher and urban populations. value-added value chains (or segments of value chains), including services for export. Specifically, they will need New analytical work is needed to identify the to reduce the costs of imports and exports, which can be bottlenecks to expanding Cambodia’s export and FDI done in several ways including: (i) modernizing procedures base. This could include exploring models for economic in trade facilitation; (ii) developing logistics infrastructure; diversification from the point of view of trade diversification (iii) simplifying border processes; (iv) providing new types of and GVC participation, including in the garments and tourism incentives that target specific investments; and (v) creating sectors (Box 6.2). It will be particularly important to identify quality assurance facilities. Policymakers also need to the current workforce constraints and skills needed in the explore ways to maintain Cambodia’s preferential access to future since these investments take time to mature. Any key markets for its new labor-intensive exports, especially policies aimed at fostering Cambodia’s transition into such as the end-dates of the US and European preferential global value chains as those in the garments and hospitality access treaties draw nearer. sectors, which tend to be important for female employment, should also support inclusive jobs. In the short run, the following policy directions should be prioritized: The policies that we identified for diversifying exports and FDI into higher value-added value • Shifting away from tax holidays towards streamlined chains correspond to the Overarching Environment incentives aimed at attracting private investment in within Phase IV of the Rectangular Strategy. The main machinery, such as tax credits on capital investment and implementing agencies would be the Ministry of Economy and technological transfers. These incentives will help firms Finance, the Council for the Development of Cambodia, the upgrade their production processes, thereby increasing Ministry of Commerce, the Ministry of Agriculture, Forestry investment in the domestic economy and increasing and Fisheries, and the General Department of Agriculture. labor productivity. • Reviewing and streamlining customs requirements and 2. Streamline Procedures and Reduce the the clearance process for export and import licenses. Costs of Establishing and Expanding SMEs This will help to save time and reduce costs for traders, thus making Cambodia’s exports more competitive in Firm sales are generally required to grow for firms global markets. to hire new workers. This points to the need for a set of policies directed toward enhancing firm profitability, • Signing, finalizing, and enforcing international trade which is then expected to have spillovers on job creation. A and investment agreements with Cambodia’s key range of policies can support the creation of new firms and trade partners to diversify across destinations as well their expansion. Four sets of policies that are particularly as products. Securing preferential market access will important for creating more jobs and better jobs (in terms of help diversification. reduced labor turnover) are identified. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 165 TECHNICAL REPORT CHAPTER 6: POLICY FOR FUTURE JOBS CONSISTENT WITH CAMBODIA’S VISION 2050 Box 6.2: Policies to Move Up the Value Chain in Cambodia’s Tourism Sector Although the tourism sector in Cambodia generates good jobs and has significant spillover effects on and links to the domestic economy, its jobs potential is not being fully realized. The sector generates jobs and opportunities along its value chain, including in the primary sector, where most of the Cambodia’s poor are engaged (World Bank 2017). Tourism can be a way to develop jobs in regions where opportunities for industrial development are often limited. However, in spite of Cambodia being home to some of the more archeologically valuable sites in the world, its tourism sector is generally low-value and with weak or only informal links to the domestic economy. While it creates many jobs, they are of a lower quality and are less inclusive than the tourist sector has the potential to create. The World Bank has identified the following recommendations for supporting the continued development of Cambodia’s tourism sector (see World Bank 2017): • Improve infrastructure and connectivity within Cambodia. Growing urbanization and tourism mean that the country needs to have better infrastructure and transportation in place to accommodate the needs of locals and tourists, curb potential traffic issues, and build resilience of the sector. • Develop new tourism markets, products, and services. For example, policymakers could look into encouraging the development of specific segments of the tourism value chains, including protected area tourism, domestic tourism, and business and events tourism. It will also be important to improve management of these tourism destinations if these markets are to be tapped. • Help producers and providers of food, crafts, and other local goods and services to access tourism value chains. This might include supporting enterprises to make Cambodian-made products that meet market demand or establishing associations among farmers and food processors to supply domestically produced foods to hotels. This would help to increase the overall benefit of tourism to featured destinations. • Expand and upgrade the skills of local workers, while strengthening public-private partnerships to provide tourism-related training. This might include fostering public-private partnerships between Government and training providers to provide and increase access to quality training for youth and women from poor communities to acquire the relevant skills and knowledge for the tourism industry, thus maximizing their potential employability. Greater collaboration among local workers, farmers, small and medium-sized enterprises, and tourism-related associations will also help expand and upgrade skills. First, reduce the cost of doing business for domestic • Improving commercial dispute resolution by reviewing firms. The policies from the Cambodia Investment Climate the existing case management system at the courts Assessment Report that are particularly important for job and considering the development of an electronic case growth include: management system for judges and lawyers; and • Reducing business registration costs and simplify • Facilitating the implementation of the insolvency legal procedures by: (i) reviewing and reducing the fees for framework by establishing the insolvency administration starting a business and consider implementing a flat fee profession and improving the capacity of the public and that reflects the cost of service and (ii) introducing a private sector related to insolvency matters. single platform for business registration by integrating the incorporation process with registration at the MoC, Second, develop an ecosystem for domestic firms to General Department of Taxation and Ministry of Labor; increase business skills and knowledge. To deepen Phnom Penh’s nascent ecosystem and broaden it to the rest • Enhancing the quality of land administration by finalizing of the country, the government should prioritize: the issuing of land titles in the country and developing an integrated online registry system for land, including • Instituting formal processes to encourage public and ownership and cadaster information; private dialogue (PPD) for the private sector to raise their concerns and discuss possible solutions; and CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 166 TECHNICAL REPORT CHAPTER 6: POLICY FOR FUTURE JOBS CONSISTENT WITH CAMBODIA’S VISION 2050 • Providing more public support to entrepreneurship and increase with urbanization—they also employ many innovation structures, including incubators, tech hubs, vulnerable workers. Since only 6 percent of household and research institutes. Funding to these structures enterprises hire workers from outside the household, policies could be made conditional upon results. aimed at fostering the growth of household enterprises should focus on upgrading job quality (productivity or Third, increase access to finance through a grant incomes) rather than expanding the number of jobs that program and fiscal incentives. This includes: they provide. As older workers are more likely than younger workers to be employed in household enterprises, improving • Financing a Jobs Fund to support different segments job quality would increase the inclusivity of Cambodia’s of firms with high job potential through technical jobs, since older workers tend to have lower participation in assistance and matching grants for three windows: (a) good jobs than young workers. While little is known about the consolidation of fragmented activities/small-sized the nature of these jobs—which underscores the urgent firms so that they can compete in national (and possibly need for more research in this area—we have developed international) markets (such as agro-processing firms); some policy recommendations based on the analysis in (b) increased support to potential domestic suppliers of this report, recent data analysis of household enterprises in existing and nascent FDI-dominated value chains, such Vietnam (Pasquier-Doumer et al. 2017), and lessons from as apparel and electronics; and (c) domestic investment policy interventions in other countries designed to support projects with a high potential for job creation; and household enterprises. • In a more long-term approach, considering lowering First, more research must be completed to the size of the investment determining the eligibility understand the nature of household enterprises in of domestic firms for fiscal incentives, as these Cambodia and how to best support them. Only a few requirements are a potential barrier to entry for SMEs. countries have collected data on household enterprises, and these data have revealed that these enterprises have Fourth, increase the ability of domestic firms to hire very different motivations, aspirations, and challenges and retain workers by: than do firms that are driven by a profit motive and that • Developing and strengthening intermediation are not an extension of the household. Since the number mechanisms (job matching platforms and the National of household enterprises is likely to grow as Cambodia Employment Agency, see below); and urbanizes and becomes wealthier, it is crucial to understand this sector better. • Subsidizing some of the basic benefits (such as health care coverage and a retirement pension) that are usually Second, it will be important to “professionalize” provided to workers by their employers. This could be household enterprise owners so that they have the accompanied by an information campaign educating tools to run their enterprises as a business rather domestic firms about the importance of providing these than as an offshoot of their households. benefits to workers and increasing worker awareness of In the short run, policymakers can: their legal rights. • Provide training in basic accounting and marketing The policies that we identified to streamline business practices to household enterprises to help them make procedures and reduce the costs of establishing and sound financial decisions and to expand their production. expanding SMEs correspond to Rectangle III (Private While experience in other countries has shown that Sector and Job Development) Side 2 (Promoting such training programs improve business practices, the SMEs and Entrepreneurship) within Phase IV of the Rectangular Strategy. The main implementing agencies evidence is weaker about their effect on profitability, would be the Ministry of Economy and Finance, the Ministry probably because of the short duration of such programs of Industry and Handicrafts, the Ministry of Commerce, and in other countries. the Ministry of Labor and Vocational Training. • Match younger firms with more established firms that can mentor them and help them to upgrade their skills. A 3. Support Household Enterprises to Enhance study of how mentoring affected Kenyan microenterprises Their Productivity showed that participating enterprises saw their profits increase (Brooks et al. 2016). Other mentoring programs Not only are household enterprises a significant have successfully been implemented in the United States source of job creation—a phenomenon that is likely to and India. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 167 TECHNICAL REPORT CHAPTER 6: POLICY FOR FUTURE JOBS CONSISTENT WITH CAMBODIA’S VISION 2050 In the medium term, the government might consider: 4. Support the Development of Links between FDI Firms and Domestic Input- • Creating a one-stop online shop to provide household enterprises with information and technical assistance, supplying Firms including basic training programs. This would reduce Strengthening links between exporting FDI firms the transaction costs involving business upgrading and and domestic SMEs can be a new way to foster SME create a virtual community of household enterprises. expansion, technology transfer, and the creation of Third, policymakers should foster the use of IT indirect export jobs. Cambodian exporters source very solutions to expand household enterprises’ access to few of their inputs locally. Experience in other countries, markets. Many household enterprises rely on their personal such as Thailand, Malaysia, and China, has shown that these network to supply inputs. Data from Vietnam showed that backward links can be a rich source of new good jobs. more than 80 percent of household enterprise inputs were However, these connections often do not emerge purchased from other household enterprises (Pasquier- naturally. Most of Cambodia’s domestic producers do Doumer et al. 2017). This can be a significant barrier not meet the quality standards and/or productivity levels to efficient operations, particularly in rural and remote to qualify to supply FDI exporters. In surveys, FDI firms areas where there are fewer and more limited informal have cited a lack of industry-ready suppliers and of both networks available. To use the hypothetical example of a technical and soft skills as the main barriers to buying retail shop, large wholesalers in cities find it very costly to from domestic input producers in Cambodia (World Bank collect information about demand from retail shops in rural 2018). Therefore, foreign investors do not actively seek out villages and similarly the owners of these village shops find domestic suppliers. Also, the absorptive capacity of these it costly to travel to the cities to purchase limited amount local suppliers is weak, so they do not benefit from GVC of goods that they need for their small shops. So, they spillovers in knowledge and technology. each buy from and sell to those closest to them, usually at higher prices and for goods of lower quality than if they The right policies can help build links between had full information across all locations. IT can overcome exporting firms and domestic input-supplying firms. this problem by aggregating demand from rural and remote These policies include: areas and make it worthwhile for large suppliers, or other intermediaries, to engage productively with small-scale • Developing a database or directory of potential suppliers household enterprises. that have adequate technology and capacity to partner with multinational FDI firms to overcome information To assist household enterprises in accessing markets, failures; policymakers can: • Establishing local supplier development programs to • Support the development and use of IT solutions to help domestic enterprises to become qualified suppliers improve the business and management practices of or subcontractors for FDI firms; low-skilled, small-scale producers. Some household enterprises in Cambodia are already using technology • Simplifying and improving the administration of VAT to reach new customers, such as PassApp or Facebook. reimbursement to ensure that exporting firms can Targeted apps to reach customers and new apps receive refunds in a timely manner; for business management are already being used in other countries. However, the evidence of their • Requiring the Cambodia Investment Board to find impact on business and management practices is out from existing investors: (i) what constraints and still weak. challenges they face in sourcing their inputs locally; (ii) what their future investment and expansion plans are; • Expand ICT infrastructure to ensure nationwide access. and (iii) whether and how they would benefit from the relocation of international suppliers to Cambodia (World The policies that we identified to help household Bank 2018); and enterprises to enhance their productivity correspond to Sides 1 and 3 of Rectangle II (Economic • Strengthening the domestic absorptive capacity of Diversification) within Phase IV of the Rectangular the local private sector and workforce and supporting Strategy. The main implementing agencies would be the and promoting entrepreneurship and SMEs as well as Ministry of Posts and Telecommunications and the Ministry workforce development (these policies are discussed of Commerce. elsewhere but are relevant here). CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 168 TECHNICAL REPORT CHAPTER 6: POLICY FOR FUTURE JOBS CONSISTENT WITH CAMBODIA’S VISION 2050 The policies that we identified for strengthening • Expand innovative models for increasing secondary links between FDI firms and domestic input- school completion rates, such as the bridge program, supplying firms correspond to Sides 2 and 4 of access to computer-aided learning, and incentives to Rectangle III (the Promotion of Private Sector encourage households to invest in schooling. Development and Employment) as well as the Overarching Environment of Phase IV of the • Provide incentives for women to continue in secondary Rectangular Strategy. The main implementing and post-secondary education, particularly in the STEM agencies would be the Ministry of Economy and fields. Data show that women attaining higher levels of Finance, the Council for the Development of Cambodia, education and being employed in speci c sectors and the Ministry of Commerce, and the Ministry of Industry occupations are key to reducing the gender wage gap and Handicrafts. and expanding employment options for women. • Introduce results-based financing (RBF) for TVET 5. Build a Skills Development System that institutions with good results being defined as the Will Attract Higher-value FDI and Increase successful integration of students in the job market, Productivity across the Economy as well as expanding the provision of short courses (those that are short and flexible and that teach Cambodia’s lack of a strong skills development practical skills in high demand) to serve the working system presents policymakers with a chance to build adult population. a system geared towards 21st century jobs that will increase the skills level of the workforce, the returns • Incentivize the enterprise sector to play a larger to education, and productivity levels throughout the and more structured role in providing, guiding, and economy. There are two elements to building such a system. advocating for a demand-driven skills development First, the education system needs to provide students with a system. This would involve: (i) co-funding skills strong foundation in a range of knowledge, behavioral, and development programs in partnership with the digital literacy skills. Second, the technical and vocational enterprise sector with the financing being contingent training (TVET) system should be agile, flexible, and market- on graduation, job placement, and job promotion responsive so that workers can regularly upgrade their skills outcomes; (ii) subsidizing infrastructure and technology throughout their working lives to keep pace with evolving for training centers within special economic zones technology and the knowledge economy. Building this skills (SEZs) to take advantage of economies of scale (similar development system is such a massive undertaking—as to the Penang Institute in Malaysia); and (iii) developing highlighted in the government’s Education Strategic Plan incentive mechanisms to systematize the dialogue and 2014-2018 and the TVET Strategy: 2016-2021—that in this cooperation between employers and skills program report we only discuss the policies for making the education providers. Additionally, creating a standardized testing and training system more responsive to the labor market. and certification framework for skills, combined with incentives based on results, can encourage private We have identified seven policy priorities: providers of skills training to play an active role. • Incorporate digital literacy, socio-behavioral skills • Collect and disseminate user-friendly information to (creativity, innovation, interpersonal, teamwork, students, jobseekers, education and training institutes, and leadership), and higher-order cognitive skills and employers to enable them to make skills development (mathematics, logic, critical thinking, complex problem choices that are aligned with market demand. This solving, reasoning) into the school curriculum with will require collecting, analyzing, and disseminating cumulative learning throughout a student’s primary information on emerging jobs and associated wages, and secondary school career. Experiences in a range the skills that are in greatest demand, and the quality of countries have shown the viability of teaching these of skills development programs as measured by job additional skills in the classroom without imposing placement rates. undue burdens on instructors.1 1 See Cunningham et al. (2016) and Guerra et al. (2014) for a review of methods for teaching socio-behavioral skills in pre-school, primary, secondary, and post- secondary schools. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 169 TECHNICAL REPORT CHAPTER 6: POLICY FOR FUTURE JOBS CONSISTENT WITH CAMBODIA’S VISION 2050 Additional analytical work is urgently needed, as rights of migrants, the Cambodian government well. First, carrying out a systems assessment will give should continue its negotiation of bilateral migration the Ministry of Labor and Vocational Training the tools to agreements. However, given that most migrants do not prioritize the long list of policy reforms, identify international use these formal channels, the government could also: best practices that could be considered in the design of the proposed skills development system, and set a baseline • Link Cambodia’s labor market information system to the and an M&E system to monitor the effectiveness of reforms. labor market information systems in countries offering In addition, future analytical work will need to be done to better jobs; assess employer-driven training programs in Cambodia and elsewhere in the world. • Support pre-departure training services for workers searching for work abroad, online resources for potential The policies that we identified for implementing a or actual migrants, and incentives and processes to skills development system that will attract higher- facilitate the repatriation and re-integration of these value FDI and increase productivity across the workers into Cambodia’s labor market; and economy correspond to Sides 1 and 2 of Rectangle I (Human Development) as well as Side 1 of Rectangle • Include sufficient questions on internal and international III (the Promotion of Private Sector Development and migration in the next Census to inform the development Employment) within Phase IV of the Rectangular of policies to support workers in going abroad to boost Strategy. The main implementing agencies would be the their skills and earnings and add questions to the Ministry of Labor and Vocational Training and the Ministry of Cambodia Socio-economic Survey to track jobs-related Education, Youth, and Sports. migration patterns. The proposed policies are aligned with side 1 of 6. Promote Ef cient Labor Mobility and Job Rectangle III (Private Sector and Job Development). Matching The key actions focus on “providing work orientation training and expanding job recruitment and advertising services; as To enhance the allocative efficiency of labor, workers well as streamlining labor market information both inside need information and support to find and access the and outside the country in order to increase employment best possible jobs for their skill sets. In Cambodia, opportunity in an equitable manner and reduce risky work workers usually find jobs through personal contacts, whether migration.” It also aims to strengthen the governance of the job is in their locality or in another country. While this is migrants’ work, with the aim of improving the quality jobs a common practice, if workers have access to information and migration experiences, as well as to support returning about job vacancies, occupations, and skills needs, then migrant workers to invest in their home communities. they can find more appropriate jobs—especially if they are also provided with support in applying for those jobs. 7. Regain Macroeconomic Independence and Policymakers can enhance job matching in Exchange Rate Flexibility Cambodia by: Cambodia should take steps to re-gain exchange • Expanding and deepening the operations of the National rate flexibility and monetary policy independence Employment Agency (NEA) to collect more and more to shield jobs from fluctuations in the US dollar and timely data on job vacancies, develop outreach programs other macroeconomic risks. Cambodian policymakers to populations that tend to use informal job search could consider developing a market-based de-dollarization mechanisms, and expand the information used by and strategy to progressively regain policy autonomy. The the role of job counselors to guide jobseekers in the job development of a domestic bond market would also help search process; and both to foster domestic savings and to progressively increase exchange rate exibility to support job creation • Investing in hardware and software for the labor market going forward. Policymakers also need to encourage private information system managed by the NEA. investment and credits to invest in the tradable sector instead of construction and real estate, sectors that are To support more meaningful job searches and safer prone to instability and have fewer good job opportunities. international migration and to safeguard the human than the exporting sector. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 170 TECHNICAL REPORT CHAPTER 6: POLICY FOR FUTURE JOBS CONSISTENT WITH CAMBODIA’S VISION 2050 This points to a range of macroeconomic and Perhaps the biggest challenge for the implementation fiscal policies to support job-creating industries. of these policy areas is in coordinating the reforms Policymakers can: across the government, yet these proposed policies must be coordinated to ensure that they collectively • Institute a campaign to publicly recognize private sector result in the creation of more, better, and more institutions that pay salaries in Khmer riel. This short- inclusive jobs. The fact that the policies presented in this term action would support the use of local currency. report cut across all aspects of Phase IV of the Rectangular Strategy demonstrates the complexity of the jobs agenda • Adopt and gradually implement a larger market-based while highlighting the risk that some crucial jobs policies de-dollarization strategy by: (i) gradually increasing might not be implemented or be implemented in a reserve requirements for deposits in dollars and (ii) coordinated manner. bringing down the riel benchmark interest rates to promote borrowing in the national currency. To ensure that these agency-speci c actions are sufficiently coordinated, the government can • Develop a domestic debt market (both public and private) undertake two processes. First, policymakers could and starting to issue sovereign debt to mobilize domestic mobilize stakeholder support for a Jobs Strategy. This and foreign savings. This will help to increase investment strategy would: (i) identify the strategic jobs goals that are and capital stock accumulation to sustain high economic yet not defined in Phase IV of the “Rectangular Strategy; (ii) growth. It would also help the corporate sector to access specify those sectoral policies from the Rectangular Strategy sources of long-term funding and establish alternatives that are also crucial for creating more, better, and more to investing in construction and real estate to mitigate inclusive jobs; and (iii) set up a system for monitoring of some of the current macro-financial risks. jobs-specific policy actions across the many different • Strengthen the public Investment management framework ministries, agencies, and other stakeholders involved in to foster capital formation and introduce investment tax implementing the strategy. Second, they could appoint credits for the acquisition of machinery and equipment. a Jobs Champion with the leadership skills to oversee the reform across the entire government. As noted in Phase IV of the Rectangular Strategy, inter-institutional 6.3 Policy Implementation or ministerial coordination is an important key to the successful implementation of the Rectangular Strategy and The policies that we identified for macroeconomic will be similarly crucial to the successful implementation and exchange rate management correspond to of the Jobs Strategy. This can be achieved by creating a the Overarching Environment of Phase IV of the coordinating body that will both guide and hold accountable Rectangular Strategy. The main implementing agencies all government and private sector actors involved in the would be the National Bank of Cambodia and the Ministry of drive to create more, better, and more inclusive jobs for Economy and Finance. Cambodia’s future growth and prosperity. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 171 TECHNICAL REPORT CHAPTER 6: POLICY FOR FUTURE JOBS CONSISTENT WITH CAMBODIA’S VISION 2050 Box 6.3: Policy Implications for Inclusive Jobs The report proposes analysis and policy recommendations for better and inclusive jobs. Indeed, “inclusiveness” is streamlined through the policy recommendations. However, it is useful to explicitly identify how these jobs may improve the inclusiveness of future jobs. First, diversifying the export and FDI sector will benefit women and young workers. Women are the employee of choice in FDI and exporting firms today (Policy Area #1). Countries with more diversified exports and FDI-sectors also see a high share of women working in industries related to electronics, food processing, and service exports. However, for Policy Area #1 to truly work for women, the government can make particular efforts to attract the kinds of exporting industries that tend to employ women. Further, these workers tend to be young, so this policy will provide new jobs for young workers. Support to improving the productivity of household enterprises (Policy Area #3) will disproportionately benefit the poor, women, and older workers. Productivity enhancing mechanisms that can be easily accessed from home, such through technology will be particularly beneficial to women, who tend to need to combine home and work duties, and to people with disabilities who may be mobility constrained. Further, efforts to link domestic (including household) enterprises with FDI (Policy Area #4) can benefit these small producers. While a skills development system will benefit everyone (Policy Area #5), women, older workers, and the poor will particularly gain. Women still have lower levels of education than their male counterparts so there is a basic need to level the playing field across genders. The expansion and enhanced quality (measured as relevance) of TVET services will be crucial since the majority of today’s workforce does not have the skills needed today, let alone in the future. At the same time, younger workers will need to develop a broad set of skills to lead Cambodia into Industrial Revolution 4.0. Young people, the unemployed, and those trying to climb the career ladder will benefit from an expanded labor market information system and an even stronger National Employment Service (Policy Area #6). These services will permit them to better prepare for, and find, the best jobs for their skills and interests. On a related note, improved information and processes for migration may provide better work experiences for young men and women (who tend to migrate) and may draw them home again. Finally, Policy Area #7 on macroeconomic independence and exchange rate flexibility will be particularly beneficial to young women, who are over-represented in export-oriented jobs. Additional actions can be undertaken to support women, youth, older, low-skilled, and rural workers. These additional policies should co-exist with the seven policy areas defined in this report to meet the goal of more inclusive jobs. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 172 TECHNICAL REPORT CHAPTER 6: POLICY FOR FUTURE JOBS CONSISTENT WITH CAMBODIA’S VISION 2050 6.4 References Bloom, Nicholas, Benn Eifert, Aprajit Mahajan, David McKenzie, and John Roberts. 2013. “Does management matter? Evidence from India.” The Quarterly Journal of Economics 128(1): 1-51. Brooks, Wyatt, Kevin, Michael Donovan and Terence Johnson. 2016. The Dynamics of Inter-Firm Skill Transmission among Kenyan Microenterprises. Washington, D.C: World Bank. Bruhn, Miriam and Zia Bilal. 2013. “Stimulating Managerial Capital in Emerging Markets: The Impact of Business Training for Young Entrepreneurs” Journal of Development Effectiveness 5(2): 232-266. Cunningham, Wendy, Pablo Acosta and Noël Muller. 2016. Minds and Behaviors at Work: Boosting Socioemotional Skills for Latin America’s Workforce. Washington, D.C.: World Bank. Guerra, Nancy, Kathryn Modecki, Wendy Cunningham (2014). “Developing social-emotional skills for the labor market : the PRACTICE model.” Policy Research Working Paper No. 7123. Washington, D.C.: World Bank. Pasquier-Doumer, Laure, Xavier Oudin, and Nguyen Thang. 2017. The Importance of Household Businesses and the Informal Sector for Inclusive Growth in Vietnam. Vietnam Academy of Social Sciences and the French National Research Institute for Sustainable Development. Taglioni, Daria, and Deborah Winkler. 2016. Making Global Value Chains Work for Development. Washington, D.C.: World Bank. World Bank. 2017. Cambodia Economic Update. Washington, D.C.: World Bank. World Bank. 2018. Cambodia Investment Climate Assessment. Washington, D.C.: World Bank. CAMBODIA’S FUTURE JOBS: LINKING TO THE ECONOMY OF TOMORROW 173 TECHNICAL REPORT The World Bank Cambodia Country Of ce Exchange Square Building Floor 10th IBRD and 11th IFC Streets 51-61 and Streets 102 -106 Sangkat Wat Phnom, Khan Daun Penh Phnom Penh, Cambodia Website: www.worldbank.org/cambodia