Türkiye in Transition Generation Human Capital Next-­ Investments for Inclusive Jobs POLICY NOTE © 2022 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522- 2625; e-mail: pubrights@worldbank.org. Photo credits: p. viii © syekteese / Shutterstock; p. xi © klenger/ Shutterstock; p. xiii © Tatevosian Yana / Shutterstock; p. xiv © el_cigarrito / Shutterstock; p. 1 © Ancapital / Shutterstock; p. 6 © Nejdet Duzen / Shutterstock; p. 7 © SvetlanaKokorina / Shutterstock; p. 19 © seyephoto / Shutterstock; p. 48 © asliuzunoglu / Shutterstock; p. 49 © Yasemin Yurtman Candemir / Shutterstock; p. 69 © Yavuz Sariyildiz / Shutterstock; p. 81 © franz12 / Shutterstock; p. 92 © Merih Salmaz / Shutterstock; p. 93 © thomas koch / Shutterstock. Further permission required for reuse. Cover photos: © temizyurek / iStockphoto. Further permission required for reuse. Design: Will Kemp, GCS, World Bank Group Contents Acknowledgements ix Overview xi 1 Motivation: Reshaping Human Capital for New Jobs 1 Introduction 1 Conceptual Framework 2 Methodology and Data 3 2 Taking Stock: Overall Human Capital Expenditure and Employment Trends 7 Introduction 7 Total Expenditure and Main Components 7 International Benchmarking and the Effects of Shocks 9 Aggregate Human Capital Outcomes 10 Pandemic Preparedness and Response 12 Institutional Governance and Political Economy 16 Key Implications 17 3 Targeting Job Creation: The Shift to Green 19 Introduction 19 Aggregate Job Growth and Productivity 20 Labor Force Dynamics 29 Gender, Informality and Opportunity 33 Determinants of Employment during Shocks 40 Examining the Potential for Green Jobs 44 Key Implications 47 4 Modernizing Skills and Job Training: Redirecting Investments 49 Introduction 49 Youth Inclusion 49 Education Investments over Time 52 Infrastructure, Coverage and Enrolment 57 Learning and Distributional Outcomes 59 to-­ School-­ Work Transition and Returns to Education 61 Skills and Green Growth 64 Key Implications 67 5 Facilitating Labor Market Entry and Mobility: Harmonizing for Just Transitions 69 Introduction 69 Wage Subsidies 69 The-­ Adult On-­ Job and Skills Training 70 Unemployment Benefits 73 Job Search 74 Key Implications 80 6 Mitigating Risk and Improving Resilience: Integrating Social Insurance 81 Introduction 81 Labor Productivity Risks 82 Overall Social Insurance Coverage 83 Pensions 84 Unemployment Insurance 85 Health Insurance 87 Social Assistance 89 Simulating Fiscal Space for Integrating Labor and Social Protection Schemes 90 Key Implications 91 Government Approach for 7 Outlook: Developing a Whole-­of-­ Transforming Human Capital 93 Introduction 93 Addressing Challenges, Maximizing Opportunities 94 of-­ Towards a Whole-­ Government Human Capital Platform 95 Conclusion 98 Tables Table 2.1. Türkiye’s vaccination progress as of November 5, 2021, shown to reflect status during peak phase.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Table 3.1. Key labor force indicator trends, 2014 versus 2021 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Table 3.2. Key employment indicators by sector and gender, 2014 versus 2021. . . . . . . . . . . . . . 22 Table 3.3. Informality by sector, 2017–2020, among females (first panel) and males (second) . . . . 37 Table 3.4. Reasons for job quits by gender, 2020. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Table 6.1. Detailed SPL expenditures by program, 2015–2019 . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Table 6.2. Health service delivery indicators, 2018–2020 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Table 7.1. Proposed policy reform matrix for boosting human capital and jobs towards inclusive, green transformation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Figures Figure 1.1. Conceptual framework for assessing human capital investments for boosting jobs. . . . . 3 Figure 2.1. Social Expenditures by type as a percent of total public expenditures (first panel) and of GDP (second), Türkiye. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Figure 2.2. Social expenditures by type as a percent of GDP, OECD (social protection and labor, first; education, second) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Figure 2.3. Social expenditure trends, OECD average growth overall, 1990–2018 (first panel) and by instrument (second). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Figure 2.4. Global Human Capital Index correlations as a function of social expenditures (first panel), labor force participation rate among adults (second) and youth (third). . . . . . 11 Figure 2.5. Public health systems versus COVID management, global comparisons (policy stringent index, first panel; universal health coverage, second). . . . . . . . . . . . . . 15 Figure 2.6. Human Capital policy aims for short-­to long-­ COVID recovery in Türkiye. . . . 18 term post-­ Figure 3.1. Jobs by major sector, 2014–2020. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Figure 3.2. Sectoral employment trends as share of GDP (first panel) and annual change (second), 2019 versus 2020 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Figure 3.3. OECD: GDP versus share of wages by country, 2019 (first panel) versus 2020 (second). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Figure 3.4. OECD: Sectoral share of Gross Value Added (GVA) versus labor share by country, 2020, for industry (first panel), services (second) and agriculture (third) . . . . . . . . . . . . 25 Figure 3.5. Türkiye: Sector GDP growth versus earnings growth over time, 2007–2019, for agriculture (first panel), industry (second), construction (third) and services (fourth) . . . 26 Figure 3.6. Net job growth by sector, 2015–2020, overall (first panel) and among females (second). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Figure 3.7. Average earnings by sector, 2006–2020, in terms of absolute levels (first panel) and annual change over time (second). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Figure 3.8. Labor force participation rate, total versus youth, relative to unemployment rate (first panel) and total labor force (second panel), 2014–2021. . . . . . . . . . . . . . . . . . . . . . 30 Figure 3.9. Labor force participation rate by gender and education, among all ages 15+ (first panel) versus youth (second), 2014–2021. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Figure 3.10. Employment rate by region and education for ages 15+ among females (first panel) and males (second), 2019 versus 2020 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Figure 3.11. Overall employment levels versus informal share by sector, 2014–2021. . . . . . . . . . . . . 33 Figure 3.12. Female Labor force participation rate (LFPR) versus Human Capital Index, Türkiye (left), and Global LFPR comparisons (right). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Figure 3.13. Overall employment levels versus female share, 2014–2020, aggregate (first panel) and by sector (second) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Figure 3.14. Average wages by type and gender, 2020, in absolute (top panel) and relative terms (second). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Figure 3.15. Informality versus formality by gender, age, education and sector, 2020 . . . . . . . . . . . . 37 Figure 3.16. Informality levels by educational level and gender, 2019. . . . . . . . . . . . . . . . . . . . . . . . . 39 Figure 3.17. Informal share of total employment versus GDP per capita, global, 2019 . . . . . . . . . . . . 39 Figure 3.18. Employment vulnerability index during COVID by sector versus employment share (first panel) and work-­ home amenability (second), Türkiye . . . . . . . . . . . . . . . . . . . 41 from-­ Figure 3.19. Probability of inactivity (left) and unemployment (right) by gender and education, 2020. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Figure 3.20. Probability of inactivity by gender as a function of past occupation, 2020 . . . . . . . . . . . 43 Figure 3.21. Probability of inactivity as a function of past sector among females (first panel) and males (second panel), 2014–19 versus 2020. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Figure 3.22. Demand for labor by total annual sectoral GHG emissions, 2019–2020. . . . . . . . . . . . . 45 Figure 3.23. Distribution of jobs by gross GHG emission category, 2010–2020. . . . . . . . . . . . . . . . . 46 Figure 3.24. Net change in job creation by gross GHG emission category, 2010–2020 relative to base year 2010. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Figure 4.1. Youth not in education, employment or training (NEET), ages 15–24, by gender (first panel) and education (second), 2014–2020. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Figure 4.2. Probability of NEET by education (first panel) and among post-­ secondary students by educational background (second), 2020. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Figure 4.3. Türkiye: Total public expenditure on education as share of GDP, 2011–2019 . . . . . . . . 52 Figure 4.4. Figure 35 Public expenditure by educational level, OECD countries, 2013 (first panel) versus 2017 (second) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Figure 4.5. Relative change in education spending by level, selected countries, 2013 versus 2017. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Figure 4.6. Secondary education expenditure versus PISA learning outcomes for Reading, Science and Math. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Figure 4.7. Public and private education expenditures, 2011–2019, and by household quintile 2002–2019, Türkiye . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Figure 4.8. Evolution of household expenditures by type as share (first panel) and wealth gap (second), 2002 versus 2019 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Figure 4.9. Schooling rates and Teacher coverage by educational level and regional zones, 2019, Türkiye. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Figure 4.10. General adult digital skills ranking, OECD+ countries, 2018. . . . . . . . . . . . . . . . . . . . . . 59 Figure 4.11. Secondary school learning outcomes (PISA) by theme, regional zones in Türkiye (2006–2018), and by household income, selected countries (2009–2018). . . . . . . . . . 60 Figure 4.12. Stylistic directional (positive versus negative) associations between economic investments and PISA Science, Math and Reading scores, illustrative regression, Türkiye, 2018–2019. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Figure 4.13. School career counseling coverage by type and school socioeconomic level, selected countries, 2018. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Figure 4.14. Rate of return to education by educational level and scope . . . . . . . . . . . . . . . . . . . . . . 63 Figure 4.15. Green skills labor intensity versus GDP per capita, global, 2019. . . . . . . . . . . . . . . . . . 65 Figure 4.16. Green jobs as share of employment versus GDP per capita, global, 2019. . . . . . . . . . . 66 Figure 4.17. Educational attainment versus green job typology (first panel) and green skills index (second). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Figure 5.1. Labor programs coverage by type over time and by gender and educational level, 2007–2020. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Figure 5.2. Jobseeker demand for On-­ Job Training (OJT, ISKUR), 2010 versus 2020. . . . . . . . . 73 the-­ Figure 5.3. Total number of job vacancies, overall (first panel) and for manufacturing (second) and construction (third), seasonally adjusted values, 2015–2021. . . . . . . . . . . . . . . . . . . 75 Figure 5.4. Total number of job applications and job applications per position, overall, seasonally adjusted values, 2015–2021. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Figure 5.5. Total number of job applicants by gender (first panel) and age (youth younger than 24 years and adult 25+, second panel), 2015–2021. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Figure 5.6. Public employment service job placement rate (ISKUR) versus share of unfilled vacancies per year (first panel) and distribution of job placements by age and gender, 2020 (second) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Figure 5.7. Efficiency of public service job placement, 2010 versus 2020: (left to right) total vacancies by sector, unfilled vacancy rate and change in female placement rate. . . . . . 78 Figure 5.8. Mapping of sectoral job vacancies, unfilled rate and female placement, 2020 (three panels) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Figure 6.1. Labor productivity components by income quintile (first panel) and country comparisons (second), 2020 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Figure 6.2. Overall coverage rates of social insurance and ALMPs by income quintile, 2010–2020. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 Figure 6.3. Social insurance subsidies for universal coverage by income quintile, illustrative analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Figure 6.4. Health care personnel, global, 2015–2019. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 communicable disease by health care expenditure (first panel) and Figure 6.5. Burden of non-­ cause of death (second), global . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 Figure 6.6. Fiscal scenarios for integrated, universal social risk protection system, illustrative. . . . . 91 Boxes Box 1.1. Data sources for Human Capital Expenditures analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Box 2.1. Mitigating Shocks to Human Capital in Türkiye: Public Policy Response During COVID. 17 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Acknowledgements This policy note was prepared as part of the Türkiye Human Capital Policy Dialogue and the multisectoral Public Financial Review (2019–2021) with additional conceptual and methodological detail. The work was led by Heba Elgazzar (Program Leader ,Human Development) and benefited from helpful discussions with and inputs by Joel Reyes (Senior Institutional Development Specialist), Laurent Bossavie (Senior Economist), Daniel Garotte (Economist), Mattia Makovec (Senior Economist), Efsan Ozen (Economist), Sirma Seker (Senior Economist), Mirey Ovadiya (Senior Social Protection Specialist), Mauro Testaverde (Senior Economist), Siddharth Hari (Economist), Aysenur Acar (Economist, Consultant), Derya Barlak (Program Assistant), Secil Paker (Program Assistant), Safir Sumer (Human Development Specialist), Nadwa Rafeh (Senior Health Specialist), Aaron Buchsbaum (Senior Knowledge Management Officer), Husein Abdul-Hamid (Senior Education Specialist), Tunya Celasin (Senior External Affairs Officer), Indira Chand (Senior External Affairs Officer), Mustafa Alver (Senior Country Officer), Eavan O’Halloran (Country Program Coordinator), Hans Beck (Program Leader, Equitable Growth, Finance and Institutions), Laurent Debroux (Program Leader, Sustainable Development) and Stephan Garnier (Program Leader, Infrastructure). The team sincerely appreciates the excellent design and editing led by Will Kemp (Graphic Designer and Visual Storyteller) and the World Bank’s Global Corporate Solutions publications team. The work accompanies the World Bank’s Türkiye Human Capital skills and jobs knowledge-­ sharing series (2019–2021), including Women, Business and the Law in Turkey: Unlocking Jobs (March 2021); Labor Returns to Education and the Green Transition (June 2021); Informality, Productivity and Risk-­Sharing for a Green Economy (September 2021); Supporting Firms in Creating Formal Jobs: Emerging Opportunities for Turkey (September 2021); and the associated components of the analysis released during preparation, including Much ado about nothing? The next generation of human capital investments, skill; and labor in Turkey and beyond (October 2020)1; and Back to the future: Harnessing women’s capital for new growth in Turkey (March 2021)2. The paper focuses on a long-­ term, retrospective review of human capital expenditure and key outcomes related to inclusive labor markets, primarily covering 2011–2021 based on data available at the time of writing (2019–2021), followed by preparation for release during 2021–2022. The work assesses historical trends until the outbreak of the novel coronavirus-­ time 2019 (COVID). As such, it is not intended to necessarily capture real-­ aspects by the time of publication, which future work may evaluate. As of 2022, “Türkiye” has become the official country name and all efforts have been made to replace “Turkey” in this note where possible, with any possible errors or omissions remaining inadvertent. The findings reflect analysis conducted by the authors based on data and information available 1 Available at https://blogs.worldbank.org/europeandcentralasia/the-­next-­generation-­of-­human-­capital-­ investments-­skills-­and-­labor-­in-­Turkey-­and-­beyond 2 Available at https://blogs.worldbank.org/europeandcentralasia/back-­future-­harnessing-­womens-­capital-­new-­ growth-­turkey Acknowledgements ix at the time of writing, adopting standard definitions and approaches developed by the World Bank described in the remainder of this work. The team thanks the Government of Türkiye for helpful discussions and the availability of survey and administrative data published routinely by national agencies, notably the Turkish Statistical Institute (TURKSTAT/TUIK), Ministry of Treasury and Finance, Ministry of National Education, Ministry of Labor and Social Security, Turkish Employment Agency (ISKUR), Turkish Social Security Institute (SGK), Ministry of Family and Social Services, and other relevant agencies. The team also appreciates non-­ governmental national and international stakeholders for invaluable discussions and inputs as part of the World Bank’s human development program in Türkiye. The work benefited from helpful guidance by Humberto Lopez (Country Director, Türkiye), Auguste Kouame (former Country Director, Türkiye), Fadia Saadah (Regional Director, Europe and Central Asia), Michal Rutkowski (Senior Global Director, Social Protection and Jobs), Cem Mete (Practice Manager, Social Protection and Jobs), Harry Patrinos (Practice Manager, Education), Ian Walker (Manager, Jobs Group), Hana Brixi (Global Director, Gender Group), Iffath Sharif (Manager, Human Capital Project), and global initiatives by the World Bank’s Social Protection and Jobs Global Practice, Human Capital Project, and Gender Group. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. x Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Overview Türkiye’s long-­standing human capital achievements can propel it to the next generation following COVID. Similar to comparable countries emerging from COVID, Türkiye’s readiness to address three key challenges and opportunities to human capital investments will prove key to an inclusive, green recovery: equitable coverage, fiscal capacity, and, importantly, adaptability to a changing labor market. The aim of this policy note is to examine the relative efficiency and equity of public expenditures and policies for human capital and labor markets over the past decade in Türkiye in preparing for a post-­ COVID, inclusive, green transition. The work adopts a thematic conceptual framework that frames social expenditures holistically along the life cycle as investments in human capital for boosting equity and productivity. The adapted framework takes a holistic approach and focuses on selected inter-­ related dimensions linking the cycle of social investments, human capital and jobs. The first section of the note includes the motivation, followed by a second section analyzing aggregate social expenditures, human capital, and life-­ cycle skills investments, from basic education to adult training. While Türkiye spends nearly 16 percent of GDP on human capital, by contrast the OECD average accounts for nearly 25 percent as of 2020. As these sections show, compared to most OECD countries, Türkiye’s spending on education, health and active labor market programs is relatively modest. In the OECD, the drop in economic growth during 2007–2009 of around 6 percentage points (2.7 to –3.5 percent) was met by an increase in social expenditures of nearly 3 percentage points of GDP on average (17.7 percent of GDP to 20.7 percent of GDP). This level has been maintained since, owing to demographic changes, and long-­ lasting impacts on jobs and consumption in general. Similarly, due to COVID-­19, most advanced countries mobilized emergency social measures that accounted for 1–2 of GDP over 2020; Türkiye’s support packages targeting households and workers’ wages and benefits were relatively modest and estimated to account for up to 0.5 percent of GDP. Overview xi Despite Türkiye’s strong human capital foundations, the paper suggests that at an aggregate level, for its level of social expenditures, the allocative efficiency of Türkiye’s spending is lower than comparable countries. In terms of average Human Capital Index (HCI), at 0.65 as of 2018 (most recent figure), Türkiye’s outcomes are somewhat lower than comparable countries. HCI-­ utilizaton, or overall labor force participation rate (LFPR), is also lower than expected for its level of social spending, indicating ineffiicent spending and likely broader factors such as demand-­ side (private investment) and social dynamics. Similarly, youth LFPR, similar to youth unemployment, is lower in Türkiye than expected for its level of HCI. Examining higher-­ level aspects of human capital further reveal inefficiens regarding public spending in three main areas: (i) boosting competitive skills, (ii) faciliating labor market entry and (iii) matching to the demand side. The third section focuses on growth and job creation trends including productivity, green jobs and sectoral distribution, complemented by the fourth section on the distribution of educational attainment, skills, competency and school-­ work transition patterns. to-­ These sections indicate that labor force participation and employment in Türkiye since 2018 have been volatile, although some recovery has been since the start of COVID recovery. Over the pre-­ vaccination pandemic period, the labor force shrank by nearly 4 percent from November 2019-­ November 2020, the equivalent of nearly 1.1 million workers, of whom over 60 percent were women exiting. This represents a contraction of approximately 3.9 percent and a loss of over 1.1 million workers compared to November 2019, with 61 percent of the losses being borne by women. The bulk of losses at over 750 thousand jobs were in the service, largely informal sector, accounting for over 60 percent of losses, although a rebound has started as of early 2021 in some sectors. Informal jobs continue to prevail among approximately 30 percent of the Turkish population, particularly in agriculture. Employment among younger workers -­ male and female, which was hit the hardest during -­ the crisis, has been growing faster than among adult workers, but the analysis shows that younger, lower-­ educated workers remain vulnerable to climate-­ related job transformation without adequate re-­ skilling. skilling and up-­ The fifth section highlights coverage and beneficiary profile of key active labor market programs, including passive and active mechanisms, as well as selected labor regulations regarding unemployment benefits and wage subsidies. The six section further discusses risk mitigation polices for vulnerable poor households, informal workers and formal sector employees, including an analysis of scenarios for integration under contributory social an umbrella of universal social security. Importantly, integrating non-­ assistance with contributory social insurance schemes allows a more rational and equitable approach to offering subsidies based on capacity, facilitation and incentives to work. Building on Türkiye’s well-­designed social information management systems, greater harmonization across schemes likewise allows outreach to informal and poor households, testing and contribution/savings management. including identification, profiling, means-­ Finally, the policy note concludes with a discussion of selected policy implications within the framework of a whole-­ government approach, fostering close inter-­ of-­ ministerial harmonization for key human capital reforms over the short-­to long-­ term. Measures to facilitate the green economy in Türkiye will require segmenting the current and expected xii Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs future stock of workers and households by need. Policies would entail: (i) expanding green the-­ skills for basic and higher education, on-­ job training and re-­skilling; (ii) targeting job creation through climate finance incentives; (iii) addressing informality and labor costs through social insurance consolidation, including harmonization of benefit levels, registry expansion and targeting streamlining across non-­ contributory and contributory social transfers; and improving demand-­ driven active labor market programs in terms of outreach and results-­based financing towards green jobs. Overview xiii Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs 1 Motivation: Reshaping Human Capital for New Jobs Introduction Türkiye3 is emerging from two decades marked by different growth and human capital outcomes, at a crossroads for which path the upcoming decade will take. The first decade (2000–2010) bore relatively low inflation, rapid growth driven by deepening industry and global trade, and impressive gains in human capital in terms of preventative health, basic education and labor force participation. Yet the second decade (2011–2021) has seen heightened risks to all three domains simultaneously due to global shocks and domestic factors. Despite these challenges, Türkiye’s potential to reach new heights still remains high, owing to historically strong public policy foundations and institutions. Its path will depend on if these foundations can be activated to meet three broad evolving challenges: boosting equitable growth; fostering a climate-­ friendly “green” transition; and reinforcing resilience to future shocks. As Türkiye advances on reigniting growth following the novel Coronavirus (COVID) outbreak, the need to ensure human capital investments keep pace will be critical. To respond to these challenges, Türkiye’s current Eleventh National Development Plan (with the Twelfth expected by 2023–2024) and 2021 Economic Reform Program (ERP), coupled with its 2021 Climate Action Plan, set a vision for boosting economic recovery and expanding opportunity, complemented by emergency near-­ term measures adopted since March 2020 to stem the impact of COVID. Key areas in the NDP and ERP include 3 Note: as of 2022, “Türkiye” has become the official country name and all efforts have been made to replace “Turkey” in this note where possible, with any possible errors or omissions remaining inadvertent. Motivation: Reshaping Human Capital for New Jobs 1 strengthening labor market functioning, quality of education for better learning outcomes and adaptability to emerging jobs, and social insurance for all4. The Government of Türkiye’s term COVID support (covering a three-­ short-­ month period since April 2020) includes access to finance for vulnerable firms and social and labor measures to provide emergency income support to poor and low-­ income households and workers. While these measures are necessary, an analysis of mid-­to long-­term policy responses and fiscal space is needed. Türkiye’s readiness to address three key challenges to human capital investments as the pertain to inclusive growth and jobs will prove key: equitable coverage, fiscal capacity, and, importantly, adaptability to a changing labor market. Conceptual Framework The aim of this policy note is to examine the relative efficiency and equity of public expenditures and policies for human capital and labor markets over the past decade COVID, inclusive, green transition. The work adopts a in Türkiye in preparing for a post-­ thematic conceptual framework5 that frames social expenditures holistically along the life cycle as investments in human capital for boosting equity and productivity. For purposes of this thematic paper, the conceptual framework is based on the World Bank’s Human Capital and Public Finance6 framework and the World Development Report Jobs (2013) approach. The adapted framework takes a holistic approach and focuses on selected inter-­ related dimensions linking the cycle of social investments, human capital and jobs: • Targeting job creation: addressing specific demand-­and supply-­side constraints jointly oriented investments and access to finance incentives. through labor intensive-­ • Modernizing skills and job training: enhancing content, delivery and provider performance incentives within education and employment training policies, linking both, and greater coordination with evolving and future growth, notably digital, green and regional investments. • Accelerating labor market entry and mobility: narrowing school-­ to-­work transitions and more strategically aligning active labor market programs, labor regulations, labor costs, contracts, dismissal and minimum wage policies, and wage subsidies. • Adapting risk mitigation and resilience: dynamically adapting and integrating social insurance, non-­ contributory-­ contributory social assistance, and progressive subsidies for expanding coverage and equality of opportunity. 4 See Türkiye Eleventh National Development Plan (Sept 2019), Section 2.1.7, “Social Security System and Financing”, and other relevant sections under public finance. 5 Informed by WDR Jobs (2013), WDR Changing World of Work (2019), World Bank SPJ Risk Sharing in a specific lessons learned. Diversifying World of Work (2019), global and Türkiye-­ 6 World Bank (2021). Investing in Human Capital for a Resilient Recovery: The role of public finance. Washington DC: World Bank. 2 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 1.1.  Conceptual framework for assessing human capital investments for boosting jobs Targeting job creation Adapting risk Modernizing skills mitigation and and job training resilience Accelerating labor market entry and mobility Source: World Bank staff authors. Methodology and Data The methodology entails quantitative policy and fiscal analysis of the main public policies and programs at the national level in Türkiye, with benchmarking globally where applicable. The analysis focuses on the following areas: (a) programmatic expenditures over time at the central and regional levels, benchmarking against a selected set of comparable countries; (b) trends in growth and job creation by sector, with special attention to the evolution of green jobs; (c) education trends and occupational skills in demand by an evolving labor market, including sectoral, socioeconomic and geographic trends; (d) active labor market policies and selected labor regulations closely linked to facilitating entry and enhancing incentives for productivity; (e) risk-­ mitigation policies in terms of contributory and non-­ contributory social protection/insurance (including social, health and labor-­related); and (f) fiscal and policy scenarios pertaining to improving the performance and, importantly, adaptability of human capital policies for addressing ongoing and future challenges. Detailed, microanalysis of all sector-­ specific issues is beyond the scope of this analysis, which focuses specifically on selected cross-­ sectoral strengths and gaps regarding allocative efficiency that complement more detailed sector-­ specific analyses elsewhere7 7 World Bank (forthcoming). Education Sector Analysis; World Bank (forthcoming). Pandemic Preparedness and Response Assessment; World Bank (forthcoming). Occupational Skills and Labor Market Programs Assessments. Motivation: Reshaping Human Capital for New Jobs 3 Box 1.1.  Data sources for Human Capital Expenditures analysis The analysis of intersectoral human capital Ministry of Health; TURKSTAT; and compared expenditures in this paper is based primarily on with data from the OECD, EUROSTAT and available data published by the Government WDI data review where needed. The Turkish of Türkiye and does not include a detailed Statistical Institute TURKSTAT/TUIK provides the analytic evaluation of different methods or national, majority of the data used for the intra-­ sources. Total public expenditure reflects all main cross-­provincial analysis. public spending at a societal level including all contributory and non-­ contributory social security In addition, regarding the methodological benefits, wage subsidies and spending beyond approach and interpretation of results, given the central government, compiled across sources Türkiye’s strong advances in human capital over from the central government and individual the past four decades, the analysis is based institutions. Where cross-­country comparisons on cross-­country comparisons for the sake of are made, data from both international and identifying next-­generation key opportunities national sources are referenced for the sake to progress even further. Based on inputs and of highlighting general ranges and trends discussions with the authorities, the analysis most relevant to policy challenges at a macro acknowledges that cross-­ country comparisons perspective. Historically, Türkiye has had a may be wrought with complications interpretation relatively strong framework for compiling, in some cases, as policy context varies and analyzing and publishing data on public finance hence comparing relative efficiency and equity and service delivery indicators in the social of public spending may not be done in a vacuum. sectors. Primary data have been sourced from For this reason, these comparisons should be the Ministry of Treasury and Finance; Social taken as illustrative and areas for further work Security Institution (SGK); Ministry of National rather than as the basis of definitive or absolute Education; Ministry of Labor and Social Security, policy conclusions. ISKUR; Ministry of Family and Social Services; Source: World Banks staff, authors. (see Box 1.1 for a methodological discussion on data sources). The chapter adopts a macro perspective and does not cover specific education or health care service delivery organization or investment projects.8 8 The data used for the analysis is based mainly on international databases for the case of cross-­ country comparisons, the Ministry of Treasury and Finance and associated line ministries, and data provided on the TURKSTAT for regional and household-­ specific indicators. While there may be modest differences in data sourced from different agencies or by year given recent updates after the time of writing, the macro trends and associated policy implications discussed in the chapter largely remain unchanged. Further work and analytic work can address further policy questions based on discussions with the authorities and stakeholders as needed. 4 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs The main data used draws upon household survey and administrative beneficiary and program data at the national level. This includes statistical survey data; administrative data sourced from line ministries9; international survey and globally compiled national accounts data (based on World Development Indicators, OECD, ILO and other relevant sources); and qualitative appraisals by World Bank staff based on national policy and program documents, discussions with key stakeholders from the public sector, and insights from key non-­ state agencies, think tanks and international partners cited throughout the report where included. The note is structured as follows. The current, first section includes the motivation, followed by a second section analyzing aggregate social expenditures, human capital, cycle skills investments, from basic education to adult training. The third section and life-­ focuses on growth and job creation trends including productivity, green jobs and sectoral distribution, complemented by the fourth section on the distribution of educational attainment, skills, competency and school-­ work transition patterns. The fifth section to-­ highlights coverage and beneficiary profile of key active labor market programs, including passive and active mechanisms, as well as selected labor regulations regarding unemployment benefits and wage subsidies. The six section further discusses risk mitigation polices for vulnerable poor households, informal workers and formal sector employees, including an analysis of scenarios for integration under an umbrella of universal social security. Finally, the policy note concludes with a discussion of selected policy implications within the framework of a whole-­ government approach, fostering close inter-­ of-­ ministerial harmonization for key human capital reforms over the short-­to long-­ term. 9 For more detailed Turkey education expenditures used for the remainder of this work, see TURKSTAT and MoNE sources available at: TURKSTAT: https://data.tuik.gov.tr/Bulten/Index?p=Education-­Expenditure-­Statistics-­2019–33670&dil=2#:~:text=The%20 74.0%25%20of%20education%20expenditure%20in%20Turkey%20in%202019%20was,expenditure%20by%20 households%20was%2020.8%25.&text=While%20the%20education%20expenditure%20per,thousand%20769%20 TL%20in%202019 MoNE: https://sgb.meb.gov.tr/meb_iys_dosyalar/2021_09/10141326_meb_istatistikleri_orgun_egitim_2020_2021.pdf Motivation: Reshaping Human Capital for New Jobs 5 6 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs 2 Taking Stock: Overall Human Capital Expenditure and Employment Trends Introduction Overall public spending in human capital in Türkiye, encompassing the broad spectrum of social policies and programs to support households and workers, have remained constant over the past ten years, yet remain modest in the face of post-­ COVID challenges to welfare and jobs. The cost of managing the impact of COVID-­ related shocks on households and workers while addressing underlying vulnerabilities is expected to related shock also be significant, with a need to ensure efficiency and equity. The COVID-­ provides an important opportunity to reform economic and social welfare policies early and implement “early alert adaptive” systems. The choice of instruments is key to striking a balance between supporting immediate needs and building long-­ term resilience. Total Expenditure and Main Components Türkiye spends a considerable share of GDP on social investments at approximately 16 percent as of 2020, which has remained stable for over a decade. Total social expenditures represented the single largest share of public expenditures at approximately 40 percent as of 2020, up from 38 percent in 2007, indicating that the share has been remained relatively constant. As a share of GDP, social expenditures saw a spike of nearly Taking Stock: Overall Human Capital Expenditure and Employment Trends 7 Figure 2.1.  Social Expenditures by type as a percent of total public expenditures (first panel) and of GDP (second), Türkiye Social expenditure, % of total public expenditure Social expenditure as percent of GDP 45% 20% 40% 18% 16% 35% 14% 30% 12% 25% 10% 20% 8% 15% 6% 10% 4% 5% 2% 0% 0% 07 08 09 10 11 12 13 14 15 16 17 18 19 08 09 10 11 12 13 14 15 16 17 18 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Education Social assistance (cash/in-kind) Wage subsidies 20 Almp-Excl wage subsidies Unemployment benefits Survivor benefits Health general govt exp-excl health insurance Health insurance Pensions Source: World Bank staff calculations; Data from Ministry of Treasury and Finance; Social Security Institution (SGK); Ministry of National Education; Ministry of Labor and Social Security, ISKUR; Ministry of Family and Social Services; Ministry of Health; TURKSTAT; and OECD, EUROSTAT and WDI data. Total public expenditure reflects all main public spending at a societal level contributory social security benefits, wage subsidies and spending beyond the central including all contributory and non-­ government, compiled across sources from the central government and individual institutions. 2 percentage points over 2007–2009 (11 to 13 percent), subsequently decreasing somewhat until COVID-­19, when the Government’s fiscal stimulus benefiting households and workers is estimated to have been the equivalent of 0.5 to 1 percent of GDP given modest amounts and coverage. Türkiye’s human capital expenditures by component are generally in line with comparable countries, with some differences. As of 2020 to present, pensions has generally accounted for the bulk of public social expenditures (5.6 percent of GDP). This is followed by education (4.5 percent of GDP), health (3.4 percent), survivor benefits (1.4 percent), wage subsidies (0.6 percent), unemployment benefits (0.2 percent), active labor market programs (0.2 percent), and non-­ kind contributory social assistance and in-­ services (0.6 percent, which increased to 1.36 percent by 2021)10. Türkiye’s contributory social insurance policies are financed mainly through employer contributions, employee contributory programs such as social assistance contributions, and public transfers (for non-­ and health insurance subsidies). 10 World Bank staff estimates based on detailed definitions of programmatic and thematic expenditures irrespective of agency implementation, using Government of Türkiye data and OECD Social Expenditures database, 2020. 8 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 2.2.  Social expenditures by type as a percent of GDP, OECD (social protection and labor, first; education, second) Public social protection and labor expenditure, 2017–2019 Total education spending as % GNI, 2017–2019 Mexico Romania Korea Greece Chile Costa rica Italy Türkiye Türkiye Colombia Lithuania Ireland Bulgaria Lithuania Switzerland Spain Latvia Slovakia Iceland Portugal Netherlands Czechia Israel Australia Serbia Estonia Latvia Slovak republic Malta Canada Germany United states Hungary Slovenia Czech republic Poland New zealand Europe average Oecd Switzerland United kingdom Poland Estonia Luxembourg France Slovenia Netherlands Japan Portugal Austria Spain United kingdom Greece Cyprus Germany Luxembourg Norway Sweden Belgium Austria Finland Italy Denmark Denmark Sweden Belgium Finland Norway France Iceland 0 5 10 15 20 25 30 35 0 2 4 6 8 10 Pensions (old age and survivors) Health Income support to the working Social service excl health age population Source: World Bank staff using WDI data and OECD Social Expenditures data. Note: “health” included as part of OECD standard indicators largely to refer to health insurance as part of conceptual broad analysis of social risk mitigation, not necessarily as health expenditures. standard definition of “social protection and labor” typically referring to non-­ International Benchmarking and the Effects of Shocks While Türkiye spends nearly 16 percent of GDP on human capital, by contrast the OECD average accounts for nearly 25 percent as of 2020. Compared to most OECD countries, Türkiye’s spending on education, health and active labor market programs is relatively modest. In the OECD, the drop in economic growth during 2007–2009 of around 3.5 percent) was met by an increase in social expenditures of 6 percentage points (2.7 to -­ nearly 3 percentage points of GDP on average (17.7 percent of GDP to 20.7 percent of GDP). This level has been maintained since, owing to demographic changes, and long-­ lasting impacts on jobs and consumption in general. Similarly, due to COVID-­ 19, most advanced countries mobilized emergency social measures that accounted for 1–2 of GDP over 2020; Türkiye’s support packages targeting households and workers’ wages and benefits were relatively modest and estimated to account for up to 0.5 percent of GDP. Taking Stock: Overall Human Capital Expenditure and Employment Trends 9 Figure 2.3.  Social expenditure trends, OECD average growth overall, 1990–2018 (first panel) and by instrument (second) OECD average public social expenditure and Average OECD social expenditures by GDP growth, 1990–2018 instument (as % of GDP), 2019 25 20 19.0 Public social spending (% of GDP) 18 20 16 15 14 12 % of GDP 10 10 % 8 7.0 5 GDP growth (annual %) 5.3 6 0 4 2.0 1.9 2 0.9 0.7 0.5 0.4 0.3 –5 0 19 4 96 20 8 00 20 2 20 4 06 20 8 20 0 20 2 20 4 16 18 19 0 19 2 e lth ca ly Su ity s t og r m s m et g l ta en or pr u ea 9 0 1 ag 9 0 0 1 9 sin 1 9 i m rk 20 m c H es To ea 19 20 20 19 ne rviv m pa r ra a Fa ou la cy a ld H y O lo so mp In i ol bo a lp U ci e tiv Ac er th O Source: World Bank staff using WDI data and OECD Social Expenditures data. Aggregate Human Capital Outcomes Türkiye has achieved considerable gains concerning fundamental or first-­ order human capital outcomes, but blind spots have emerged on higher-­ order human capital “plus” areas needed for an evolving economy. The human capital index (HCI) shows that first-­ order needs have largely been met in terms of basic literacy and infant and maternal mortality, although HCI skills outcomes lag in Türkiye relative to comparable countries. In addition, the HCI labor dimension, or utilization, also shows that overall employment levels are relatively lower and gender gaps wider than elsewhere. Türkiye’s average HCI, at 0.65 (on a scale of 0 to 1) as of 2018, is on par with many middle-­income countries but lags others and bears socioeconomic gaps particularly regarding education11. HCI by income level shows minimal differences in life expectancy but vulnerabilities in health status and learning, made worse during COVID. In Türkiye, the percentage of children in the top 20 percent of households who are not stunted is 96 percent while it is 69 percent among the poorest 20 percent, a gap of 27 percentage points. This gap is larger than the typical gap across the 50 countries (19 percentage points). 11 World Bank (2019). Türkiye -­Insights from Disaggregating the Human Capital Index. Human Capital Project October 2, 2019, brief. Washington DC: World Bank Group. 10 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 2.4.  Global Human Capital Index correlations as a function of social expenditures (first panel), labor force participation rate among adults (second) and youth (third) HCI vs SOCX, 2017–2020 0.9 0.8 HCI, 2017 (0 to 1) 0.7 0.6 TUR 0.5 0.4 0.3 0 5 10 15 20 25 30 35 SOCX % GDP LFPR 2019 vs HCI 90 80 70 60 R2 = 0.0921 LFPR (%) 50 TUR 40 30 20 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 HCI Youth LFPR 15–24 vs HCI 1.0 0.9 0.8 R2 = 0.085 HCI, 2017 0.7 TUR 0.6 0.5 0.4 0.3 20 30 40 50 60 70 80 90 Youth LFRP, 15–24 (%) Source: World Bank staff calculations, World Development Indicators, 2017–2020. Taking Stock: Overall Human Capital Expenditure and Employment Trends 11 Students from the richest 20 percent of households in Türkiye score 521 (out of a learning index ranging from 300 to 625) while those from the poorest 20 percent score 426, a gap of 94 points. This gap is somewhat wider than that observed across the 50 countries (55 points) assessed. At an aggregate level, for its level of social expenditures, the allocative efficiency of Türkiye’s spending is lower than comparable countries. In terms of average HCI, Türkiye’s outcomes are somewhat lower than comparable countries. HCI-­ utilizaton, or overall labor force participation rate (LFPR), is also lower than expected for its level of social spending, indicating ineffiicent spending and likely broader factors such as demand-­ side (private investment) and social dynamics. Similarly, youth LFPR, similar to youth unemployment, is lower in Türkiye than expected for its level of HCI. Examining higher-­ level aspects of human capital further reveal inefficiens regarding public spending in three main areas: (i) boosting competitive skills, (ii) faciliating labor market entry and (iii) matching to the demand side. Pandemic Preparedness and Response The high uncertainty associated with COVID exacerbates Türkiye’s economic and demographic challenges, which are particularly evident since the downturn in economic growth experienced in 2018. Taking a retrospective look since the beginning of the pandemic12, COVID cases have drastically dropped since the first case was officially declared in March 2020, owing to broad public measures adopted and a ramping up of vaccination efforts. As of December 2021, cases still remain somewhat high given a considerable share of the total population remain unvaccinated, although serious illness and mortality have significantly dropped. Since Spring 2021 when national vaccination efforts were expanded beyond initial high-­ risk populations, cases have been marked by intermittent periods of slowdown and spikes in line with global patterns, public measures, vaccination progress and the spread of new variants. Regarding positive cases per 1M population, Türkiye remains high, at 26th highest in the world as of Sept 21, 2021 (325/1M), despite uneven slowdown and uptick since April 2021, significantly higher than March 2020 (.03/1M)[1]. This put Türkiye at a lower relative case load than Serbia, Israel, Malaysia, Georgia, UK, Lithuania, USA, but higher than Albania, Greece, Russia, Ukraine, and France, among others. From early-­ July 2021, case levels began to accelerate particularly in eastern provinces. By August 2021, case numbers reached around 25,000 per day, considerably lower than the 60,000 cases per day seen in the third wave in May 2021. However, the latest data as of late September 2021 show that cases were an upward trend, hovering at approximately 30,000 cases per day, dipping to a weekly average of around 20,000– 24,000 per day by early December 2021. 12 Since the time of writing this note, the public health and epidemiological profile has continued and will continue to evolve; this section is not intended to reflect the precise status by the time of publication, but rather a retrospective analysis of key issues. For accurate information on the current status, see Ministry of Health, Republic of Türkiye, COVID-­19 Dashboard: https://covid19.saglik.gov.tr; Our World in Data Dashboard, Oxford University: https://ourworldindata.org/coronavirus 12 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Table 2.1.  Türkiye’s vaccination progress as of November 5, 2021, shown to reflect status during peak phase. Indicator Outcome 19 vaccine doses administered Total number of COVID-­ 117.2 million Total number of people fully vaccinated (two doses) 49.1 million As a share of total population 61.4 percent As a share of adult population 18+ 79.0 percent Source: World Bank staff using official Ministry of Health data. Data shown as of November 2021 to depict COVID management over pandemic peak. Türkiye was among the first countries to initiate relatively swift vaccinations for health care workers and the older population in January 2021, expanding to a significant share of the population by summer 2021. Although the pace of the vaccination roll-­ out had slowed by March, Türkiye has now taken steps to regain momentum as vaccination is plateauing. As of November 5, full vaccination coverage stands at nearly 6 percent of the BionTech). Provinces in the east total population (primarily CoronaVac, followed by Pfizer-­ and southeast have the lowest vaccination rates at 40–50 percent. Türkiye has also been advancing domestic vaccine R&D, with two vaccines, TURKOVAC and Vaccine Like Particle (VLP) vaccines, have completed Phase 2 (proof of principle) and are about to apply for Phase 3 trials for human use approval. For its population size, Türkiye’s case fatality rate has remained relatively low during the pandemic, reflecting its historically steady investment in public health services and health insurance. Mortality halved between mid-­ 2020 to mid-­2021 with the increase in vaccination rollout and is among the lowest 40 countries globally, at approximately 1 percent. This is likely due to a relatively young population, relatively high quality of tertiary health care, and a combination of COVID public measures taken to control infection rates over the past year. As of September 2021, Türkiye relaxed most public measures (such as school closures, social venues, partial/full weekly curfews) and has gradually been increasing requirements to show COVID-­ free status and/or vaccination to frequent certain venues. Likewise, despite a slow start and some regional gaps, Türkiye’s public health system has demonstrated broad coverage in terms of vaccination, particularly relative to other comparable economies. After a relatively gradual roll-­ risk out among older and at-­ populations between January and June 2021, vaccination significantly accelerated since July 2021, particularly in Ankara, Istanbul, and Aegean and Mediterranean provinces, Taking Stock: Overall Human Capital Expenditure and Employment Trends 13 where rates exceed 80 percent13. The share of the total population fully vaccinated as of early December 2021 stood at approximately 61 percent, up from 50% in September 2021, when it compared to Poland 51%, Sri Lanka 52%, Spain 77%, and Chile 73%, putting Türkiye at a ranking of approximately 70th highest fully vaccinated rate in the world. As a share of the adult population, by early December 2021, Türkiye had reached over 80 percent, up from 68 percent in September 2021, with just over 10 million adults having received a third dose (less than 10 percent of the total population). Vaccination lags behind in interior and eastern provinces, hovering near 50 percent of the total population, largely due to vaccine hesitancy and misinformation. To further boost its public health system, Türkiye’s renewed investments in pandemic preparedness for the future remains critical. Its National Pandemic Action Plan includes prevention, detection, and response measures, launched during the COVID pandemic14. An inter-­ ministerial coordination mechanism exists between key ministries (health, interior, treasury and finance, industry and technology, and trade) and non-­ state stakeholders (international agencies, bilateral development partners and international financial institutions). Similarly, Provincial Health Directorates are also responsible for carrying out determined plans at the provincial level. Nonetheless, future pandemic preparedness efforts will likely require strengthening early alert systems and inter-­ agency coordination more efficiently given lessons learned from COVID. For example, with the reclassification of several facilities during COVID as ‘pandemic hospitals’, Türkiye will need to increase Level 3 intensive care unit capacity15 for greater pandemic preparedness in line with the international and national guidance.16 There are around 25,000 adult intensive care beds in Türkiye, of which 11,171 belong to MoH. There are 28.6 ICU beds per 100,000 population in Türkiye. Countries like Germany which has better ICU bed capacity (47.7 ICU beds per 100,000 population) continues to increase its existing supply of ventilators by increasing the number by 50 percent (from 20,000 to 30,000).17 The mobilization of additional mechanical ventilators is extremely important to increase the number of Level 3 ICU beds in bigger cities like İstanbul (which has almost 60 percent of the total cases in Türkiye) where there are only 14 ICU beds for per 100,000 population. 13 Publicly available data remains limited in Türkiye regarding detailed information on COVID response and management. Türkiye initially began mainly rolling out the CoronaVac vaccine as of January 2021 and began rolling out BionTech at scale as of approximately April 2021. Since early summer 2021, data on dose procurement by type Pfizer-­ has not been made public by the Government, nor has disaggregated data at the national or province level by detailed demographics regarding case load or access to vaccines, for example. Türkiye has also been advancing progress on R&D on domestic vaccine capacity. Two domestic vaccines, TURKOVAC and Vaccine Like Particle (VLP) vaccines, have completed Phase 2 (proof of principle) and are about to apply for Phase 3 trials for human use approval. 19 Health Project. Report No. PAD3897. Washington DC: World 14 World Bank (2020). Türkiye Emergency COVID-­ Bank. 15 Level-­3 Intensive Care Units require invasive hemodynamic monitors and ventilators for each ICU bed. https://www.saglik.gov.tr/TR,10979/yogun-­bakim-­unitelerinin-­standartlari-­genelgesi-­200853.html 16 https://www.nice.org.uk/guidance/ng159/resources/covid19-­rapid-­guideline-­critical-­care-­in-­adults-­ pdf-­66141848681413 http://www.istanbulsaglik.gov.tr/w/sb/ozeltedk/belge/8_ek_madde.pdf 17 https://www.ft.com/content/d979c0e9–4806–4852-­a49a-­bbffa9cecfe6 14 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 2.5.  Public health systems versus COVID management, global comparisons (policy stringent index, first panel; universal health coverage, second) COVID Mitigation policy stringency index versus prevalence 100000 Log case ratio (cumulative cases per 100,000 population) EST CZE USABHR SRB GBR GEO HRV SYC URY SVK MNG PRT ARG ESP IRL LTU BEL SVN HUN MDA LBN PAN CHL BRA BGR BIH DNK LUX CYP MLT LVA NLD COL ISR FRA ITA AUT GRC 10000 BWA SWE JOR POL CPV RUS OMN MYS BHS AZE ALB CHE PER IRN DEU BLR NAM TUR PRY CUB IRQ SGP ZAF ROUCRIUKR BLZGUY BRB SUR FJI TUN NOR ISL BOLLBY DOM MUS CAN KAZ KGZ THA ECU TTO MAR PHL BRN FIN MEX GTM HND SWZ SLV TLS IND GAB LKA NPL JAM VEN BGD ZMB JPN SAU VNM AUS y = –3.0148x2 + 310.95x – 1520.5 1000 MRT RWA IDN R2 = 0.0568 PAK KOR LSO KHM LAO GMB SEN UZB ZWE MMR AFG DZA EGY MOZ CIV KEN GHA BTN SYR HTI BDI CAF CMR COG TGO ETH MWI BEN PNG NZL TJK SDN LBR GIN UGA NIC MDG SOM 100 SSD NGA BFA SLE MLI TCD COD TZA YEM NER CHN 10 SLB VUT 1 0 10 20 30 40 50 60 70 80 90 100 COVID public measures stringency index (0–100) Universal health coverage versus COVID vaccination coverage levels y = 0.0137x2 – 0.2747x – 3.8305 90 R2 = 0.6816 CHL MLT PRTCUB 80 KHM URY DNK ISL KOR JPN Log share of total population fully vaccinated (%) BEL SGP ESP SYC FIN ITA AUS CAN BTN MYS IRL NOR FRA CHN ISR NZL MUS BHR CYPBRN NLD 70 MDV SAUARG DEU SWE LKA ECUAUT LUX SLV MNG LTUGRCCRI BRA GBR FJI MAR HUN CZE PAN CHE 60 TUR ATGEST USA LVA PER POL THA WSM IRN DOM MEX TKM COL SVN 50 BLZ OMN HRV TTO CPV KAZ BRB TON SRB TUN SVK LAO VNM AZE MKD SURROU RUS 40 BWA FSM HND PRYGUY VEN BHS IDN MNE JOR TLS BOLGRD NIC IND ALB 30 NPL COM PHL UZB LSO LCATJK BLR PAK BGD BGR MDA GTM BIH SWZ LBN KIR ZAF 20 ZWE RWA GEOUKR VCT MRT GNQ MMR EGY MOZ STP JAM DZA GMB LBR TGO SLBVUT AFGCOG IRQ NAM KGZ 10 SLE SENCIV ARM GIN LBY CAF ETH BEN GNB PNG BFANGA TZA YEM UMWI SDN GHA CMR GA GAB KEN SOM MDG MLI COD BDI SSD NER HTI ZMB SYR 0 TCD 0 10 20 30 40 50 60 70 80 90 100 Universal health coverage index (0 to 100) Source: World Bank staff analysis using Our World in Data as of early 2022 available at https://ourworldindata.org/coronavirus prone countries, compounding In addition, Türkiye is one of the world’s most disaster-­ pandemic risks. Its population and economy, especially its cities due to the concentration of people and assets, are highly exposed and vulnerable to earthquakes, floods, landslides, and other hazards, having experienced multiple magnitude 5 or higher earthquakes over the past two decades. It ranks 45th among the 191 “high risk group of countries” in the Global Risk Index for Risk Management. Reviewing and or updating existing protocols and Taking Stock: Overall Human Capital Expenditure and Employment Trends 15 contingency plans for pandemic hospital during the COVID-­ 19 response will be important to integrate any preparedness measures and/or supplies required to mitigate the impacts of disasters that may occur during the pandemic outbreak. It is estimated 33 % of schools in the country are potentially seismically vulnerable.18 Türkiye is also particularly vulnerable to the impacts of climate change, which have already manifested in an increase in annual mean temperature, changes in the precipitation regime, and increasing numbers of climate-­ related hazards such as floods and droughts (such as catastrophic floods seen in 2019). Climate-­induced hazards and extreme weather events will continue to affect the safety and welfare of hundreds of thousands of people and will also likely alter demand patterns and cause substantial damages to energy, waste, and transport infrastructure. The country’s emergency preparedness and resilience to climate change urgently requires further strengthening, especially as regards protecting the most vulnerable. Institutional Governance and Political Economy Overall, Türkiye has long benefited from relatively well-­ developed institutions that have helped set the foundations for human capital and labor markets, with opportunities to enhance further for meeting future needs. Its governance system has provided a strong basis for policymaking since the founding of the Republic in 1923, resulting in well-­ developed ministerial policy strategies, broad service delivery and regular, actionable national development plans. Historically, for example, the country was among the first nations to ratify laws that empowered women -­from suffrage to the civil code. At the same time, political institutions have been in transition notably since constitutional changes introduced over the last two decades, which have shifted certain policy-­ making roles from sectoral to central structures. Moving forward, these changes can potentially result help support whole-­ government approaches to policies and services, which will depend on of-­ aligning cross-­ cutting objectives, measures and data-­ sharing more closely. Two key global challenges faced by most countries lie in fostering periodic tri-­ partite dialogue between government, citizens and the private sector, as well as ensuring collective, transparent policymaking. These challenges will be particularly needed for reforming critical human capital educational needs, labor market policies regarding wage-­ setting and firms’ incentives, youth inclusion, and gender parity. Türkiye has been placed as part of the OECD and broader global dialogue in sharing best practice well-­ regarding institutional development, experience that will prove invaluable to safeguarding participatory policymaking to meeting these challenges. 18 See World Bank-­ administered Global Facility for Disaster Reduction and Recovery (GFDRR), Türkiye Profile: https://www.gfdrr.org/en/turkey 16 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Key Implications As noted, COVID has put the spotlight on the need for public expenditures to refocus on households and workers that are at greatest risk of being left behind. Key policy aims include the need to: (i) continuing to stem immediate pandemic transmission as epidemiology evolves, enhance disease management and boost preparedness for ensuring access to therapy and treatment; (ii) enhance jobs opportunities, employment retention, and safeguard and restore wages and benefits for vulnerable workers; (iii) preserve human capital in terms of learning and enhance skills among children, young adults and vulnerable workers for changing economic realities; and (iv) compensate loss of income and mitigate the impact of high prices on consumption, food security and health care costs. Box 2.1.  Mitigating Shocks to Human Capital in Türkiye: Public Policy Response During COVID Globally since the COVID crisis and within two providing short term wage coverage for eligible to four weeks from the start of the outbreak, 45 firms (reaching an estimated 240,000 firms and countries introduced new measures or adapted over 3 million workers over 2020); increasing the existing social protection and labor programs minimum pension; allocating TL 1,000 as one off to mitigate the impacts, as did Türkiye to an cash support to social assistance beneficiaries extent19. Most of the programs have focused and eligible vulnerable households (reaching on direct income support (30 programs), wage up to 5.5 million households over 2020); and subsidies (11), subsidized sick leave through provisions for outreach social services for older employers (10) and other forms of subsidies persons. or deferrals to social security contributions and social pensions, as well as unemployment The health system was boosted to leverage benefits. new community health services, manage COVID hospitalizations and expand vaccination roll-­ out. Türkiye’s COVID response was measured and COVID World bank support under the Health Pre-­ gradually expanded to address the health, Systems Strengthening and Support Project20 education, purchasing power and employment helped established over 207 Healthy Living effects of economic closures during 2020. In disciplinary health Centers, innovative inter-­ terms of welfare, the government rolled out centers supporting Family Medicine Centers measures to mitigate the shocks on firms, designed to bring preventative and primary workers and households. Early during the care closest to citizens and boost the delivery outbreak, it announced an initial economic of professional continuing training for health package totaling approximately TL 100 billion care workers. In addition, the new Emergency (US$15 billion) to stem the impact on firms and COVID-­ 19 Health Project included the acquisition targeted households, including deferral of social of over 2,000 ventilators nationwide, personal security and payroll tax on the part of firms, protective equipment and, under the pre-­ COVID Time Review of Country 19 World Bank/ Gentilini U et al (2020). Social Protection and Jobs Responses to COVID: A Real-­ Measures. Washington DC: World Bank. 20 World Bank (2015). Türkiye Health Systems Strengthening and Support Project. Project Appraisal Document. Report No: PAD1294. Washington DC: World Bank. Taking Stock: Overall Human Capital Expenditure and Employment Trends 17 project, planned technical and equipment innovations, materials, and teacher training support to two national vaccine production (through expansion of digital education platform, centers. virtual classrooms, and provision of digital/ TV materials and courses). Key results to date Türkiye was also among the first countries to include: (i) increased the number of concurrent adopt one of the largest distance education users of the online distance education from systems to mitigate learning loss, capitalizing 300,000 during baseline (March 2020) to based learning among on online and television-­ 1,000,000; (ii) simultaneous Virtual Classroom 12 students. Türkiye was also the first country K-­ sessions increased from 30,000 to 255,000; and to request World Bank support to respond to (iii) student usage of digital platform from 26% to education needs early on during COVID, with 68%. In addition, more than 2,000 digital/video supported Education-­ the first World Bank-­ materials were prepared, as well as adapted Digital Development project mobilized, the Safe for sight-­and hearing-­impaired students. MoNE Schooling and Distance Education Project21. is requesting restructuring of some of the The project has been supporting Türkiye project sub-­ components mainly to re-­ assess the during COVID distance learning, recent school development of a new digital education platform re-­ term digital learning opening and longer-­ and to focus more strongly on blended education investments. These include creating a system needs of vulnerable students (early childhood, to improve “blended” (face-­ face/EdTech) to-­ rural, dispersed, migrant workers). Source: World Bank staff. 21 World Bank (2020). Türkiye Safe Schooling and Distance Education Project. Project Appraisal Document. Report No: PAD3962. Washington DC: World Bank. Figure 2.6.  Human Capital policy aims for short-­to long-­term post-­COVID recovery in Türkiye Pandemic health response: Early alert to long-run preparedness Jobs: Activation to job creation Response: scale-up testing and Education: Continuing to transforming skills vaccination Response: protect key capabilities, HR, firms’ operating and Resilience: Short- to long-term facilities and goods wage expenditures Response: facilitate risk-sharing and equipment Outlook: expand jobs' emergency Outlook: upgrade incentives, financing connectivity, preparedness and Response: boost and labor policy in parent-teacher support income support for long-run services essential future Outlook: restructure poor households, sectors particularly for learning modalities, pensioners, and women and youth infrastructure, skills vulnerable and labor activation for self-employed new sectors Outlook: revitalize universal, integrated social insurance Source: World Bank staff. 18 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs 3 Targeting Job Creation: The Shift to Green Introduction Growth in Türkiye, despite the pandemic, has continued to be strong, although gains have not always translated into poverty reduction or broad job creation since 201822. As of 2020, annual growth was higher than in comparable countries that averaged 1.8 percent, second-­ fastest among G20 countries following China, projected to increase between 4.5 to 9 percent over 2021. Ongoing monetary stabilization, inflation and a widening deficit remain challenges over the short-­to mid-­ term in Türkiye. Growth has mainly been driven by credit and stocks, followed by exports and investments, particularly within industry and services. Further, since 2018 and exacerbated by COVID, poverty simulations have revealed more households have likely fallen or been pushed deeper into poverty as a result of earnings’ shocks23 , with an increase in poverty by up to 2.1 percentage points by the end of 2020. This is the equivalent of nearly 1.6 million more individuals, bringing the total to nearly ten million individuals, or an increase of 40 percent relative to the 2018 level. Despite policy measures taken during the pandemic to ban layoffs, subsidize social security premiums, and allocate household emergency cash assistance, labor market losses and poverty increased nonetheless given limited coverage and the sheer magnitude of the pandemic’s impact. 22 World Bank (2021). Türkiye Economic Monitor, August 2021. 23 Measured as the share of households with consumption (based on income-­ based simulations) levels below the Middle Income Countries ($5.50 per person per day in constant 2011 purchasing World Bank’s poverty line for Upper-­ power parity prices). Targeting Job Creation: The Shift to Green 19 Aggregate Job Growth and Productivity Building on trends over 2020 and early 2021, key sectors have started to regain 2018 job levels24. Monthly labor force surveys show that aggregate employment has been showing active recovery in 2021. Nearly 3 million jobs were regained between December 2020 and June 2021, which corresponds on average to 500,000 jobs added every moth. When compared to June 2020, the economy created 2.4 million jobs, an increase of 8.8%. Job 2018 crisis level. figures in June show that total employment is back to the pre-­ The demand for labor has evolved over the past two decades in Türkiye, increasingly shifting towards technology-­and service-­ based occupations expected to keep pace moving forward25. The labor market in Türkiye has exhibited a strong performance in the 15 years before the pandemic, creating over 8.4 million jobs between 2004 and 2019 until COVID. The economic sectors which drove job creation over this period were manufacturing (17%), health (12%) and education (12%), accommodation and food services (11%), followed by wholesale and retail (10%), public administration and defense (9%), and administrative and support services (9%). The agriculture sector remains sizable, accounting for 18 percent of total employment in 2019 (vs. 30 percent in 2004) and still employing over 5 million workers. Most job creation in the period considered occurred therefore in sectors with average or below-­ average productivity. Further, high productivity sectors, with few exceptions, have not experienced on average higher employment growth than lower sized. This evidence suggests that the productivity sectors, and remain relatively small-­ process of reallocation of labor across sectors over time might not have been fully efficient. Productivity has varied widely across sectors with similar educational attainment, suggesting that capital-­ intensity and the quality of competencies, irrespective of educational attainment, matter. Productivity is not necessarily correlated with traditional educational attainment, making it important to distinguish between occupational skills in demand and educational attainment. Certain high-­ productivity sectors (finance and ICT) also tend to comprise a relatively higher share of workers with tertiary education than other relatively productive sectors such as transport, utilities, mining and real estate, where there is a relatively higher share of workers with a lower educational attainment. By the same token, large sectors like manufacturing and wholesale trade, absorb a below-­ than average share of tertiary educated workers. The trends are consistent with the nature of these sectors and the manual for higher-­ order cognitive skills needed. At the same time, relatively sluggish productivity in manufacturing and agriculture suggests that the skill mix and level, irrespective of educational attainment, may be hindering value-­ added growth. 24 World Bank (2022). Türkiye Economic Monitor. up simulations of scenarios under various policy options to be assessed in World Bank (forthcoming). 25 Follow-­ Green Growth Technical Note, Jobs Impacts. 20 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Table 3.1.  Key labor force indicator trends, 2014 versus 2021 Total (15+) Female (15+) Male (15+) Change Change Change Indicator 2014 2021/Q3 (%) 2014 2021/Q3 (%) 2014 2021/Q3 (%) Labor Force (Thousand) 28786 33564 16.6% 8729 11022 26.3% 20057 22542 12.4% Illiterate 1170 886 –­24.3% 809 601 –­25.7% 361 285 –­21.1% Illiterate without diploma 1145 924 –­19.3% 476 376 –­21.0% 669 548 –­18.1% Primary school/education 14955 14519 –­2.9% 3792 3962 4.5% 11163 10557 –­5.4% Targeting Job Creation: The Shift to Green High school 2977 4092 37.5% 795 1193 50.1% 2182 2899 32.9% Vocational high school 2848 3793 33.2% 664 933 40.5% 2184 2860 31.0% Higher education 5691 9350 64.3% 2193 3957 80.4% 3498 5393 54.2% Employment (Thousand) 25933 29652 14.3% 7689 9327 21.3% 18244 20324 11.4% Employment Rate (%) 45.5 46.4 2.0% 26.7 28.9 8.2% 64.8 64.3 –­0.8% Labour Force Participation Rate (%) 50.5 52.6 4.2% 30.3 34.2 12.9% 71.3 71.4 0.1% Not In Labour Force (Thousand) 28200 30274 7.4% 20112 21222 5.5% 8089 9052 11.9% Among 15–24 years 6936 6695 –­3.5% 4236 3986 –­5.9% 2700 2709 0.3% Unemployment (Thousand) 2853 3912 37.1% 1040 1694 62.9% 1813 2218 22.3% Unemployment Rate (%) 9.9 11.7 18.2% 11.9 15.4 29.4% 9 9.8 8.9% Among 15–24 years 17.9 22.1 23.5% 20.4 30.2 48.0% 16.6 17.6 6.0% Informal employment 9069 9057 –­0.1% 3724 3564 –­4.3% 5344 5493 2.8% Formal employment 16864 20594 22.1% 3964 5763 45.4% 12900 14831 15.0% Source: World Bank staff calculations, TURKSTAT data. 21 22 Table 3.2.  Key employment indicators by sector and gender, 2014 versus 2021 Total Female Male Change Change Change Indicator 2014 2021/Q3 (%) 2014 2021/Q3 (%) 2014 2021/Q3 (%) Agriculture (Registered) 970 796 –­17.9% 143 134 –­6.3% 827 662 –­20.0% Agriculture (Not-­Registered) 4500 4636 3.0% 2390 2234 –­6.5% 2110 2402 13.8% Share of total employment 21.1% 18.3% –­13.1% 32.9% 25.4% –­22.9% 16.1% 15.1% –­6.4% Share informal 82.3% 85.3% 3.7% 94.4% 94.3% 0.0% 71.8% 78.4% 9.1% Industry (Registered) 4239 5303 25.1% 836 1226 46.7% 3403 4076 19.8% Industry (Not-­Registered) 1077 898 –­16.6% 400 293 –­26.8% 677 605 –­10.6% Share of total employment 20.5% 20.9% 2.0% 16.1% 16.3% 1.3% 22.4% 23.0% 3.0% Share informal 20.3% 14.5% –­28.5% 32.4% 19.3% –­40.4% 16.6% 12.9% –­22.1% Construction (Registered) 1212 1227 1.2% 69 86 24.6% 1143 1141 –­0.2% Construction (Not-­Registered) 700 629 –­10.1% 10 5 –­50.0% 690 624 –­9.6% Share of total employment 7.4% 6.3% –­15.1% 1.0% 1.0% –­5.1% 10.0% 8.7% –­13.6% Share informal 36.6% 33.9% –­7.4% 12.7% 5.5% –­56.6% 37.6% 35.4% –­6.1% Services (Registered) 10443 13268 27.1% 2916 4316 48.0% 7527 8952 18.9% Services (Not-­Registered) 2791 2894 3.7% 924 1032 11.7% 1867 1862 –­0.3% Share of total employment 51.0% 54.5% 6.8% 49.9% 57.3% 14.8% 51.5% 53.2% 3.3% Share informal 21.1% 17.9% –­15.1% 24.1% 19.3% –­19.8% 19.9% 17.2% –­13.4% Source: World Bank staff calculations, TURKSTAT data. Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 3.1.  Jobs by major sector, 2014–2020 Jobs by sector 15246 15774 15873 13235 13892 14617 15061 Services Construction 1911 1914 1987 2095 1993 1549 Industry 1538 5316 5332 5297 5383 5675 5561 5498 Agriculture 5470 5483 5304 5464 5296 5097 4716 2014 2015 2016 2017 2018 2019 2020 Source: World Bank staff calculations, TURKSTAT data. Figure 3.2.  Sectoral employment trends as share of GDP (first panel) and annual change (second), 2019 versus 2020 Sectoral share of GDP vs employment 56.2% 56.5% 19.6% 19.4% 20.5% 19.8% 21.9% 23.6% 17.6% 18.2% 6.6% 6.3% 6.0% 6.5% 5.7% 5.5% GDP share Employment GDP share Employment GDP share Employment GDP share Employment share share share share Agriculture Industry Construction Services 2019 2020 Annual percentage change, GDP and average earnings, by sector 16.7% 12.3% 10.3% 5.9% 3.3% 3.2% 3.1% 2.4% 1.4% –1.0% –0.9% –3.5% –3.0% –5.5% –5.9% –8.6% GDP share Earnings GDP share Earnings GDP share Earnings GDP share Earnings Agriculture Industry Construction Services 2019 2020 Source: World Bank staff calculations, TURKSTAT data. Targeting Job Creation: The Shift to Green 23 Figure 3.3.  OECD: GDP versus share of wages by country, 2019 (first panel) versus 2020 (second) GDP (billion PPP $) vs labor share (%), 2019 $25,000 70% 60% $20,000 50% $15,000 40% 30% $10,000 20% $5,000 10% $0 0% Es land xe La nia Sl ou a ov Li ove rg ak thu nia pu ia n c H ree d De ng e nmary Po rw k rtu ay Is al h Ire ael pu nd c Sw Sw stria e en Coelgi d th om m la a Au ola s s nd na ia Sp da K ain r a M kiye Ki Ita o ng ly Fr om Ge us e rm sia Un Eu Jap y ite o A an St a es P nd Fi bli Au bli an No ar b i er bi Tü ore d re u c R c ic G lan B rlan m tv Re an g Catral Ne l u at an itz d Re la ex to r d e e Ic r d Lu ec ite Cz Un Sl Gross domestic product (GDP), current PPPs, billions US dollars Compensation of employees, percentage of gross value added (GVA) GDP (billion PPP $) vs labor share (%), 2020 $25,000 70% 60% $20,000 50% $15,000 40% 30% $10,000 20% $5,000 10% $0 0% Es and xe La a bo a Sl urg ak hu ia pu ia Fi blic Gr and un e nm ry No ark rtu y h ela l pu d Au lic Sw we ria er n Ne Bel nd er m Po ds na d Sp a Ko n Tü rea e Fr om R ce Eu rma a ro ny ea ng ly ec Ir ga Po rwa ni m vi d i H eec itz de ai iy Re n Ca an ov Lit en Re an Ge uss Ki Ita th giu De ga b n an Ar S t la t rk to d s la el nl l ov Ic d Lu ite Cz Un Sl Gross domestic product (GDP), current PPPs, billions US dollars Compensation of employees, percentage of gross value added (GVA) Source: World Bank staff calculations, OECD data, growth and employment https://data.oecd.org 24 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 3.4.  OECD: Sectoral share of Gross Value Added (GVA) versus labor share by country, 2020, for industry (first panel), services (second) and agriculture (third) Industry as share of gross value added vs industry labor share of GVA, 2020 50% 50% 45% 45% 40% 40% 35% 35% 30% 30% 25% 25% 20% 20% 15% 15% 10% 10% 5% 5% 0% 0% G rg te Fr ce N ing ce he om n ly lg s I c um La d Po tvia Sp al Sw ain en ro rk Sw Est a er a F i nd th d Au nia un ia er y ak or y pu y Po lic Tü nd Sl kiye ec R nia pu a Ire lic nd Be nd G gar ov N an Re wa e itz oni Re si an Li an De Ita g H str Eu ma u b b e d an ar ed la la la h us ua e rtu i et d bo m rla re el nl r ov m K xe Lu ni Cz Sl U Gross value added, industry, percentage of total activity Compensation of employees, industry, percentage of GVA Services as share of gross value added vs services labor share of GVA, 2020 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Tü nd ec R iye p a ov c Po ia ak n d pu y N blic th y Fi nia er d Au ny Sw Est ia er a rla e Ic nd d ro ia ea nm y Sp k Sw ain Po den lg l Lu ng e m m g N Gre m te F ds Be ga Sl ubli Re gar Li rwa De Ital ar Re si itz oni he c Ki nc ov Hu lan G lan an ur en r Eu Latv iu xe do n a e ar st la la h us k ua rtu bo m d ra el e r Ire n o et ni Cz Sl U Gross value added, services, percentage of total activity Compensation of employees, services, percentage of GVA Agriculture as share of gross value added vs agriculture labor share of GVA, 2020 10% 10% 9% 9% 8% 8% 7% 7% 6% 6% 5% 5% 4% 4% 3% 3% 2% 2% 1% 1% 0% 0% Be and Ire y nm ia Sw ark N ro n he ea itz dom pu c N blic ay Po nia er m ov y Au nd Fi nia Po nd S d un a Ru ry La ia ak Fra ds ec ep ce G d e Sw ing g Es gal th in Ic via Tü ce an Re li Sl tal H ani Eu ede iy n K ur an De str ss Li pa G iu ga h ub w n Cz R n e et ar a la la t rk e to rtu o I m rla lg re l nl el or u er te mb U uxe d L ov ni Sl Gross value added, agriculture, forestry and fishing, percentage of total activity Compensation of employees, agriculture, forestry and fishing, percentage of GVA Source: World Bank staff calculations, OECD data, growth and employment https://data.oecd.org Targeting Job Creation: The Shift to Green 25 Figure 3.5.  Türkiye: Sector GDP growth versus earnings growth over time, 2007–2019, for agriculture (first panel), industry (second), construction (third) and services (fourth) Agriculture GDP growth vs earnings growth 30% 25% 20% 15% 10% 5% 0% –5% –10% –15% 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 GDP (PPP $ billion percentage change, %) Earning PPP $ percentage change, %) Industry GDP growth vs earnings growth 25% 20% 15% 10% 5% 0% –5% –10% –15% 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 GDP (PPP $ billion percentage change, %) Earning (PPP $ percentage change, %) Contruction GDP growth vs earnings growth 40% 30% 20% 10% 0% –10% –20% –30% 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 GDP (PPP $ billion percentage change, %) Earning (PPP $ percentage change, %) Services GDP growth vs earning growth 20% 15% 10% 5% 0% –5% –10% 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 GDP (PPP $ billion percentage change, %) Earning (PPP $ percentage change, %) Source: World Bank staff calculations, TURKSTAT data. 26 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 3.6.  Net job growth by sector, 2015–2020, overall (first panel) and among females (second) Net job growth overall, annual percentage change (%) 5.0% 5.2% 5.4% 4.3% 5.4% 3.8% 3.6% 3.5% 2.7% 3.0% 0.3% 2.2% 2.0% 1.6% 0.2% 0.2% 0.6% –1.1% –0.7% –0.7% –2.0% –2.3% –3.3% –3.1% –3.8% –4.5% –4.9% –5.1% –7.5% –22.3% 2015 2016 2017 2018 2019 2020 Agriculture Industry Construction Services Total Net job growth among females, annual percentage change (%) 16.9% 10.1% 8.9% 6.3% 6.7% 6.6% 4.8% 5.0% 3.1% 3.6% 3.1% 3.3% 3.0% 0.6% 0.9% –0.4% 0% –1.0% –0.2% –1.2% –1.0% –1.8% –4.8% –4.8% –4.7% –5.7% –6.9% –10.1% –15.6% –18.3 2015 2016 2017 2018 2019 2020 Agriculture Industry Construction Services Total Source: World Bank staff calculations, TURKSTAT data. Targeting Job Creation: The Shift to Green 27 Figure 3.7.  Average earnings by sector, 2006–2020, in terms of absolute levels (first panel) and annual change over time (second) Average earnings by sector (PPP $) 25,000 20,000 15,000 10,000 5,000 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Agriculture Industry Construction Services Annual percentage change (%), average earnings by sector (PPP $) 35.0% 30.0% 25.0% 20.0% 15.0% 10.0% 5.0% 0.0% –5.0% –10.0% –15.0% 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Agriculture Industry Construction Services Source: World Bank staff calculations, TURKSTAT data. 28 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Labor Force Dynamics Labor force participation and employment in Türkiye since 2018 have been volatile, although some recovery has been since the start of COVID recovery. At the peak of the COVID pandemic, Türkiye’s labor force stood at 31.1 million workers as of November 2020, of whom 27.1 million were employed, the lowest levels since 2016–201726, with a modest rebound seen within the first quarter of 2021. Over the pre-­ vaccination pandemic period, the labor force shrank by nearly 4 percent from November 2019-­ November 2020, the equivalent of nearly 1.1 million workers, of whom over 60 percent were women exiting. This represents a contraction of approximately 3.9 percent and a loss of over 1.1 million workers compared to November 2019, with 61 percent of the losses being borne by women. The bulk of losses at over 750 thousand jobs were in the service, largely informal sector, accounting for over 60 percent of losses. Informal jobs continue to prevail among approximately 30 percent of the Turkish population, particularly in agriculture. Employment among younger workers -­ male and female, which was hit the hardest during the crisis, has -­ been growing faster than among adult workers. The recovery has helped to make up for the job losses recorded since June 2020. 22 percent of the jobs regained have been filled by young workers. Taking a retrospective view, COVID job losses seen through early 2021 have added to losses that had started during 2018, when over 650,000 jobs were lost overall, and which have yet to fully recover. Employment declines were seen across all age groups and more markedly for the very young (15–24). The extent of job losses experienced during 2020 resulted in massive outflows from the labor force towards inactivity, both from formerly employed and from unemployed who stopped searching for work (total unemployment declined by around 400,000 units). This happened across all age groups and particularly for the youth 25–34. The pool of inactive people increased by over 2.7 million between 2019 and 2020, implying that substantial flows into inactivity occurred also from the cohorts who were just entering the labor market at the onset of the crisis. Employment measures adopted in Türkiye during COVID contributed to largely maintaining unemployment levels. These measures entailed restricting layoffs and subsidizing wages, protecting jobs for nearly 4 million formal sector workers. As a result, unemployment rates have not increased sharply in 2020 in comparison to previous years. However, the sharp decline in labor force participation rates have also masked otherwise increases in unemployment. Female labor force participation rates declined from 34 percent in 2019 to 31 percent by the end of 2020, whereas the male participation rates declined from 72 percent to 68 percent during the same period. Male labor force declined from 21.8 million by the end of 2019 to 20.0 million in 2020 whereas female labor force declined from 10.2 million by the end of 2019 to 9.4 million in 2020. 26 Turkish National Statistical Institute (TUIK), February 2021 Labor Force Statistics Quarterly Release reflecting data as of November 2020. Targeting Job Creation: The Shift to Green 29 Figure 3.8.  Labor force participation rate, total versus youth, relative to unemployment rate (first panel) and total labor force (second panel), 2014–2021 LFPR (%) and unemployment rate (%), total 15+ vs youth 15–24, 2014–2021/Q3 60 51.3 52 52.8 53.2 53 52.6 50.5 49.3 50 44 44.4 44.2 42 42.4 43.3 40.8 39.1 40 30 25.4 25.3 20.8 22.1 17.9 18.5 19.6 20.3 20 10 13.7 13.2 9.9 10.3 10.9 10.9 11.0 11.7 0 2014 2015 2016 2017 2018 2019 2020 2021/Q3 Labour force participation rate (%), 15+ Labour force participation rate (%), 15–24 Unemployment rate (%), 15+ Unemployment rate (%), 15–24 Total labor force (000s) vs youth share (%), 2014–2021/Q3 34,000 17.0% 33,000 16.5% 32,000 16.0% 31,000 15.5% 30,000 15.0% 29,000 14.5% 28,000 27,000 14.0% 26,000 13.5% 2014 2015 2016 2017 2018 2019 2020 2021/Q3 Labour force (thousand), 15+ Share of total labor force, 15–24 (%) Source: World Bank staff calculations, TURKSTAT data. 30 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 3.9.  Labor force participation rate by gender and education, among all ages 15+ (first panel) versus youth (second), 2014–2021 LFPR (%) by gender and education, ages 15+ years 100 84.7 83.3 80.1 77.9 90 71.0 68.6 66.4 65.6 80 70 60 40.7 37.0 50 33.4 29.9 40 30 20 10 0 2014 2015 2016 2017 2018 2019 2020 2021-Q3 2014 2015 2016 2017 2018 2019 2020 2021-Q3 2014 2015 2016 2017 2018 2019 2020 2021-Q3 2014 2015 2016 2017 2018 2019 2020 2021-Q3 2014 2015 2016 2017 2018 2019 2020 2021-Q3 2014 2015 2016 2017 2018 2019 2020 2021-Q3 (15+) Female (15+) Female (15+) Female (15+) Male (15+) Male (15+) Male (high school) (vocational school (higher education) (high school) (vocational school (higher education) at high school at high school level) level) LFPR (%) by gender and education, youth 15–24 years 82.6 90 77.1 72.5 69.3 67.0 66.5 80 70 60 45.7 39.5 41.2 50 36.3 40 25.5 20.6 30 20 10 0 2014 2015 2016 2017 2018 2019 2020 2021-Q3 2014 2015 2016 2017 2018 2019 2020 2021-Q3 2014 2015 2016 2017 2018 2019 2020 2021-Q3 2014 2015 2016 2017 2018 2019 2020 2021-Q3 2014 2015 2016 2017 2018 2019 2020 2021-Q3 2014 2015 2016 2017 2018 2019 2020 2021-Q3 (15–24) Female (15–24) Female (15–24) Female (15–24) Male (15–24) Male (15–24) Male (high school) (vocational school (higher education) (high school) (vocational school (higher education) at high school at high school level) level) Source: World Bank staff calculations based on monthly LFS data. Targeting Job Creation: The Shift to Green 31 Figure 3.10.  Employment rate by region and education for ages 15+ among females (first panel) and males (second), 2019 versus 2020 Employment rate (%) by region and education, females 15+ years, 2019 vs 2020 70 60 50 40 30 20 10 0.4 0.3 0 –2.0 –2.0–3.5 –1.1 –2.0 –2.8 –3.0 –1.5 –2.4 –10 –2.9 –3.9 –1.9 –3.9 –2.7 –6.7 –4.0 –4.7 –5.4 –3.7 –4.8 –3.1 –2.9 –3.2 –3.6 –20 e an lia lia a a an ia ia lia a a ul Se ar Se ar iy ol ol nb to to to ge ne rk m m at at na na na ta k ck Tü rra ar ar Ae An An ac İs la lA tA tA M tM ite Bl tB st st ra st as es es ed st Ea Ea es Ea nt lE W Ea W M Ce W th h ra ut or nt So N Ce (15+) Female (less than high school) change (15+) Female (higher education) change (15+) Female (less than high school) 2019 (15+) Female (higher education) 2019 (15+) Female (less than high school) 2020 (15+) Female (higher education) 2020 Employment rate (%) by region and education, males 15+ years, 2019 vs 2020 90 80 70 60 50 40 30 20 10 2.5 1.1 1.6 1.1 0.1 0.7 1.0 0.8 2.2 0.1 0.7 0.2 0.4 2.7 0.6 0.7 0 –0.5 –1.6 –1.7 –0.8 –0.5 –0.8 –0.2 –1.3 –0.9 –0.8 –10 e n ia ia a a an lia ia lia a a l bu Se ar Se ar iy a ol ol ol to to ge ne rk an m m t t at na na a na k ck Tü rra ar ar Ae An An ac t İs la lA tA tA M tM ite Bl tB st st ra st as es es ed st Ea Ea es Ea nt lE W Ea W M Ce W th th ra or u nt So N Ce (15+) Male (less than high school) change (15+) Male (higher education) change (15+) Male (less than high school) 2019 (15+) Male (higher education) 2019 (15+) Male (less than high school) 2020 (15+) Male (higher education) 2020 Source: World Bank staff calculations, TURKSTAT data. 32 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 3.11.  Overall employment levels versus informal share by sector, 2014–2021 Employment and informality by sector 35,000 100 90 30,000 80 25,000 70 60 20,000 50 15,000 40 10,000 30 20 5,000 10 0 0 2014 2015 2016 2017 2018 2019 2020 2021 Services (000s) Total informality (%) Construction (000s) Agriculture informality (%) Industry (000s) Non-agriculture informality (%) Agriculture (000s) Source: World Bank staff calculations, TURKSTAT data. Gender, Informality and Opportunity Similar to historical trends, female labor force participation continues to lag that of males, with the bulk of labor force losses since COVID borne largely by women, youth ages 15–24 and semi-­ skilled workers. Overall, labor force participation losses over the period of November 2019 to November 2020 were nearly equivalent to the gains in employment since 2016. The labor force participation rate (LFPR) decreased to 49.3 percent, versus 52.5 percent in November 2019. LFPR remains highest among the most highly educated (higher education), and lowest among the low-­to mid-­ skilled (high school or just below high school). Female labor force participation continues to be less than half that of males, or 30.6 percent compared to 68.4 percent, down from 34 percent in 2019 and relatively constant around this rate since 2015, lower than comparable regional averages such as Central Europe and the Balkans (45.2 percent) and Latin America and the Caribbean (41.5 percent). Similar to NEET challenges in Türkiye, the share of female labor force participation, at 30 percent is much lower despite its HCI among females, or 0.658. Targeting Job Creation: The Shift to Green 33 Figure 3.12.  Female Labor force participation rate (LFPR) versus Human Capital Index, Türkiye (left), and Global LFPR comparisons (right) Female LFPR and female HCI LFPR by gender, global, 2019–2020 100 90 Labor force participation rate, female, 90 80 80 70 70 R2 = 0.1835 60 2019–2020 (%) 60 50 50 % 40 40 30 30 TUR 20 20 10 10 0 0 0.3 0.5 0.7 0.9 1.1 Ce bb s ra n on ia th C ras ed ia M nia G ico Rw ece da ol ly iL a er rk a go e na & ari er Sr dov k nt ea ze iy Ita H l As ac at an vi an ex o du pe C mb M ro re Human Capital Index, 2020 (0 to 1) T ü ro a & e M Eu ric D m H e C Am OE d or an N ia sn Bo tin La Female Male Source: World Bank staff using Turkstat data and (for female labor force comparisons) and World Development Indicators data. The unemployed continue to be women and first-­ time job seekers. Unemployment stands at 13.1 percent as of November 2020, compared to 13.6 percent in November 2019, however, unemployment rates mask, on the one hand, substantive labor force exits and, on the other, job protections provided for the formal sector during COVID. Unemployment previously increased significantly during 2007–2009 from 9.2 percent to 13.1 percent and has essentially remained high since. Unemployment stands at 15 percent among women compared to 12.2 percent among men, and 25 percent among youth. Younger and first-­ time job seekers tend to lack the breadth of skills and networks needed to adapt or transfer between jobs and sectors. Data from the formal sector demonstrates younger workers aged 15–24 years were disproportionately more likely to drop out from the job search compared to workers 25 years or above. In addition, COVID job protections do not apply to workers in the informal sector who, prior to COVID, comprised over thirty percent of the workforce. The lack of employment support to these workers accentuates labor market and welfare segmentation which unemployment rates tend to mask. Females exiting the labor force also continue to cite household responsibilities as the main cause. Exit from the labor force was dominated by household responsibilities accounting for 31 percent and driven by women; 46 percent of women who leave the labor force cited this as the factor, compared to zero percent among men. This represents 9.8 million work-­able women; were these women to work, this would represent an increase of nearly 30 percent of the labor force. Discouragement and retirement showed an increase COVID levels, notably among women. relative to pre-­ 34 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 3.13.  Overall employment levels versus female share, 2014–2020, aggregate (first panel) and by sector (second) Total jobs vs female share 29000 32.0% 28500 31.5% 28000 31.0% 27500 27000 30.5% 26500 30.0% 26000 29.5% 25500 25000 29.0% 24500 28.5% 2014 2015 2016 2017 2018 2019 2020 Total jobs Female % Total employment: share by sector vs share of females 60 56.5 56.2 54.1 54.9 52.2 53.7 Services sector share 51.0 50 46.3 46.1 45.2 44.9 44.4 44.0 40.1 40 Agriculture % female 32.1 33.1 33.2 33.3 Services % female 30.4 31.5 29.0 30 Industry % female 23.3 23.1 23.4 23.7 24.0 24.3 24.1 21.1 Industry sector share 20.6 19.5 19.4 19.7 19.8 20.5 20 20.5 20.0 19.5 19.1 Agriculture sector share 18.4 18.2 17.6 10 7.4 7.2 7.3 7.4 6.9 5.7 Construction sector share 5.5 Construction % female 4.1 3.7 4.2 3.9 4.1 4.3 4.5 0 2014 2015 2016 2017 2018 2019 2020 Source: World Bank staff, TURKSTAT data. As seen elsewhere, the COVID pandemic had led to significant job and labor force term scarring in the absence of targeted participation losses in Türkiye, foster long-­ policy measures. Employment fell nearly by 1.5 million between December 2019 and 202027. Employment rates declined from 63 percent in 2019 to 60 percent in 2020 for men, and from 29 percent in 2019 to 26 percent in 2020 for women. The pandemic was especially hard on the traditionally disadvantaged groups such as the youth, women and informal workers. The employment losses experienced during the pandemic exacerbated a tendency of labor market stagnation in place already after the currency crisis of 2018, as term structural challenges such as high rates of NEET among young well as several long-­ cohorts of labor market entrants, low female labor force participation, and a large pool of inactive and unused labor force potential, particularly among women. 27 Data based on Turkish Statistical Institute (TUIK) Labor Force Survey annual and quarterly monitoring. Targeting Job Creation: The Shift to Green 35 Figure 3.14.  Average wages by type and gender, 2020, in absolute (top panel) and relative terms (second) 140,000 127,908 Average annual wages (Turkish Lira) 120,000 105,767 100,000 80,000 60,000 43,974 37,392 35,255 40,000 19,434 21,516 20,000 10,572 0 Female Male Female Male Female Male Female Male Casual employee Employer Regular employee Self employed Wages, gender gap (percentage females vs males) 10% 0% –10% –20% –30% –40% –50% –60% –70% 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Casual employee Employer Regular employee Self employed Source: World Bank staff calculations, TURKSTAT data. While informality seems to have declined during 2020 and more so for women, this may be due to losses to labor force participation rates among women. For women, the sharpest decline is in accommodation and food services, wholesale and retail, administrative support and manufacturing (between 3 to 7 percentage points decline in informality rates) For men, the sharpest decline is in construction, manufacturing, accommodation and food services, wholesale and retail and arts, entertainment and recreation (between 3 to 10 percentage points decline. The decline in informality during the pandemic could be due to several reasons. On the on hand, the reduced demand for labor in sectors with high informality due to depressed consumer demand may have led to major exit by workers, particularly in services which tend to be less productive. On the other hand, firms may have formalized previously informal workers to benefit from pandemic-­ related subsidies, although the extent that this occurred has yet to be evaluated. 36 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 3.15.  Informality versus formality by gender, age, education and sector, 2020 100% 8% 13% 90% 21% 17% 20% 28% 27% 27% 80% 37% 36% 35% 41% 48% 70% 60% 77% 81% 88% 50% 92% 87% 40% 79% 83% 80% 72% 73% 73% 30% 63% 64% 65% 59% 52% 20% 10% 23% 19% 12% 0% 4 4 e + ry ol s ity re ry e g s) e e n e 65 al al ad –2 –6 io t rin ar ho da st ra tu ic rs m M ct ye du Tr tu te 15 25 rv ul ve sc on Fe tru ric ac Se Li In ni (5 c h ns Se uf Ag U ig ol Co an H ho M sc y ar im Pr Gender Age Education Economic sector Formal Informal Source: World Bank staff calculations based on LFS. Table 3.3.  Informality by sector, 2017–2020, among females (first panel) and males (second) Sector (Females) 2017 2018 2019 2020 Agriculture 94% 91% 96% 94% Mining 0% 0% 2% 3% Electricity, Gas, Water Sewerage 38% 29% 21% 30% Manufacturing 32% 31% 27% 23% Construction 12% 11% 8% 5% Administrative Support 31% 41% 44% 41% Finance, Insurance, Real Estate and Other. 11% 10% 11% 10% Accommodation and Food Services 26% 27% 28% 21% Wholesale, Retail Trade, Transportation and Storage 26% 24% 26% 19% Public Administration, Education, Human Health Ser. 18% 18% 18% 16% Arts, Entertainment and Recreation 30% 24% 25% 24% Other 52% 52% 55% 45% Targeting Job Creation: The Shift to Green 37 Sector (Males) 2017 2018 2019 2020 Agriculture 74% 76% 79% 76% Mining 4% 5% 5% 5% Electricity, Gas, Water Sewerage 35% 29% 29% 28% Manufacturing 15% 17% 17% 14% Construction 37% 35% 39% 36% Administrative Support 5% 7% 8% 8% Finance, Insurance, Real Estate and Other. 13% 16% 17% 16% Accommodation and Food Services 33% 33% 34% 29% Wholesale, Retail Trade, Transportation and Storage 28% 28% 31% 27% Public Administration, Education, Human Health Ser. 4% 3% 3% 2% Arts, Entertainment and Recreation 40% 40% 46% 36% Other 39% 40% 43% 38% Source: World Bank staff calculations based on LFS. Looking ahead, a key challenge lies in supporting informal, low-­ skilled workers who lack skills and social protection. Adaptability to new jobs among the existing work-­ able labor force until the next generation will mean being equipped with foundational skills, adaptability to learn new skills measured by competency, age, and access, which may be hindered by barriers such as cost, awareness and supply of training. Despite modest progress over the past two decades, informal jobs have persistently hovered at one third of all jobs, corresponding to over 9 million workers. Informal jobs are typically low-­productivity jobs in low-­ employment activities and are less likely to be productivity firms and self-­ exposed to R&D and innovation or benefit from technological change. Informal employment reaches 86 percent in the agriculture sector, and is higher among women than men (42 percent versus 31 percent, respectively). Unskilled workers with less than secondary education represent 80 percent of total informal workers, a total of 7.7 million workers. By being “off the grid” and detached from a formal social registry, educational or job search system, these workers are at highest risk of missing out on opportunities to gradually adopt the skills needed or access new green jobs as a result. Furthermore, labor force dropouts remain sizeable and as such a vital target for reskilling need to boost the post-­ COVID and green transition. After accounting for the population in education/training, and for the retired, 28% of the population 15+ is inactive, representing a large unused labor potential, exacerbated during COVID. Being detached from the labor market, the inactive are prone to rapid skills erosion, requiring greater investment and time training the share who exhibit adaptability competencies. for upskilling or re-­ 38 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 3.16.  Informality levels by educational level and gender, 2019 Number of informal workers (2019) 8000 7000 6000 5000 4000 3000 2000 1000 0 Below secondary Secondary Tertiary Men Women Source: World Bank staff calculations based on TUIK LFS. Figure 3.17.  Informal share of total employment versus GDP per capita, global, 2019 100 BDI COD TCD BFA BEN MLI BGD MOZ HTI MDG CMR 90 KHM MRT TGO TZA SEN CIV IND LBR UGA SDN NIC AGO BOL MWI RWA COG HND LAO 80 PAK IDN NER ZWE NPL MMR GTM GMB GHA ZMB PRY 70 SLV PER Share of informal employment LSO VNM BWA EGY LKA COL THA CPV ECU BRB 60 SWZ ALB MEX NAM LBN DOM KGZ JOR GEO PAN 50 BRA ARG ARM 40 MNG ZAF CRI TUR 30 MDA CHL BIH URY 20 R = 0.6568 2 SRB MKD 10 0 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5 11.0 Log GDP PC PPP (2019) Source: World Bank staff calculations using World Bank World Development Indicators and ILOSTAT. Targeting Job Creation: The Shift to Green 39 Determinants of Employment during Shocks The resilience of workers as a result of COVID and other economic shifts will depend on adaptability to evolving jobs, skills and technology. An analysis of the vulnerability of employment28 across sectors in Türkiye shows similar patterns to those seen in other countries29. Most notably, employment is most vulnerable in textile and apparel, food and accommodation, leather, agriculture, and furniture sectors vulnerable to shocks. Interestingly yet not surprisingly given the propensity for independent work, employment in some sectors such as electrical and computer are less vulnerable than the sector, which may be relatively higher in vulnerability. COVID has shown the vulnerability of key sectors to adapt to telecommuting. Out of 32 sectors30, information and communication technology (ICT), finance, and health sectors have the highest potential for working from home, accounting for 10 percent of total employment at best, assuming all workers in these sectors had the skill level, digital competencies and ICT infrastructure needed to work from home as expected from surveys Sectors with the least potential for working from home, as expected, include wood and paper products, rubber and plastic, metal, hospitality (accommodation and food), textile and apparel, agriculture, furniture, wholesale and retail, and food, beverage and tobacco, collectively accounting for nearly 51 percent of all employment. These are mostly the sectors with lower productivity due to several reasons such as low skills and high informality. In addition, while the digital economy is gaining momentum particularly among the services sector, on average only a minority of Turkish adults are digitally-­savvy. 6.9 percent of all Turkish working-­ age adults have a medium-­ level proficiency of problem-­solving skills in technology-­ rich contexts, compared to the OECD average of 24.7 percent, showing similar relative scores for numeracy and literacy31. The low likelihood for most workers to work from home leaves a large share of households vulnerable to income shocks should the COVID pandemic persist. 28 Based on WBG Background note, “Türkiye: Covid-­ 19 and Employment Vulnerability Index by Sector and Potential for Home-­ Based Work”, May 2020, unpublished. Methodology is based on the Employment Vulnerability Index (EVI), which uses the sector vulnerability index, and expands it to investigate which sectors are more likely to face job-­related vulnerabilities inspired from Bazillier et al, 2016, and the Australian Employment Vulnerability Index to the Turkish context. Employment is defined as the number of jobs and vulnerability is an index using the Principal Component Analysis (PCA) method. The index includes nine indicators related to the average economic, financial and sectoral viability of the sector (i.e. sector vulnerability index); average degree of protection of workers (self and unpaid workers, part-­ time and informal employment, overtime work); average level of difficulty for workers to generate income (minimum wage non-­ compliance); average level of education and skills among workers, and amenability to working from home. 29 Amenability to working from home analysis is based on the methodology developed by Hatayama, Viollaz and Winkler (Hatayama, Maho; Viollaz, Mariana; Winkler, Hernan. 2020. Jobs’ Amenability to Working from Home: Evidence from Skills Surveys for 53 Countries. Policy Research Working Paper; No. 9241. World Bank, Washington, DC). The methodology uses OECD PIAAC (Survey of Adult Skills) 2014/2015 wave which has over 5 thousand respondents. Job is less amenable to working from home if it has high physical and manual intensity, requires to-­ significant face-­ face interaction, there is low ICT use at work, and low ICT infrastructure at home. More information will be available in the WBG Background note, “Türkiye: Covid-­ 19 and Employment Vulnerability Index by Sector and Potential for Home-­ Based Work”, May 2020, unpublished. 30 Due to the data constraints, certain sectors are excluded from the analysis such as Services of households as employers. 31 OECD Survey of Adult Skills (PIAAC), 2018. 40 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Te Employment Vulnerability Index (0 to 1) Te Employment Vulnerability Index (0 to 1) xt xt il e, ile 0 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 Ac a Ac ,a co ppa co pp m m are . & rel .& l fo fo o Le d Le od Ag th a Ag th a r e r e Co icu r Co icu r W n ltu W n ltu oo st r oo st r ru e ru e Fo d & c Fo d & c od p Fu tion od p Fu tion O , a r O a th , be per rnit PIAAC dataset are not included. th be per nitu er ve p r er ve p ur no ra rod e no ra rod e n- ge u n- ge u m , t cts m , t cts et ob et ob a a Fa llic acc Fa llic acc br m o b m o Ru ica in Ru rica in Targeting Job Creation: The Shift to Green bb te era bb te era Tr er d m l Tr er d m l M an an e M an an e sp d tal sp d tal Ar oto ts r v o p Ar oto ts r v o p , e eh Wh rt & last , e eh Wh rt & last nt ic ole st ic nt ic ole st ic er le o er le o ta s & sal rag ta s & sal rag in e in e m tra & e m tra & e en n re en n re t a sp tai t a sp tai nd ort l nd ort l Employment Vulnerability Index (0 to 1) El re ve El re ve ec c h ec c h tri B rea . tri B re . t Employment Vulnerability Index (0 to 1) M ca as i i o M ca as atio i ac l an c m n ac l a n c m n hi d e hi d e ne co tal ne co tal ry m s ry m s an pu an pu (first panel) and work-­from-­home amenability (second), Türkiye O d e ter O d e ter th th El er qui El er qui se pt. se pt. Employment vulnerability across sectors ec ec t rv t rv Pr ric ic Pr ric ic of ity e of ity e .a , w Mi s .a , w Mi s dm n i dm n i at n g at n g in er in er & & Employment vulnerability and work from home amenability & & su ga su ga pp H s pp H s or ea or ea Share in total employment (%) t s lth t s lth Pu erv Pu erv bl ice bl ice Amenability to working from home (%) ic ic ad s a s Ed mi Ed dmi uc n uc n at at io io n n I IC Fi CT Fi T na na Figure 3.18.  Employment vulnerability index during COVID by sector versus employment share nc nc e e 4% 8% 2% 6% 5% 0% 0% 14% 18% 12% 16% 15% 10% 10% 30% 25% 20% 20% Source: World Bank staff using Türkiye Labor Force Survey 2018, and PIAAC 2014/2015 for Türkiye. Calculations from Seker, Ozen Share in total employment (%) and Acar (2020). Note: Size of the bubbles show the size of employment in each sector. Sectors with less than 10 observations in the Amenability to working from home (%) 41 The determinants of employment further reveal that the shock created by the pandemic has led to a “she-­ cession” in terms of the jobs’ recession hitting females acutely in Türkiye. Although there were significant labor force losses among Turkish men, a closer look at the composition of inactive individuals by gender, skilled women were much more likely to drop out the labor force in 2020 compared to any other education group by men and women. Labor force participation rates by women aged (25+) with college degrees fell from 71 percent in 2019 to a staggering 65 percent in 2020. Unemployment rates for women seem to have declined in 2020, but only because many women dropped out of the labor force in 2020. The determinants of employment and labor force participation rates show that losses have been driven by key sectors, differentiated by gender. The loss in skilled female based classification). Transition labor is evident in the analysis by occupation groups (ISCO-­ into inactivity in 2020 was strongest for women in professional occupations, among technicians and clerical support (reference category: elementary occupations). Looking at men who dropped out of labor force in 2020 but who were employed in 2019, 34 percent were employed in wholesale, retail, transportation and storage, accommodation and food services while 18 percent were employed in construction and 19 percent were employed in manufacturing. As for women, 27 percent of the women who are inactive in 2020 but had a job in 2019 were employed in wholesale, retail, transportation and storage, accommodation and food services, while 18 percent were employed in public administration, education and human health services and 20 percent were employed in manufacturing. Figure 3.19.  Probability of inactivity (left) and unemployment (right) by gender and education, 2020 Change in probability of inactivity Change in probability of unemployment 0.02 0.10 0.01 0.05 0 0 –0.05 –0.01 –0.10 –0.02 Primary High Vocational Tertiary Primary High Vocational Tertiary school school school degree school school school degree Male Female Male Female Source: World Bank staff calculations based on a linear regression. Education categories are interacted with year 2020. The reference category is individuals with no degree. 42 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 3.20.  Probability of inactivity by gender as a function of past occupation, 2020 Change in probability of inactivity by past occupation, 2020 (ref. group = elementary occupations) Managers Professionals Technicians Clerical support Services and sales Skilled agricultural Craft and artisans Machine operators –0.10 –0.05 0 0.05 0.10 0.15 Male Female Source: World Bank staff calculations based on a linear regression. Occupations that individuals reported that they were working at year ago are interacted with year 2020. The reference category is elementary occupations. Figure 3.21.  Probability of inactivity as a function of past sector among females (first panel) and males (second panel), 2014–19 versus 2020 Probability of inactivity by sector last year, females (ref. cat = manufac. & mining) Agriculture Construction Wholesale, transport., accommodation Information and communication Financial and insurance actives Real estate activities Administrative and support services Public administration and defense, education, health Other –0.15 –0.10 –0.05 0 0.05 0.10 0.15 2014–2019 2020 Probability of inactivity by sector last year, males (ref. cat = manufac. & mining) Agriculture Construction Wholesale, transport., accommodation Information and communication Financial and insurance actives Real estate activities Administrative and support services Public administration and defense, education, health Other –0.05 0 0.05 2014–2019 2020 Source: World Bank staff calculations based on a linear regression. Occupations that individuals reported that they were working at year ago are interacted with year 2020. The reference category is manufacturing and mining. Targeting Job Creation: The Shift to Green 43 Table 3.4.  Reasons for job quits by gender, 2020 Reason cited Female Male Job was temporary, came to an end 41% 46% Was working seasonally 9% 4% Dismissed/Liquidated/Bankrupted 16% 21% Not satisfied with job 10% 9% Own illness or disability 5% 3% Looking after children or incapacitated adults in the family 5% 0% Her spouse’s request / Due to marriage 1% 0% Education or training 2% 3% Retirement (including early retirement) 1% 3% Military service 0% 1% Other 11% 10% Source: World Bank staff calculations based on monthly LFS data. Examining the Potential for Green Jobs The pace of job creation in increasingly green sectors and occupations will need to significantly increase in Türkiye to keep up with advanced economies. Over half of greenhouse gas emissions (GHGs) are dominated by five key sub-­ sectors. Electricity/ gas, air transport, manufacturing (metallic and non-­ metallic products), and land transport collectively account for less than 10 percent of employment but over half of total GHGs. At the same time, 75 percent of Türkiye’s jobs today found in medium-­and high-­ emission sub-­sectors32,33. A closer look at categories of GHG emissions and net aggregate jobs over time34 shows that while employment in high-­ emissions sectors has decreased over the past decade from 33 percent to 25 percent, the share in low emissions sectors has remained persistently low at 23–25 percent. 32 Methodological note: In order to correlate more directly job creation within and across sectors with indicators of pollution by GHG, the methodology combines sector-­ level information on GHGs emission intensity (defined as total GHG emissions in the sector over the value of production in the sector) obtained from the Global Trade Analysis Project (GTAP) database (https://www.gtap.agecon.purdue.edu/resources/res_display.asp?RecordID=5537 ), with sector level employment data from the Turkish labor force survey. In GTAP, sector level information of GHG emissions, is available in some cases at quite disaggregate levels (e.g., NACE-­ 3 digit), especially in manufacturing and in sub-­sectors responsible for large shares of GHG emissions. However, in other cases, the information is available only at very aggregate level: for instance, for the wholesale and trade sector, emissions data are not further broken down, and are actually available only at the NACE-­ 1 digit level. For these reasons, the methodology used here constructs a mapping between GTAP sectors and Turkish LFS NACE 2 -­ digit sectors, identifying 45 sectors (with intermediate aggregation between NACE 1 and 2 digits) relevant for our analysis. 33 For detailed methodology, see: Makovec, M and Garrote Sanchez D (2021). Green Jobs and Green Skills in Europe and Central Asia, mimeo; Vona, F, Marin G, Consoli D and Popp D (2018). Environmental Regulation and Green Skills: An Empirical Exploration. Journal of the Association of Environmental and Resource Economists Volume 5, Number 4; Vona, Francesco, Giovanni Marin, Davide Consoli, and David Popp (2015) Green Skills, NBER Working Paper 21116. 34 The methodology identifies the top 15 sectors for emission density as “high emission sectors”, the bottom 15 as “low emission sectors”, and the mid 15 as intermediate emission sectors. 44 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 3.22.  Demand for labor by total annual sectoral GHG emissions, 2019–2020 Sectoral employment and GHG Emission intensity 800 Electricity, gas, steam, air conditioning 700 600 GHG emission intensity (per production) 500 400 Air transport 300 Manufacturing of non-metallic min prod 200 Manufacturing of Crop and animal metallic prod production Wholesale and retail 100 Land transport trade 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Share of employment (%) Source: World Banks staff calculations based on TUIK LFS and GHG emissions data, Global Trade Analysis Project. Calculations from Makovec, M and Garrote Sanchez D (2021). Green Jobs and Green Skills in Europe and Central Asia, mimeo. Over the past decade, Türkiye has had limited job creation in low-­ emissions sectors due to several factors, among which are human capital and firms’ demand for capital versus labor. The cost of labor, for which Türkiye is generally somewhat below average in terms of competitiveness relative to OECD averages, is multi-­ faceted and addressed elsewhere35. Human capital attainment comprises three key dimensions: the years of education, the type of skills (irrespective of educational attainment) and the quality (level) of skills. To begin with, over time, the demand for education across sectors by GHG levels shows that high-­ emissions sectors tend to comprise the lowest levels of highly educated (tertiary education) workers, with low-­ emissions sectors comprising the highest share, typically in technology and professional services. 35 World Bank (forthcoming). Public Financial Review, Human Capital Expenditures. Targeting Job Creation: The Shift to Green 45 Figure 3.23.  Distribution of jobs by gross GHG emission category, 2010–2020 Share of employment by sectoral levels of GHGS emissions 100 Share of total employment 80 33 33 32 32 29 28 27 27 26 26 25 60 45 45 45 45 48 49 49 43 43 44 44 40 20 23 24 24 25 26 27 28 28 26 25 25 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Low emissions (bottom 15 sectors) Mid emissions (mid 15 sectors) High emissions (top 15 sectors) Source: World Banks staff calculations based on TUIK LFS and GHG emissions data, Global Trade Analysis Project. Calculations from Makovec, M and Garrote Sanchez D (2021). Green Jobs and Green Skills in Europe and Central Asia, mimeo. Figure 3.24.  Net change in job creation by gross GHG emission category, 2010–2020 relative to base year 2010 Changes in employment by sectoral levels of GHGS emissions (2010 = 100) 160 150 140 130 120 110 100 90 80 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Low emissions (bottom 15 sectors) Mid emissions (mid 15 sectors) High emissions (top 15 sectors) Source: World Banks staff calculations based on TUIK LFS and GHG emissions data, Global Trade Analysis Project. Calculations from Makovec, M and Garrote Sanchez D (2021). Green Jobs and Green Skills in Europe and Central Asia, mimeo. 46 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Key Implications Targeted policies will be needed to even out the jobs recovery over the next phase of COVID, coupled with an advancing green, digital transformation in Türkiye. Learning from COVID-­ related jobs measures in Türkiye and elsewhere, a better balance is needed between supporting workers already well-­ integrated in the formal sector and those COVID include a latent demand for labor in excluded due to multiple barriers. Barriers post-­ key sectors, uneven job-­skills matching and lagging female youth labor force participation among both vocational students and the highly killed exiting for household reasons. The post-­ COVID reality, with a greater reliance on technology and green sectors in light of Türkiye’s vision to boost to climate investments, will require an integrated demand-­ and supply-­ side approach to promote new jobs. Moving forward, measures would include: (i) expanding temporary income support during a transition period conditional upon active job search in key sectors; (ii) expanding eligibility and financing for skills upgrade by expanding on-­ job training that is firm (demand) driven; (iii) broadening the-­ job matching mechanisms and counseling to reach informal sector workers; and (iv) linking these measures to finance for labor-­ intensive firms, including occupations boosted during COVID in manufacturing, construction, professional services, logistics, health care equipment manufacturing, and renewable energy. Over the next phase, more structural reforms to barriers to the demand for labor and job search will mean tackling Türkiye’s policies regarding labor regulations in terms of sectors restricting female participation; flexible work arrangements in terms of remote support and technology solutions; expansive wage subsidies that have tended to benefit selected firms and retaining formal workers; social insurance benefits across the spectrum and linkages to severance pay policies that may impede hiring; and finally, optimizing minimum wage policy and labor costs in Türkiye relative to OECD and comparable countries. Overall, the time is ripe for Türkiye to launch a renewed national social dialogue policy reform program and lay the groundwork for an equitable COVID transition to green, digital jobs. Targeting Job Creation: The Shift to Green 47 48 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs 4 Modernizing Skills and Job Training: Redirecting Investments Introduction Trends emerging from uneven employment gains during COVID reveal a greater need to stimulate the demand for labor and increase the inclusion of women and vulnerable youth not in employment, education or training (NEET). Given the drivers of employment losses in key sectors such as professional services, manufacturing and industry, leveling the playing field will depend on the extent to which investments favor labor-­ intensive technologies for stimulating job growth. In addition, labor policies will play an important role in protecting and promoting job growth, as evident during COVID for selected workers. This section probes the demand for labor and measures to stimulate an even recovery over the next phase, marked by green, digital transformation. Youth Inclusion Youth NEET has been increasing over time in Türkiye, with determinants revealing different drivers depending on age group and educational status. Due to past education reforms, more and more young women graduate from high school and tertiary education. The share of female NEET (15–24) who reported that they were in education one year ago was 10 percent in 2014 whereas the same share was 18 percent in 2019 and 22 percent in Modernizing Skills and Job Training: Redirecting Investments 49 2020. The share of female NEET who reported that they were busy with household chores a year ago was 74 percent in 2014 whereas the same figure is 54 percent in 2020. We see similar trends for individuals in NEET status who are aged 15–29. In terms of reasons for NEET, “being a housewife” is listed as the main reason of inactivity for female NEET but its share declined from 69 percent in 2014 to 49 percent in 2020. By contrast, the share of female NEET who listed “transition to education” as a reason for inactivity increased from 4 percent in 2014 to 13 percent in 2020. The probability of being in NEET status appear to be significantly higher for women and men with vocational and tertiary degrees in comparison to other education categories as of 2020 for the 15–24 cohort. As for 15–29 cohorts, the probability of being in NEET status was higher for women with vocational and tertiary degrees but not for men in 2020. Taking a closer look at the individuals with tertiary degrees who are in NEET status, graduates of social sciences and languages seem to be particularly at risk of NEET in comparison to engineering degrees. Figure 4.1.  Youth not in education, employment or training (NEET), ages 15–24, by gender (first panel) and education (second), 2014–2020 NEET (ages 15–24) 40% 30% 20% 10% 0% 2014 2015 2016 2017 2018 2019 2020 Male Female All NEET (ages 15–24) 100% 80% 60% 40% 20% 0% 2014 2015 2016 2017 2018 2019 2020 Illiterate Less than high school High school University Source: World Bank staff calculations based on LFS. 50 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 4.2.  Probability of NEET by education (first panel) and among post-­secondary students by educational background (second), 2020 Change in probability of NEET status (15–24) in 2020 0.10 0.05 0 –0.05 –0.10 Primary school High school Vocational school Tertiary degree Male Female Probability of NEET by field of study in comparison to engineering Education Arts Humanities Languages Social sciences Journalism Business Law Biology Physics Maths and stat ICT Manufacturing Architecture Agriculture Veterinary Health –0.1 0 0.1 0.2 0.3 Male Female Source: World Bank staff calculations based on LFS. Modernizing Skills and Job Training: Redirecting Investments 51 Education Investments over Time Education spending over time shows strengths in terms of basic literacy and challenges in terms of twenty-­ first century competitiveness for a new economy. Public expenditure on education as percent of GDP has largely been stable in Türkiye over the past decade (2011–2019), reaching approximately 4.3 to 4.6 percent of GDP as of 2020. Although spending levels are not always directly comparable, this level is lower than the OECD average of approximately 5.6 percent and selected smaller countries in terms of geographic territory and population such as Costa Rica at 7.4 percent and Tunisia at 6.6 percent. Nearly 80 percent of spending finances wages and benefits of the teacher workforce, with the remainder spent on capital and other current expenditures. In terms of expenditure components, emphasis on secondary education has been decreasing while that on tertiary education has been increasing since 2013, although most OECD countries have seen the reverse trend. Further, in line with comparable countries that have increased investment in early childhood education significantly, Türkiye has gradually started to boost resources in this area to avoid lagging behind.. In addition, for its level of secondary education spending, noting that the relationship between expenditure and PISA scores is complex, PISA scores appear to be average; however, other comparable countries appear to have higher PISA scores for the same or less secondary education expenditure. This trend shows that factors beyond the level of spending on secondary school and beyond the scope of this analysis likely play a critical role in maximizing the returns to investment, such as organizational, human resource and curriculum design. Figure 4.3.  Türkiye: Total public expenditure on education as share of GDP, 2011–2019 Total public EDU EXP as % GDP 6.0% 5.0% 4.0% Percent of GDP 3.0% 2.0% 1.0% 0% 2011 2012 2013 2014 2015 2016 2017 2018 2019 Source: World Bank staff, TURKSTAT and Ministry of National Education data. 52 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 4.4.  Figure 35 Public expenditure by educational level, OECD countries, 2013 (first panel) versus 2017 (second) 2013 10 9 8 7 Share of GDP (%) 6 5 4 3 2 1 0 Tü ania e* G ly Sw way N Au m m e ia N and nm n Fi ark Lu Be nd m m te C urg ng s rla a Fr ds Eu Sw sto e ro tz nia er d Po ge ov d er ia M y La ta Se ia Cz rbia rtu a ov l Sp a Bu ain th ia Sl ga Ki ru an he tri Po chi i De ede Ro ec E c Av lan Sl an Ita G en tv ak Li gar an do xe lgiu al iy n an a a d yp et s bo m re rk el nl l or e u pe er l Ic i ni U 2017 9 8 7 Share of GDP (%) 6 5 4 3 2 1 0 or d Fi ark Lu Be nd m m te C urg ng s N Au m rla a Fr ds Eu Sw sto e ro tz nia er d Po ge ov d er ia M y ov l th ia Tü ania Sw way nm n La ta Se ia rtu a Sp a Bu ain e* G ly m e ia Cz rbia Sl ga Ki ru an he tri Po chi i De ede E c Ro ec an Av lan Sl an Ita G en tv ak Li gar an xe lgiu do al iy n an a a d yp et s bo m re rk el nl l e u pe er l Ic N i ni U Tertiary education (levels 5–8) Upper secondary and post-secondary non-tertiary education-vocational (levels 35 and 45) Upper secondary and post-secondary non-tertiary education (levels 3 and 4) Primary and lower secondary education (levels 1 and 2) Early childhood education Source: World Bank staff, OECD Education data, TURKSTAT and Ministry of National Education data. Note: * refers to staff country comparability used for purposes of this analysis. calculations for cross-­ Modernizing Skills and Job Training: Redirecting Investments 53 Figure 4.5.  Relative change in education spending by level, selected countries, 2013 versus 2017 Relative percentage change from 2013 to 2017 by component, Türkiye and selected countries 200 150 Relative percentage change 100 50 0 –50 –100 Türkiye* Estonia Poland Europe average Lithuania Early childhood education Primary and lower secondary education (levels 1 and 2) Upper secondary and post-secondary non-tertiary education (levels 3 and 4) Upper secondary and post-secondary non-tertiary education vocational (levels 35 and 45) Tertiary education (levels 5–8) Total Source: World Bank staff, OECD Education data, TURKSTAT and Ministry of National Education data. While generally maintaining the level of public expenditures, private expenditure by households on education has doubled over the past two decades since 200236. This pattern may reflect changes in preferences and/or the relatively lower public expenditure on secondary education as tertiary education has absorbed more resources over the same period. While education has comprised a modest part of household consumption expenditure, the level has been increasing over time, compared to health which has accounted the same share over time, at approximately 2.2 percent as of 2019 (most recent of-­ available data at the time of writing). By contrast, out-­ pocket expenditure on education has gone from 1.3 to 2.5 percent between 2002 and 2019, driven by higher-­ income households. The highest-­ income households have gone from spending 2.2 percent of total household expenditure on education to 4.2 percent, while the lowest-­ income households have hovered at 0.1 to 0.2 percent of total expenditure. The socioeconomic gradient in out-­ pocket education spending is also greater than that of health. By 2019, while the lowest-­ of-­ income households spent nearly half on health as that of their wealthest counterparts, they spent only 3 percent of what the wealthiest spent on education, primarily secondary education. These differences may partially explain the socioeconomic differences in learning outcomes such as PISA in Türkiye and its higher HCI-­ health outcomes as compared education relative to comparable countries. to HCI-­ 36 TURKSTAT Household Budget Survey, 2011–2019. 54 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 4.6.  Secondary education expenditure versus PISA learning outcomes for Reading, Science and Math Secondary EXP and PISA reading Secondary EXP and PISA science 3.0 3.0 BRN NOR BRN 2.5 CYP FIN 2.5 CRI CYP AUT NLD FIN Secondary public as % GDP Secondary public as % GDP NOR ISL NLD NZL NZL CRI ARG TUR AUT MDA SWE 2.0 ARG KAZ ISL GBR 2.0 KAZ MDA SVN AUS MYS ISR GBR NLD MYS HUN LVA KOR TUR ESP ISR SVK MEX CHL KOR LTU EST COL COL AUR POL 1.5 POL CZE UKR 1.5 SVK JPN EST IRI PER ROU MEX JPN JOR PER JOR LTU EST ROU UKR CZE HKG 1.0 CHL HKG 1.0 ALB LVA ALB 0.5 0.5 R2 = 0.0329 R2 = 0.0004 0.0 0.0 300 350 400 450 500 550 600 300 350 400 450 500 550 600 Reading PISA Science PISA Secondary EXP and PISA math 3.0 CYP 2.5 FIN Secondary public as % GDP NLD BRN AUT SWE NZL 2.0 KAZ TUR GBR MDA MEX HUN LVA LTU CZE PSL ISR MYS POL SVK 1.5 KOR SVN COL ESP CHL UKR AUS HKG JOR 1.0 EST PER IRL ROU ALB 0.5 R2 = 0.0237 0.0 300 350 400 450 500 550 600 Math PISA Source: World Bank staff calculations, World Development Indicators and OECD PISA Scores, 2018. Modernizing Skills and Job Training: Redirecting Investments 55 56 Figure 4.7.  Public and private education expenditures, 2011–2019, and by household quintile 2002–2019, Türkiye Public education EXP Private education EXP 1.80 0.35 General upper 1.60 secondary, 0.30% 0.30 Lower secondary, 1.40 0.29% Tertiary, 1.29% 0.25 Tertiary, 0.27% 1.20 Primary, 0.21% 1.00 Lower secondary, 0.20 0.82% 0.80 Primary, 0.79% 0.15 EXP as % of GDP EXP as % of GDP Vocational and 0.60 technical upper Vocational and secondary, 0.66% 0.10 technical upper 0.40 General upper secondary, 0.10% secondary, 0.40% 0.05 0.20 Pre-primary, Pre-primary, 0.04% 0.29% 0 0 1 12 13 14 15 16 17 18 19 11 12 13 14 15 16 17 18 19 201 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 4.5 4.2 4.0 3.5 3.2 3.0 2.5 2.5 2.2 2.0 2.0 1.5 1.3 1.0 0.5 0.4 0.2 Share of household expenditure (%) 0.1 0 2002 2010 2019 Total Quintile 1 (poorest 20%) Quintile 2 Quintile 3 Quintile 4 Quintile 5 (richest 20%) Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Source: World Bank staff calculations, TURKSTAT Household Budget Survey. Figure 4.8.  Evolution of household expenditures by type as share (first panel) and wealth gap (second), 2002 versus 2019 Share of household expenditure components (%), 2002 vs 2019 30 25 20 15 10 5 0 ge - cc s, w nd nt rv m s lth n n ltu nt ic l te t ic d ra on se ho se rv na tio io ho an ba e rv o re cu e s o r ic e re es ls es ea ot a to ag ea at se go d inm ve d n re nd ou se atio d ur rta fo ing d es H ic d er an sta an ca a , h d s be an po un an rta uc an ev an iou h e ot Re g ns m lic d te nc ur Ed b sin r Cl ho oo m Va a En ia it tte ic Tr pl urn Co e ou re ol co F ga oh H ap F ci lc 2002 2019 A al Relative difference in consumption poorest vs wealthiest, 2002 vs 2019 150% 133% 134% 120% 100% 77% 46% 40% 50% 4% 0% –50% –40% –41%–37% –24% –39% –43% –54%-51% –100% –68% –71% –72% –66% –79% –80% –84% –97% –95% –150% ge - cc s, w nd nt rv m s lth n n ltu nt ic l te t ic d ra on se ho se rv na tio io ho an ba e rv o re cu e s o r ic e re es ls es ea ot a to ag ea at se go d inm ve d n re nd ou se atio d ur ta fo ing d es H ic d er or an sta an ca e a e, h d s be an un an ta uc an ev an iou h sp ot Re r g m lic d te nc ur Ed b an sin r Cl ho Foo m Va En ia it tte ic Tr pl urn Co ou re ol ga oh H ap F ci lc co A al 2002 2019 Source: World Bank staff calculations, TURKSTAT Household Budget Survey. Infrastructure, Coverage and Enrolment Basic education is predominately public with broad nationwide outreach, with only 4 percent of all primary students in Türkiye attending private schools. The public teacher workforce is significant at one million, covering nearly 18 million students from early education to secondary school, with teacher-­ student ratios generally within a similar margin across province. Primary education is wide covering at a 94 percent schooling rate, while secondary education hovers at 85 percent with wide regional variation ranging from 72 percent to 92 percent37. Lower ratios are partially explained by somewhat lower female secondary schooling ratios. In Southeastern Anatolia and Middle east Anatolia regions, girls’ secondary enrollment was 4 percent and 2 percent lower than that of boys, respectively. Lower secondary schooling rates are also generally lower in regions with higher secondary student-­to-­teacher ratios. Secondary ratios range from 11 to 15 students per teacher nationwide for general secondary education and 8 to 13 students per teacher for 37 TURKSTAT Education Indicators, as of 2019. Most recent data at the time of writing. Modernizing Skills and Job Training: Redirecting Investments 57 technical and vocational secondary education. While this illustrative analysis does not clearly show a strong association between student-­ teacher ratios and PISA outcomes, it suggests that other investments and the nature of spending, such as teacher effectiveness, performance incentives and school management, play a more important role at this stage of Türkiye’s skills path. Given student-­ teacher ratios and PISA outcomes by region to-­ described earlier at the macro level, further analysis is needed to determine the role of different factors in influencing learning at the micro level at this stage of Türkiye’s skills path, such as teacher effectiveness, performance incentives and school management. Figure 4.9.  Schooling rates and Teacher coverage by educational level and regional zones, 2019, Türkiye Schooling rate, primary (%) Student: Teacher ratio, primary (number) 100 94 92 92 92 94 95 93 94 95 93 94 94 92 25 20 21 80 20 17 17 17 15 15 15 16 14 14 14 14 60 15 40 10 20 5 0 0 M niz Do ara ul Do ti A ara u z Ba Ist e yd An ye do Ana u An lu Ak lu ra u An ra Do ara ul M niz iz u Do i A ara rta u A olu Ba Ist e z yd An ye An lu u ra u An a M niz ni ni ol Eg ar l Ka dol ol Eg ol Ka dol n K nb o O ogu ado gu do K nb gu do a ki ki de de e gu de de gu de ad ad ad d ad Ba arm O arm Ba arm O arm Ku do Tür Ku do Tür Ba k d a gu na a a do na gu na A M rta ze gu rta ze gu ti ti t g ti ti Ba o rta ey ey ün ün O G G Schooling rate, secondary (%) Student: Teacher ratio, general secondary (number) 100 89 89 91 91 91 92 16 15 85 86 88 88 14 14 14 13 13 13 13 80 72 73 74 12 12 12 11 11 11 11 60 10 8 40 6 20 4 2 0 0 M niz Do ara ul Do ti A ara Do ara ul Do i A ara u Ba Ist e z iz z u Ba Ist e yd An ye do Ana u An lu rta u A olu An lu ra u yd An ye Ak lu An ra u An a ra u M niz M niz ni ni ar ol Eg ol Eg l Ka dol ol Ka dol n K nb K nb O ogu ado o gu do gu do a ki ki de de e de gu de gu de ad ad d ad ad Ba arm Ba arm O arm O arm Ku do Tür Ku do Tür Ba k d a a a do na gu na gu na A M rta rta ze gu ze gu ti ti t g ti ti Ba o ey rta ey ün ün O G G Schooling rate, secondary, female:male ratio (%) Student: Teacher ratio, technical and vocational secondary (number) 5.0 4.2 14 13 12 11 11 3.0 10 10 10 10 10 9 9 9 9 8 8 1.0 0.4 0.7 0.5 0.6 0.1 0.2 8 –1.0 –0.4 –0.5 –0.4 –0.4 6 –1.8 4 –3.0 2 –5.0 –4.0 0 Do ara ul iz u Ba Ist e Do i A ara z An lu yd An ye do Ana u An lu Ak lu M niz An ra u Ba Ist e Do ara ul M niz Do ti A ara ra u z yd An ye rta u A olu u ra u An a M niz ni ni ar ol Eg l ol Eg Ka dol ol Ka dol n o K nb K nb gu do O ogu ado gu do a ki ki e de de gu de de gu de ad ad ad d ad B a ar m O arm Ba arm O arm Ku do Tür Ku do Tür Ba k d a a gu na do na a gu na A M ze g u rta rta ze gu ti ti t g ti ti Ba o ey rta ey ün ün O G G Source: World Bank staff calculations, TURKSTAT data. 58 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Learning and Distributional Outcomes While Türkiye’s overall HCI prior to COVID was on par with most middle-­ income countries, challenges were already apparent prior to COVID regarding skills. Based on PISA38 2018, Türkiye, though lagging the OECD average, was one of the top improvers, having increased reading scores relative to 2015 by 37 PISA points, roughly equivalent to a year of schooling; yet significant losses to learning accrue with each month that passes. While the digital economy is gaining momentum particularly among the services sector, savvy; by way of comparison, on average only a minority of Turkish adults are digitally-­ 6.9 percent have a medium-­ level proficiency of problem-­ rich solving skills in technology-­ contexts, compared to the OECD average of 24.7 percent, showing similar relative scores for numeracy and literacy39. These patterns leave most workers vulnerable to exclusion from digitally based learning and jobs over the near-­ term. Regional disparities in learning outcomes have grown since 2006, closely correlated with regional GDP and public investment. PISA scores show a gap of over 60 points between the best-­performing regions and worst-­ performing regions in Türkiye for Science, a pattern repeated in Math and Reading scores. This gap is the equivalent of low-­ income and high-­income scores globally. Socioeconomic differences are also evident in Türkiye, where the wealthiest quartile of the income distribution of households outperforms the lowest-­income quartile. While similar socioeconomic differences are seen throughout the OECD, what is striking in Türkiye is that the relative gap had been narrowing until 2015 and increased again to 2012 levels by 2018. These trends indicate that learning inequalities are sensitive to economic shocks, such as that witnessed in Türkiye since 2016 and evident in regional growth disparities. Figure 4.10.  General adult digital skills ranking, OECD+ countries, 2018 Percentage of adults scoring level 2 in problem solving in technology-rich environments (ICT proficiency, levels 1–3) (%) 40 34 34 35 35 35 32 32 33 28 29 29 29 29 29 30 25 26 26 27 27 24 25 25 25 22 23 23 20 20 22 20 15 15 16 15 12 12 10 9 7 4 5 5 0 ve n 2 P r Tü eru M iye G ico Ka C ce kh ile Po tan th nd Fe I nia ra el ak Ire nia pu d Es blic N Sta ecd un nia e y ni Ir 2 e St d 4 20 ) pu n Ko lic rs Au rea el ria gl rm ) Si d ( y ap ) Ca ore s a nm lia ew in k N Ze and rla d o s ed y en h Ja 17 es K En Ge ium ng UK do N nd th s v ar an an Sw rwa N F ar Au nad U rn 201 rag Re pa d an 01 Re lan he an de sra at (U Sl tio b De tra za h e (B st la rk ex ua d O H to or te A g s ua te el /2 re l et al g o Ec L i ec ov de n Cz ia Sl e an ss Fl Ru te ni U Source: OECD Survey of Adult Skills (PIAAC) 2018 data. Similar trends for numeracy and literacy. 38 OECD (2019). Programme for International Student Assessment, Results for 2015 and 2018. 39 OECD Survey of Adult Skills (PIACC), 2018. Modernizing Skills and Job Training: Redirecting Investments 59 Figure 4.11.  Secondary school learning outcomes (PISA) by theme, regional zones in Türkiye (2006–2018), and by household income, selected countries (2009–2018) 530 PISA science by region, 2006–2018 510 490 470 450 430 410 390 370 350 2018 2015 2012 2009 2006 Bati Marmara Doğu Marmara Bati Anadolu Istanbul Bati Karadeniz Ege Akdeniz Türkiye Doğu Karadeniz Orta Anadolu Güneydoğu Anadolu Kuzeydoğu Anadolu Ortadoğu Anadolu 140 (top 25% vs bottom 25% of hh income quartiles) Socioeconomic gap PISA in reading score 120 100 80 60 40 20 0 Estonia Hungary Lithuania Poland Türkiye OECD Argentina average-35a PISA 2009 PISA 2012 PISA 2015 PISA 2018 Source: OECD PISA Survey, 2006–2018. Regional differences for PISA for Math and Reading are similar to Science. Socioeconomic breakdown available for PISA Reading for 2009–2018. These trends indicate that learning inequalities are sensitive to economic shocks, such as that witnessed in Türkiye since 2016 and evident in regional growth disparities. An illustrative analysis of the determinants of PISA scores across regions shows that basic education infrastructure (number of teachers, schools, and associated student-­ resource ratios) explains only a modest part of learning outcomes. Despite limited data and a lack of disaggregated data at the province or school level, PISA scores (2018) in Türkiye are closely correlated with economic activity and the size of school, a proxy for population density and potentially the availability of specialized resources such as advanced teacher training, digital skills development and curriculum. The level and nature of public investments therefore influences learning outcomes in a synergistic way. 60 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 4.12.  Stylistic directional (positive versus negative) associations between economic investments and PISA Science, Math and Reading scores, illustrative regression, Türkiye, 2018–2019 Coefficients P-value Directional associations, PISA science 100 50 Percent 0 –50 –100 Exports by economic Total electricity Total number of hospital Number of students per Workforce status of the activities (ISIC, Rev.4) consumption beds per one hundred general secondary non-institutional population (1000 USD) general per capita (KWh) thousand people education teacher [5+ years]: Non-agricultural total per capita unemployment rate (%) Coefficients P-value Directional associations, PISA math 100 80 Percent 60 40 20 0 Exports by economic Total electricity Total number of hospital Number of students per Workforce status of the activities (ISIC, Rev.4) consumption beds per one hundred general secondary non-institutional population (1000 USD) general per capita (KWh) thousand people education teacher [5+ years]: Non-agricultural total per capita unemployment rate (%) Coefficients P-value Directional associations, PISA reading 100 50 Percent 0 –50 –100 Exports by economic Total electricity Total number of hospital Number of students per Workforce status of the activities (ISIC, Rev.4) consumption beds per one hundred general secondary non-institutional population (1000 USD) general per capita (KWh) thousand people education teacher [5+ years]: Non-agricultural total per capita unemployment rate (%) Source: World Bank staff calculations using TURKSTAT data and OECD PISA Survey, 2018. Based on limited sample of sub-­ national PISA scores (at regional zone level) and associated available economic and public investment indicators (n=12). Above horizontal axis = positive association; below horizontal axis = negative association. School-­to-­Work Transition and Returns to Education As an indicator of investments in the quality of secondary education and the school-­ work transition, despite advances, there remain gaps in early career orientation, to-­ particularly in light of a rapdily changing global economic context. Türkiye relies mostly on existing teachers to facilitate school-­ work opportunities and orient students to to-­ careers, through a system of in-­ school career guidance and research centers particularly for special needs students.40 Routine evaluations of the returns to work and the quality of these services would be useful for future analysis. However, using internationally-­ comparable time career counselors), indictors of exclusive services (i.e., offered by dedicated, full-­ only 30 percent of schools in Türkiye report the availabilty of career guidance couselors, 40 The Ministry of National Education has developed a range of career guidance evaluations and has embedded career development within the national education strategy, Education Vision 2030. Modernizing Skills and Job Training: Redirecting Investments 61 Figure 4.13.  School career counseling coverage by type and school socioeconomic level, selected countries, 2018 Career guidance resources 100 Percent of students receiving career 90 80 70 guidance (%) 60 50 40 30 20 10 0 Estonia Poland OECD average Mexico Türkiye Teachers for career guidance - all students Career guidance counselor - all students By school’s socio-economic profile (ESCS) career guidance counselor - disadvantage schools By school’s socio-economic profile (ESCS) career guidance counselor - advantage schools Source: World Bank staff, OECD PISA, 2018. dominated by schools in higher-­ income provinces. Türkiye is an outlier among OECD countries in terms of dedicated job counseling in secondary schools. Its level compares to half that of the OECD average, or 64 percent of schools. Based on a detailed analysis using international methods41, the rate of returns to general secondary education may be lower than that of vocational and technical education or tertiary education in Türkiye, especially for females. While higher education generally reaps higher returns to education in terms of wages at a global level, the effects are more pronounced in Türkiye. Most recently, when considering the cost-­ benefit of public expenditures with respect to years of schooling, demographics and labor market earnings, the social returns are also lowest for general secondary education. In Türkiye, the private rate of return is estimated at 16 percent for higher education, with a social return of 10 percent, with an overall average rate of return of 8.8, just above the global average. When controlling for having children younger than 15 years, the returns to education for females are higher than those for males. Contrary to patterns seen in other comparable countries, the private returns to those working in the public sector are higher than those in the private sector in Türkiye. Based on this illustrative analysis, results may imply that job prospects for general secondary education graduates are more limited than for vocational training, which is often demanded by more productive sectors in Türkiye such as manufacturing and industry. At the same time, these outcomes may also be driven by broader economic shifts impacting the demand for labor. 41 See Patrinos et al (2021). Private and Social Returns to Investment in Education: the Case of Turkey with Alternative Methods. Applied Economics 53(14): 1638–1658. 62 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 4.14.  Rate of return to education by educational level and scope Earnings by age and educational level Rate and return to public investment in education 18000 18 15.8 16000 16 14000 14 13.0 12000 12 11.4 10.4 10000 10 8000 8 6.0 6000 6 5.2 4000 4 2000 2 0 0 Private rate of return (%) Social rate of return (%) 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 57 61 63 65 Primary Secondary Higher None Primary Secondary Higher Source: Patrinos et al (2021). While COVID has potentially hampered returns to education as a result of learning losses anticipated with long-­ term school closures, Türkiye has responded by expanding virtual digital education systems. Learning losses among adults, youth and children are expected to retard recovery unless targeted measures are taken to safeguard skills. On average, for example, one year of school closures in Türkiye has been estimated to lead to a 40-­ point decrease in reading and mathematics PISA scores, reversing recent gains and harming future productivity42. The effects are expected to be higher for vulnerable low-­ income households and non-­ native Turkish speakers. Over the long-­ term, lost learning is expected to lead to HCI and growth losses43, particularly for poor and low-­ income households. During the COVID-19 school closures, Türkiye has been delivering distance education services through its online Digital Education System, EBA (Eğitim Bilişim Ağı), which comprises public education lessons delivered through television and, for students with access, mobile and computer technology. Teachers, students, and parents have access to the EBA learning environment and interface which can be customized for student-specific learning, including calendar, supportive publications and library resources. The public school system is obliged to use EBA, while it is optional for private schools. Further investing in expanding and strengthening the EBA system can lay the foundation for future digital learning in-­ classroom and outside, boosting resilience and equity. 42 World Bank (2020). Türkiye Safe Schooling and Distance Education Project Appraisal Document, Annex 4 10 Learning Loss Assessment. Report No: PAD3962. Washington DC: World Bank. COVID-­ 43 World Bank (2020). Human Capital Project Report, Türkiye Snapshot. Washington DC: World Bank. Modernizing Skills and Job Training: Redirecting Investments 63 Skills and Green Growth The extent to which Türkiye’s current workforce comprises occupations utilizing skills most highly in demand for the green economy, or green skills index, shows Türkiye lags somewhat below the ECA average. Using methodologies44 to (a) create an index that measures the intensity in the usage of green skills across occupations aggregated at the national level, or ‘Green Skills Index”, and (b) categorize current jobs by their different exposure to the greening process, trends show that Türkiye lags other countries in the OECD. From the standpoint of “green skills intensity”, ECA countries tend to exhibit relatively similar indexes across levels of development (the index varies between 0.33 and 0.38). Türkiye’s “green skills intensity” index falls slightly below the ECA average. In other words, the share of occupations currently in Türkiye utilizing skills needed for green jobs is somewhat lower than the ECA average. Categorizing jobs based on the different impact of the greening process confirms that Türkiye lags behind other comparable countries in ECA. The analysis focuses on three types of jobs in relation to the greening process: (i) “green jobs/tasks” (share of time spent in an occupation working on tasks complying with environmental sustainability): these jobs are expected to remain in demand; (ii) “brown jobs” (share of jobs in most polluting 44 The methodology is based on Vona et al. (2018) and Makovec and Garrote-­ Sanchez (2021) to identify green jobs through green skills and tasks in Türkiye. There is no standard definition of what green jobs are and how to classify them (ILO,2011; CEDEFOP and OECD, 2014). Some studies define green jobs as those in businesses and industries that produce goods or provide services that benefit the environment or conserve natural resources (the output approach). Another definition categorizes jobs as green when the core business entails changes in the production process to be more environmentally friendly (the process approach). The US Bureau of Labor Statistics combines the output and process approach and, surveying US firms, found that about 2.4% of jobs can be considered green (Deschenes, 2013). Therefore, BLS green jobs definition focuses on the output or consequences that result from a given occupation on the green economy. However, these definitions are problematic, as they assume proportionality between outcomes (goods and services produced) and processes (technologies and production methods). Nevertheless, there can be jobs that do not produce green products but engage in energy efficient production methods, and the other way around. The ‘Green Economy’ program developed by the Occupational Information Network (O*NET) identifies the skill content of green jobs, emphasizing the impact of the green economy on occupations themselves. It classifies occupations in three groups: (i) existing occupations that are expected to be highly in demand due to the greening of the economy (green demand); (ii) occupations that are expected to undergo significant changes in task content due to the greening of the economy (green-­ enhanced); and (iii) new and emerging occupations in the green economy (green new) (See Dierdoff et al, 2009; 2011). Note that many of the occupations that ONET categorize as “green” are not fully so (Vona et al., 2018;). First, the green demand occupations do not necessarily involve green specific tasks. And only part of the tasks performed in occupations in the other two categories (green-­ enhanced and green demand) can be labeled as green. Vona et al (2018) propose that there are skills that are more highly correlated with green jobs, but workers in non-­ green occupations can have these green skills to a different degree, which would determine the human capital transition costs to work in green jobs. There are relatively few specific “green skills” while most green jobs require a mix of general skills that are also present in other similar non-­ green occupations (OECD, 2012). Based on regression analysis explaining what skills are used more intensely in green jobs (compared to non-­ green jobs), Vona et al. (2018) find that the most important skills in green jobs can be categorized in four main categories: (i) “engineering and technical skills”, (ii) “science skills”, (iii) “operation management skills” (adaptative management and capacity to use products and processes relevant for the environment), and (iv) “monitoring skills” (assessing the observance of legal standards). Based on these assumptions they create a Green General Skills index. This index thus measures the green skills workers have in their occupations (regardless of whether they work in green or non-­ green jobs) and can shed light on the human capital transition costs to green jobs. The methodology used here follows this approach to estimate a Green Skills Index for Türkiye and to compare it with other countries in ECA. 64 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 4.15.  Green skills labor intensity versus GDP per capita, global, 2019 0.40 R2 = 0.0228 0.39 0.38 ALB EST BLR POL CHE GEO SVN 0.37 FIN NOR LTU SWE O*net green skill index (0–1) ROU CZE HUN DEU 0.36 KAZ LVA FRA GBR AUT IRL BGR SVK BEL LUX BIH HRV 0.35 KGZ TUR PRT XKX MKD SRB DNK ITA NLD TJK GRC ESP 0.34 MDA MNE CYP 0.33 0.32 0.31 0.30 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 LOG GDP PC PPP (2019) Source: World Bank staff calculations using WDI, ECA LFS, O*NET, based on Vona et al (2018), from Makovec, M and Garrote Sanchez D (2021). Green Jobs and Green Skills in Europe and Central Asia, mimeo. industries): these jobs are expected to be significantly changed and/or replaced; and (iii) “jobs requiring upskilling”: expected to remain but require retraining45. The share of green jobs in Türkiye is lower than the average in ECA, with a higher share of current brown jobs and jobs requiring upskilling for the green transition. This finding highlights the need to invest in green skills in order to smooth the green transition and mitigate the risks of increasing structural unemployment. 45 The methodology also follows Vona et al (2018) to classify jobs as “green” based on the intensity of green tasks, and “brown” on the basis of the pollution intensity of certain industries where such jobs are prevalent. Additionally, the methodology uses O*NET categorization of jobs “requiring upskilling” due to the greening process based on their task content. In particular: (i) Green jobs/tasks capture the share of specific tasks in each occupation that are considered green according to ONET surveys. This indicator can be interpreted as a proxy of the time spent working on tasks based on (or complying with) environmental sustainability. (ii) Brown jobs are defined as those occupations in industries in the 95 percentile of pollution intensity (according to 6 air pollutants and CO2 emissions). (iii) Jobs requiring upskilling: According to ONET surveys, these are occupations that are expected to undergo significant changes in task content due to the greening of the economy. The essential purposes of the occupation remain the same, but tasks, skills, knowledge, and external elements, such as credentials, have been altered. This classification does not measure how green the tasks of workers in a given occupation are, but rather whether those tasks (regardless of what they are) are likely to change due to the green transition. These different categorizations are not mutually exclusive, that is, a job could be considered “brown” and at the same time it could have a higher share of tasks considered green. As a result, there can be an overlap between the different categories. Modernizing Skills and Job Training: Redirecting Investments 65 Figure 4.16.  Green jobs as share of employment versus GDP per capita, global, 2019 7% R2 = 0.2306 EST 6% LVA LTU BLR SVK GBR CHE Share of green tasks among workers 5% FRA SWE HUN CZE HRV POL SVN FIN NOR BEL DEU IRL 4% PRT DNK TJK XKX BGR NLD MKD CYP ESP AUT LUX MNE ROU ITA 3% MDA BIH SRB GRC KAZ GEO TUR 2% KGZ ALB 1% 0% 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 LOG GDP PC PPP (2019) Source: World Bank staff calculations using WDI, ECA LFS, O*NET, based on Vona et al (2018), from Makovec, M and Garrote Sanchez D (2021). Green Jobs and Green Skills in Europe and Central Asia, mimeo. In Türkiye, the share of green versus brown jobs has remained relatively stable over the past decade. The share of jobs requiring upskilling has increased, and, similarly, the Green Intensity Skills index has somewhat decreased.46 In other words, the lag in the share of occupations with skills needed for the green transition, and therefore the need for upskilling, has increased, likely due to a lag in green skills in growing green sectors, exacerbated by a lag general learning outcomes, shift to capital-­ intensive technologies, and dips in labor force participation due to economic stagnation and COVID since 2016. Although the three job categories are not necessarily perfectly distinct, up to 18%, or nearly one in five, of all workers in Türkiye today may be vulnerable to job loss due to job destruction in “brown jobs” and/or lags in upskilling towards greener competencies and tasks within occupations, in the absence of measures taken well in advance, particularly, but not limited to, within the manufacturing sector. The manufacturing sector in particular is characterized by a high number of sub-­ sectors with high share of brown jobs. 46 The reduction in green skills is due to the sectoral changes in employment, as the majority of sectors that contributed the most to employment growth in the last decade (trade, hospitality, public sector, education and healthcare and social work) have lower than average green skills, while agriculture, which is the only sector that saw employment reductions, has more green skills than the average sector in the country. 66 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 4.17.  Educational attainment versus green job typology (first panel) and green skills index (second) 100 16% 0.38 13 90 23 22 14% 0.37 80 41 25 12% 0.36 70 21 23 Green skill index (0–1) 10% 0.35 60 Share of jobs 50 8% 0.34 22 40 6% 0.33 30 62 56 55 4% 0.32 20 37 2% 0.31 10 0 0% 0.30 Total Brown Green Jobs need Low Middle High jobs jobs jobs upskilling Education level Low education Mid education High education Brown jobs Green jobs Need upskilling Green skills (RHS) Source: World Bank staff calculations using WDI, ECA LFS, O*NET, based on Vona et al (2018), from Makovec, M and Garrote Sanchez D (2021). Green Jobs and Green Skills in Europe and Central Asia, mimeo. In terms of profile, 48 percent of “brown jobs” employ workers below 34 years old, and 42% between 35 and 50. This means that for “brown jobs” early retirement policies alone might suffice to tackle the bulk of labor market effects, and that either unemployment support measures via income or retraining, safety net programs and pre-­ graduate education system skills programs need to be in place for workers and pre-­ graduates at young and prime ages. Workers with basic and secondary education form the bulk of the labor force in current brown jobs who will likely need to be targeted for retraining. 62% of workers in brown jobs and 55% of workers in jobs that need upskilling have attained basic or secondary education. Key Implications Overall, while educational physical infrastructure is generally distributed across regions, disparities in learning suggest allocative inefficiency within the education system, requiring a redesign of education spending and incentives for boosting 21st century skills. Modernizing educational systems in Türkiye will pave the way for building back a Modernizing Skills and Job Training: Redirecting Investments 67 better workforce with a focus on key areas: (i) competitive skills curriculum, (ii) incentives for demand-­ driven teacher training and school performance, and (iii) early career counseling in line with evolving labor demand. The need to boost connectivity and training of teachers on technology curriculum due to COVID is high, especially with a view towards digital and low-­ setting in place education in emergency models at the same time. Greater fiscal space is needed for teacher training for expanding twenty-­ first century skills such as innovation, problem-­ solving and digital competencies. Additional investment in secondary education and early job training is also needed to improve the school-­ work transition and reduce to-­ NEET rates. Finally, enhancing performance-­ based incentives for teachers and school governance to flexibly respond to rapidly evolving local and national labor market demand can boost competitiveness for an increasingly digital and green economy. 68 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs 5 Facilitating Labor Market Entry and Mobility: Harmonizing for Just Transitions Introduction In Türkiye, a range of labor market regulations and active labor market policies and programs are available to protect vulnerable workers in the formal sector, but on aggregate, these have had limited impact for informal workers, youth, and women’s labor outcomes. Key labor market programs assessed in this section include wage subsidies, active labor market programs, and unemployment benefits47. Wage Subsidies Wage subsidies (employment incentives) have tended to expand following shocks, 2008 and post-­ such as the post-­ 2018 periods. Over thirteen employment subsidies operate in Türkiye, targeting different populations and firms, with a range of different parameters regarding duration and benefits. While some of these programs were introduced before 2008, several were introduced following the 2008 global financial 47 This analysis is based on data available at the time of writing, covering through early 2021. This analysis will be updated as needed in future work outside the scope of this paper, as the main emphasis of the policy note is trends and implications over time which remain largely unchanged in spite of modest changes in data and indicators. Facilitating Labor Market Entry and Mobility: Harmonizing for Just Transitions 69 crisis. Among the registered unemployed, subsidies to cover wages and social security contributions are afforded to apprentices, interns and trainees not covered by full-­ time job contracts, amounting to an estimated 1.5 million individuals as of first quarter 2020. Of the employment incentives afforded for full-­ time jobs, one main subsidy scheme predominates, Scheme 5510 (or Five-­ Points Scheme), benefiting nearly 70 percent of all firms receiving SGK employment subsidies. Employment subsidies are primarily financed through the Unemployment Insurance Fund and the Ministry of Treasury and Finance, costing approximately 0.5 percent of GDP as of 2017. The four largest schemes reached a total of over 1.5 million firms (out of an estimated 3.5 million active SMEs, representing 99.8 percent of all registered enterprises)48 and 9 million workers in 2019 (out of 28 million)49. COVID-­ associated expansions of employment subsidies in terms of wage protection and social security premia have further broadened the scope of beneficiary firms and workers. In terms of the impact of employment subsidies on employment, previous work has shown that formal employment in small firms has tended to increase through the formalization of existing informal jobs and tend to be larger in sector such as construction and manufacturing50. Targeting to the most vulnerable workers, notably first-­ time job seekers and women, would further improve efficiency and employment impacts, although the duration of impacts depends on productivity gains and labor costs over the mid-­ term. Adult On-­The-­Job and Skills Training Active labor market programs (ALMPs) in terms of training programs Türkiye covered 15 percent of the registered unemployed51 in 2019 and 14 percent in 2020, coverage which has generally increased anti-­ cyclically since 2007. Until 2019, ALMP coverage has typically been dwarfed by wage subsidies, followed by unemployment benefits, and expansion of wage subsidies in 2019 and other job protections (layoff freeze during COVID have increased coverage of wage subsidies further, while unemployment benefit coverage has decreased. The decrease in unemployment benefits is closely correlated with the significant decrease in labor force participation, early retirement, and the layoff freeze during 2020. In 2019, ALMPs accounted for over 568,000 beneficiaries (15 percent of the unemployed), compared to 1.013 million recipients of unemployment benefits (26 percent of the unemployed). 48 Union of Chambers and Commodity Exchanges of Türkiye (TOBB), https://www.tobb.org.tr/KobiArastirma/ Sayfalar/Eng/SMEsinTurkey.php. Accessed March 24, 2020. 49 World Bank (forthcoming), Evaluation of employment subsidy schemes, from progress reviews for June 2019 and February 2020. 50 Betcherman et al., 2020; World Bank, forthcoming. 51 Registered unemployed defined as those registered with the Turkish national employment agency, ISKUR. On average, ISKUR data capture approximately 80 percent of the total unemployed estimated through national labor force surveys conducted by the Turkish national statistics institute, TUIK. 70 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 5.1.  Labor programs coverage by type over time and by gender and educational level, 2007–2020 Coverage of labor programs Unemployment benefit coverage vs demand 70% 2,500,000 60% 2,000,000 50% 40% 1,500,000 30% 1,000,000 20% No. of workers 10% 500,000 0% 0 1 Share of registered unemployed (%) 7 8 9 10 1 1 2 13 1 4 15 16 17 18 19 0 1 00 00 00 20 02 07 08 09 10 1 12 13 14 15 16 17 18 19 20 2 2 2 20 20 20 20 20 20 20 20 20 2 20 20 20 20 20 20 20 20 20 20 20 20 20 20 % ALMP % UB % Wage subs Applicants Beneficiaries ALMP coverage by type, gender and education, 2020 Unemployment benefit coverage by gender and education, 2020 250,000 1,40,000 1,20,000 200,000 1,00,000 150,000 80,000 100,000 60,000 Facilitating Labor Market Entry and Mobility: Harmonizing for Just Transitions 40,000 50,000 20,000 0 0 Male Female Male Female Male Female te e nn e se r e ra rat tioio at te at a c oci c en as tor Vocational training On-the-job All ALMPS ite L ite uc u l at and Li M c Ill t) ss courses training programs ed ed o en A Do r y ry cho ival a da h s qu im Pr c on hig e ( Doctorate License Se Male Female Post graduate Associate degree Secondary education (high school and equivalent) Primary education Literate Illiterate Source: World Bank staff using Turkish National Employment Agency (ISKUR) and National Social Security Institution (SGK) Statistics, most recent data through December 2020 as of January-­March 2021. 71 As of 2020, of the over 423,000 beneficiaries enrolled in ALMPs, the national On-­ the-­Job Training Program (OJT) remained the dominant choice (80 percent), with the Vocational and Technical Courses Program (VT) accounting for 20 percent. Among the nearly 1,400 VT courses on offer, clothing and textiles was the most common occupational skill in demand, accounting for nearly one out of three beneficiaries (26 percent). Among nearly 34,000 OJT programs on offer, nearly one in three beneficiaries were in sales related occupations or retail occupations (26 percent), followed by clothing and textile-­ (14 percent), with the remainder split nearly equally across trades (metallurgy, furniture), hospitality, and other services. Over the past decade, the demand for OJT by occupation has evolved, with a shift towards more skilled manufacturing workers at mid-­ level and client services. ALMPs tend to cover younger, less-­ skilled workers and serve as a pathway to re-­ skilling and facilitating the transition to new jobs and sectors. Most ALMP beneficiaries are young adults under the age of 34 years (77 percent), heavily concentrated among 20–24-­ year-­ olds (33 percent). The majority of ALMP beneficiaries continue hold a primary or secondary education (71 percent), although vocational courses are skewed towards primary-­ schooled workers than secondary (51 versus 27, respectively). By gender, while no major differences are seen overall and among on-­ job training, vocational course enrollment is skewed the-­ towards females relative to males (69 versus 31 percent, respectively). These patterns have been generally constant over time. The impact of ALMPs is tied to how responsive they are to the demand by firms for certain occupational skills, and shifts expected as a result of COVID will heighten the need for demand-­ driven training. Administrative and online job vacancy data highlights the need for social as well as technical skills in the formal sector and across regions52. The demand for skills may have shifted pre-­and post-­ COVID, as the need for service sector workers has declined and that for construction, for example, has increased, although it is unclear whether the latter is specifically due to COVID. Occupations including routine tasks (such as machine operators, call center information clerks and product graders and testers) and occupations requiring non-­ routine manual tasks (such as customer service) have historically been in high demand, with wide variations across provinces. Increasingly, related and social skills (such as software knowledge, communication and teamwork IT-­ skills) and professionalism (discipline, time management) also tend to be in high demand, particularly in regions with higher economic activity. As greater attention is given to building back better and the green economy as part COVID-­ recovery, targeting skills retraining to youth and first-­time job seekers, particularly females, may be especially cost-­ effective. 52 World Bank, forthcoming; Turkish Employment Agency data. 72 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 5.2.  Jobseeker demand for On-­the-­Job Training (OJT, ISKUR), 2010 versus 2020 OJT, 2010 OJT, 2020 Manual worker 14.1% Marketers 12.3% Computer manager 6.9% Manual labor, manufacturing/readymade 7.9% Janitor 3.8% Receptionist/ front office staff 6.0% Business manager 2.9% Mechanic, sewing 3.5% Machine operator, sewing/clothing 2.7% Sewing machine operator-clothing 3.4% Labourer, manufacturing/readymade 2.4% Customer service officer/ assistant 3.2% Office staff 2.2% Patient advisor 3.1% Marketers 2.2% Call center personnel 2.7% Mechanic, sewing 1.8% Office worker 2.6% Patient advisor 1.8% Waiter 2.5% Costumer service officer 1.7% Market personnel 1.6% Cooking assistant 1.7% Laborer, Designer, web plastic/manufacturing fitter 1.5% 1.5% Mechanic, Basic manufacturing/ 1.3% machinery and assembly personnel 1.4% electronics/computer Designer, computer-aided 1.0% Assembler, metal products 1.2% Source: World Banks staff calculations, ISKUR data. Unemployment Benefits Unemployment benefits also tend to go to less-­ skilled workers, although they are generally older than ALMP beneficiaries. Unemployment benefits are still relatively nascent in Türkiye, having gone from 221 thousand beneficiaries at their introduction in 2007 to 841 thousand by 2018 as a result of the economic downturn, expanding to 1 million applicants in 2019, with coverage hovering at 26 percent of all registered unemployed. Over Facilitating Labor Market Entry and Mobility: Harmonizing for Just Transitions 73 2020, coverage decreased by nearly half to 509 thousand, as labor force exits increased and the layoff ban took effect. Between 2012 to 2018, coverage nearly doubled; at the same time, ALMP coverage remained constant over this period. Unemployment benefits remain concentrated among workers with primary education (50 percent of beneficiaries), followed by secondary (27 percent), suggesting the program appears to be relatively progressive by educational level. Most beneficiaries tend to be younger at 25–44 years (over 70 percent), with 20 percent aged between 45–54 years, and the remainder older. As a policy instrument, unemployment benefits have gradually assumed a greater focus in Türkiye, but the effectiveness of labor market instruments as a whole at improving labor force participation appears mixed. Job Search Along with learning disparities, insufficient early job placement and preparation can delay labor market entry, evident in shifting trends regarding the demand for on-­ the-­ job training, but also job matching. While public social investments matter for growth, the nature of spending will be key to effective recovery and long-­ term inclusive growth. GDP level) is positively associated with student ratio and population density, per capita (province-­ as well as ISKUR programs (job placements, associated with OJT) and public social sector infrastructure more broadly. The demand for labor in the formal sector based on analysis of vacancies53 shows that job growth is driven both by latent demand for workers and an uptake of active job search. The increase in the total number of job vacancies continued from late 2016 until early 2018, and has been almost steadily decreasing since then even before COVID-­ 19 hit. Total positions decrease by about 40% between January to March 2020 as a first response to the COVID-­ 19 induced lockdowns. More recently though, seasonally adjusted data show that the number of positions have reached levels closer to those in early 2018 once again. Job application figures from also shows recovery in labor markets in 2021 although the total applications are yet to surpass the levels in 2018. Recovery seems to be particularly strong in manufacturing and construction sectors based on an analsis of labor force statistics and vacancy analysis. According to seasonally adjusted data, vacancies in manufacturing decreased since early 2018, with a large shock at the beginning of COVID-­19, and picked up to reach the 2018 levels by September 2021. Jobs in construction have been steadily decreasing since mid-­2018, and while there has been an uptick in labor demand during 2021, recovery so far has not reached the levels of 2018. 53 World Bank staff analysis based on data from Türkiye’s Kariyer Net Job Portal, the main private sector jobs portal for online vacancies in the formal sector. 74 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 5.3.  Total number of job vacancies, overall (first panel) and for manufacturing (second) and construction (third), seasonally adjusted values, 2015–2021 Total positions 1,60,000 1,40,000 1,20,000 1,00,000 80,000 60,000 40,000 20,000 0 01 06 11 04 09 02 07 12 05 10 03 08 01 06 11 04 09 15 20 17 15 18 20 17 19 16 21 19 17 15 16 18 20 21 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Total positions (manufacturing) Total positions (construction) 70,000 12,000 60,000 10,000 50,000 8,000 40,000 6,000 30,000 4,000 20,000 10,000 2,000 0 0 15 1 20 06 1 1 16 4 20 09 20 702 20 707 1 2 20 805 20 1810 19 3 8 20 1 20 06 2 1 21 4 09 15 1 20 06 1 1 16 4 20 09 20 702 20 707 1 2 20 805 1 0 19 3 8 20 1 20 06 2 1 21 4 09 20 200 20 150 20 151 20 201 20 90 20 60 20 10 20 0 20 171 20 150 20 151 20 200 20 201 20 60 20 90 20 10 20 0 20 171 20 181 1 1 1 20 1 1 20 Source: World bank staff calculations based on data from Kariyer.net, courtesy of Ozen, E and Demir Sekir, S, mimeo. Figure 5.4.  Total number of job applications and job applications per position, overall, seasonally adjusted values, 2015–2021 Total applications Applications per position 15,000,000 160 140 120 10,000,000 100 80 60 5,000,000 40 20 0 0 2 01 16 4 1 9 1 2 20 707 1 2 20 805 0 3 8 15 1 20 06 1 20 006 1 21 4 09 15 1 20 06 1 1 20 604 20 609 20 702 20 07 20 1712 20 05 20 810 20 03 20 908 20 1 20 06 21 1 20 04 09 20 150 20 151 20 201 20 50 20 151 20 00 20 01 20190 20160 20210 20 171 20 90 20 70 20 0 20 181 20 0 17 19 18 21 2 1 1 2 2 1 1 1 1 20 20 Source: World bank staff calculations based on data from Kariyer.net, courtesy of Ozen, E and Demir Sekir, S, mimeo. Facilitating Labor Market Entry and Mobility: Harmonizing for Just Transitions 75 Job search data shows that the number of applications per position has been decreasing since 2019 for various reasons. Despite a brief peak in June 2020 when the COVID-­ induced lockdown measures were eased, the decreasing trend is still visible as of September 2021. Looking at patterns and profiles of job search applicants (noting that one applicant may apply to more than one job), seasonally adjusted data show a similar pattern of job search increases and decreases over time by gender. The overall number of applicants is still below the 2018–2019 levels for both men and women. As observed in general labor force statistics, persistent gender gaps in labor force participation are reflected in the trend that 6 out of 10 of jobs were filled by male workers as of 2020. Despite the recovery, female labor force participation still remains below the 2018 levels. As of August 2021, seasonally adjusted female labor force participation rate was 32.3 percent. Overall job search declines may be driven by declining labor force participation rates, a mismatch between available competencies and occupational skills most in demand over time and regional changes in the demand for jobs. Figure 5.5.  Total number of job applicants by gender (first panel) and age (youth younger than 24 years and adult 25+, second panel), 2015–2021 Applicants by gender 6,00,000 5,00,000 4,00,000 3,00,000 2,00,000 1,00,000 0 20 01 20 04 20 7 20 10 20 01 20 04 20 7 20 10 20 01 20 4 20 7 20 10 20 01 20 04 20 7 20 10 20 01 20 04 20 7 20 10 20 01 20 04 20 07 20 10 20 01 20 04 07 0 0 0 0 0 0 15 15 16 16 17 17 18 18 19 19 20 20 21 16 15 17 18 19 20 21 18 15 16 17 19 20 21 20 Female Male Applicants by youth 8,00,000 7,00,000 6,00,000 5,00,000 4,00,000 3,00,000 2,00,000 1,00,000 0 01 20 04 20 7 20 10 20 01 20 04 20 7 20 10 20 01 20 4 20 7 20 10 20 01 20 04 20 7 20 10 20 01 20 04 20 7 20 10 20 01 20 04 20 07 20 10 20 01 20 04 07 0 0 0 0 0 0 15 15 16 16 17 17 18 18 19 19 20 20 21 15 16 17 18 19 20 21 15 16 17 18 19 20 21 20 20 Adults Youth Source: World bank staff calculations based on data from Kariyer.net, courtesy of Ozen, E and Demir Sekir, S, mimeo. 76 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs In terms of job search among youth, a marked decrease was seen due to COVID, with recovery generally quicker among older youth. The number of young job applicants (aged 24 or below) decreased about 29 percent due to the immediate effect of COVID-­ 19, between March and April 2020, similar to those aged 25 and above. Recovery in the number of young applicants seem to be faster than that for those aged 25 and above, with an increase of close to 40 percent in the number of young applicants, compared to a little above 20 percent for those aged 25 and above between March 2020 and September 2021. Figure 5.6.  Public employment service job placement rate (ISKUR) versus share of unfilled vacancies per year (first panel) and distribution of job placements by age and gender, 2020 (second) Job vacancy rate vs share unfilled 120% 100% 80% 60% 40% 20% 0% 07 08 09 10 11 12 13 14 15 16 17 18 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Vacancies as % of reg unemployed % Unfilled vacancies Distribution of job placements by age group and gender (ISKUR), 2020 100% 80% 60% 40% 20% 0% 15–24 25–34 35–44 45–54 55–64 60+ Total Share of placements by age (%) Female share of placements (%) Source: World Bank staff calculations, ISKUR data. Facilitating Labor Market Entry and Mobility: Harmonizing for Just Transitions 77 78 W W ho ho le le sa sa le le /re /re Ad ta Ad ta m il/ M m il/ M in au an in au an ist to uf ist to uf ra m ac ra m ac –150 –100 –50 0 50 100 150 200 250 0 5 10 15 20 25 30 35 40 45 50 Ac tiv o t Ac tiv o t co P / C tive urine co P / C tive urine m ro u on re g s m ro u on re g s W m fe p st p p m fe p st p p od ss o ru ai od ss o ru ai ho r r r r le Tr atio ion t se ctio Tr atio ion t se ctio sa an n al/ rv n an n al/ rv n le/ sp /fo te ic sp /fo te ic 0 50 100 re M or od ch es or od ch es ta ta s nic ta s nic il/ an To H H 2010 au uf ta ea O tion erv al ea O tion erv al Ad to ac l lth th /st ice lth th /st ice m m tu El Pu /s er or s El Pu /s er or s in ot rin ec bl oc se ag ec bl oc se ag ist iv e g tri ic ia rv e tri ic ia rv e ra tiv Co rep ci ty Ag dm a l s ic e e ci ty Ag dm a l s ic e e ,g r ,g r e/ ai i i Ed rvi s i i Ed rvi s 2020 su nst r as cu nis uc ces as cu nis uc ces Ac P pp ruc ,s te l t u t r at ,s te l tu tr at co rof o t io re atio io re atio io m es rt n am /fo n n am /fo n n m s s W , ai M re /de IC W , ai M re /de IC od ion erv at al i at r c in str fe T er on in y/ nc at r c in str fe T er on in y/ nc io /te ces Tr n ch su di g/q fis e su di g/q fis e an /fo n ic pp tio ua hin ly nin rr g pp tio ua hin ly nin rr g Source: World Bank staff calculations, ISKUR data sp od Ar Ar or se al ts /w g yi ts /w g yi ta r ,e as su ng ,e as su ng n F t n F t Distribution of vacancies tio vic 2010 n es H ter ina R e M pply ou ta n e G H ter ina R e M pply ou ta n e G Change, share of all vacancies (%) H O /sto th se in cia al M m l se in cia al M m l ea ra change in female placement share e ho e /in es T ho e /in es T g Change, placement share female (%) lth rs e ld nt su tat ld nt su tat Change in distribution of vacancies vs /s go /re ra e go /re ra e oc erv ia ic od cre nc od cre nc e 2020 ls s s/ at e se io s/ at e se io Pu er rv n rv n bl vi ic ic ic Ed ces es es ad uc W W El ec m at ho ho Ag in io tri r i s n le le ci ty ic tra sa sa ,G ul tu tio le le as re n/ ICT /re /re de Ad ta Ad ta ,S /fo re fe m il/ M m il/ M te in au a in au a –150 –100 –50 0 50 100 150 200 250 –10 0 10 20 30 40 50 60 70 80 90 am st nc ist ist ,A M r y/ e ra to nu T ra to nu T in fi Ac tiv m fa ot o c Ac tiv m fa ot o c ir in sh W co g/ i n co P e/ C tiv tur al co P e/ C tiv tur al at qu g m ro su o e r ing m fe pp ns ep m ro su o e r ing m fe pp ns ep er ndit ar su io ry od ss o tr a od ss o tr a in Tr ati ion rt s ucti ir Tr ati ion rt s ucti ir pp nin g g 2010 ly an on al er on an on al er on /w su sp /fo /te vic sp /fo /te vic as p or od ch es or od ch es te ply H ta s n H ta s n Ar M ea O tion erv ical ea O tion erv ical G Share of all vacancies (%), 2020 ts R M lth th /s ic lth th /s ic , e Fin T El ec P ub /s er tor es El ec P ub /s er tor es an eal vacancies by sector, unfilled vacancy rate and change in female placement rate nt 2020 es tri lic oc s a tri lic oc s a er ci H t a l t a ci ty ia erv ge ci ty ia erv ge ou ai n / i ns t e ,g Ag adm l s ic e ,g Ag adm l s ic e u r r Share of female placement vs distribution of vacancies, 2010 vs 2020 se me as ic in Ed rvi es as ic in Ed rvi es ho nt/ ran ,s ul ist t uc ce ,s ul ist t uc ce ld re ce te ur ra at s te ur ra at s am e/ io t io am e/ io t io go cre at f n f n od W , ai M ore n/d IC W , ai M ore n/d IC s/ io at r c in st ef T at r c in st ef T se n er o in ry en er o in ry en rv su ndi g/ /fis ce su ndi g/ /fis ce ic pp tio qu hi pp tio qu hi es Ar ly nin arr ng Ar ly nin arr ng Figure 5.7.  Efficiency of public service job placement, 2010 versus 2020: (left to right) total ts /w g yi ts /w g yi Unfilled vacancy rate ,e ,e Change, unfilled vacancy rate (%) n F as su ng te p n F as su ng te p H te ina H te ina Change, placement share female (%) ou rta n Re M ply ou rta n Re M ply Change in unfilled vacancies vs se in cia al GM se in cia al GM change in female placement share ho me l/i es T ho me l/i es T ld nt nsu ta ld nt nsu ta go /re ra te go /re ra te od cr nc od cr nc s/ eat e s/ eat e se io se io rv n rv n ic ic es es Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 5.8.  Mapping of sectoral job vacancies, unfilled rate and female placement, 2020 (three panels) Share of all formal job vacancies (ISKUR) vs rate unfilled, 2020 100% Household goods/services Rate of unfilled job vacancies (%) Real estate Professional/technical Other services 50% ICT Financial/insurance Electricity, gas, steam, air conditioning supply Manufacturing Water supply/waste MGMT Wholesale/retail/automotive repair Arts/entertainment Transportation/storage Construction Mining/quarrying Accommodation/food services Education Administrative/support services Health/social services Agriculture/forestry/fishing 0% 25% 50% Share of all vacancies (%) Share of all formal job vacancies (ISKUR) vs female placement rate, 2020 100% Health/social services Female placement rate (%) Education Financial/insurance 50% Agriculture/forestry/fishing Wholesale/retail/automotive repair ICT Arts/entertainment Professional/technical Real estate Accommodation/food services Manufacturing Other services Administrative/support services Household goods/services Transportation/storage Water supply/waste MGMT Mining/quarrying Construction Electricity, gas, steam, air conditioning supply 0% 25% 50% Share of all vacancies (%) Facilitating Labor Market Entry and Mobility: Harmonizing for Just Transitions 79 Rate of unfilled vacancies vs female placement rate, 2020 100% Health/social services Female pacement rate (%) Education Agriculture/forestry/fishing Financial/insurance Wholesale/retail/automotive repair 50% Arts/entertainment ICT Professional/technical Accommodation/food services Manufacturing Real estate Administrative/support Transportation/storage Other services services Household goods/services Water supply/waste MGMT Electricity, gas, steam, air conditioning supply Construction Mining/quarrying 0% 50% 100% Rate of unfilled vacancies (%) Source: World Bank staff calculations, ISKUR data Key Implications Taken together, labor market program investments in Türkiye will need significant strengthening to address the relatively high level of informality, high youth NEET and low female labor force participation rates, exacerbated due to COVID. Key areas include: (i) targeting reforms to more transparently identify and include excluded vulnerable informal workers and women in vocational, on-­ the-­job training and wage subsidy programs, particularly in more productive and green sectors; (ii) expanding job matching services and partnerships with the private sector through incentives and regional, routine outreach services; (iii) developing integrated labor market case management services to register and provide routine job counseling to poorer households and vulnerable informal workers; and (iv) consolidating and harmonizing benefit levels across wage subsidy and unemployment benefit programs to ensure more equitable and efficient investments towards boosting job outcomes. 80 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs 6 Mitigating Risk and Improving Resilience: Integrating Social Insurance Introduction As an upper middle-­ income country, Türkiye’s demographic profile is fast approaching its older, more urban counterparts throughout the OECD. Urbanization has been increas- ing rapidly, with approximately 68 percent of the population living in urban areas across 81 provinces.54 Integrated indices of socioeconomic development vary across provinces, with northwestern areas faring better compared to southeastern areas55. The proportion of elderly population (65 years and over) is 10.2 percent as of 2019 (most recent data).56 22.6 percent of the population is younger than 15 years (0–14 years), and 67.3 percent between 15–64 years. With an aging population, the proportion over 65 years is expected to rise to 16.3 percent in 2040 and 25.6 percent in 2080 according to population projec- tions.57 By province, the provinces that had the highest proportion of the elderly population were Sinop (18.3 percent), Kastamonu (17.1 percent) and Artvin (15.7), concentrated along the Black Sea. 54 World Bank (2018). Türkiye Systematic Country Diagnostic. Washington DC: World Bank. 55 Turkstat Index of Well-­Being, developed by the national statistical institute TUIK, is a composite index covering eleven domains (housing, work life, income and wealth, health, education, environment, safety, civic engagement, access to infrastructure services, life satisfaction), analyzed using 41 indicators. The index value ranges from 0 to 1, with values approximating to 1 state a better level of well-­ being. See http://www.turkstat.gov.tr/PreHaberBultenleri. do?id=24561 56 Turkish Statistical Institute (2020). 57 TurkStat (2018) Elderly Statistics 2018. Ankara: TurkStat. Mitigating Risk and Improving Resilience: Integrating Social Insurance 81 Labor Productivity Risks Labor vulnerabilities are particularly evident among the lowest income over the past decade. Over a ten-­ year period (2008–2019), the likelihood of formal employment, all else held equal, was significantly highest among highly skilled males, adults older than 25 years, and heads of households, and significantly lowest among divorced and widowed workers. employment versus informal Similar determinants are seen with respect to formal self-­ employment. In terms of economic sector, all else held equal, informality is significantly self-­ highest in agriculture. For these reasons, the determinants of labor income over a decade show that earnings’ growth among the lowest-­ income households are attributed to labor productivity and hours worked. Earnings declined with working hours and increased with productivity, favoring high-­income households. All deciles witnessed a modest boost from switching to more productive sectors and to formal jobs. Likewise, changes in labor income over the same period showed that low-­ income households were highly sensitive to changes in productivity and working hours. COVID has had a significantly higher impact on employment and earnings, all else held equal, on women, youth, unskilled and informal sector workers, regardless of sector. Given labor productivity risks, Türkiye’s labor market is relatively more vulnerable to shocks than in comparable countries. The Future of Work will also increasingly require a holistic view of labor and social risk mitigation.58 With lagging productivity across sectors with high shares of employment and the exclusion of informal workers, financial inefficiency and inequity are growing in the system. Labor costs in Türkiye at over 50 percent as a share of commercial profits, or overall productivity, are relatively higher than in France, Mexico and Chile, which has among the lowest share of labor costs. Productivity in Türkiye is also low relative to its level of minimum wage. Labor costs in Türkiye are related to social security contributions, including unemployment insurance, old-­ age pensions, health insurance coverage (with subsidies for informal workers) and implicit costs associated with dismissal procedures. Payments by employers to dismissed workers are relatively generous in Türkiye in terms of the equivalent number of months’ wages received compared to Brazil, Malaysia or France. The structure of the system imposes an economic burden on employers, limits portability of benefits, and leads to negative incentives and distortions for hiring and firing. A reformed system should look to decrease costs to employers and increase coverage for workers, within the scope of integration with social insurance. 58 See World Bank (2019). Protecting All: Risk Sharing for a Diverse and Diversifying World of Work. Washington DC: World Bank; and Palacios and Robalino (2020). Integrating Social Insurance and Social Assistance Programs for the Future World of Labor. IZA DP 13258. Bonn: IZA Institute of Labor Economics. 82 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 6.1.  Labor productivity components by income quintile (first panel) and country comparisons (second), 2020 1.6 1.4 1.2 Variance (%) 1.0 0.8 0.6 0.4 0.2 0 1 2 3 4 5 6 7 8 9 10 Household income decile Employment Participation Informality Sectors Productivity Hours Türkiye France Mexico Chile 0 10 20 30 40 50 60 % labour force cost Source: World Bank staff calculations based on World Bank (2019) and Palacios and Robalino (2020). Overall Social Insurance Coverage The Turkish social security fund (SGK) currently provides health insurance and protection from old-­age risk to approximately 22 million workers and 15 million pensioners in the public and private sector (9 million excluding public sector).59 This amounts to nearly 68 percent of the labor force and 86 percent of the older population. By household income, social insurance coverage ranges from 24.9 percent among the lowest-­ income quintile to 41.3 percent among the highest-­ income quintile.60 The demand for social insurance is increasing rapidly in the face of both a young population, a persistent share of the labor force working informally and an increasing cohort of over 65-­ olds. Over the medium-­ year-­ term, given the loss in wages expected due to COVID, the financial viability of SGK will need to be rapidly assessed and financing revisited to account for a diversifying jobs landscape. 59 Republic of Türkiye Social Security Institution (SGK) (2019). Organizational Profile and Overview of the Social Security System in Türkiye. Ankara: SGK. Republic of Türkiye and World Bank (2017). 60 World Bank WDI, ASPIRE Social Protection and Jobs database. Mitigating Risk and Improving Resilience: Integrating Social Insurance 83 Figure 6.2.  Overall coverage rates of social insurance and ALMPs by income quintile, 2010–2020 Coverage of social protection & labor programs by quintile (% of population) 60 Coverage in 5th quintile (richest) (%) - all social insurance, 41.3 Percent of population enrolled (%) 50 40 Coverage in 1st quintile (poorest) (%) - all social insurance, 24.9 30 20 Coverage in 5th quintile (richest) (%) - all labor market, 1.2 10 Coverage in 1st quintile (poorest) (%) - all labor market, 0.6 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: World Bank World Development Indicators and ASPIRE Database. Estimates for 2016–2020. Pensions Türkiye’s pay-­ you-­ as-­ go pension system has recently consolidated separate schemes, with an effort to align contributions and benefits more closely. The pension fund is financed by a pay-­roll contribution of 21 percent (9 percent employees and 11 percent employers). Workers can retire at age 61 if women or 63 if men and replace 2 percent of their last salary for each year of contribution to the system. There is also a minimum pension of TL 1,402 which represents approximately 42 percent of average earnings, which was raised to TL 1,500 in 2020. However, only the last year of salaries is included in the calculation of the pension; the accrual rate is not linked to the level of the contribution rate, life expectancy at retirement, and the sustainable rate of return of the system; and financing for minimal pensions is effectively implicit. This generates regressive financing, where higher-­ income workers benefit to a greater extent, potentially reducing incentives to create formal jobs. The system is therefore accumulating unfunded liabilities that harm long-­ term financial sustainability. The pensions system also offers survivorship and disability pensions and other benefits such as maternity leave, compensation for temporary incapacity, and insurance against work accidents and occupational hazards. New non-­ contributory social pensions schemes for widows and families of soldiers serving compulsory military service were implemented in 2012 and 2013, respectively, and following COVID, the Government has announced new social assistance programs, gas and electricity bills support for the poor and transfer for families with multiple births. Integrating across divergent pensions schemes and harmonizing with a unique, explicit subsidy may help ensure equity across beneficiaries moving forward. 84 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Unemployment Insurance Expenditures on unemployment insurance, administered by the Turkish Employment Agency, bear benefits that are not closely aligned to contributions; in this case, they may income workers from seeking more productive employment as design. adversely affect low-­ Türkiye has a standard unemployment insurance scheme for formal wage employees that covers around 22.7 percent of the unemployed at a cost of 0.19 percent of GDP. The scheme is financed by a contribution rate of 4 percent (1 percent employees, 2 percent employers, and 3 percent the government). The unemployment benefit is equal to 40 per- cent of the salary and has a duration of 4 months. Without incentives to maintain jobs, search for new jobs or retrain, the system risks moral hazard and inefficiency. Figure 6.3.  Social insurance subsidies for universal coverage by income quintile, illustrative analysis Unemployment insurance and implicit subsidies 9 8 7 6 Percent 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 10 Equilibrium contribution Subsidized contribution System contribution Pensions, inequity and regressive implicit subsidies 100 0.35 90 0.30 80 70 0.25 60 0.20 Percent 50 0.15 40 30 0.10 20 0.05 10 0 0 1 2 3 4 5 6 7 8 9 10 Equilibrium contribution Covered pop Effective contribution Source: World Bank staff calculations based on World Bank (2019) and Palacios and Robalino (2020). Mitigating Risk and Improving Resilience: Integrating Social Insurance 85 Table 6.1.  Detailed SPL expenditures by program, 2015–2019 Pensions 2015 2016 2017 2018 2019 Insured 20,913,338 21,272,012 22,421,748 22,215,231 22,137,342 Pensioners 7,944,373 8,207,381 8,493,984 8,822,664 9,079,479 Coverage (% LF) 70% 70% 71% 69% 68% Coverage (% 60+ pop) 88% 88% 87% 87% 86% Expenditures (Millions TL) 166,234 201,754 229,671 266,395 298,615 Pension payments 133,515 162,139 184,984 214,133 240,032 Survivor payments 32,719 39,615 44,687 52,262 58,583 Expenditures (% current GDP) 7.1% 7.7% 7.4% 7.2% 7.0% Pension payments 5.7% 6.2% 5.9% 5.7% 5.6% Survivor payments 1.4% 1.5% 1.4% 1.4% 1.4% Unemployment Insurance 2015 2016 2017 2018 2019 Insured 14,462,847 15,006,103 16,054,439 15,800,234 15,777,952 Beneficiaries 43,745 64,499 54,958 91,011 1,013,056 Coverage (% LF) 48.7% 49.1% 50.7% 49.0% 48.4% Coverage (% unemployed) 1.4% 1.9% 1.6% 2.6% 22.7% Expenditures (Millions TL) 2,193 3,683 3,834 4,824 7,985 Expenditures (% current GDP) 0.09% 0.14% 0.12% 0.13% 0.19% Active Labor Market Programs (ALMP) 2015 2016 2017 2018 2019 ALMP Expenditures (Millions TL) 15,024 19,550 23,748 28,598 32,308 Non-­Wage Subsidies 3,564 5,725 5,085 6,252 6,625 Wage Subsidies 11,459 13,825 18,664 22,346 25,682 Expenditures (% current GDP) 0.6% 0.7% 0.8% 0.8% 0.8% Non-­Wage Subsidies 0.2% 0.2% 0.2% 0.2% 0.2% Wage Subsidies 0.5% 0.5% 0.6% 0.6% 0.6% Health Insurance 2015 2016 2017 2018 2019 Total Covered Individuals 77,402,060 77,611,638 80,622,172 80,851,993 82,169,815 Individuals Not Paying Health Insurance 53,808,985 54,450,366 55,877,740 56,243,675 57,647,859 Individuals Paying Health Insurance 23,593,075 23,161,272 24,744,432 24,608,318 24,521,956 Coverage of paying individuals (% LF) 79.5% 75.9% 78.2% 76.2% 75.3% Coverage of covered individuals (% LF) 99.0% 97.9% 100.4% 99.3% 99.5% Expenditures (Millions TL) 80,463 91,330 103,077 121,444 146,904 Health Insurance Payment 59,356 67,993 77,632 91,512 110,697 (contributory & subsidized) Health General Government Expenditures (MoH) 21,107 23,337 25,445 29,931 36,206 Expenditures (% current GDP) 3.4% 3.5% 3.3% 3.3% 3.4% Health Insurance Payment 2.5% 2.6% 2.5% 2.5% 2.6% (contributory & subsidized) Health General Government Expenditures (MoH) 0.9% 0.9% 0.8% 0.8% 0.8% 86 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Social Assistance 2015 2016 2017 2018 2019 Coverage of General Cash Programs 3,271,121 3,493,505 3,580,483 3,694,411 3,694,411 Coverage (% all Households) 14.9% 14.9% 14.9% 14.9% 14.9% Expenditures (Millions TL) 12,828 17,537 20,819 22,874 26,288 Expenditures (% current GDP) 0.55% 0.67% 0.67% 0.61% 0.61% Source: World Bank staff calculations; Data from SGK; Ministry of Labor and Social Security, ISKUR; Ministry of Family and Social Services; Ministry of Health; TURKSTAT; Ministry of Treasury and Finance; and OECD, EUROSTAT and WDI data review. Health Insurance Türkiye’s burden of disease has been shifting to non-­ communicable diseases, COVID notwithstanding, although health expenditures have not increased significantly on the whole. NCDs account for 89 percent of all deaths61, with certain underlying risk factors associated with the COVID disease burden. Underlying risk factors among adults for NCD-­ attributed mortality include relatively high rates of tobacco use (28 percent; nearly twice as high among males than females), raised blood pressure (20 percent), diabetes (raised blood glucose (13 percent) and obesity (32 percent; nearly twice as high among females than males). Türkiye’s overall advanced level of HCI in terms of health has been due to strong investments in public health services and financial protection, which also helped mitigate some of the serious health consequences associated with COVID. As part of its national Health Transformation Program since 2003, infant and maternal mortality rates have drastically dropped, the quality of tertiary health care improved, and its research and development capacity have been strengthened, given it aims to boost the sector as an engine of growth. At the same time, health care coverage in terms of human resources, remains below the average for the region and comparable economies, suggesting potential capacity constraints to address pandemics especially in high-­ density areas. The number of physicians and nurses per capita, for example, is nearly 30 to 60 percent less than the average for the Europe and Central Asia (ECA) region and that of the OECD as of 2015 (most recently available comparative data)62, despite recent increases in nurse ratios. Türkiye has 536 persons per physician, with a total of over 153 thousand physicians nationwide63. The Ministry of Health Strategic Plan emphasizes the importance of increasing the number of the primary health care (PHC) workforce and sets higher targets for 2030.64 61 World Health Organization (2018). Türkiye: World Health Organization Noncommunicable Diseases (NCD) Country Profiles, 2018. Geneva: World Health Organization. Most recent data. 62 World Bank World Development Indicators (WDI). 63 TurkStat, 2018. Most recent data. 64 World Bank (2019). Building an improved primary health care system in Türkiye through care integration. Washington DC: World Bank. Mitigating Risk and Improving Resilience: Integrating Social Insurance 87 While Türkiye has made important strides in subsidizing universal health insurance for income households, the challenge is to link financing to the evolving cost of care low-­ equitably. Although the subsidized system links contributions to the average per capita cost of the package of health services, the contributory system relies on pay-­ roll taxes. This implies that workers contributions are not linked to the expected cost of the health services they receive, which depends on age and family structures and health risk profiles. Hence, some workers might be paying more than the expected costs (an implicit tax) while others pay less (an implicit subsidy). This implicit redistribution can be regressive and can also reduce incentives to enroll, particularly among low-­ risk population such as youth who are single. Figure 6.4.  Health care personnel, global, 2015–2019 Comparative ratio of health care personnel per 100,000 population 1,000 800 Personnel per 100,000 600 400 200 0 Middle income Türkiye Upper middle OECD members Europe & income Central Asia Physicians Nurses and midwives Source: World Development Indicators, 2015–2019. Table 6.2.  Health service delivery indicators, 2018–2020 Indicator Total Total No. Health Facilities 34,559 No. Inpatient Facilities 1,534 No. of Outpatient Facilities 33,025 Total No. Hospital Beds 231,913 No. Hospital Beds per 1000 Persons 2.83 Total No. of Doctors 153,128 No. of Persons per Doctor 536 No. of Patient Hospital Visits per Doctor 5,110 Source: TurkStat and Ministry of Health, 2018–2020. 88 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 6.5.  Burden of non-­communicable disease by health care expenditure (first panel) and cause of death (second), global Burden of disease (diabetes, non-communicable) Cause of death due to non-communicable diseases vs health spending (% of total) 6,000 100 91.5 88.8 89.4 Current health expenditure per captia, ppp (current inl $) ,2016 90 87.6 OECD 84.1 5,000 80 74.4 70 4,000 60 3,000 50 Central 40 Europe & 2,000 Baltics Latin America & 30 Caribbean Europe & 20 1,000 Central Asia Türkiye 10 East Asia & Pacific 0 0 0 2 4 6 8 10 12 Latin East OECD Europe & Türkiye Central Diabetes prevalence America & Asia & Central Europe & (% of population ages 20 to 79), 2019 Caribbean Pacific Asia Baltics Source: World Development Indicators, 2019. Social Assistance Türkiye’s social assistance schemes are fairly well-­ developed and include 45 programs, with cash assistance the dominant form across these schemes, collectively accounting for 1.36 percent of GDP65 as of 2020–2021. In 2021, 5.9 million households (number of households including pandemic social assistance) receive social assistance (cash or in-­ kind). The ratio of the resources allocated to cash aids in all aids is 94 percent. The Integrated Social Assistance Information System (ISAS), managed by the Ministry of Family, and Social Services (Social Assistance Directorate), links applicant data across over 120 databases from 28 institutions, using real time information to assess eligibility. Over 17 million house- holds (57 million individuals) are registered in ISAS’s Single Registry. During 2014–2019, the coverage rate for social assistance (notably cash transfers) in Türkiye was estimated to be approximately 60 percent among the lowest quintile based on an analysis of household survey data. 65 World Bank, Turkey Social Assistance Review (forthcoming). Mitigating Risk and Improving Resilience: Integrating Social Insurance 89 Türkiye has in place a broad set of policies to support the poorest households beyond cash transfers, including in-­ kind social care services for older needs, disabilities and children, with cash transfers thereby covering a relatively specific profile. Türkiye’s coverage of cash transfers is relatively lower than the coverage rate for the poorest quintile in Chile (93.8 percent), Russian Federation (83.4 percent), Romania (74.7 percent) and Mexico (88.3 percent)66. In addition, social assistance benefits are generally low, typically approximately 10 percent of household expenditure needs although some cash and in-­ kind benefits provide higher benefits. In general, however, this level is lower than in comparator countries that provide at least 15–20 percent. Nonetheless, Türkiye has taken an important step in its social assistance system with the “Family Support Program” expected over the coming period, building on its well-­established institutional, policy and delivery systems foundation through one of the most sophisticated information systems used to monitor, target and deliver benefits nationwide. Simulating Fiscal Space for Integrating Labor and Social Protection Schemes Greater integration among Türkiye’s social insurance programs, together with social assistance, can potentially improve coverage, efficiency and equity. Stylized, illustrative analysis shows that while greater fiscal space is needed for an integrated scheme, equity, incentives for productivity and administrative simplification improve for beneficiaries over term. Currently, costs are expected to rise from 12 to 15 percent of GDP, but the long-­ regressive financing, inequitable benefits and thirty percent informality would remain. Increasing spending by 2 percent of GDP allows a more progressive alignment of contributions and benefits, but coverage remains the same. For both fiscal harmonization and universal coverage to be achieved to address post-­ COVID recovery and begin a just, green transition, an increase of nearly 4 percent of GDP would likely be needed over the mid-­ term to invest in human capital and jobs gaps, essentially putting Türkiye on part with OECD countries. The current social insurance system would need to adapt in the following ways: (i) similar to health insurance, creating a simplified social insurance means-­ testing approach to collecting contributions and providing targeted subsidies to informal workers, rendering minimum pension and unemployment benefit-­setting explicit; (ii) integrating non-­contributory social assistance transfers into the contributory social insurance scheme, effectively rendering them as tiered subsidies (tapered basic income, i.e., TBI) to participate in a national risk pool; and (iii) harmonizing unemployment insurance and active labor market program parameters to strengthen incentives for labor force participation for individuals able to work. 66 World Bank Social Protection and Jobs ASPIRE Database, as of 2018–2021. 90 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Figure 6.6.  Fiscal scenarios for integrated, universal social risk protection system, illustrative 25% 20% 15% EXP as % GDP 0% 5% 0% Base case Reform (A) equity, coverage Reform (B) equity, rates remain constant universal coverage Workers contributions Conribution subsidies Active labor market programs Rest of the population Unemployed Elderly Social assistance (TBI) Contribution subsidies Unemployment insurance Payments to cover unfundes liabilities Contribution subsidies Pensions Source: World Bank staff calculations based on World Bank (2019) and Palacios and Robalino (2020). Key Implications Developing a universal risk-­ sharing scheme is intended to treat all households, regardless of income-­ earning status or type, are treated equitably within a unified policy framework. Contributions are more closely linked to benefits, with progressive and appropriately designed subsidies based on capacity-­ pay, regardless of sector or status. to-­ Informal workers and the self-­ employed would be able to enroll in the same system as formal wage workers, similar to how Türkiye’s reformed universal health insurance scheme operates. The key feature is to define benefits, contributions and subsidies more explicitly, a shift from the current system, enabling a more managed financing formula to adjust to evolving demand and unit costs over time. Importantly, integrating non-­ contributory social assistance with contributory social insurance schemes allows a more rational and equitable approach to offering subsidies based on capacity, facilitation and incentives to work. Building on Türkiye’s well-­ designed social information management systems, greater harmonization across schemes likewise allows outreach to informal and poor households, testing and contribution/savings management. including identification, profiling, means-­ Mitigating Risk and Improving Resilience: Integrating Social Insurance 91 92 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs 7 Outlook: Developing a Whole-­ of-­Government Approach for Transforming Human Capital Introduction Overall, COVID has exacerbated underlying human capital vulnerabilities in Türkiye, calling for renewed investments to boost allocative efficiency. While a more rigorous analysis of allocative efficiency is needed to quantify the determinants of human capital more substantively over time in Türkiye, key trends emerge. Social expenditures in some dimensions appear increasingly more efficient at boosting outcomes to date, largely due to having been better adapted to future needs early. Human capital as an aggregate index, basic learning outcomes, NEET rates and labor force participation among low-­ income adults, females, and youth are sensitive to multiple simultaneous factors in Türkiye, but certain investments and policies may boost these outcomes more than others. At an aggregate level and over the last decade, health spending in Türkiye appears to be the most efficient as compared to education, social protection and labor expenditures towards meeting sector-­ related HCI specific aims. This is evidenced by the relatively high health-­ components and the relatively lagging education-­ related components. Further, despite spending on social transfers, poverty has nonetheless increased, suggesting that targeting, benefit levels and the choice of instruments may need to be rebalanced. To improve efficiency towards poverty-­reduction goals, broadening related support to female youth, poorer households and informal employment-­and skills-­ workers, as opposed to increasing income support alone or expanding support to the formal Outlook: Developing a Whole-­of-­ Government Approach for Transforming Human Capital 93 sector, would likely boost efficiency of social spending. At this stage in Türkiye’s trajectory, accelerating the rate of investment in modernizing curricula, technology-­ based learning and teaching practices would boost learning outcomes to a greater extent than accelerating investments in basic infrastructure. Addressing Challenges, Maximizing Opportunities A robust, long-­ term vision of integrated, joint private sector and human capital investments can put Türkiye ahead of the curve as the digital and green economy accelerates. As of early 2021, Türkiye has started its post-­ pandemic recovery, though underlying challenges facing inclusive growth and human capital have resurfaced. The share of the population fully vaccinated has reached nearly 80 percent, students have returned to in-­person schooling, and the first half of 2021 has seen a rebound in some of the labor losses witnessed early during the pandemic. As vaccination coverage has increased and economic sectors have reopened, employment levels have begun resuming on aggregate largely in construction, followed by industry. However, low female labor force participation, a high rate of youth not in education, employment or training (NEET), poverty has deepened, and relatively high informality remains, making Türkiye stand out within Europe and Central Asia. While recovery appears to have started, the pandemic may have long-­ lasting consequences on these excluded groups. To avoid an uneven recovery particularly within the context of uncertain economic volatility, targeted policies for human capital would help foster inclusive, green growth. Given particular challenges and opportunities facing Türkiye’s green transition, fine-­ tuning ongoing investments on the demand and supply side simultaneously would help pave the way for a more robust, inclusive green transition. This analysis has demonstrated the limits of addressing the green transition challenges through silo-­based policies that tackle structural change, the demand for labor and human capital separately. While an exhaustive discussion of all aspects of the quality of public services is beyond the scope of this discussion, key policy measures come to the forefront over the short to long-­ term. Addressing the youth, especially female youth, employment challenge critical for recovery. Both demand and supply-­ side constraints need to be addressed in Türkiye to improve labor market entry and continuity for women and youth. Among multiple factors to be addressed, labor market dynamics reveal three specific constraints that should be addressed over the short-­to medium-­ term. First, on the demand side, the growth of firms and their hiring practices suggest that job growth has not kept up with labor force growth. intensive activities and lagging productivity in job-­ The mix of labor-­versus capital-­ creating sectors in agriculture and services have hampered job growth for low-­and high-­ skilled workers67. 67 World Bank (2019). Country Economic Memorandum: Firm Productivity and Economic Growth in Türkiye. 94 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Second, on the supply-­ side, in certain sectors such as manufacturing, industry and, increasingly, services particularly with the onset of COVID-­ 19, firms report a lack of adequate skills among applicants. Nearly 20 percent of firms find an inadequately skilled workforce is a major constraint to doing business.68 The need for a digitally skilled workforce that responds to the demand of firms as the economy becomes greener and more knowledge-­ based is key. Third, at the cross-­ roads between the demand-­and supply-­ side, informality has remained relatively high at approximately 30 percent. This suggests that incentives within the labor market for firms to continue employing workers informally and for undeclared self-­ employment to persist remain high. This may be due to the perceived rate of return of labor costs in Türkiye. The hike in minimum wage in 2016, for example, in the absence of other inducing measures, was shown to hamper firm growth and hiring notably among small labor-­ and medium enterprises (SME)69; the recent hike in 2021 will need to be evaluated in the future in light of recovery aims for investment and job creation. Government Human Capital Towards a Whole-­of-­ Platform Setting a new trajectory for inclusive growth in Türkiye will mean better targeting human capital investments to more specific aims for private sector-­ led growth. Public spend- ing over the past decades has played a strong role in reducing population-­ wide infectious disease, averting premature mortality, boosting basic literacy, and creating opportunities to allow a middle-­class to emerge. But skills needed for a 21st century economy are ineq- uitably distributed, especially digital, and advanced non-­ cognitive skills. In terms of social protection, one-­third of Türkiye’s total labor force of 34 million remain uncovered in terms of income and old-­ age risks as informal workers. Despite expanding health insurance cov- erage, addressing pandemic preparedness and the rapid increase in non-­ communicable morbidity remains increasingly challenging. The demand for social expenditures is increasing rapidly in the face of both a young population and an increasing cohort of over 65-­ olds, who are likely to remain active in the labor force for far longer than previous year-­ generations. Taken together, these factors put households’ ability to find gainful employ- ment, cope with shocks and boost productivity, into a vicious cycle, as highlighted by COVID. Moving forward, a whole-­ government, or integrated, approach to simultaneously of-­ boosting green sectors while redesigning human capital policies will boost equality of opportunity. Leveraging Türkiye’s past gains, reaching new human capital and jobs heights will likely imply deepening outreach to lagging regions and among female youth 68 WBG Enterprise Survey 2019. 69 World Bank (Bossavie, L; Acar, A; Makovec M) (2019). Do Firms Exit the Formal Economy after a Minimum Wage Hike? World Bank Policy Research Working Paper No. 8749. Washington DC: World Bank. Outlook: Developing a Whole-­of-­ Government Approach for Transforming Human Capital 95 for the greatest impact. Greater fiscal space and adaptation to evolving needs for social expenditures is needed to boost post-­ COVID recovery efficiently, notably in the following areas: 1. Over the short-­ term: stepping up incentives for closing the skills gap in key areas by: (a) substantively increasing investment in digital and green curricula and training, particularly in key lagging regions with lower learning scores; (b) despite adequate basic school infrastructure and the number of teachers, in line with the first aim, introducing new competency assessments and incentives for teacher training and performance to level the playing field across regions, in close cooperation with the private sector and line agencies responsible for industry, trade, agriculture, environment and others; 2. Over the mid-­ term, closing the quality of the school-­ work gap in green sectors by: to-­ (a) strengthening investments in secondary education through dedicated job counseling to-­ and training early on to facilitate the school-­ work transition; (iv) expanding coverage, new mechanisms for targeting and a more strategic design of demand-­ driven active labor market training programs and wage subsidies to address female labor force participation, digital transformation, and green jobs. 3. Over the long-­ term, unifying risk-­ sharing mechanisms and coverage for excluded groups by: (a) harmonizing benefits, parameters and financing approaches to social assistance, unemployment benefits, old-­ age (pensions) and health insurance programs within a unified, holistic social insurance system; (b) reforming eligibility for the poorest and informal workers to labor market programs; and (c) modernizing delivery systems through greater mobile outreach counseling, technology-­ based payment, identification, registration and training systems for the most vulnerable households, particularly lagging regions. These measures will likely require Türkiye to enhance public financing over the short-­to mid-­term as the labor force shifts to low-­and medium-­level carbon emissions sectors, including women and youth more substantively. Compared to most OECD countries, Türkiye’s spending on education, health and active labor market programs is relatively modest. By way of benchmarking the likely fiscal space needed to address the green 2008 economic shock and the COVID pandemic provide transition, learnings from the post-­ an indication of the types of measures and expenditures that will be needed. In the OECD, the drop in economic growth during 2007–2009 of around 6 percentage points (2.7 to –3.5 percent) was met by an increase in social expenditures of nearly 3 percentage points of GDP on average (17.7 percent of GDP to 20.7 percent of GDP)70. This level has been maintained since, owing to demographic changes, and long-­ lasting impacts on jobs and consumption in general. Similarly, because of COVID-­19, most advanced countries mobilized emergency social measures that accounted for 1–2 of GDP over 2020; Türkiye’s support packages targeting households and workers’ wages and benefits were relatively modest and estimated to account for up to 0.5 percent of GDP. 70 World Bank staff calculations using OECD Social Expenditures database and TUIK. 96 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Table 7.1.  Proposed policy reform matrix for boosting human capital and jobs towards inclusive, green transformation Policy Aims Short-­Term Reforms Long-­Term Reforms Targeting job creation • Strengthen access-­to-­finance • Enhance ISKUR job placement incentives for job creation among performance-­ based models with joint vulnerable and lagging contexts private sector, transparent mechanism • Align MoL labor contract regulations • Develop MoL-­ Financial Sector for hiring and dismissal, minimum Framework for scaling up targeting wage and severance pay policies job creation, entrepreneurship, innovation twinning and coached • Boost fiscal space and reforms to seed financing strengthen and expand coverage of MoNE secondary and tertiary school dedicated job counseling/OJT and placements Modernizing skills and • Boost fiscal space for targeted ISKUR • Boost fiscal space for region s ­ pecific job training OJT and VT performance incentives MoNE twenty-­ first skills curriculum and training, including governance • Boost fiscal space for expanding reforms for more rapidly adapting IKSUR to targeted MoFSS SA modernized curriculum development beneficiaries, multi-­dimensional, adapted targeting criteria to enhance • Enhance and adapt school autonomy female coverage, and outcome-­ models oriented case management • Boost fiscal space for teacher performance incentives Accelerating labor • Reform eligibility among youth • Target subsidies to support green job market entry and mobility NEET and informal sector for ISKUR investments and vocational training programs, • Strategically reform severance pay based policies including revisiting age-­ and unemployment benefits for • Consolidate wage subsidies, harmonization and equity targeting, eligibility and exit criteria Adapting risk mitigation • Reform social insurance structure • Consolidate contributory and non-­ and resilience for new forms of work, contracts, contributory policies, benefits, contributions, benefits-­setting targeting and coverage (SGK and and targeted subsidies for more MoFSS) progressive financing and greater • Develop policies for informal sector sustainability coverage and independent work/non-­ contractual Source: World Bank staff in consultation with Government of Türkiye and stakeholders. Outlook: Developing a Whole-­of-­ Government Approach for Transforming Human Capital 97 Given particular challenges facing vulnerable females, youth and southeastern regions, an integrated approach to social investments based on performance and outcomes can help set a new trajectory. This analysis has demonstrated the limits of addressing based policies. High female youth NEET rates, high female their challenges through silo-­ labor force exclusion and wide regional disparities in learning and jobs persists. While an exhaustive discussion of all aspects of the quality of public services is beyond the scope of this discussion, key policy measures come to the forefront over the near-­ term, combined with broader measures below. Emerging programs in Türkiye are already experimenting with conditioning and incentivizing financing on performance. Historically, performance-­ based financing has gained momentum and evolved over the past two decades worldwide, from poverty-­ reduction through cash transfers (conditioned on investing in girls’ human capital) to on-­ the-­ job training incentives to firms and vocational training (conditioned on females’ inclusion). Specifically, performance-­ based policies can be expanded across three priority areas facing the greatest lag in outcomes, although other programs can follow suit: (i) basic education in lagging regions to retain girls in school; (ii) across MoNE vocational training and ISKUR on-­ job training in key sectors with low female participation (manufacturing, industry and the-­ technology-­ based services); and (iii) within access to finance policies for firms, such as that being implemented currently by the Development and Investment Bank of Türkiye (TKYB).71 Conclusion Future detailed scenarios of the coverage and fiscal space by category of household and workers can help to further inform how early and how long to introduce targeted measures72. Measures to facilitate the green economy in Türkiye will require segmenting the current and expected future stock of workers and households by need. Policies would entail: (i) expanding green skills for basic and higher education, on-­ job training and the-­ skilling; (ii) targeting job creation through climate finance incentives; (iii) addressing re-­ informality and labor costs through social insurance consolidation, including harmonization of benefit levels, registry expansion and targeting streamlining across non-­ contributory and contributory social transfers; and improving demand-­ driven active labor market programs in terms of outreach and results-­ based financing towards green jobs. 71 World Bank (2020). Project Appraisal Document for a Formal Employment Creation Project (P171766). Report No PAD3491. Washington DC: World Bank. 72 See World Bank (forthcoming). Türkiye Public Finance Review, Human Capital Expenditures. 98 Türkiye in Transition: Next-­Generation Human Capital Investments for Inclusive Jobs Türkiye in Transition: Generation Human Capital Next-­ Investments for Inclusive Jobs A PUBLIC EXPENDITURE FRAMEWORK  |  2011–2021