DIAGNOSTIC TURKEY Page 1 of 114 © 2019 International Bank for Reconstruction and Development / The World Bank. 1818 H Street NW, Washington, DC 20433, USA. Telephone: 202-473-1000; Internet: www.worldbank.org. Some rights reserved 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 of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be 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 This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http://creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: Aysenur Acar and Ximena Del Carpio. 2019. “Turkey Jobs Diagnostic” World Bank, Washington, DC. License: Creative Commons Attribution CC BY 3.0 IGO. Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses 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. Images: © World Bank Tajikistan. Further permission required for reuse. Page 2 of 114 ACKNOWLEDGMENTS The report was prepared by Aysenur Acar Erdogan and Ximena V. Del Carpio. The team is grateful to the following colleagues for their excellent contributions: Reyes Aterido, Laurent Loic Yves Bossavie, Mattia Makovec, Efsan Nas Ozen, Habib Rab, Adrian Scutaru, Sirma Demir Seker, Pinar Yasar, [add after review meeting]. The team is grateful to the Government of Turkey more broadly for supporting this work. Numerous colleagues from the Ministry of Family, Labor and Social Security, the Ministry of Industry and Technology, World Bank, development partners, private sector, and the broad academic community supported the team in conceiving this study. The team received general guidance from Cem Mete (Practice Manager; Social Protection and Jobs). The team is grateful for the excellent advice provided by the peer reviewers Achim Daniel Schmillen, Metin Nebiler and Pinar Yasar. The team also thanks Gozde Yilmazturk and Derya Barlak for providing excellent administrative support. The report has been edited by Graham Colin-Jones. Page 3 of 114 CONTENTS EXECUTIVE SUMMARY ............................................................................................. 10 Economic Growth and Jobs ...................................................................................... 10 Workforce ............................................................................................................... 11 Recommendations to facilitate job creation in Turkey ................................................ 13 INTRODUCTION ....................................................................................................... 15 1. ECONOMIC GROWTH AND JOBS ....................................................................... 16 2. WORKFORCE ..................................................................................................... 24 Demographics, Labor Force Participation and Unemployment ................................... 24 Employment and Job Types ..................................................................................... 28 Profile of the workers ............................................................................................... 31 Inactivity................................................................................................................. 36 Education, Skills and Labor Market Outcomes .......................................................... 40 Labor force skills as measured by PIAAC, PISA, and TIMSS ....................................... 44 Technological change and skill composition in the labor market ................................. 46 Labor market returns ............................................................................................... 51 Workforce skills in relation to job creation and productivity growth ............................. 54 Job quality............................................................................................................... 56 3. LABOR DEMAND IN THE FORMAL SECTOR ...................................................... 61 Profile of formal sector labor demand ........................................................................ 61 The relationship between productivity and employment ............................................. 71 Firm perceptions of obstacles to performance ............................................................ 76 1. Informal competition and informality ............................................................... 76 2. Excessive labor regulation ............................................................................... 77 3. Inadequately educated and skilled workforce .................................................... 79 4. RECOMMENDATIONS TO FACILITATE JOB CREATION IN TURKEY ................. 80 1. Details of actions recommended ...................................................................... 84 APPENDIX ............................................................................................................... 102 BIBLIOGRAPHY........................................................................................................ 113 Page 4 of 114 ABBREVIATIONS EIS Entrepreneurship Information System EU European Union FDI Foreign Direct Investment GDP Gross Domestic Product GII Global Innovation Index GNI Gross National Income ICT Information and Communications Technology JQI Job Quality Index LFP Labor Force Participation LFS Labor Force Survey NEET Not in Education, Employment, or Training OECD Organization for Economic Co-operation and Development PIAAC Programme for the International Assessment of Adult Competencies PISA Programme for International Student Assessment PPP Purchasing Power Parity SME Small and Medium-sized Enterprise SuTP Syrian under Temporary Protection TFP Total Factor Productivity TIMMS Trends in International Mathematics and Science Study TUIK Agency of Statistics under the Republic of Turkey UN United Nations WB World Bank WDI World Development Indicators Page 5 of 114 LIST OF FIGURES Figure 1 Turkey’s GDP per capita (PPP), 1990-2017 ................................................................................16 Figure 2 Real GDP annual growth, 2000-2017 ...........................................................................................16 Figure 3 Poverty rate in Turkey, 2002-2016.................................................................................................17 Figure 4 Key labor market indicators, compound annual growth rate, 2005-2017 ...............................17 Figure 5 Export, FDI and Remittances, 2005-2017 ...................................................................................18 Figure 6 GDP and employment growth, 2005-2017 ..................................................................................19 Figure 7 Employment-GDP elasticities, quarterly, 2005 Q1- 2018 Q2 ...................................................19 Figure 8 Total employment and net increase in employment by sector ..................................................20 Figure 9 Shapley decomposition of per capita value-added growth.........................................................20 Figure 10 Employment and value added by major sector, 2005-2017 .....................................................22 Figure 11 Sectoral contributions to per capita value-added growth and employment growth, 2005- 2017 ....................................................................................................................................................................22 Figure 12 Labor productivity, 2001-2016 .....................................................................................................23 Figure 13 Demographic window for Turkey ...............................................................................................24 Figure 14 Profile of the working age population in Turkey (2017) ..........................................................25 Figure 15 Labor force participation rates, by gender and age group .......................................................26 Figure 16 Unemployment rates over time, all sectors ................................................................................26 Figure 17 Employed, unemployed, LFP in the population, 2017.............................................................27 Figure 18 Employed, unemployed, LFP in the population, by gender, 2017 .........................................27 Figure 19 Country comparison in terms of labor force participation rates, 2016 or the latest year ....28 Figure 20 Employment status by year and gender ......................................................................................29 Figure 21 Evolution of informality in Turkey .............................................................................................30 Figure 22 Evolution of informality rate in Turkey, by gender (in percent) .............................................30 Figure 23 The share of informal workers in sectoral total employment by employment status and sector, 2017 .......................................................................................................................................................31 Figure 24 Formality and informality rates across regions, 2017 ...............................................................32 Figure 25 Formality and informality rates by gender, age, education and economic sector, 2017 ......34 Figure 26 Number employed by contract type and informality (indexed to 2005=100) ......................35 Figure 27 Hours worked per week by contract type and informality (2005=100) .................................36 Figure 28 NEET (%) in selected countries, aged 15–29, 2016 .................................................................37 Figure 29 NEET rates (%) in Turkey over time, by gender and education level completed ...............37 Figure 30 Characteristics of NEETs (2017).................................................................................................38 Figure 31 Inactivity rates.................................................................................................................................39 Figure 32 Reasons for inactivity among the NEET, 2017.........................................................................39 Figure 33 Labor force participation by education level and type of employment by education level age 25-64, 2017........................................................................................................................................................40 Figure 34 Highest diploma obtained, by age cohort and gender, 2017 ...................................................40 Figure 35 Employment levels and growth by education level, age 25-64, 2017 .....................................41 Figure 36 Employment levels and growth by gender and education level, age 25-64, 2017 ................42 Figure 37 Shares of employment by education level and gender, age 25-64, 2017 ................................43 Figure 38 Informality by education level and gender, age 25-64, 2017 ..................................................43 Figure 39 Proficiency gaps by age group (2015, difference between mean proficiency scores) ..........44 Figure 40 Proficiency gaps between Turkey and OECD...........................................................................44 Figure 41 PISA scores (2003-2015)...............................................................................................................45 Figure 42 PISA scores and GDP per capita ................................................................................................45 Page 6 of 114 Figure 43 TIMSS Mathematics scores and ranks (Turkey, Grades 4 and 8, 2011 and 2015) ...............46 Figure 44 TIMSS Science scores and ranks (Turkey, Grades 4 and 8, 2011 and 2015) ........................46 Figure 45 Trends in labor shares in output every 10 years since 1975 ....................................................46 Figure 46 Distribution of occupational employment over the probability of computerization...........47 Figure 47 Global Innovation Index (2018, selected countries, score) .....................................................48 Figure 48 Employment trends by type of occupation in Turkey, 2009-2017 .........................................49 Figure 49 Employment composition by type of occupation in Turkey, 2010-2017 ..............................50 Figure 50 The percentage of firms offering formal training for their permanent, full-time employees (2018, selected countries, score) ....................................................................................................................51 Figure 51 Daily real wage for formal and informal sectors, 2007-2017 ...................................................52 Figure 52 Evolution of the daily minimum wage (both nominal and real) .............................................52 Figure 53 Evolution of real hourly wage, by occupation, 2009–2017......................................................52 Figure 54 Average monthly real wages in the formal and informal sector (2003=100) ........................53 Figure 55 Wage premium over time in Turkey............................................................................................54 Figure 56 Evolution of nominal minimum wage in Turkey ......................................................................54 Figure 57 Employment composition by type of occupation and firm size (2016) ...............................55 Figure 58 Job creation (2014-2016) by type of occupation and firm size ..............................................55 Figure 59 Employment composition by firm type based on skill intensity (in percent) .......................56 Figure 60 Distribution of Job quality ............................................................................................................57 Figure 61 Job quality index (JQI) by contract type, 2009-2017 ................................................................57 Figure 62 Scores by component of the JQI, 2014-2017 ............................................................................57 Figure 63 Job quality of wage employment, 2017 .......................................................................................58 Figure 64 Distribution of formal firms by province, 2016 ........................................................................62 Figure 65 Distribution of employment by province, 2016 ........................................................................62 Figure 66 Annual growth in the number of firms between 2007-2016 (in percent) ..............................62 Figure 67 Annual growth in employment between 2007-2016 (in percent) ...........................................63 Figure 68 Distribution of firms and employment share by firm size, 2007-2016 ..................................63 Figure 69 Employment accounted for by largest firms ..............................................................................64 Figure 70 Revenue accounted for by largest firms......................................................................................64 Figure 71 Distribution of firms and employment share by firm’s age, 2007-2016 ................................65 Figure 72 Ownership status and firm size 2015–2016 ...............................................................................66 Figure 73 Gender composition of ownership and management ..............................................................67 Figure 74 Distribution of firms share by sector, 2007-2016 .....................................................................67 Figure 75 Distribution of employment share by sector, 2007-2016 .........................................................68 Figure 76 Annual change in number of firms and employment, by sector, 2007-2016 ........................68 Figure 77 Job flows in Turkey........................................................................................................................71 Figure 78 Average firm productivity .............................................................................................................72 Figure 79 Value added per worker, by firm size..........................................................................................73 Figure 80 Value added per worker, by firm age ..........................................................................................73 Figure 81 Change in value added per worker, by sector ............................................................................73 Figure 82 The ratio between capital and labor, by sector ..........................................................................74 Figure 83 The ratio between capital and labor cost ....................................................................................74 Figure 84 Concentration of workers around the minimum wage by sector, 2016 .................................75 Figure 85 Percentage of firms identifying various obstacles to their performance................................76 Figure 86 Informal competition, by region, 2015–2016 (percentage of firms reporting it as a top obstacle).............................................................................................................................................................77 Figure 87 Labor regulation, by region, 2015–2016 (percentage of firms reporting it as a top obstacle) ............................................................................................................................................................................78 Page 7 of 114 Figure 88 Inadequately educated workforce, by region, 2015–2016 (percentage of firms reporting it as a top obstacle)...................................................................................................................................................79 LIST OF TABLES Table 1 Key factors of good job creation ........................................................................... 59 Table 2 Transition Matrix................................................................................................ 71 Table 3 Challenges in the Turkish labor market, as identified through this diagnostic ............. Error! Bookmark not defined. Table 4 Employment and job creation, Turkey’s 10th National Development Plan ...................... 82 Table 5 Summary of the actions proposed to address Turkey’s labor market challenges ............... 83 APPENDIX App 1 Nuts Classification of Turkey ................................................................................ 102 App 2 Manufacturing NACE Rev.2 2digit Subsectors .......................................................... 103 App 3 Construction NACE Rev.2 2digit Subsectors ............................................................ 104 App 4 Service Nace Rev.2 2-digit Subsectors and Sector Classifications .................................. 104 App 5 Variables, definitions and formulas used in Labor demand chapter ............................... 106 App 6 The percentage of the number workers at around minimum wage in sectoral employment, Accommodation sector, by province, 2016 ........................................................................ 107 App 7 The percentage of the number workers at around minimum wage in sectoral employment, Construction and Specialised construction activities sector, by province, 2016 .......................... 107 App 8 The percentage of the number workers at around minimum wage in sectoral employment, Education sector, by province, 2016................................................................................. 107 App 9 The percentage of the number workers at around minimum wage in sectoral employment, Architectural and engineering activities, technical testing and analysis sector, by province, 2016 ... 108 App 10 The percentage of the number workers at around minimum wage in sectoral employment, Entertainment and recorded media sector, by province, 2016 ................................................ 108 App 11 The percentage of the number workers at around minimum wage in sectoral employment, Food and beverage service activities sector, by province, 2016 .............................................. 108 App 12 The percentage of the number workers at around minimum wage in sectoral employment, Forestry and fishing, by province, 2016............................................................................. 109 App 13 The percentage of the number workers at around minimum wage in sectoral employment, Legal, accounting, financial and insurance sector, by province, 2016 ....................................... 109 App 14 The percentage of the number workers at around minimum wage in sectoral employment, Manufacture of food products sector, by province, 2016 ...................................................... 109 App 15 The percentage of the number workers at around minimum wage in sectoral employment, Manufacture of furniture sector, by province, 2016 ............................................................. 110 App 16 The percentage of the number workers at around minimum wage in sectoral employment, Manufacture of leather sector, by province, 2016 ................................................................ 110 App 17 The percentage of the number workers at around minimum wage in sectoral employment, Manufacture of textiles sector, by province, 2016................................................................ 110 Page 8 of 114 App 18 The percentage of the number workers at around minimum wage in sectoral employment, Manufacture of wearing apparels sector, by province, 2016 ................................................... 111 App 19 The percentage of the number workers at around minimum wage in sectoral employment, Manufacture of wood sector, by province, 2016 ................................................................. 111 App 20 The percentage of the number workers at around minimum wage in sectoral employment, Transportation sector, by province, 2016 .......................................................................... 111 App 21 The percentage of the number workers at around minimum wage in sectoral employment, Wholesale and retail sector, by province, 2016.................................................................... 112 Page 9 of 114 EXECUTIVE SUMMARY Economic Growth and Jobs Turkey’s economy has grown significantly since the early 2000s, though not always at a steady pace. Between 2002 and 2006, the annual GDP growth rate exceeded 7 percent. Turkey was then hit hard by the global financial crisis, with output declining by almost 5 percent in 2009. But unlike many other countries around the world, the economy bounced back quickly, with growth reaching about 9 percent in 2010 and 2011. Since then, GDP growth has slowed significantly to about 3 percent in 2016 and was 7 percent in 2017. The poor have benefited from this strong growth. Between 2002 and 2016, the poverty rate declined from 32 percent to 10 percent. Nevertheless, there were still 8 million poor in the country in 2016, with about 10 percent of the population below the poverty line of US$5.5 per day (2011 PPP). The inflow of foreign capital through exports, FDI and personal remittances has been decreasing at constant 2009 US$ prices. While the inflows were significant at current prices between 2005 and 2017, these indicators have not been performing so well in terms of constant prices in recent years. After a peak around 2008, a decreasing trend has become evident. The period of strong economic growth was accompanied by significant employment growth and an expansion of the formal sector. Employment grew by 6 percent from 2005 to 2009. During the period of recovery from the 2008 global financial crisis (2009–2012), employment increased by 17 percent. However, the economic slowdown in recent years has negatively affected the pace of job creation. Employment increased by 4 percent between 2012 and 2014, while from 2014 to 2017, employment and formal employment grew 9 percent and 10 percent, respectively. Rapid job creation has been accompanied by flat productivity. Such a phenomenon, which is not unique to Turkey,1 indicates that workers and firms are not producing goods and services in the most efficient manner, which affects Turkey’s competitiveness. One of the important ingredients to increasing productivity is having a labor force that is adaptable to changing conditions and flexible enough to gain new skills for changing occupational demands. Thus, low productivity could be a symptom of the inadequacy of skills, which reduces economic growth, limits employment opportunities, and exacerbates social pressures (Arias and Sanchez-Paramo 2014; EU Commission 2012). About half of the jobs created in 2015–2016 are in sectors with low or medium levels of productivity. Similarly, as the shift of employment out of low-productivity agriculture into higher- productivity industry and services has slowed down, associated productivity gains are likely to decline unless workers and firms in the manufacturing and services sector become more productive. Productivity stills well below OECD average, but it has been the main driver of GDP growth since 2005. Per capita value added has grown by about 3.5 percent per year. Improvements in productivity accounted for about 63 percent of the variation in value added growth; the labor force participation rate grew annually by 1.1 percent, accounting for 31 percent of the change in per capita value added. In other words, had everything else stayed the same, the change in the participation rate 1 In the United Kingdom, for instance, this is explained by a dramatic fall in real wages after the crisis, which essentially led to a substitution effect in favor of labor but runs the risk of creating longer-term competitiveness problems once real wages recover. Page 10 of 114 would have generated growth equivalent to 31 percent of the actual observed growth. The working- age population grew by 0.3 percent per year between 2005 and 2017, accounting for 7 percent. The employment rate decreased by 0.03 percent annually on average between 2005 and 2017; its contribution to the growth was negative. A relatively rapid increase in employment can reflect a negative contribution to growth when slow value-added growth is observed, since the growth is measured in per capita terms. Growth in overall labor productivity was driven by the industry sector. The industry sector made the largest contribution of 28 percent to labor productivity, although industry employment did not change. The services sector contributed positively to the change in both the employment rate (14 percent) and productivity (18 percent). On the other hand, the agriculture sector made only a limited contribution (5 percent) to productivity and contributed negatively to employment (-15 percent). Workforce The rate of job creation has not kept pace with the number of labor market entrants. This is particularly important given the country’s demographic trends. According to TUIK, the Turkish population is expected to peak around the year 2050. Until about 2025, Turkey will have a falling dependency ratio; after that, rising old-age dependency will outweigh falling youth dependency. For Turkey to reap a demographic dividend, it needs to create jobs fast enough to accommodate the youth and women entering the workforce. Each year, about 700,000 new jobs are needed (World Bank 2014). Turkey has a chance to grow rich before it gets old, but for this to happen, structural reforms to boost job creation are needed. With many new entrants to the labor market (not all skilled), the economic slowdown has resulted in increased unemployment (the increase was 1 percentage point~600 thousand unemployed people from 2014 to 2017). Turkey’s human capital is underutilized, exacerbated by the high rate of NEET. As of 2017, about 19.5 million of the working-age population are not in education, employment or training (NEET). Of the 59 million working-age population, about 28 million are not in the labor force. While about 9 million of the latter are in school or training, the remaining 19 million are not in school. This means that they are not contributing to the economy now or improving their own skills to be able to contribute in the future. Over 50 percent of NEET are involved in domestic work, while the highest concentration of NEET is among people with less than high school completion. Between 2014 and 2017, overall labor force participation rates increased slightly (2 percentage points). The female labor force participation (LFP) rate increased from 30 percent to 34 percent, while the male LFP rate increased from 71 percent to 72 percent. The unemployment rate was highest among the younger cohorts (especially for the 20–24 age group). The LFP rate of young people (20– 24) was 58.6 percent, and their employment rate was 46 percent in 2016. Despite a substantial decrease in the informality rate, the informal sector remains a large source of jobs for workers in Turkey. Between 2005 and 2017, informality in Turkey decreased by 14 percentage points (from 48 percent to 34 percent). However, in recent years, the decrease has been slight or even stagnated. In non-agriculture, informality gradually decreased until 2015, after which it started to increase slightly. As of 2017, 22.1 percent of employment in non-agriculture was informal. Informality among wage and casual workers in non-agriculture has remained remarkable (13.3 percent), showing only slight decreases since 2012. Page 11 of 114 Informal workers (including unpaid family workers) tend to be female, older and less educated and work in agriculture, construction or low and medium technology services. The economic sectors requiring low or medium technology capabilities are more likely to employ low skill and informal workers. This confirms the expectation that informality in the market decreases as educational attainment increases. Premiums for university educated workers increased from 2009 to 2012 and then fell from 2012 to 2016. The decline in 2016 was particularly sharp. The wage premium remained roughly at the same level for secondary school educated workers compared to primary educated workers. The increase in the supply of better educated workers might be running ahead of increases in demand. As the education level has been steadily increasing in Turkey, increases in the relative supply of better educated workers may have put pressure on the earnings premiums, leading to a sharp decline. High school graduates are sought after as indicated by the high growth of real wages and wage premium, and the large influx of better-educated workers that is easily absorbed by the labor market. However, firms continue to demand workers with low and medium skill levels, implying that wage premium increased substantially for them. Labor demand in the formal sector Like many developed countries, there are clear differences in labor market dynamism across regions in Turkey. Formal firms are concentrated in economic centers in the west such as Istanbul, Ankara, Aegean, and East Marmara. Such subregional asymmetry is exacerbated by differences in firm characteristics, obstacles faced, and opportunities encountered, and overall firm performance. As such, obstacles to doing business are more likely to be accentuated in provinces where the density of firms is higher than in provinces with fewer formal firms. Labor demand has been affected by two key factors. First, a large subset of firms has not expanded over time, even though one in four small firms has been operating for at least 20 years. This finding could signal that smaller firms face obstacles to grow beyond a certain threshold, limiting their potential to serve as an engine of employment growth. Second, nominal wage increases of more than 100 percent since 2006 have contributed to stagnant labor productivity, probably adversely affecting the country’s ability to create good jobs. As labor costs (namely wages) increase, firms face the risk of losing their competitive advantage. Expanding high-technology production and moving to more profitable activities within the value chain2 are critical steps to fast-tracking economic development and the widespread creation of good jobs. Firms state three main obstacles to doing business and creating more jobs: (i) unfair competition from firms that avoid taxes and employ workers informally, (ii) high labor costs, and (iii) inadequately skilled workers. Employers find that various dimensions of the doing business and regulatory environment remain unfavorable and constrain them from creating jobs. For instance, labor regulation and costs, and persistent informality that results in uneven competition among stakeholders, are just some examples of what formal firms perceive as hindering their ability to create quality jobs. Turkey’s minimum wage is high and binding in the formal sector. The cost of hiring additional workers is particularly high when put into perspective with the relatively low and diminishing marginal product of labor—that is additional workers who may not have the required 2 This includes, for example, research and development (R&D), design, marketing, specialized logistics, or after-sale service. Page 12 of 114 skills. To adjust to changes in demand, firms therefore prefer to hire additional workers informally— and stay small when counting formal workers. Hence, facilitating the creation of formal jobs partly means altering the cost of informality relative to the marginal product of labor. While the government has already put in place certain measures to fight informality, it now needs to introduce incentives for smaller firms to employ formal workers. Such incentives could include measures to increase the marginal product of labor to improve the employability of workers, especially those with a low skills base. Employment activation policies are part of the answer, but automation and job de-routinization will likely reduce the demand for low-skilled (formal) workers in the long run. Ensuring that Turkey’s workforce can engage in productive jobs requires sustained and systematic efforts to improve the country’s education and training system and educational outcomes. Recommendations to facilitate job creation in Turkey Actions to Address Labor Market Challenges 1 Strengthen the investment climate through policy predictability and transparency; lower administrative burden on businesses; an efficient tax system; access to longer-term finance; and agglomeration economies to accelerate longer-term productive investments including in R&D. 1.1 2 Strengthen the efficiency of fiscal support for businesses, including through better targeting of Growth and support to firms with high growth and/or large job creation potential, winding down fiscal productivity support that keep inefficient firms in business. 3 Support firms to adopt technology by reforming labor market policies that disincentivize them from doing so and increasing worker support measures to obtain complementary skills. 1 Reductions in excessive employment costs, elimination of employment barriers, and increase the capacity for monitoring compliance with (enforcement of) labor regulations. 2 Increase flexibility of types of formal employment so firms can adjust to changing markets while increasing worker protections through a reformed unemployment insurance system, social 1.2 assistance with a strong graduation emphasis, and re-employment support services. Formalization 3 Streamline new business setup procedures and exit procedures, including targeting support (financial and non-financial) to new firms entering the economy while reducing financial support to inefficient firms that rely on subsidies to stay in business. 4 Promote the positive perception and practice of formality in society and strengthen informality prevention measures. 1 Rationalize existing employment support programs by targeting them to support disadvantaged workers (youth, women, disabled, refugees), ensure their effectiveness in the long term by accompanying them with structural reforms that improve employability (e.g. education and skills, regulations that reduce youth and women's competitiveness). 2 Improve employment related services, including operational efficiency and using technology and data-based tools to improve service delivery. Continue to monitor and address occupations 1.3 and skills requirements and consider matching jobseekers to jobs based on skills not occupations. Job creation 3 Review and reform financial support programs to firms, to increase their impact on productivity and sustain jobs by targeting their support to economically sound firms with high growth potential and plans for longer-term productive related investments. Shift support from (old and stagnant) SMEs and toward younger start-ups and vibrant SMEs that bring innovation and create new jobs. 4 Strengthen system of incentives (monetary and non-monetary) to support fast growing firms and sectors and those that have high export potential. 1 Facilitate improvements in the provision of quality affordable childcare, especially among the 1.4 most vulnerable women, for whom it should be strongly subsidized. Reform policies to allow Labor force flexible working times and improve parental benefits to prevent gender discrimination. participation 2 Rationalize employment support programs by targeting them to support disadvantaged workers in firms that are productive and economically sound and can sustain employment once the subsidy Page 13 of 114 is done. (Follows recommendation 1.3.1) Introduce accompanying non-financial measures to promote economic activity. 3 Improve employment related services to increase their effectiveness placing disadvantaged groups into sustainable jobs. Customize service delivery and strengthen referral system for expert services (beyond employment) to address distinct needs that disadvantaged groups may have. 1 Modernize vocational and lifelong learning systems, provide flexible adult learning options which offers modular courses, which lead to certifications and improvements in qualifications. 1.5 2 To improve the skills of existing workers and promote upskilling of the workforce, consider Workforce financial support to employers for allowing (lower-skill and at-risk of dismissal) staff to take time skills and off to train in recognition of the costs they will incur and to promote lifelong learning. talent 3 Facilitate the establishment of employer-led sector skills councils that focus on skills development that address the needs in the economy. Councils should ensure occupational standards, learning provision and qualifications meet sector standards. Page 14 of 114 INTRODUCTION The main challenge for achieving a country’s development goals is the creation of more, better and inclusive jobs, as economic growth needs to be accompanied by job growth if the poor are going to benefit to any significant extent. Turkey’s rapid economic growth since the early 2000s has been studied by the World Bank and others, highlighting the role of comprehensive reforms that promoted the country’s integration into the global economy, facilitated structural change, and catalyzed job creation. However, the big question is: Is Turkey making progress in achieving its development goals? The answer depends on the extent to which Turkey’s economic growth is able to generate employment and earning opportunities on a sustainable basis. Therefore, the objective of this report is to present a comprehensive Jobs Diagnostic for Turkey with a view to promoting inclusive economic growth and poverty reduction. The report also aims to lay the evidence base to support policy dialogue in the country and provide input for the development of a national jobs strategy for the delivery of more, better, and inclusive jobs in Turkey. Understanding the factors that influence the creation of more, better and inclusive jobs essentially requires a multisectoral approach. Therefore, this Jobs Diagnostic aims to assess the relationships between supply- and demand-side factors. The report utilizes a variety of data sets to provide a comprehensive analysis on the labor market in Turkey; these are summarized in Box 1. Box 1 Summary of key sources used in the Turkey Jobs Diagnostic • Household Labor Force Surveys (LFSs): The LFS is a large household-based survey providing results on the labor participation of people aged 15 years and older, as well as people outside the labor force. • Entrepreneurship Information System (EIS): The EIS has been collected by Ministry of Industry and Technology and includes enterprise level data that explores a number of firm characteristics and performance variables in the period 2007-2016. Enterprises report on their total revenue and employment, their location and a detailed sector code (4-digit code, but excludes military service, banks and financial institutions) and worker information (such as age, gender, wage, number of days worked and occupations). • Enterprise Survey 2015/16 (World Bank): The Enterprise Survey is a firm-level survey of a representative sample of registered firms operating mainly in the private sector. The survey covers a broad range of business environment topics including access to finance, corruption, worker skills, infrastructure, competition, and performance measures. The rest of the report is organized in four chapters, with a number of topical sections in each chapter. Chapter 1 provides an overview of Turkey’s economic development and broad structural changes in terms of job creation and labor productivity growth. Chapter 2 presents the results of the analysis of labor supply, mainly based on the Turkish Household Labor Force Survey for several years until 2017. It looks at: 1) demographics and labor force participation; 2) employment and types of jobs; 3) profile of the workers; 4) inactivity 5) education, skills and labor market outcomes; and 6) the role of skills. Chapter 3 focuses on labor demand and labor productivity using firm-level data from the Administrative Records (EIS, 2007-2016). The chapter discusses: 1) the profile of labor demand; 2) firm growth and job creation; 3) productivity and jobs and 4) firm perceptions of obstacles to performance. Finally, Chapter 4 outlines a set of policies and programs to address the country’s main challenges in terms of job creation, labor productivity and earnings growth, and access to jobs. Page 15 of 114 1. ECONOMIC GROWTH AND JOBS Turkey’s economy remains relatively strong due to significant achievements in several economic fields over the past decades. While many factors have contributed to Turkey’s successful transformation to a more diversified, higher quality jobs sector, a more dynamic private sector will be crucial to even greater success. Thanks to sound public financial management and structural reform of the banking sector, Turkey was able to achieve single-digit inflation after the 2001 economic crisis. GDP per capita was around 23 thousand (Figure 1) in 2007. Although these improvements placed Turkey in a better position compared to many other countries during the global financial crisis, it still experienced negative growth during 2008 and 2009. But the recovery was also strong, with growth reaching about 8 percent in 2010 and 11 percent 2011 (Figure 2). Since then, output growth has slowed significantly to about 3 percent in 2016 and reached to 7 percent in 2018. Figure 1 Turkey’s GDP per capita (PPP), 1990- Figure 2 Real GDP annual growth, 2000-2017 2017 12% 30,000 10% 8% 25,000 Constant 2011 INTERNATIONAL US$ Real GDP growth, % 6% 4% 20,000 2% 15,000 0% 2004 2000 2001 2002 2003 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 -2% 10,000 -4% -6% 5,000 -8% ECA GDP growth (annual, %) 0 Turkey GDP growth (annual, %) Source: Authors’ elaborations based on World Development Indicators The poor have benefited from high growth performance in the last decade. Between 2002 and 2016, the poverty rate declined from 32 percent to 10 percent (Figure 3). Although Turkey has witnessed significant poverty reduction during the 2000s, there were still about 8 million poor in the Page 16 of 114 country in 2016. About 10 percent of the population remains below the poverty line of $5.5 per day (2011 PPP). Figure 3 Poverty rate in Turkey, 2002-2016 40 35 Percent of population 30 25 20 15 10 5 0 2010 2002 2003 2004 2005 2006 2007 2008 2009 2011 2012 2013 2014 2015 2016 $1.90 a day (2011 PPP) (% of population) $5.50 a day (2011 PPP) (% of population) Source: Authors’ elaborations based on World Development Indicators Turkey has benefited significantly from a demographic transition through steady growth in the working-age population. Between 2005 and 2017, the working-age population grew by 1.8 percent per year (Figure 4). The participation rate grew annually by 1.1 percent, bringing millions of women into the labor force (the female employment rate increased 2.8 percent per year over this period). At the same time, the unemployment rate increased for individuals aged 15 years and over (0.2 percent per year) and for those aged between 15 and 24 years (0.4 percent per year). Figure 4 Key labor market indicators, compound annual growth rate, 2005-2017 0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% Total population 1.5% Total population of working age 1.8% Total labor force 2.9% Total number of employed 2.9% Total number of employed (non-agricultural) 3.6% Share of population of working age 0.3% Employment rate 1.1% Employment rate (non-agricultural) 1.6% Male employment rate 0.3% Female employment rate 2.8% Labor force participation rate 1.1% Unemployment rate 0.2% Youth unemployment rate (15-24) 0.4% Source: Authors’ elaborations based on TUIK data Inflow of foreign capital through exports, foreign direct investment (FDI) and personal remittances has been decreasing at constant 2009 US$ prices. Despite an impressive Page 17 of 114 performance in the export of goods and services, FDI and personal remittances at current prices between 2005 and 2017, these indicators have not been performing so well in terms of constant prices in recent years (Figure 5). After reaching a peak in 2008, they reflect a decreasing trend at constant 2009 US$ prices. Figure 5 Export, FDI and Remittances, 2005-2017 200,000 0.4 3,000 180,000 0.3 2,500 160,000 140,000 0.3 2,000 Percentage 120,000 millions millions 0.2 100,000 1,500 80,000 0.2 60,000 1,000 0.1 40,000 0.1 500 20,000 0 0.0 0 2007 2012 2005 2006 2008 2009 2010 2011 2013 2014 2015 2016 2017 2005 2007 2009 2011 2013 2015 2017 Exports of goods and services (BoP, constant 2009 Personal remittances, received (constant US$) 2009 US$) Foreign direct investment, net inflows (BoP, constant 2009 US$) Personal remittances, received (% of GDP) Source: Authors’ elaborations based on World Development Indicators Economic growth in the last decade has translated into significant job creation. Between 2005 and 2017 the economy added more than 8 million jobs. During this period, real GDP grew by an average of 6.9 percent per year, as employment increased by 3.4 percent per year (Figure 6). This implies an average employment to growth elasticity of around 0.49 (Figure 7). Page 18 of 114 Figure 6 GDP and employment growth, 2005-2017 Figure 7 Employment-GDP elasticities, quarterly, 2005 Q1- 2018 Q2 8% 0.90 CHANGE IN EMPLOYMENT (ANNUAL) 2005- 7% 0.80 6% 0.70 5% 0.60 4% 2017 0.50 3% 0.40 0.81 2% 0.30 1% 0.51 0.20 0.42 0% -10% -5% 0% 5% 10% 15% 0.10 -1% GDP GROWTH RATE (ANNUAL), 2005-2017 0.00 2005 Q1 - 2009 Q4 - 2014 Q2 - 2009 Q3 2014 Q1 2018 Q2 Source: Authors’ elaborations based on TUIK data In the last few years, employment in Turkey has shifted away from agriculture and industry towards service activities. The share of employment in agriculture declined by 10 percentage points between 2004 and 2017, indicating a structural transformation in the labor market. Nonetheless, 19 percent of those who are employed continue to work in agriculture, which is the sector with low labor productivity and high informality. During the same period, the share of those employed in industry decreased from 20 percent to 19 percent, while the share in construction increased from 5 percent to 7 percent. The services sector occupies an increasingly important position in the Turkish economy, with its share rising from 46 percent to 54 percent (Figure 8). Page 19 of 114 Figure 8 Total employment and net increase in employment by sector 30,000 800 600 25,000 400 20,000 200 15,000 0 10,000 -200 -400 5,000 -600 0 -800 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Other services Agriculture Wholesale and trade Industry Transportation and communication Construction Construction Transportation and communication Industry Wholesale and trade Agriculture Other services In Turkey, labor productivity, measured as value added per worker, has been the main driver of GDP growth since 2005.3 Per capita value added has grown by about 3.5 percent per year, with the factors contributing to this outcome being shown in Figure 9. • Improvements in productivity accounted for about 63 percent of the variation in value added growth; • The labor force participation rate grew annually by 1.1 percent, accounting for 31 percent of the change in per capita value added. In other words, had everything else stayed the same, the change in the participation rate would have generated growth equivalent to 31 percent of the actual observed growth; • The working-age population grew by 0.3 percent per year between 2005 and 2017, accounting for 7 percent; and • The employment rate decreased by 0.03 percent annually on average between 2005 and 2017. Its contribution to the growth was negative. A relatively rapid increase in employment can reflect a negative contribution to growth when slow value-added growth is observed, since the growth is measured in per capita terms. Figure 9 Shapley decomposition of per capita value-added growth 3 In the analysis, Shapley decomposition was used to understand how growth is linked to changes in employment, labor force, labor productivity and working-age population at the aggregate level and by sector. Growth in aggregate per capita value added can be described by the growth in productivity and employment relative to demographics. Page 20 of 114 6.0 Annual Change (percentage points) 5.0 4.0 Productivity 3.0 Employment Rate 2.0 Participation Rate Demographic Change 1.0 Total Period 0.0 -1.0 -2.0 2005-2017 2005-2008 2008-2012 2012-2017 y=3.5% y=2.91% y=3.3% y=4.01% y = Annual Growth Per capita Value Added (Percent) Source: Authors’ elaborations based on TUIK data, World Bank staff calculations using WBG Growth Decomposition Tool The highest change in value added by sector was in industry. Despite the highest share of services in total employment, growth in value added in services remained modest compared to the industry sector, which has greater value added to the Turkish economy than other sectors. As for agriculture, the agriculture sector is still dominant in the economy (with a share of 19.4 percent of total employment). The change in labor productivity in the agriculture sector was consistently positive over time. The services sectors were the main contributors to job creation. Between 2005 and 2017, employment expanded by an annual average of 3.4 percent, driven by the services sectors. Official TUIK data reveals that approximately 8.2 million jobs were created during this period, of which 69 percent were created in the services sectors (5.6 million), 14 percent in industry (1.2 million), and 12 percent in construction (Figure 10). Page 21 of 114 Figure 10 Employment and value added by major sector, 2005-2017 Employment (in thousands) 20000 11.5 8.0% Industry Logharitm of Value added per worker 11 Annual average value added growth, 2005-2017 15000 7.0% 10.5 Services 10000 10 6.0% 9.5 5000 5.0% 9 4.0% Agriculture 0 8.5 1 2 3.0% Agriculture 2.0% Industry Services 1.0% Agriculture - value added (2nd axis) 0.0% Industry - value added (2nd axis) -1.0% 0.0% 1.0% 2.0% 3.0% 4.0% 5.0% Annual average employment growth, 2005-2017 Services - value added (2nd axis) Source: Authors’ elaborations based on TUIK data Note: Bubble size represents sectoral employment in 2005. Growth in overall labor productivity was driven by the industry sector. The industry sector made the largest contribution of 28 percent to labor productivity, although industry employment did not change. The services sector contributed positively to the change in both the employment rate (14 percent) and productivity (18 percent). On the other hand, the agriculture sector made only a limited contribution (5 percent) to productivity and contributed negatively to employment (-15 percent) (Figure 11). Figure 11 Sectoral contributions to per capita value-added growth and employment growth, 2005-2017 35% 30% 25% 20% Contributions 15% 28% 10% 18% 5% 13% 14% 5% 0% 0% Industry Services etc. Industry Services etc. Agriculture Agriculture Intersectoral Reallocation -5% -15% -10% -15% Effect -20% Change in Productivity Change in Employment rate Page 22 of 114 Source: Authors’ elaborations based on TUIK data, World Bank staff calculations using WBG Growth Decomposition Tool Labor productivity in Turkey has improved over the past 15 years, but remains far behind the levels seen in OECD and EU28 countries; the low productivity is related to the structure of Turkey’s labor demand. Figure 12 shows labor productivity in terms of GDP per hour worked for Turkey, the OECD and EU28 countries between 2001 and 2016. Turkey has been performing worse than both the OECD and EU28 throughout this period; the situation has not improved over the years. In fact, it is particularly worrisome that labor productivity in Turkey stagnated from 2015 to 2016. Together, micro firms (less than 5 workers) and small firms (between 5 and 19 workers) account for about 90 percent of all (formal) firms in Turkey.4 This fact is discussed in chapter 3, which focuses on labor demand. Smaller firms tend to operate in low-value-added sectors of the economy, offering jobs with less favorable conditions such as lower wages, fewer benefits, and limited career growth options (Paula and Scheinkman 2007). They are more likely to face obstacles that limit their opportunities to innovate and grow. In Turkey, smaller businesses are much more limited in accessing finance or benefiting from trade liberalization and international capital flows (Sener et al. 2014). With limited access to capital, smaller firms are more likely to use suboptimal technologies, which in turn can negatively affect labor productivity, employment growth, and economic growth. Figure 12 Labor productivity, 2001-2016 140 GDP per hour woked (US dollars) 120 100 80 EU28 60 OECD TUR 40 20 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Source: Authors’ elaborations based on OECD data Note: GDP per hour worked is a measure of labor productivity. It measures how efficiently labor input is combined with other factors of production and used in the production process. Labor input is defined as total hours worked of all persons engaged in production. Labor productivity only partially reflects the productivity of labor in terms of the personal capacities of workers or the intensity of their effort. The ratio between the output measure and the labor input depends to a large degree on the presence and/or use of other inputs (e.g. capital, intermediate inputs, technical, organizational and efficiency change, economies of scale). This indicator is measured in USD (constant prices 2010 and PPP) and indices. 4 Enterprise Survey 2015–2016. Page 23 of 114 2. WORKFORCE Demographics, Labor Force Participation and Unemployment The demographic window is open for Turkey. Until about 2025, Turkey will have a falling dependency ratio; after that, rising old-age dependency will outweigh falling child dependency (0-14 years old) (Figure 13). The Turkish population is expected to peak around the year 2050. For Turkey to reap a demographic dividend, it needs to create jobs fast enough to accommodate youth and women entering the workforce. Turkey has a chance to grow rich before it gets old but for this to happen, structural reforms to boost job creation are needed. Figure 13 Demographic window for Turkey 100 80 Change in dependency ratio 60 40 20 0 1965 2034 1950 1953 1956 1959 1962 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013 2016 2019 2022 2025 2028 2031 2037 2040 2043 2046 2049 2052 2055 2058 2061 2064 2067 2070 2073 Child dependency ratio Old-age dependency ratio Total dependency ratio Source: TUIK Turkey’s human capital is underutilized. Turkey has a working-age population, defined as 15+ years old, of about 59 million people who can contribute productively to the economy. A breakdown of this population, as of 2017, shows that almost 19.5 million are not in school, meaning that they are neither investing in their own human capital nor contributing economic activity (Figure 14). Turkey needs to find a way to make better use of its human capital. The breakdown shows: • Of the working-age population of 59 million, about 31 million (53 percent) are involved in the labor force, while 28 million are not. Page 24 of 114 • About 28 million (89 percent) of the labor force are employed, while about 3 million (11 percent) are unemployed. • Of the 28 million not in the labor force, just under 9 million are in school or training, while almost 19.5 million are not in employment, education or training. Figure 14 Profile of the working age population in Turkey (2017) WORKING AGE POPULATION (15+) (IN THOUSANDS) 59,894 (74%) NOT IN LABOR LABOR FORCE FORCE 31,643 (53%) 28,251 (47%) EMPLOYED UNEMPLOYED IN SCHOOL NOT IN SCHOOL 28,189 (89%) 3,454 (11%) 8,815 (31%) 19,436 (69%) SELF- PAID UNPAID FAMILY EMPLOYED EMPLOYEE WORKER AND EMPLOYER 18,960 (67%) 3,134 (11%) 6,094 (22%) Source: Author’s calculations and illustration. Female participation in the labor market has been increasing in recent years, but remains at a lower level than male participation. Between 2014 and 2017, overall labor force participation rates increased slightly (2 percentage points). A gender breakdown shows that the female labor force participation (LFP) rate increased from 30 percent to 34 percent, while the rate for males increased from 71 percent to 72 percent. Thus, despite more women entering the labor market than men during this period, there are still far fewer women in the labor market than men (Figure 15). Such low levels of participation have a negative effect on the availability of workers who can fill vacancies. It also limits the talent pool, especially when people are skilled but exit the labor force, thus imposing difficulties for firms to find the workforce they need to grow and succeed. As for labor force participation by age group, while youth’s participation increases from 41 percent in 2014 to 43 percent in 2017; participation of population aged between 25 and 64 decreased from 41 percent to 38 percent in the same period. The slight increase in youth’s labor force participation seems to be related increasing unemployment among youth, which will be elaborated in the next figure. Page 25 of 114 Figure 15 Labor force participation rates, by gender and age group 80% 71% 72% 72% 72% 60% 52% 53% 51% 51% 70% 50% 60% 52% 53% 41% 40% 39% 38% 51% 51% 40% 50% 40% 32% 33% 34% 30% 30% 30% 20% 41% 42% 42% 43% 20% 10% 10% 0% 0% 2014 2015 2016 2017 2014 2015 2016 2017 ALL FEMALE MALE ALL AGE 15-24 AGE 25-64 Source: Authors’ elaborations based on LFS Women’s increased participation could not be matched by their employment, implying an increase in women’s unemployment. Figure 17 shows the overall unemployment rate since 2005, as well as a breakdown by gender and youth. As of 2017, the unemployment rate was 10.9 percent, having peaked at around 14 percent during the 2009 crisis. Not only has unemployment for women been higher than for men for many years, but of greater concern is the fact that the gap has been widening since the 2009 crisis, especially after 2016. Youth unemployment is a major challenge that the country has been struggling to address. Even though youth unemployment decreased significantly from 25.3 percent during the 2009 crisis to 18.4 percent during the high growth period (2010-2011), it has been increasing gradually since 2014, reaching 20.8 percent in 2017 (Figure 16). Figure 16 Unemployment rates over time, all sectors 27.0 25.0 25.3 23.0 21.7 21.0 20.8 19.9 20.0 20.5 19.6 19.0 19.1 18.7 18.4 17.5 17.9 18.5 17.0 15.0 14.1 13.0 11.0 10.9 9.0 9.4 7.0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 All Men Women Youth (15-24) Source: Authors’ elaborations based on LFS The unemployment rate is the highest among the younger cohorts (especially for the 20–24 age group). It is useful to look at more disaggregated age groups for key labor market indicators to Page 26 of 114 understand the areas that need to be improved. Figure 17 compares employment, unemployment and LFP rates for various age groups. The LFP rate for young people (aged 20–24) was 58.6 percent, and their employment rate was 46 percent in 2017. As expected, employment and labor participation rates increase until age 49; however, they decrease sharply thereafter, reflecting the early exit patterns among Turkish workers. Figure 17 Employed, unemployed, LFP in the population, 2017 80 70 60 50 (%) 40 30 20 10 0 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 Unemployment rate Labor force participation rate Employment rate Source: Authors’ elaborations based on LFS The LFP rate for women is lower than for men across all age groups. Turkey exhibits substantial variation in LFP and employment figures by age cohort and gender. For males, LFP and employment figures show a reversed U-shape, with LFP peaking at middle age for men (Figure 18). In contrast, females show a gradual increase until their early 30s, with a lower level of LFP than males in all age groups. At the older end of the distribution, for both males and females, participation sharply decreases near pension eligible ages. Figure 18 Employed, unemployed, LFP in the population, by gender, 2017 Male Female 100.0 100.0 50.0 50.0 % % 0.0 0.0 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 Unemployment rate Unemployment rate Labor force participation rate Labor force participation rate Employment rate Employment rate Source: Authors’ elaborations based on LFS Page 27 of 114 Turkey has continued to perform below OECD and EU-28 in terms of LFP rates. Despite improved LFP rates over time, the gap between Turkey and the OECD and EU-28 countries is 7 and 5 percentage points, respectively (Figure 19). Efforts are needed to further improve Turkey’s LFP rates. Figure 19 Country comparison in terms of labor force participation rates, 2016 or the latest year 80% 73% 70% 60% 58% 60% 53% 50% 40% 30% 20% 10% 0% NZL AUS HUN AUT USA DNK DEU LUX TUR MEX EU28 MLT RUS CHE IRL JPN CHL CRI BGR NOR LTU CZE SVN GRC EST LVA G7M SVK ESP SWE CAN OECD FRA FIN ITA PRT POL BEL HRV Source: OECD Employment and Job Types Most of the labor market consists of wage and casual workers, but with a high rate of unpaid family workers. Figure 20 shows the employment status with gender breakdown since 2014. As of 2017, wage and salaried workers made up 60 percent of the labor force, the self-employed and employers constituted 19 percent, and unpaid family workers 10 percent. Since Turkey has not yet completed its modernization of agricultural production, about 3.5 million of the working-age population are engaged in unpaid family work in the agricultural family business. A majority of unpaid family workers were informal (91 percent) in 2017. Men are most likely to be either employed as a wage and casual worker or self-employed; unfortunately, women still constitute a substantial proportion of unpaid family workers. Page 28 of 114 Figure 20 Employment status by year and gender All Sectors 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2014 2015 2016 2017 2014 2015 2016 2017 2014 2015 2016 2017 All Male Female Wage and casual Employer Self-employed Unpaid family worker Source: Authors’ elaborations based on LFS Despite a substantial decrease in the informality rate, the informal sector remains a large source of jobs for workers in Turkey.5 Another way to gauge the status of the labor market is to measure the share of people engaged in informal work in the labor market. Over the years, the Government of Turkey has conducted a series of campaigns against informality, including the introduction of legislation. As a result, informality decreased by 14 percentage points (from 48 percent to 34 percent) between 2005 and 2017. However, in recent years, the decrease has been slight or even stagnated. Not surprisingly, the informality is higher in the agriculture sector, and therefore, it would be more meaningful to look at informality in non-agricultural sectors. In fact, informality in non- agriculture gradually decreased until 2015, after which it started to increase slightly. As of 2017, 22.1 percent of employment in non-agriculture was informal. Informality among wage and casual workers in non-agriculture has remained high (13.3 percent), showing only slight decreases since 2012 (Figure 21). 5 The definition of informality used here is: “workers not registered in the social security system through their main job.” Page 29 of 114 Figure 21 Evolution of informality in Turkey 60% 48% 47% 50% 45% 44% 44% 43% 42% 39% 40% 34.3% 34.1% 37% 35% 34% 33% 34% 32.3% 29.8% 30.1% 29.1% 27.8% 30% 24.5% 22.4% 22.3% 21.2% 21.7% 22.1% 20% 22.0% 22.0% 20.9% 18.7% 18.4% 18.2% 17.9% 10% 15.7% 14.3% 14.2% 13.4% 13.3% 13.3% 0% 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Total informality rate Non-agricultural informality rate Informality rate for Wage and paid workers in Non-agriculture Source: Authors’ elaborations based on LFS Women are most likely to be employed in informal jobs in Turkey. Figure (22) plots the evolution of informality by gender in Turkey. The overall informality rate declined from around 48 percent in 2005 to 34 percent in 2017. Informality is clearly much higher among females than men for all years, and it is a concern that the gap has not been narrowing over time. As of 2017, the informality rate for women was 44.6 percent, as opposed to 29.2 percent for men. From 2011 to 2015, informality decreased for both genders, but the picture appears to have changed after 2016. Figure 22 Evolution of informality rate in Turkey, by gender (in percent) 67.0 62.0 57.0 52.0 47.0 44.6 42.0 37.0 34.0 32.0 29.2 27.0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 All Men Women Source: Authors’ elaborations based on LFS Page 30 of 114 Workers in the agriculture sector are generally hired as informal, impeding efforts to reduce informality. In general, informality is a more serious problem for wage and casual workers, for whom the informality rate was 12 percent in 2017 (Figure 23). However, given the current structure of the Turkish labor market, it would be more informative to make a distinction between agriculture and non-agricultural sectors. More than 80 percent of the workers employed in the agriculture sector are informal, with 45 percent being unpaid family workers. It is clear that Turkey has not been using modern techniques for agricultural production. As a result, family enterprises are still common, which creates a potential barrier against efforts to reduce informality. As for the non-agricultural sector, the informality rate is higher for wage and casual workers (14 percent). Figure 23 The share of informal workers in sectoral total employment by employment status and sector, 2017 50% 45% 40% 28% 30% 20% 14% 12% 11% 10% 8% 10% 6% 1% 0% 1% 2% 0% Self-employed Self-employed Self-employed Wage & casual worker Wage & casual worker Wage & casual worker Employer Employer Employer Unpaid family worker Unpaid family worker Unpaid family worker All Agricultural Non-agricultural Source: Authors’ elaborations based on LFS Profile of the workers The Eastern regions of Turkey suffer more from high informality. The figure below shows the informality rates in 2017 across NUTS2 level regions.6 The region with the lowest informality rate is Ankara (18 percent), which is the region with the second highest proportion of public employment (Figure 24). The region with the highest informality rate is Agri (Kars-Agri-Ardahan-Igdir) (68 percent). As abovementioned, in Eastern regions, agricultural production still prevalent, due to its genuine structure, agriculture stands as a critical barrier against efforts to reduce informality in Turkey and these are the same regions with high concentrations of low-educated workforce, youth inactivity, and low LFP. Informality is significantly higher in agriculture than in the non-agricultural sectors. It is interesting that Gaziantep and Mardin regions have substantially lower informality rates compared to other East and Southeast regions of Turkey. This is interesting especially given that the influx of Syrians seeking refuge in Turkey, which started after 2011 and peaked in 2015, is largely concentrated in the south-eastern parts of Turkey (Box 2 provides more details). Istanbul has the highest population 6 For NUTS classification of Turkey, please see App 1. Page 31 of 114 in Turkey and its informality rate is 21 percent. Like in the south-eastern provinces of Turkey, Istanbul hosts over half a million Syrians under temporary protection, and most of them are working informally. The share of skilled workers (secondary graduates and above) working informally provides an indication of the types of employment opportunities available. For instance, if labor markets are not dynamic and few jobs are available, workers have few choices of employment locally. But if the worker has few skills to offer employers, despite having a relatively high level of education, his or her choices are limited locally (and likely elsewhere). For instance, in Ankara, only 7 percent of skilled workers are employed in an informal activity, whereas in the subregions of Van-Mus-Hakkari-Bitlis, and Agri-Kars-Igdri-Ardahan, about 30 percent of skilled workers work informally. Figure 24 Formality and informality rates across regions, 2017 100% 90% 21% 24% 23% 29% 18% 33% 34% 32% 31% 80% 39% 39% 42% 40% 42% 38% 44% 45% 51% 48% 39% 40% 55% 49% 70% 63% 62% 68% 60% 50% 40% 79% 76% 77% 71% 82% 67% 61% 66% 61% 68% 69% 30% 58% 60% 58% 62% 56% 55% 49% 52% 61% 60% 45% 51% 20% 37% 38% 32% 10% 0% Kastamonu (Kastamonu-Cankiri-… Manisa (Manisa-Afyonkarahisar-… Hatay (Hatay-Kahramanmaras-… Kirikkale (Nevsehir-Aksaray-… Samsun (Samsun-Tokat-Corum-… Kocaeli (Kocaeli-Sakarya-Duzce-… Trabzon (Trabzon-Ordu-Giresun-… Malatya (Malatya-Elazig-Bingol-… Zonguldak (Zonguldak-Karabuk-… Gaziantep (Gaziantep-Adiyaman-… Mardin (Siirt-Mardin-Batman-… Erzurum (Erzurum-Erzincan-… Ankara (Ankara) Aydin (Denizli-Aydin-Mugla) Balikesir (Balikesir-Canakkale) Antalya (Antalya-Isparta-Burdur) Agri (Kars-Agri-Igdir-Ardahan) Istanbul (Istanbul) Izmir (Izmir) Sanliurfa (Diyarbakir-Sanliurfa) Bursa (Bursa-Eskisehir-Bilecik) Konya (Konya-Karaman) Kayseri (Kayseri-Sivas-Yozgat) Tekirdag (Edirne-Tekirdag-Kirklareli) Adana (Adana-Mersin) Van (Van-Mus-Bitlis-Hakkari) Formal Informal Source: Authors’ elaborations based on LFS Page 32 of 114 Box 2 Syrian refugees7 in Turkey and their impact on the labor market Turkey is the largest refugee-hosting country in the world with 3.6 million Syrians. Syrian refugees started coming to Turkey in 2011 and the numbers increased substantially between 2013 and 2016. There are at least 300,000 more people from other countries (e.g. Afghanistan, Iraq) seeking refuge in Turkey. This rapid increase poses a labor supply shock to the Turkish economy; but the shock is mostly absorbed by the informal economy. Given Turkey’s challenges with economic dynamism in the last few years, the presence of nearly 4 million foreign people, most of them minimally skilled, puts pressure on the Turkish economy (especially on low-skilled Turkish workers engaged in the informal economy). Data shows that most refugees (namely Syrians) are overwhelmingly employed informally in the Turkish labor market. Apart from having low skills, most refugees have no easy access to work permits, which makes their entry into formal work very difficult. There is a growing literature on the impact of refugees in Turkey. Del Carpio and Wagner (2015) analyze the impacts of Syrian refuges on Turkish labor market. The refugee shock causes large-scale displacement of Turkish out of informal employment for all categories of workers (male, female, young and older, and by educational attainment). On average the impact of the inflow of refugees on the formal employment of Turkish workers is positive. An increase in the supply of informal labor increases the demand for formal jobs for Turkish workers. However, that is not true for all types of Turkish workers. There is no increase in formal employment for either women or high-skilled Turkish workers due to the inflow of refugees. At the same time, consistent with occupational upgrading, there are increases in formal employment for the Turkish. Tumen (2016), Bagir (2017), Ceritoglu et al. (2017) find that refugees penetrate labor the labor market through informal manual jobs and displace natives informally employed in those jobs. Tumen (2018) analyzes the impact of Syrian refugees on high school enrollment rates of Turkish youth and finds that high school enrollment rates increased 2.7 to 3.6 percentage points among Turkish youth in refugee-receiving regions. Also, a one-percentage point increase in the refugee-to-population ratio in a region generates around 0.4 percentage point increase in native’s high school enrollment rates. Informal workers (including unpaid family workers) tend to be female, older and less educated and work in agriculture, construction or low and medium technology services. The light blue bars in Figure 25 depict informality rates for different groups in the population. The figure clearly indicates that informality is always higher for females compared to male counterparts since females are mostly employed in agricultural sectors as unpaid family workers. As for age groups, the informality rate among the working group, 25-64 years old, was 30 percent in 2017, but significantly higher for youth (15-24 years old) at 46 percent and for the old (aged 65+) at 85%. Informality in the market decreases as educational attainment increases. Since low educational attainment brings the risk of being informally employed, the economic sectors requiring low or medium technology capabilities are more likely to employ low skill and informal workers. Besides agriculture, the services sectors are more likely to hire informal workers. Besides service sectors, construction is the sector with the highest informality rate. 7Instead of ‘SuTP and/or ASuIP’ this document uses ‘Refugees’ to make it in line with international terminology and ease the reading. Thus, the ‘Refugees’ referred to in this document shall be understood as ‘Syrian people under temporary protection’ and other ‘Asylum Seekers under International Protection’ according to Turkish legislation. According to DGMM and UNHCR, as of December 2018 there are 3.9 million refugees in Turkey. This total includes Syrians (3.6 million), Afghanis (170,000), Iraqis (142,000), Iranians (39,000), Somalians (5,700), and other nationalities (11,700). Page 33 of 114 Figure 25 Formality and informality rates by gender, age, education and economic sector, 2017 100% 4% 5% 7% 10% 90% 20% 23% 29% 30% 27% 28% 31% 31% 80% 39% 36% 45% 46% 49% 70% 60% 85% 81% 83% 50% 96% 95% 93% 90% 40% 80% 77% 71% 70% 73% 72% 69% 69% 30% 61% 64% 55% 54% 51% 20% 10% 15% 19% 17% 0% Literate but not completed… Female University 15-24 25-64 65+ Male Secondary Mfg (low/med) Primary school (5 yrs) High school Construction (low/med) Mfg (high) Ag, forest, fishing (low) Mining (low/med) Services (low/med) Services (high) Services, transport (low) Wholesale-retail (med) Accom, Food (med) Gender Age group Education Economic sector, level of VA Formal Informal Source: Authors’ elaborations based on LFS Informal employment for both part- and full-time work has increased in recent years. There is a diverging trend for part-time employment for formal and informal contracts after 2013. However, especially after 2016, the number employed for part-time formal work showed a remarkable decrease and the opposite trend was seen for part-time informal employment (Figure 26). This might be the result of the increase in the minimum wage that occurred in 2016. Between 2005 to 2017, part-time employment increased significantly and is the source of 9 percent of the increase of 40 percent in total employment, whereas the remaining 31 percent is from full-time jobs. Page 34 of 114 Figure 26 Number employed by contract type and informality (indexed to 2005=100) 350 300 250 200 150 100 50 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Full time, Formal Full time, Informal Part time, Formal Part time, Informal Source: Authors’ elaborations based on LFS The number of hours worked per week for both full-time and part-time employment in Turkey has declined significantly since 2005. This decrease is evident from the left side of Figure 27, which depicts the hours worked per week from 2005 to 2017. However, the type of contractual arrangement that a worker holds matters. Hours worked per week for full-time employment has been decreasing since 2015, whereas the trend is flat for part-time employment. The significant increase in the minimum wage in 2016 might have led firms to change workers’ contract type, and raises the question of the role played by informality in relation to the number of hours worked. The right side of Figure 29 plots hours worked per week by contract type and informality status. Even though hours worked per week for both part-time formal and part-time informal have increased since 2015, the increase for part-time formal was higher. As the share of full-time employment in total employment is higher compared to the share of part-time employment, decreases in total hours worked per week were due to the full-time employment. As of 2017, the average hours worked per week in Turkey was around 45, with full-time employment at 47 hours per week and part-time employment at 17 hours. Page 35 of 114 Figure 27 Hours worked per week by contract type and informality (2005=100) 101 101 99 99 97 97 95 95 93 93 91 91 89 89 87 87 85 2009 2005 2006 2007 2008 2010 2011 2012 2013 2014 2015 2016 2017 85 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Hours worked per week (Full-time, Formal) Hours worked per week (total) Hours worked per week (Full-time, Informal) Hours worked per week (full time employment) Hours worked per week (Part-time, Formal) Hours worked per week (part time employment) Hours worked per week (Part-time, Informal ) Source: Authors’ elaborations based on LFS Inactivity Youth idleness, i.e. youth neither employed nor in education or training, is high by international standards. Young people are more likely to delay their participation to the labor market and become inactive or tend to stay in education without being active in the labor market. Therefore, while unemployment and employment rates are most appropriate way to analyze how people are participation the labor market, the proportion of young people neither in employment nor in education or training (NEET) would provide a better sense about the difficulties at the jobs search stage of young people. Figure 28 shows that 28.2 percent of those aged between 15 and 29 in 2016 were neither in employment nor in education or training (NEET), meaning that they were not contributing to the economy or improving their own skills to be able to contribute to the economy in the future. On the other hand, the NEET rate among OECD countries (13.9 percent) is less than half of Turkey’s. Turkey is lagging not only behind OECD countries but also behind its peers such as Italy, Greece and Mexico. Page 36 of 114 Figure 28 NEET (%) in selected countries, aged 15–29, 2016 30.0 28.2 25.0 20.0 13.9 15.0 10.0 5.0 0.0 DNK TUR NZL AUT LUX CRI MEX HUN USA RUS AUS DEU CHE COL OAVG SVK NLD GRC LVA GBR CZE SVN ESP FRA EST LTU NOR POL ISR CAN SWE ISL ITA PRT FIN BEL Source: Authors’ elaborations based on OECD data. As of 2017, NEET rate among university graduates is 34 percent. In 2017, 76 percent of the youth aged between 15 and 25 were either in school or employed, or both (Figure 29). However, 34 percent of females are neither economically active nor in education, compared with 15 percent of males. The NEET rates decrease slightly (less than 1 percentage point) in recent years. The highest concentration of NEETs is among people with less than high school completion (including illiterates), but there is a sizeable proportion of NEET among youth with a university degree (34 percent in 2017). These people are skilled but not attached to the labor market. Even though it merits further research, there might be possible reasons. Educational levels provide important insights into the skill level of Turkey’s workforce, but a more nuanced evaluation of the composition of these skills (taking into account the skills content of occupations) needs to be considered. Occupations in Turkey have been increasing (albeit slightly) their use of non-routine cognitive (analytical and interpersonal) and routine cognitive skills, commonly associated with better quality of jobs. Analysis also shows that the reliance on lower level manual skills has declined over time; both non-routine manual physical and routine manual physical skills are becoming less dominant in the Turkish economy. However, compared to the United States and other benchmark countries, and given its level of development, occupations in Turkey are still relying too much on manual skills, and too little on cognitive (routine and non-routine) skills. Firms in Turkey continue to demand workers with low and medium skill levels. Figure 29 NEET rates (%) in Turkey over time, by gender and education level completed Page 37 of 114 15-24 years olds not employed and not in NEET (15-24 YEARS OLD) school/training Illeterate Less than high school 80.0% High school University 70.0% 81% 78% 78% 77% 60.0% 35.0% 33.8% 33.5% 34.0% 50.0% 40.0% 30.0% 14.7% 14.3% 14.6% 14.6% 34% 20.0% 31% 30% 28% 24.9% 24% 24.0% 24.0% 24.2% 23% 23% 22% 10.0% 22% 21% 21% 21% 0.0% 2014 2015 2016 2017 All Male Female 2014 2015 2016 2017 Source: Authors’ elaborations based on LFS A review of the characteristics of NEETs points to the importance of education. The left side of Figure 30 indicates, as of 2017, very similar percentages of male (51 percent) and female (49 percent) NEETs. However, the right side, which focuses on the level of education (highest education level obtained), shows that 54 percent of NEETs have not completed high school; they are primary or secondary school graduates. 26 percent of the NEET completed high school. 6 percent of the NEET were illiterate. Among university-educated young people, the NEET rate is 14 percent in 2017. Since their education level is already high, the problem for this group of people is not on the side of not being in education but on the side of not being in employment. Figure 30 Characteristics of NEETs (2017) Illiterate 14%6% Less than Male high school 49% 51% 26% Female 54% High school University Source: Authors’ elaborations based on LFS NEET rates among youth have proven to be persistent over time. In the past 4 years, the share of NEETs among youth (15-24 years old) decreased by only 1 percentage point, from 25 percent to 24 percent (Figure 31). However, the size of the economically inactive population is also high among those aged 25-64. Page 38 of 114 Figure 31 Inactivity rates 100% 2% 1% 1% 1% 90% 80% 34% 34% 34% 32% 37% 36% 39% 38% 70% 60% 5% 5% 6% 6% 25% 24% 24% 24% 50% 40% 7% 8% 8% 9% 30% 54% 55% 55% 56% 20% 34% 34% 34% 34% 10% 0% 2014 2015 2016 2017 2014 2015 2016 2017 Age 15-24 Age 25-64 Not in employment, but in education/training Neither in employment nor in education/training Unemployed Employed Source: Authors’ elaborations based on LFS An analysis of the reasons for inactivity among NEETs suggests the need for better management of the school-to-work transition. The analysis shows that 50 percent of NEETs are involved in domestic work (Figure 32). It seems that potentially the most productive group of the population in terms of contribution to the economy remains idle. One of the aspects of the key strategy for reducing the number of people who are neither in employment nor in education is to improve the management of the school-to-work transition. To improve the transition from school to work, education systems should aim to ensure that there is a match between individuals’ skills and labor market needs. Public investment in education, during recession times in a country, might be an option to counterbalance unemployment by matching labor market’s skill needs (OECD 2015). Figure 32 Reasons for inactivity among the NEET, 2017 Not seeking a job but available to start, Discouraged Not seeking a job Personal or family but available to reasons start, Other 14% 6% 9% 8% 0% Disabled, ill etc. Working seasonally 13% Other 49% Housewife Source: Authors’ elaborations based on LFS Page 39 of 114 Education, Skills and Labor Market Outcomes Employment rates are positively correlated with educational attainment, while the majority of low-educated people are neither in employment nor in education. For 25-64 year-old adults with higher education, 75 percent are employed (Figure 33). On the other hand, inactivity is negatively correlated with educational attainment, as 49 percent of 25-64 year-old adults with only primary education are doing nothing (i.e. neither in employment nor in education). Being an unpaid family worker is more common among the population with only primary education. In 2017, while 87 percent of those who have completed higher education are working as wage and salaried workers in the labor market, 51 percent of 25-64 year-old adults who have primary education are employed as unpaid family workers (Figure 33). Self-employment is also more common among those with only primary education. Figure 33 Labor force participation by education level and type of employment by education level age 25-64, 2017 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 10% 20% 0% 10% 0% All Vocational high school Primary and less than primary High school University Secondary Primary and less than All Vocational high school High school Secondary University primary Not in employment but in education Neither in employment nor in education/training Unemployed Wage & salaried Employer Employed Self-employed Unpaid family worker Source: Authors’ elaborations based on LFS Females are less educated compared to males, even at early ages. Figure 34 shows the comparison between males and females by age cohort in terms of the highest diploma obtained by individuals aged between 25 and 64. For both genders, although the education level increases by age, females are less educated for all age groups. Figure 34 Highest diploma obtained, by age cohort and gender, 2017 Page 40 of 114 60-64 66% 10%7%6%12% 60-64 85% 4% 3% 3% 5% 55-59 61% 10%10%8%11% 55-59 81% 5% 4% 6% 5% Female age cohort Male age cohort 50-54 58% 14% 9%7%12% 50-54 78% 4% 6%6% 6% 45-49 55% 14% 8% 9% 15% 45-49 73% 7%7%4%9% 40-44 47% 12%10%12% 19% 40-44 67% 7% 8%5%12% 35-39 36% 15% 13% 14% 23% 35-39 54% 8% 10% 8% 19% 30-34 25% 19% 14% 14% 29% 30-34 43% 13% 10% 8% 26% 25-29 7% 34% 14% 12% 33% 25-29 19% 26% 13% 8% 34% 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Primary and less than primary Primary and less than primary Secondary Secondary High school High school Vocational high school Vocational high school University University Source: Authors’ elaborations based on LFS Employment has increased at all educational levels, except primary and less than primary education. Between 2014 and 2017, employment increased by 9 percent overall for the population of 25-64 year-olds. While employment growth was negative for primary or less educated people, it increased by 21.9 percent for secondary educated people and by 26.5 percent for the university educated (Figure 35). Figure 35 Employment levels and growth by education level, age 25-64, 2017 12,000 30.0% Percentage change in number of Number of employment (000) 10,000 25.0% employment (2014-2016) 20.0% 8,000 15.0% 6,000 10.0% 4,000 5.0% 2,000 0.0% 0 -5.0% Primary and less Secondary High school Vocational high University than primary school Education 2017 Growth rate, 2014-17, right axis Source: Authors’ elaborations based on LFS Educational polarization in employment is emerging between male and females. Despite educational improvements in the Turkish population, the educational composition across men and women exhibits a clear U-shape—high concentrations at the lower end and higher end of the educational distribution. Men’s U-shape composition indicates a large share of men in total Page 41 of 114 employment has less than secondary completed education level whereas women with higher education levels are considerably less likely than men to enter the labor market. University-educated women saw an increase of 23.1 percent in employment in the 2014–2017 period (Figure 36). Figure 36 Employment levels and growth by gender and education level, age 25-64, 2017 Male Female 7,000 40.0% 7,000 40.0% Percentage change in number of employment (2014-2016) Percentage change in number of employment (2014-2016) Number of employment (000) Number of employment (000) 35.0% 35.0% 6,000 6,000 30.0% 30.0% 5,000 25.0% 5,000 25.0% 4,000 20.0% 4,000 20.0% 15.0% 15.0% 3,000 10.0% 3,000 10.0% 2,000 5.0% 2,000 5.0% 0.0% 0.0% 1,000 1,000 -5.0% -5.0% 0 -10.0% 0 -10.0% Vocational high school Vocational high school Primary and less than primary High school Primary and less than primary High school Secondary University Secondary University 2017 Growth rate, 2014-17, right axis 2017 Growth rate, 2014-17, right axis Source: Authors’ elaborations based on LFS Most of the employed women aged 25-64 years are either very low-educated or very high- educated. In 2017, about 56 percent of employed people had undergone secondary or lower level of education (Figure 37). Only 21.7 percent of employed women had completed secondary or high school education, while the figure was 38.3 percent for men. On the other hand, about 29.3 percent of female workers had a university education, whereas the estimate for men was lower (22.0 percent). The main explanation for this finding is that educated female workers are considerably less likely than men to enter the labor force unless the benefits (wages) outweigh the costs of working. One of the reasons is that they complete their education at an age that coincides with their prime fertility years, and thus, they have to weigh the costs of child care (and other non-cost related factors) against the wages they can potentially earn. Page 42 of 114 Figure 37 Shares of employment by education level and gender, age 25-64, 2017 Male 39.7% 16.0% 11.0% 11.3% 22.0% Female 48.9% 7.9% 7.9%5.9% 29.3% Total 41.2% 14.3% 10.0% 9.7% 24.7% 0% 20% 40% 60% 80% 100% Primary and less than primary Secondary High school Vocational high school University Source: Authors’ elaborations based on LFS Informality decreases with educational attainment, but it is widespread among those who have less than secondary education. Informality is higher for women than men at all education levels (Figure 38). For the most part, females with low education levels are employed in agriculture and work as unpaid family workers. Also, one effect from the crisis is that employment increased among lower-educated women who went back to work in agriculture and other low or unskilled jobs, mostly in the informal sector of the economy. Figure 38 Informality by education level and gender, age 25-64, 2017 80.0% 4.00 70.0% 3.00 60.0% 2.00 50.0% 1.00 40.0% 0.00 30.0% 20.0% -1.00 10.0% -2.00 0.0% -3.00 Primary and Secondary High school Vocational high University less than school primary Female, 2017 Male, 2017 Change (male, pp, 2014-17) Change (female, pp, 2014-17) Source: Authors’ elaborations based on LFS Page 43 of 114 Labor force skills as measured by PIAAC, PISA, and TIMSS Turkey has the third-lowest average score in literacy and numeracy among the 33 countries surveyed.8 According to the PIAAC, adults in Turkey show below-average proficiency in numeracy and literacy domains. As many as 46 percent of adults in Turkey attain only Level 1 or below in literacy; this percentage is high compared to only 19 percent among OECD countries. Similarly, half of all adults (50.2 percent) attain only Level 1 or below in numeracy, compared to 22.7 percent among OECD countries. Only 8 percent of adults in Turkey attained one of the two highest proficiency levels in problem-solving in the technology-rich environments domain. Almost 40 percent of adults reported no computer experience or failed the information and communication technology (ICT) core test. Figure 39 Proficiency gaps by age group (2015, Figure 40 Proficiency gaps between Turkey and OECD difference between mean proficiency scores) 300 300 250 250 Numeracy 55-65 Literacy 200 200 Numeracy 45-54 150 150 Literacy Age groups Numeracy 100 100 35-44 Literacy 50 50 Numeracy 25-34 Literacy 0 0 Cyprus Austria Israel Norway Italy New Zealand Australia Korea Sweden Finland Slovenia United States Canada Ireland Lithuania Germany Poland England (UK) Czech Republic Russia Denmark Slovak Republic Spain OECD Singapore Netherlands Chile Japan N. Ireland (UK) France Jakarta (IDN) Turkey Greece Estonia Flanders (BEL) Numeracy 16-24 Literacy 0 20 40 60 Literacy Numeracy Problem Solving Source: Survey of Adult Skills 2015, OECD Turkey needs to improve the quality of education. According to the 2015 Programme for International Student Assessment (PISA)9 results, Turkey performed well below the OECD average in all three domains (Figures 41 and 42). • Reading: Turkey’s score is 428, the OECD average is 493; • Science: Turkey – 425, OECD average – 493; and • Mathematics: Turkey – 420, OECD average – 490. There has been a clear decline in scores in 2015, returning to the 2006 level. The percentage of students performing below proficiency levels has remained quite stable since 2006. However, the percentage of students in the top proficiency level has decreased, especially in reading (World Bank 2016a). 8 Besides the 28 OECD countries, Cyprus, Jakarta (Indonesia), Lithuania, the Russian Federation, and Singapore are included in the data set. The Surv ey of Adult Skills was conducted in Turkey from April 1, 2014, to March 31, 2015. Some 5,277 adults aged 16–65 were surveyed. 9 PISA covers 15-year-old students near the end of their compulsory education and assesses key knowledge and skills, namely science, reading, and mathematics. Page 44 of 114 Figure 41 PISA scores (2003-2015) Figure 42 PISA scores and GDP per capita 100 Averages for PISA reading scale: overall 90 1400 Luxembourg (Thousands) 80 1200 years 70 France 1000 Belgium United States 60 reading, age 15 Australia International Switzerland Sweden 800 50 Average Germany (OECD) Austria Canada Denmark 40 Finland 600 Greece United Italy 30 Kingdom 400 Israel Portugal Slovenia 20 Turkey Latvia 200 10 Mexico Spain 0 0 2003 2006 2009 2012 2015 400 450 500 550 GDP per capita, constant prices, constant PPPs, Science Reading Math OECD base year=2010 Source: Authors’ elaborations based on OECD and PISA data The mathematics and science knowledge & skills are lower in Turkey than in OECD countries. The Trends in International Mathematics and Science Study (TIMSS) is used to measure the mathematics and science knowledge and skills of 4th and 8th graders over time. In 2015, average mathematics scores of 4th and 8th grade students in Turkey were 483 and 458, respectively, and average science scores were 483 (4th grade) and 493 (8th grade) (Figures 43 and 44). Both internationally comparable assessment tools used for the analysis PISA and TIMSS indicate that skill challenges prevalent among the pipeline of future workers will likely continue in the future, when they enter the workforce, unless drastic measures are taken. Two main challenges are highlighted, one is to improve the education level of the workforce further and the second is to enhance the skills of students through better-designed curriculums which take into account the need of the labor market. Page 45 of 114 Figure 43 TIMSS Mathematics scores and ranks Figure 44 TIMSS Science scores and ranks (Turkey, Grades 4 and 8, 2011 and 2015) (Turkey, Grades 4 and 8, 2011 and 2015) 15 15 Grade 8 Grade 8 Grade 8 20 (2011) (2015) (2011) 20 Grade 8 25 (2015) 25 Ranking Ranking Grade 4 30 Grade 4 Grade 4 (2011) 30 (2011) (2015) 35 Grade 4 35 (2015) 40 40 45 45 450 460 470 480 490 500 450 460 470 480 490 500 Scores Scores Source: Authors’ elaborations based on TIMMS data (accessible at TIMSS) Technological change and skill composition in the labor market The labor share of global income has declined significantly since the early 1980s across countries and industries. Global factors, such as consumer preferences, technological upgrading, and demographic changes, play an important role in the changing skills demands. The advances in information technology, which decrease the relative price of investment goods, induce firms to shift away from labor toward capital. Most countries experienced a decline in labor shares in output, with Poland and Mexico experiencing the most significant decreases in the last decade. Although Turkey has experienced an increase in labor share in output, the rate of the increase is about 1.5 percentage points over the sample period (Figure 45). Figure 45 Trends in labor shares in output every 10 years since 1975 Source: Karabarbounis and Neiman (2013, 2014) Page 46 of 114 Note: The figure shows estimated trends in the labor share for all countries in the data set with at least 15 years of data starting in 1975. Trend coefficients are reported in units per 10 years (that is, a value of −5 means a 5 percentage point decline every 10 years). Note that countries have different time coverage. However, the units are comparable across countries, because the variable is standardized to show the change in the labor share per 10 years (for all countries). The costs of technology have fallen, while technology adoption in Turkey has increased. Factors that influence labor-technology substitution are the skill composition of the workforce, different levels of complementarity (or substitutability) with technology, cost of capital relative to labor (elasticity of substitution), technology availability for the tasks that humans perform, among others. The computerization risk estimation (used mostly in developed countries) can be useful to foresee the occupations that carry a high risk of disappearing in the future as technology adoption increases among firms in Turkey. Based on the occupational distribution in Turkey in 2017, about 62 percent of all occupations can be considered at high risk of computerization in the future (Figure 46). High-risk occupations are in sales, production (manufacturing), transport and some other services. Figure 46 Distribution of occupational employment over the probability of computerization Source: An adaptation for Turkey of Frey and Osborne 2013. Note: Arrows show the share in low, medium, and high probability categories in the United States, 2010. Turkey is in the transition from ‘efficiency-driven development’ to ‘innovation-driven development’. Productivity links innovation with growth. For Turkey, a study (Srour, Taymaz, and Vivarelli 2013) finds that domestic and imported technologies increased the demand for skilled labor much more than the demand for unskilled labor. It also stresses the importance of innovation and technological upgrading as complementary factors to an increasingly skilled labor force, and as underlying factors for productivity growth. The Global Innovation Index (2018) shows that Turkey is in the transition from ‘efficiency-driven development’ to ‘innovation-driven development’, which means that the country is transitioning from improving production processes to producing new and different goods. However, Turkey ranked 50th among 126 countries in the Global Innovation Index in 2018, falling 8 places from the 2016 Index (42nd among 128 countries) (Figure 47). Page 47 of 114 Figure 47 Global Innovation Index (2018, selected countries, score) 80 70 60 50 40 30 20 10 0 Malaysia Turkey Zambia France Iceland Kenya Georgia Indonesia Finland Norway Chile Ireland Cyprus Croatia Thailand Mexico Japan Azerbaijan Malta United Arab Emirates Yemen Switzerland Portugal Mongolia Uruguay Kazakhstan Armenia Egypt Tajikistan United Kingdom Iran, Islamic Republic of Source: Global Innovation Index 2018 Report. Note: The overall Global Innovation Index (GII) score is the simple average of the Input and Output Sub-Index scores. The GII (2018) has been co-published by Cornell University, INSEAD, and the World Intellectual Property Organization (WIPO, an agency of the United Nations). Global technological developments are affecting the level of skills most suited for employment. Clear trends of job polarization are emerging across the world as a result of routine- biased technological change, which is negatively affecting (reducing the share of) occupations requiring medium skills. Nonetheless, employment in Turkey increased between 2009 and 2017 in all three types of occupations: low-skill, middle-skill, and high-skill (Figure 48). However, a breakdown into sub- periods shows variations: • From 2009 to 2011, increases were observed in all occupation types. • From 2012 to 2013, while low-skilled occupations increased, high-skill occupations decreased. One possible explanation for the increase is that people with low levels of education entered the labor market due to the effects of the 2009 crisis and remained there during the recovery period because the government provided various incentives and subsidies. This might create an asymmetry in the market for high-skill occupations. • After 2014, employment for low- and middle-skill occupations has been stagnating, but increasing for high-skill occupations. Page 48 of 114 Figure 48 Employment trends by type of occupation in Turkey, 2009-2017 30,000 Number of employment (000) 25,000 20,000 15,000 10,000 5,000 0 2009 2010 2011 2012 2013 2014 2015 2016 2017 Low-skilled occupations (intensive in nonroutine manual skills) Middle-skilled occupations (intensive in routine cognitive and manual skills) High skilled occupations (intensive in nonroutine cognitive and interpersonal skills) Source: Authors’ elaborations based on LFS Note: Occupation classification follows Autor 2014. High-skilled occupations include legislators, senior officials and managers, professionals, and technicians and associate professionals. Middle-skilled occupations comprise clerks, craft and related trades workers, plant and machine operators and assemblers. Low-skilled occupations refer to service and sales workers and elementary occupations. The Turkish economy is increasingly using the cognitive skills associated with better quality jobs instead of the manual skills associated with lower quality jobs; encouragingly, the reliance on lower-level manual skills is declining over time. On the left chart of Figure 49, for the national sample, both non-routine manual physical and routine manual skills are becoming less dominant, a sign of decreasing dependence on the manual skills that are typical in lesser quality jobs. Non-routine cognitive (analytical and interpersonal) and routine cognitive skills are dominant, which is expected in an economy with relatively high levels of good jobs, which have increased in number in recent years. The biggest increase relates to the use of routine cognitive skills, such as those performed by bookkeepers or call center operators. Unfortunately, these jobs are at risk of mechanization. The right panel, on the other hand, paints a very different picture, which reflects the kinds of jobs held by workers with low education levels. These workers are employed in occupations with high levels of routine tasks. In the medium or longer term, as mechanization is adopted, this segment of the population is at particular risk of substitution, unemployment and obsolescence. Page 49 of 114 Figure 49 Employment composition by type of occupation in Turkey, 2010-2017 Turkey: National Sample Turkey: Low education 0.15 0.6 0.1 0.4 0.05 0.2 Z-score Z-score 0 0 2010 2011 2012 2013 2014 2015 2016 2017 20102011201220132014201520162017 -0.05 -0.2 -0.1 -0.4 -0.15 -0.6 Non-routine cognitive: Analytical Non-routine cognitive: Analytical Non-routine cognitive: Interpersonal Non-routine cognitive: Interpersonal Routine cognitive Routine cognitive Routine manual Routine manual Non-routine manual physical Non-routine manual physical Source: World Bank calculations, skills classification follows Autor (2014). Note: In the methodology, first, ISCO 88 2-digit occupations were converted to ISCO 08, and then each occupation was assigned with an average per skill (e.g., Non-routine cognitive: Analytical), which is associated with a scale such as importance. Each scale has a minimum and maximum value (e.g., Importance: 1 – 5). Then, to get standard values, z-scores were computed by year and skill. Firms see skills inadequacy in the labor force as one of the binding constraints that needs remedying to meet business objectives.10 An inadequately educated labor force was perceived to be an obstacle to doing business by 7 percent of firms (Enterprise Survey, 2015-16). But apart from the formal education and training system, firms can also provide training to upgrade the skills of the labor force. Figure 50 shows the percentage of firms offering formal training programs for their permanent and full-time employees. As shown, the percentage of firms offering formal training programs to their permanent full-time workers is highest in China and Sweden. Turkey ranks 52 among 91 countries in 2018, indicating that Turkish firms are investing in training, albeit at a lower frequency than many other countries. 10 This is the fifth biggest obstacle: others are tax rates, access to finance, informal competitors and political instability. Page 50 of 114 Figure 50 The percentage of firms offering formal training for their permanent, full-time employees (2018, selected countries, score) 120 100 80 60 40 20 0 Brazil Serbia Turkey Zambia Israel Czech Republic Albania Bangladesh Hungary China Mexico India Poland Ukraine Thailand Peru Slovakia Slovenia TFYR Macedonia Sweden Argentina Lebanon Egypt Moldova, Republic of Source: Authors’ illustration. Global Innovation Index (2018). Note: The GII (2018) has been co-published by Cornell University, INSEAD, and the World Intellectual Property Organization (WIPO, an agency of the United Nations). Labor market returns In Turkey, the daily wage increased between 2007 and 2017 in both the formal and informal sectors. People’s skills stock and their use of skills at work closely affect their labor earnings. Figure 51 shows the evolution of the daily wage (for both the formal and informal sectors), whereas Figure 52 plots the evolution of the nominal and real minimum wage. It is evident that the increase in the real daily wage for the informal sector was less than that for the formal sector. Page 51 of 114 Figure 51 Daily real wage for formal and Figure 52 Evolution of the daily minimum wage informal sectors, 2007-2017 (both nominal and real) 80.0 80.0 70.0 70.0 2003 prices in TL 60.0 60.0 2003 prices in TL 50.0 50.0 40.0 40.0 30.0 30.0 20.0 20.0 10.0 10.0 2011 2007 2008 2009 2010 2012 2013 2014 2015 2016 2017 2018 0.0 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Statutory Minimum Wage (Real, Daily, Net) All Formal Informal Statutory Minimum Wage (Nominal, Daily, Net) Source: Authors calculations using Labor Force Survey (LFS) The real wage increased between 2009 and 2017 for occupations at all three skill levels. The real hourly wage increased by 69 percent, although the increases were different across occupations. The increase in low-skill occupations (94 percent) was higher than the increase in middle-skill (78 percent) and high-skill (44 percent) occupations (Figure 53). Although high-skill occupations earn more over time, the Turkish labor market still needs the supply of labor for middle- and low-paying occupations. Figure 53 Evolution of real hourly wage, by occupation, 2009–2017 High skilled occupations (intensive in nonroutine cognitive and interpersonal skills) Middle-skilled occupations (intensive in routine cognitive and manual skills) Low-skilled occupations (intensive in nonroutine manual skills) Total 3 Log of real hourly wage (in TL) 2.5 2 1.5 1 0.5 2009 2010 2011 2012 2013 2014 2015 2016 2017 Source: Authors’ calculations using LFS. Note: Wage is ‘log of real wage per hour’ and available for wage and salary workers. Occupation classification follows Autor 2014. High-skilled occupations include legislators, senior officials and managers, professionals, and technicians and associate professionals. Middle-skilled occupations comprise clerks, craft and related trades workers, plant and machine operators and assemblers. Low-skilled occupations refer to service and sales workers and elementary occupations. Page 52 of 114 Given the current trends in wages, secondary graduates see a higher return on their educational investment than those with a lower level of education, who face low (and flat) returns. Secondary graduates working in the formal sector also expect the higher return to increase time. On the other hand, workers in the informal sector derive minimal or no benefit from achieving a higher level of education (Figure 54). Figure 54 Average monthly real wages in the formal and informal sector (2003=100) Formal jobs only Informal jobs only 3500 3500 3000 3000 2500 2500 2000 2000 1500 1500 1000 1000 500 500 0 0 2009 2010 2011 2012 2013 2014 2015 2016 2017 2016 2009 2010 2011 2012 2013 2014 2015 2017 Below high school Below high school High school High school Higher than high school Higher than high school Source: Authors’ calculations based on Labor Force Surveys. Note: Nuts 2 level CPIs were used in the calculation of real monthly wages. Premiums for university educated workers increased from 2009 to 2012 and then fell from 2012 to 2016.11 The decline in 2016 was particularly sharp. The wage premium remained roughly at the same level for secondary school educated workers compared to primary educated workers. One explanation for the trends observed in Figure 55 is that the increase in the supply of better educated workers might be running ahead of increases in demand. As the education level has been steadily increasing in Turkey, increases in the relative supply of better educated workers may have put pressure on the earnings premiums, leading to a sharp decline. High school graduates are sought after as indicated by the high growth of real wages and wage premium, and the large influx of better-educated workers that is easily absorbed by the labor market. However, firms continue to demand workers with low and medium skill levels, implying that wage premium increased substantially for them. 11 The log of real hourly wage is estimated as a function of education, work experience and labor market status. The estimated equation relates wages to education, experience, labor market status and other individual characteristics. Regional CPIs were used to obtain wages at constant prices. Education comprises the four following categories: (i) primary and less than primary education, (ii) secondary, (iii) high school (inc. vocational high school), and (iv) university degree. Labor market status consists of three categories: (i) interaction for permanent, paid and casual and formal workers, (ii) interaction for temporary, paid and casual and formal workers, and (iii) interaction for paid and casual and formal workers. The sample includes permanent wage and casual workers aged between 25 and 64. Page 53 of 114 As shown in Figure 56, the minimum wage in Turkey increased very sharply after 2016. Several studies find that minimum wages compress the distribution of wages among workers, because the minimum wage has a bigger impact on the wages in the bottom part of the distribution than on those higher up (World Bank, 2012). Figure 55 Wage premium over time in Turkey Figure 56 Evolution of nominal minimum wage in Turkey Secondary (8 yrs) High school University Level of Daily Minimum Wage (in TL) 1.2 67.7 59.3 54.9 53.4 1 46.8 41.3 43.4 0.8 36.8 30.5 33.3 29.0 32.5 24.8 27.2 26.3 0.6 22.7 21.5 24.0 19.6 17.9 0.4 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 0.2 0 Statutory Minimum Wage (Observed, Gross) 2012 2009 2010 2011 2013 2014 2015 2016 2017 Statutory Minimum Wage (Observed, Net) Source: Authors’ calculations based on LFS and official data from MoFLSS Workforce skills in relation to job creation and productivity growth Job creation has been driven mostly by large firms (at least 250 employees). Occupations that are intensive in non-routine cognitive skills represent almost 20 percent of total jobs in formal firms, while over 80 percent of jobs are occupations that are intensive in either routine or non-routine manual skills.12 Small and medium-sized enterprises (SMEs) (below 250 employees) are characterized by a larger share of jobs that are intensive in non-routine manual tasks, while larger firms present a higher incidence of employment in occupations with intensive non-routine cognitive tasks (Figure 57). Most of the jobs created during the three-year period 2014-2016 were in either non-routine manual or non- routine cognitive occupations: 60 percent and 35 percent, respectively. Job creation has been driven mostly by large firms, while SMEs have shed labor. Large firms have also created jobs in routine 12 Given the absence of information on workers’ educational attainment in the EIS data, skills of the workforce are proxied by the information available on occupations ISCO-codes. Occupations are classified therefore into three broad categories: (i) Occupations intensive in non-routine cognitive tasks and interpersonal skills, such as professionals (doctors, lawyers, architects, dentists, journalists), academics, managers; (ii) occupations intensive in routine tasks (both manual and cognitive), such as clerks, secretaries, administrative assistants, book-keepers, mechanics, call-center attendants; and (iii) occupations intensive in non-routine manual tasks, typically involving repairing, attendance, care, or some form of interaction with clients, such as plumbers, electricians, care-givers, caretakers, security personnel, sales workers, cleaners, waiters, bartenders. Page 54 of 114 occupations, while both medium-sized enterprises (between 50 and 250 employees) and smaller firms (below 50 employees) have reduced jobs in routine occupations in this period (Figure 58). Figure 57 Employment composition by type of Figure 58 Job creation (2014-2016) by type of occupation and firm size (2016) occupation and firm size 60 53.0 300000 50 46.9 42.9 250000 39.238.5 40 35.0 200000 Percentage 33.0 Job creation 29.3 150000 30 22.3 22.1 100000 20.1 20 17.7 50000 10 0 -50000 0 -100000 Below 50 [50-250) 250+ Total Below 50 [50-250) 250+ Total Non-routine manual Routine Non-routine cognitive Non-routine manual Routine Non-routine cognitive Source: Authors’ calculations using EIS data. Employment in firms characterized predominantly by non-routine manual occupations accounted for half of total employment.13 Figure 59 shows that, in 2015, (i) firms where employment was higher than the national average in non-routine cognitive skills accounted for 16.4 percent of total employment; and (ii) firms characterized predominantly by “routine” occupations absorbed almost 35 percent of total employment. The trends in employment composition by firm type appear quite smooth from 2006 to 2015. Firms that have the largest share of workers in non- routine manual occupations have increased slightly their employment share over time. On the other hand, the share of employment in firms that absorb a higher than average number of workers in non- routine cognitive occupations has slightly declined. 13To assess the relationship between firms’ skill composition and employment trends, firms were classified into three groups, based on the percentage of employees engaged in the three types of occupation-related skills: (i) firms where the share of employment in non-routine manual occupations is higher than the national average; (ii) firms where the share of employment in routine occupations is higher than the national average; and (iii) firms where the share of employment in non-routine cognitive occupations is higher than the national average. The analysis here is restricted to firms that were registered for the full period considered (2006-2015). Page 55 of 114 Figure 59 Employment composition by firm type based on skill intensity (in percent) 60.0 50.0 40.0 30.0 20.0 10.0 0.0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Non-routine manual Routine Non-routine cognitive Source: Authors’ calculation based on EIS data: firms registered over the full period 2006 -2015. Job quality The goal for Turkey is not just to create ‘more jobs’, but also to create ‘better jobs’. Unfortunately, generating enough quality employment opportunities is one of the challenges Turkey is currently facing. The distribution of job quality itself is widespread and multimodal. We observe a shift to the left of the entire job quality distribution between 2014 and 2017, resulting in lower job quality at the aggregate level. Figure 60 shows that the wide distribution of job quality in Turkey remains; such distribution is spread from as low as 0 (worst quality) to as high as 1 (best quality) in 2017. Interestingly, the left tail of the distribution has spread out further in 2017, compared to 2014, indicating that there are more bad quality jobs in 2017 than in 2014. As mentioned, there is an overall leftward shift in the 2017 distribution compared to 2014, indicating reversals in job quality at the top and the bottom end of the distribution since 2014. (Figure 60). Average job quality of wage employment improved between 2009 and 2017, with some slight reversals in recent years. There are also persistent differences between formal permanent jobs, formal temporary jobs and informal jobs; and there have been minimal increases in job quality in the informal sector between 2009 and 2017 (Figure 61). Job quality clearly plays a significant role in individuals’ well-being (Scarpetta et al. 2016). Creating a positive and safe work environment, providing competitive benefits, and facilitating flexible work arrangements all matter for worker satisfaction. Workers who are content and work with a positive mind-set have higher levels of creativity, engagement, and productivity (Achor 2012; Lyubomirsky and Diener 2005; Oswald et. al. 2015;). Thus, higher job quality promotes LFP, productivity, and overall economic performance. In 2009, the average JQI for wage-employed workers was 0.66. Aggregate job quality remained showed slight improvement until 2011 and improved somewhat in later years, reaching 0.72 in 2015. In the past 2 years, there has been decline in job quality (0.70 for 2016 and 2017) (Figures 61). Page 56 of 114 Figure 60 Distribution of Job quality Figure 61 Job quality index (JQI) by contract type, 2009-2017 5 4 3 2 1 0 0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1 JQI score 2014 2017 Source: Authors’ calculation based on LFS data Between 2014 and 2016, job quality has declined at the aggregate level, largely a result of offsetting trends at the individual component level. Figure 62 shows trends over time for all 13 components between 2014 and 2017. While resilience to individual shocks and work safety improved, lower minimum wage compliance, reduced resilience to aggregate shocks and higher risk of skills obsolescence mean that job quality declined between 2014-17 (Figure 62). Negative effects of decreased compliance with minimum wage shows that the labor market is becoming more informal, and informality is highly correlated with worse labor market conditions at work and limited opportunities for career prospects. There are improvements in terms of decreased vulnerability to idiosyncratic shocks, indicating that people have a better ability to cope with negative changes in their personal circumstances and do not face job loss because of them. Figure 62 Scores by component of the JQI, 2014-2017 Source: Authors’ calculation based on LFS data Page 57 of 114 Job quality varies substantially by sub-regions. In 2017, aggregate job quality was highest in Ankara. Other regions with relatively high scores are near Ankara, namely: Bursa, Eskisehir, Bilecik and Bolu, Duzce, Sakarya, Yalova, Kocaeli. Low quality jobs are concentrated in the south eastern part of Turkey where formal jobs are limited and of less quality than in other parts of Turkey. Ankara, where a large share of formal jobs is in public sector, has led the ranking and has had limited change in its’ score or rank since 2009. Ankara has the highest score because it has the largest share of formal wage jobs and has high standards regarding minimum wage compliance and employment engagement (or contracts). These are critical components in the index which contribute to high and improving levels of job quality (Figure 63). Figure 63 Job quality of wage employment, 2017 Kastamonu Istanbul Zonguldak Tekirdag Trabzon Kocaeli Samsun Bursa Ankara Erzurum Agri Balikesir JQI scores Kayseri Malatya (0.73,0.78] Manisa (0.70,0.73] Kirikkale Van (0.65,0.70] Izmir [0.55,0.65] Konya Aydin Hatay Mardin Antalya Sanliurfa Adana Gaziantep Source: Authors’ calculations based on LFS The score reported so far is relevant to wage employment only. When focusing on all workers, wage workers, self-employed and subsistence workers, scores decrease substantially. The focus on wage workers excludes about 20 percent of Turkey’s workers who are engaged in subsistence agriculture and informal small-scale self-employment. This is important because there are regions where the share of non-wage work can be as high as 55 percent, meaning that a large percentage of the workforce are excluded from the job quality measure. When all workers—wage workers, self- employed and subsistence workers—are included, scores decrease substantially. In fact, the impact of including all workers—many of them low-productivity and informal—reduces the score in 2014 to 0.45, compared to 0.72. The reduction is similar in 2017. Many argue that the lower score, which is based on 9 components of the index rather than 13, reflects the real picture of job quality in Turkey. Sub-national estimates of job quality for all workers also show important regional disparities. When all workers are included, the quality of employment continues to be substantially lower in the eastern part of the country. JQI estimates are lowest in sub-regions such as Agri-Kars-Ardahan-Igdir, Kastamonu-Sinop-Cankiri, Samsun-Amasya-Tokat-Corum, Erzurum-Erzican-Barbut, and Sanliurfa- Diyarbakir, with scores below 0.35. Workers in Ankara have on average the highest job quality by far, followed by workers in Istanbul. Having a better understanding of which job quality components matter most can help design effective policies to promote the creation of good jobs. The analysis identifies four relevant latent factors that matter most for job quality; these are: formality, adaptability of skills and employment resilience, wages, safety and protection. Within formality, social security benefits and employment engagement Page 58 of 114 vary together and influence job quality in a similar way. If one component improves, the other will also improve. Rigid forms of employment limit employers’ ability to react to changing labor markets (namely peaks and troughs in demand). Non-wage costs affect the cost of employment directly. Evidence for Turkey shows that high non-wage costs (high labor tax) create a burden on employers and slow down employment creation in the formal sector, while encouraging informality. Table 1 Key factors of good job creation Adaptability of Safety and Formality skills and Wage protection resilience social security minimum wage skills obsolescence work safety benefits compliance resilience to employment resilience to relative idiosyncratic engagement aggregate shocks underpayment shocks The association between skills obsolescence and resilience to aggregate shocks suggest that adaptability of skills and employment resilience matter jointly for good job creation. Adaptability to changing economic conditions, be it short-term impacts of a crisis or long-term challenges related to structural transformation, stems largely from acquiring malleable skills which can be applied to distinct tasks and occupations and which are useful across economic sectors. Achieving employment resilience to external shocks such as the global financial crisis requires the implementation of macro-prudential regulations that ensure Turkey’s financial sector is not unduly exposed to irregular financial transactions. For instance, an increase in non-performing loans is likely to result in tighter credit conditions, particularly for small and medium-sized enterprises. Negative effects on the real economy can decelerate formal sector job creation and push workers into informal or short-term temporary contract work. Similarly, the increased inflow of refugees in Turkey has direct effects on employment patterns. Effects are negative among informally employed workers. Improving compliance with the minimum wage in per hour terms also helps reducing the share of wage workers that are underpaid. Wages affect the cost of formal employment directly. Noncompliance with minimum are not unusual in the Turkish economy, especially in low productive sectors where informal work is more common. Appropriateness of wage levels cannot be determined without considering relative worker productivity. Many workers earn at or above minimum wage, but they are underpaid relative to their peers and relative to their productivity levels; underpayment is likely in labor markets where formal work is limited, and supply of suitable workers exceeds the demand for their labor. Aspects of safety and protection describe a fourth area of good job creation. Improving work safety, enforcing child labor laws and strengthening workers’ resilience to individual shocks are critical measures to generating safe jobs that can also protect workers. Weak enforcement of labor regulation, high employment costs and persistent informality in the labor market are all contributing factors to relatively high levels of work-related accidents and injuries (especially in metal, mining and construction), and the persistent utilization of children in work for hire (especially in agriculture). Working conditions in many occupations in Turkey involve distinct and severe hazards to worker health which reduces wellbeing, working capacity and even the lifespan of individuals. Some of these Page 59 of 114 trends probably intensified with the recent influx of asylum seekers in the country, many of them children that are active in the informal economy. Page 60 of 114 3. LABOR DEMAND IN THE FORMAL SECTOR Profile of formal sector labor demand Differences in labor market dynamism across regions in Turkey are also associated with the differences observed in firm characteristics. Like many developed countries such as Italy and Germany, Turkey has an unequal distribution of labor market performance. Data show clear differences in labor market dynamism across regions in Turkey. These differences are directly and indirectly related to the differences observed in firm characteristics and how they are distributed across Turkey. Some of the reasons for subnational differences are economic specialization, investment climate, ethnic and demographic composition, and geographic location. Recent studies show that regional unevenness in economic development and urbanization and demographic trends, agricultural shedding, and uneven educational attainment have exacerbated labor market dynamism. The ongoing conflict in neighboring countries (namely the Syrian Arab Republic and Iraq) affects trade and economic dynamism, and also the demographic composition of some of Turkey’s poorest and least dynamic subregions, which are hosting hundreds of thousands of people seeking refuge (Del Carpio and Wagner 2016). Some subregions are more economically dynamic than others. Such subregional asymmetry is exacerbated by differences in firm characteristics, obstacles faced and opportunities encountered, and overall firm performance. In fact, it is plausible to find that firms operating in the same (narrowly defined) industry can have distinct levels of productivity depending on their characteristics and business environment (Doms and Bartelsman 2000; Syverson 2011). Therefore, it is important to not only analyze what characteristics best determine firm performance, but also to evaluate what obstacles hinder firms from performing to their fullest potential to mitigate them. Formal firms and employment are mostly concentrated in Turkey’s large western regions such as Istanbul, Ankara, Aegean, and East Marmara. Figures 64 and 65 show the distribution of formal firms and employment across Turkey’s provinces in 2016. As expected, the majority of the formal firms and employment are located in the large provinces, such as Istanbul, Bursa, Izmir, Ankara, Antalya and Konya. However, the number of formal firms and employment is also quite significant in the southeastern regions of Turkey (such as Adana, Gaziantep, Sanliurfa), which are the provinces with the highest share of SuTPs. On the other hand, the density of formal firms is significantly lower in the eastern part of Turkey where many workers are outside the labor force, engaged in self-employment or agriculture work, or are employed in (unregistered) informal firms, which are excluded from the sample (Figure 64). The differences in density of formal firms across Turkey is an important characteristic to keep in mind when reading this chapter. For instance, various obstacles to doing business (described in detail in this chapter) are more likely to be accentuated in provinces where the density of firms is higher than in provinces with fewer formal firms, thus often resulting in counterintuitive findings. Page 61 of 114 Figure 64 Distribution of formal firms by province, 2016 Source: Authors’ elaborations based on EIS data Figure 65 Distribution of employment by province, 2016 Source: Authors’ elaborations based on EIS data Economic growth benefited Turkey’s eastern regions in particular. As shown in Figures 66 and 67, the number of firms and employment increased between 2007 and 2016 in a majority of eastern and southeastern provinces. Figure 66 Annual growth in the number of firms between 2007-2016 (in percent) Source: Authors’ elaborations based on EIS data Page 62 of 114 Figure 67 Annual growth in employment between 2007-2016 (in percent) Source: Authors’ elaborations based on EIS data Although the great majority of registered firms are small, in aggregate large firms employ the most workers. The Turkish economy is dominated by micro and small-size firms (employing less than 19 workers); they are the backbone of the Turkish economy, accounting for well around 90 percent of all firms. (Figure 68). While up to 85 percent of firms have less than 20 employees, most wage workers (about 70 percent) are in firms with at least 20 employees (Figure 68). In fact, more than 40 percent of workers in registered firms work in firms with over 100 employees. Thus, while there are relatively few large firms, their size ensures they account for a significant share of employment. In contrast, registered firms with up to 9 employees account for only 10 percent of employment. Furthermore, while the share of large firms increased from 3 percent to 5 percent between 2007 and 2016, their share of employment grew from 43 percent to 48 percent. Figure 68 Distribution of firms and employment share by firm size, 2007-2016 Source: Authors’ elaborations based on EIS data The very largest firms are mainly responsible for the growth in both jobs and revenues. The top 5 percent of firms account for around 60 percent of jobs. However, the top 1 percent of firms accounted for a growing share of employment and even greater share of revenues between 2007 and 2016 (Figures 69 and 70). While the composition of firms is shifting towards smaller firms over time, Page 63 of 114 the share of employment accounted for by the largest 1 percent of firms has increased from 37 to 40 percent in the period. The employment share of the top 5 percent and 10 percent of the largest firms has been around 60 percent and 70 percent, respectively. Figure 69 Employment accounted for by largest Figure 70 Revenue accounted for by largest firms firms Source: Authors’ elaborations based on EIS data Young firms tend to be smaller than older firms. Turkey has a relatively large share of young firms; around 50 percent were 5-years old or less in 2016. Young firms are generally smaller than older firms, and thus employ fewer workers (Figure 71). Firm age may matter for job creation because it is also associated with overall firm performance and stability; however, the direction of the relationship is not straightforward, meaning that in some contexts, young firms do better than older ones and vice versa. For instance, there are cases where older firms remain operational because they are productive and innovative, while there also cases where older and larger firms are not productive but they survive, despite having low performance, because of access to various external support (for example, inexpensive credit and subsidies) mechanisms. Older productive firms are good sources of formal employment as they provide stability and also opportunities for innovation and career progression. On the other hand, older unproductive firms that are sustained artificially through unsustainable means are less favorable for the economy because they typically misuse human resources that can be better utilized elsewhere. The firm age-productivity relationship has been studied in Turkey, but not extensively, and results are limited to few economic markets (that is, Ankara, Istanbul) and only some economic sectors. Yildiz et al. (2013) show that in Ankara firm age does not determine firm profitability; instead, other factors (for example, economic sector, access to capital, ownership type) have a more relevant effect on overall firm performance. Further analysis of this relationship in Turkey is merited to guide policy makers in their allocation of resources for private sector support to maximize sustainable employment. Page 64 of 114 Figure 71 Distribution of firms and employment share by firm’s age, 2007-2016 Source: Authors’ elaborations based on EIS data The ownership structure of the firm is also an important determining factor for firm success, especially because there is a relationship between innovation, productivity, and foreign ownership (Guadalupe, Kuzmina, and Thomas 2010)—all of which are good for job creation. As expected, micro and small firms, which employ few workers individually but engage the vast majority of workers as a whole, are 99 percent domestically owned. Large firms, especially foreign- owned firms, are more likely to engage in export-related activities. But in Turkey, few large firms export products or services. Yellow diamonds in Figure 72 show that only 24 percent of all large firms engage in export activities, while only 8 percent of all medium firms export abroad. Foreign-owned firms are more likely to be large in size and operate formally. Given their size, they typically employ large numbers of workers (over 100). Thus, they are not only important in terms of productivity but also as a source of formal work. Foreign ownership is not common across Turkish firms; only 4.3 percent of all large firms are foreign owned. This is an issue of concern because a large body of evidence shows that productivity increases when firms enter into foreign markets (through exports) because international trade facilitates the transfer of knowledge and best-practice techniques (Blalock and Gertler 2004; Fafchamps 2009). Apart from technological transfers, another reason for having a larger presence of foreign-owned firms is that these firms are more likely to import international practices, especially human resource practices, which is good for the Turkish economy. Foreign-owned firms are more likely to seek and hire talent irrespective of their gender, to amplify their talent pool (Olcott and Oliver 2014; Siegel, Pyun, and Cheon 2010). Foreign-owned firms competing in international markets face competitive pressures that force them to increase their talent pool irrespective of gender (Heyman, Svaleryd, and Vlachos 2013). This is especially important in a context where young people, women especially, are increasingly more educated than their parents and seek more opportunities for skilled formal work. In labor market contexts where female labor participation is low, even among highly educated women, having well-remunerated jobs that compensate for the opportunity cost of staying home is critical to maximize the utilization of the full talent pool. Page 65 of 114 Figure 72 Ownership status and firm size 2015–2016 Source: Enterprise Survey 2015–2016. Note: Yellow diamond depicts the percentage of firms engaging in export activities. In Turkey, the estimate of firms with some female ownership participation is less than 15 percent in 2015–2016 and 25 percent in 2013 (for all firms). Another dimension that matters for human resources in firms is the gender composition of the owners and leaders. Comparable data for some countries in East Asia and Pacific show that 50 percent of all firms have some level of female ownership (Figure 73). The percentage is about 38 percent for Latin America and 32 percent for Sub- Saharan Africa (World Bank 2011). Only the countries included in the East Asia and Pacific sample have reached parity in female firm ownership; Turkey and other countries have not been able to foster entrepreneurship among women, and have problems (likely structural) that prevent women from accessing top leadership positions. Such findings are worrisome because studies show that women make a substantial positive impact on economic growth through various channels: employment, innovation, and wealth creation (Brush et al. 2006). It is also worrisome because the gender of the owner (and key stakeholders) has an effect on the firm’s gender composition of the leadership structure. A recent study finds that firm ownership and corporate governance explain the likelihood of having a certain gender composition in upper management. It finds that firms that have a female chief executive officer are more likely owned or directed by majority female shareholders (Sekkat, Szafarz, and Tojerow 2015). This relationship likely explains why the levels of female top managers in Turkey are about 9 percent. Again, this is a concern because Turkish women have education and skills levels that can fulfill top leadership positions; however, given the current statistics, it is likely that they will hit a glass ceiling and be prevented from exerting their full potential. This in turn hurts the firm’s performance and the economy as a whole. Arguing for the importance of equal treatment of men and women in the workplace is outside the scope of this chapter. But it can be highlighted that the absence of women throughout the various managerial layers can have a negative effect on firm performance; this was the case in a recent study in Germany (Gagliarducci and Paserman 2014). There is also a relation between the number of women who sit on the board and the profitability of female top managers, where more women on the board positively affect the performance of female leaders (Flabbi et al. 2014; Sekkat, Szafarz, and Tojerow 2015). In other words, it is important to have a critical mass of women and men in management to benefit from having a gender balance in management. Page 66 of 114 Figure 73 Gender composition of ownership and management 30 25 20 Percentage 15 Turkey 2013 10 Turkey 2016 5 0 Percent of firms with female Percent of firms with female top participation in ownership manager Source: Enterprise Survey 2015–2016. The wholesale and retail trade is the sector with the highest share of firms, whereas the manufacturing sector holds the largest share of employment in the formal labor market. Figures 74 and 75 show the distribution of firm and employment shares by sector in 2007 and 2016. Even though the share of the wholesale and retail trade sector decreased by 6 percentage points from 2007 to 2016, it still has the largest share in the formal economy (38.7 percent in 2016). Manufacturing and construction have a share of about 30 percent of total formal firms. However, the employment share of the wholesale and retail trade sector decreased (from 21.9 percent in 2007 to 20.5 percent in 2016) and is relatively low compared to the manufacturing sector. The employment share of the manufacturing sector decreased from 35.7 percent in 2007 to 29.9 percent in 2016, whereas that of construction increased from 10.4 percent in 2007 to 12.3 percent in 2016. Figure 74 Distribution of firms share by sector, 2007-2016 100% Human health and social 90% work activities Education 80% Public administration and 70% defence Administra-tive and 60% support service activities 50% Professional, scientific and technical activities 40% Real estate activities 30% Financial and insurance activities 20% Information and 10% communication Accommodation and food 0% service activities 2007 2016 Transportation and storage Source: Authors’ elaborations based on EIS data Page 67 of 114 Figure 75 Distribution of employment share by sector, 2007-2016 100% Human health and social work 90% activities Education 80% Public administration and 70% defence Administra-tive and support 60% service activities 50% Professional, scientific and technical activities 40% Real estate activities 30% Financial and insurance activities 20% Information and 10% communication Accommodation and food 0% service activities Transportation and storage 2007 2016 Source: Authors’ elaborations based on EIS data While the industry sector expanded more than the other sectors, the services sector created more jobs, especially for low-skill workers. The electricity, gas and steam subsector expanded between 2007 and 2016 at an annual growth rate of 24 percent, while the water supply and sewerage subsector’s expansion was 18 percent (Figure 76). Despite this significant growth in the industry sector, education, human health and social work, construction, and the wholesale and retail trade sectors created more jobs in this period. The possible consequences of this finding on productivity merits further research. Figure 76 Annual change in number of firms and employment, by sector, 2007-2016 Human health and 12% social work activities Electricity, gas, steam (Q) (D) Annual change in number of employment 10% 8% 6% (2007-2016) Water supply, 4% sewerage etc.( E) 2% 0% -15% -10% -5% 0% 5% 10% 15% 20% 25% 30% -2% Public administration -4% and defence (O) -6% Annual change in number of firm (2007-2016) Source: Authors’ elaborations based on TUIK data Page 68 of 114 Box 3 High Performing Firms Understanding which sectors are most likely to create jobs can improve the design of policies. Based on two outcomes (employment and value added), sectors were ordered according to incidence of having HPFs. Overlapping between employment and value added HPFs is quite high. Security and investigation services, building and landscape service, air transport and manufacture basic pharmaceutical product are the sectors with the highest incidence of employment and value added HPFs. Incidence of HPFs of employment and value added Incidence of employment HPFs Incidence of value added HPFs 0% 10% 20% 30% 40% 50% 0% 10% 20% 30% 40% 50% 60% 80. Security and investigation 21. Manufacture of basic activities pharmaceutical products 81. Services to buildings and 51. Air transport landscape activities 80. Security and investigation 51. Air transport activities 91. Libraries, archives, 97. Activities of households as museums employers of domestic personnel 29. Manufacture of motor 91. Libraries, archives, museums vehicles, trailers and semi… 42. Civil engineering 29. Manufacture of motor vehicles, trailers and semi trailers 21. Manufacture of basic 62. Computer programming, pharmaceutical products consultancy and related activities 87. Residential care activities 81. Services to buildings and landscape activities 97. Activities of households as 28. Manufacture of machinery employers of domestic… and equipment n.e.c. 23. Manufacture of other non 30. Manufacture of other metallic mineral products transport equipment 26. Manufacture of computer, 26. Manufacture of computer, electronic and optical… electronic and optical products 17. Manufacture of paper and 22. Manufacture of rubber and paper products plastic products 62. Computer programming, consultancy and related… 87. Residential care activities 22. Manufacture of rubber and plastic products 86. Human health activities 86. Human health activities 20. Manufacture of chemicals and chemical products 11. Manufacture of beverages 25. Manufacture of fabricated metal products 24. Manufacture of basic 23. Manufacture of other non metals metallic mineral products 27. Manufacture of electrical 17. Manufacture of paper and equipment paper products 28. Manufacture of machinery and equipment 24. Manufacture of basic metals 20. Manufacture of chemicals 27. Manufacture of electrical and chemical products equipment Note: Top 20 sectors were seleceted for visualization Page 69 of 114 Sectoral overlap between employment and value-added HPFs: OECD definition Source: Authors’ elaborations based on EIS data Note: Among 18 different definitions for identifying a gazelle firms, the graphs used the results based on OECD gazelle definition, which requires an overall growth of 72.38 percent from period t to t+3 (i.e. the equivalent of 20% a year, but with no restriction on how that overall growth is achieved across the years). In the aggregate, the rates of job creation have increased over time. Net job creation in the formal sector has been increasing since 2009, though it declined after 2015 (Figure 77, left panel). One possible explanation for this decline could be the hike in the minimum wage in 2016. Job creation can occur through the entry of new firms or through the expansion of existing firms (i.e., incumbents). For Turkey in 2016, job creation by incumbents is 1 million 127 thousand or 87 percent of all jobs created, while entry is 6 percent (Figure 77, right panel).14 Likewise, more jobs are lost from incumbents who are downsizing (1 million 70 thousand) than from exits (though 32 percent of all jobs destroyed are due to exits, roughly 514 thousand jobs). Overall, net job creation has been positive as the wage sector has expanded in the economy. 14Figure 77 (right panel) has net job creation (job creation minus job destruction), but also total job creation by entry and incumbents and total job destruction by exit and incumbents. Page 70 of 114 Figure 77 Job flows in Turkey Source: Authors’ elaborations based on EIS data An analysis of dynamics over a short period shows that, as in most countries, a sizable share of firms changed size category. Depending on size category, 40 to 85 percent of firms changed size category over 10 years (Table 2). Very few firms grew significantly, moving from a micro or small firm to a large firm. Instead, firms that grew mostly do so by moving up one size category. However, downsizing is more common. More than 30 percent of small (10-19) and medium-sized (20-99) firms downsized, and another 20 percent went out of business. Downsizing and exit were less prevalent among larger firms. Table 2 Transition Matrix Transitions 2007-2016 in 1-9 10-19 20-99 100+ exit % Turkey 1-9 57 8 3 0 32 100 10-19 20 35 28 1 16 100 size at birth to age 5 20-99 8 10 56 12 14 100 100+ 2 1 9 77 12 100 1-9 83 8 2 0 8 100 10-19 23 49 22 0 6 100 size t to t+3 20-99 6 12 70 7 5 100 100+ 2 1 12 80 6 100 Source: Authors’ elaborations based on EIS data The relationship between productivity and employment Figure 78 illustrates the productivity levels of an average firm based on different measures of the productivity. As shown in the figure, there is increase in all productivity measures. One could argue that minimum wages can contribute to higher labor productivity through several channels. One way could be capital-labor substitution, implying that productivity increases may be the result of declines in employment due to the increase in minimum wage. In this case, firms adopt more capital-intensive production technologies and substitute capital for labor. One way could be labor-labor substitution. In order to keep same production, due to higher labor costs, firms might shed labor to reduce costs and protect profits. This might result in increasing informal labor. One also could argue that with higher minimum wage levels, there might be decline in turnover and employers can easily retain their labor Page 71 of 114 via on the job training and other efforts in response to higher wages. Thus, workers might become more productive in the long run. Figure 78 Average firm productivity Source: Authors’ elaborations based on EIS data Larger firms are more productive, on average, though productivity differences by firm size are not that notable. It is desirable to have a positive correlation between the size of a firm and productivity (static efficiency), but even more so to have the ones that are increasing their productivity to be growing (dynamic efficiency). More productive firms gaining resources and expanding will be the drivers of overall productivity growth. In the presence of scale economies, larger firms should be more productive. Figure 79 shows that there are very productive (right tail) – and unproductive (left tail) – firms in each size category, though larger firms skew more toward higher levels of productivity. The same pattern is observed for the relationship between value added per worker and firm age, implying that more mature firms tend to be more productive (Figure 80). Page 72 of 114 Figure 79 Value added per worker, by firm size Figure 80 Value added per worker, by firm age Source: Authors’ elaborations based on EIS data Over time, the cross term (the correlation of changes in productivity and changes in employment) is negative. Figure 81 shows the causes of change in a logarithm of value added per worker. The cross term is negative, implying that firms with growing productivity are shedding workers rather than hiring more workers. However, there is evidence of reallocation towards firms that are more productive (between term is positive). Entry effects seem to be important for change in value added worker. Figure 81 Change in value added per worker, by sector Source: Authors’ elaborations based on EIS data Note: “M_Other” stands for other manufacture sectors. The relationship between capital intensity and jobs differs by sector. Figure 82 and 83 reveals the ratio between capital and labor and labor cost by sector, respectively. Manufacture and services Page 73 of 114 firms tend to be more capital intensive than mining, utilities and commerce sectors. There is substantial increase in capital-labor ratio in 2016. The relationship between labor regulations and capital-labor-ratio (or capital intensity) is well documented in the literature (such as Autor et al. (2007), Calgagnini et al. (2014), Cingano et al. (2010 and 2014)). There are findings that higher labor regulations increase capital-labor ratio (such as Janiak and Wasmer (2014), Cette et al., (2017). One could argue that higher labor regulations might be increased capital-labor ratio in Turkey. Figure 82 The ratio between capital and labor, by Figure 83 The ratio between capital and labor sector cost Source: Authors’ elaborations based on EIS data Administrative records show that most of the formal workers in registered firms are earning around the minimum wage. Figure 84 indicates in blue the concentration of workers who are earning around the statutory minimum wage (between the minimum wage and 5 percent over).15 The top sectors where most workers earn around the minimum wage are construction and repairing of land transport. However, certain subsectors of the manufacturing sector with a high share of informal workers (such as the manufacture of leather goods) also pay workers around the minimum wage. These subsectors are the most vulnerable to increases in labor costs (namely a minimum wage increase) since they pay workers around the minimum wage. Maps at the province level for those subsectors are presented in Appendix. Such information provides important input for developing relevant policy tools to combat informality in Turkey. Fortunately, some subsectors like air transport, information and electricity pay workers well above the minimum wage. 15One could argue that some workers might be recorded in the books as minimum wage earners for the firm to save from payroll tax contributions. Opposite might be also the case, some wage earners who are entitled to the minimum wage are paid wages below the legal floor. Page 74 of 114 Figure 84 Concentration of workers around the minimum wage by sector, 2016 Note: dwage stands for daily wage. Page 75 of 114 Firm perceptions of obstacles to performance High tax rates and access to finance are the two top concerns in Turkey and have been for years. Firm characteristics, such as firm’s age, size presented before can affect firm performance, but there are other non-firm specific factors that are also correlated with productivity, employment, and overall firm outcomes. These are contextual and policy factors that affect the business climate. Figure 85 shows that high tax rates and access to finance have been the two top concerns in Turkey for a number of years. Nevertheless, informal competition and having an inadequately educated workforce are also major concerns of firms of all sizes. In addition, these obstacles are persistent across time. Given that the job diagnostics focuses on labor market outcomes, the focus here is on the perceptions of firm owners and managers about three contextual factors (obstacles) that negatively affect labor market performance. These are informal competition, excessive labor regulation, and inadequacy of the education and skills of the workforce. They are elaborated below in detail. Figure 85 Percentage of firms identifying various obstacles to their performance Tax rates Access to finance Informal competitors Political instability Poorly educated workers Tax administration Labor regulations Licensing and permits Electricity Crime, theft and disorder Transport Turkey 2016 Corruption Customs and trade… Turkey 2013 Access to land Turkey 2008 Courts 0% 5% 10% 15% 20% 25% 30% 35% 40% Percentage of firms Source: Authors’ elaborations based on WB Enterprise Survey 2015–2016 1. Informal competition and informality Formal firms see informal competition as a key problem in Turkey. Greater competition is often associated with productivity increases. However, this relationship holds when competition is favorable to the business environment, like deregulation and trade liberalization (Dollar, Hallward-Dreimeier, and Mengistae 2005; Verhoogen 2008). In other words, the positive relationship comes about when the playing field is equal for all, not when competition stems from unlawful actions by some firms to be more competitive (or in some cases, to survive). Almost all countries have informality in their economies, but in developing countries like Turkey, the share of the economy that is not accounted for is larger, employing a large segment of the labor force. Formal firms see informal competition as a key problem in Turkey because informal firms do not contribute to the country’s tax revenue, thus putting pressure on formal firms to increase their tax burden while having to remain competitive. Informal firms often serve as a low-cost input provider for semiformal firms, which exacerbates the unevenness of the playing field for law-abiding firms. Such unfair competition has a negative effect Page 76 of 114 on the economy as a whole because formal firms, which offer better labor conditions, are sometimes forced out of business or into informality to cope with uneven competition (OECD 2009). Formal firms across Turkey, especially the smaller ones, recognize that informal competition is a problem. But it is more commonplace in some subregions more than others. Figure 86 shows that informal competition is viewed as a more prevalent obstacle in western and central subregions; in the west: Istanbul, Edirne-Tekirdag-Kirklareli, Denizli-Aydin-Mugla, and Manisa-Afyonkarahisar-Kutahya- Usak, and in the central subregion: Adana-Mersin and Kayseri-Sivas-Yozgat. Figure 86 Informal competition, by region, 2015–2016 (percentage of firms reporting it as a top obstacle) Source: Authors’ elaborations based on WB Enterprise Survey 2015–2016 The Turkish Government has been actively seeking to reduce informality through various means. Informal labor and informality in the economy have decreased in the last decade. But informal practices are still commonplace, and labor informality remains high at about 36 percent (World Bank 2016). The government has developed various anti-informality strategies that include actions for various institutions to implement and enforce. The first campaign, which took place in 2005, was a comprehensive action plan (Struggle against Informal Employment) launched by the Ministry of Labor and Social Security. A new strategy was proposed in 2009, called the ‘Struggle against the Informal Economy Action Plan’, and another in 2011, which covered the period until 2013. These were led by the Revenue Administration and with the core participation of other government institutions. Each strategy included several actions aimed at eliminating informal competition and informalization of the economy more broadly. 2. Excessive labor regulation Excessive labor regulation tends to lead to noncompliance and thus contributes to persistent informality rates. Evidence shows that, when labor market regulations, notably those determining the flexibility of employment, are rigid, they prompt widespread noncompliance with the law. Rigid labor regulations also constrain employers from creating more formal jobs. Various studies show that employment protection legislation (EPL) in Turkey is stringent compared to other OECD countries, largely because of collective dismissal and severance payment regulations (OECD 2014). Thus, it can be argued that by easing labor market rigidities, especially through regulation aimed at reducing labor costs, formal firms in Turkey would be more likely to create new formal jobs. Page 77 of 114 The regulation on the minimum wage is also contributing to formal firms’ reluctance to create jobs. The minimum wage is relatively high and increasing. In fact, the ratio of the minimum wage to median wage was about 69 percent in 2014, the highest among all OECD countries. But this ratio has likely increased (to about 80 percent), with the January 2016 announcement of a 30 percent increase. Such an increase was not fully aligned with present socioeconomic conditions, which risks having a large share of the population earning above their level of labor productivity, which in turn affects key labor market outcomes (for example, unemployment, LFP, informality). Firms in Istanbul and along the more economically prosperous subregions in the southwest, plus the subregion where Gaziantep, Adiyaman and Kilis are located, find that labor regulation is excessive and poses an obstacle to their functioning. Figure 87 illustrates the subregional breakdown; the east and west divide is less apparent with respect to this obstacle than the previous one on informal competition. Figure 87 Labor regulation, by region, 2015–2016 (percentage of firms reporting it as a top obstacle) Source: Authors’ elaborations based on WB Enterprise Survey 2015–2016 One way to counteract the rigidity has been to introduce more flexible forms of contract. The current Labor Law, enacted in 2003, combined for the first time some measures of flexibility with regard to job security and some measures against the arbitrary termination of employment contracts by employers (Tuncay 2013). The legislation made it possible for temporary contracts to be offered to employees; this was a move toward increased flexibility in the labor market. There is initial evidence that some of the reductions in informality are due to the introduction of formal forms of temporary employment. In 2016, a new set of regulations was passed to allow for more flexible forms of employment, but these have not been beneficial to workers. The regulations included better-regulated temporary work contracts, part-time work, and home-based work (World Bank 2016). The objective of these reforms was to increase LFP and reduce the costs of employment and dismissal. However, no reforms have been made to the two more costly regulations: severance pay and unemployment insurance. The recent amendments were made mainly to improve the dynamics of the labor markets by ensuring flexibility for employers to manage their human resources, but the effects are unlikely to be positive for workers without parallel changes to mechanisms that help protect them from income loss and unemployment in the transition. The main constraint identified in the unemployment insurance system is the eligibility requirement,16 which is inconsistent with most fixed-term contracts. This has 16Employees need to be working in a job covered by social security at least 600 days in the previous three years and must pay contributions for at least 120 days before termination of the contract. Page 78 of 114 limited how workers can transition from one job to another. Thus, employers can increase their use of temporary and part-time forms of contracting for workers, which enables them to avoid dismissal measures, but many workers have limited access to any type of income support during job search periods. 3. Inadequately educated and skilled workforce An inadequately educated and skilled workforce is most salient for firms located in the western parts of Turkey. The importance of skills, beyond formal educational achievement, in fostering employment and productivity is widely documented. Across occupations, employers ask for workers to have a mix of advanced cognitive, technical, and socioemotional skills to be successful in the labor market. Such breadth of skills increases the employability of the labor force; enables workers to carry out their job more efficiently, use new technology, and innovate; and allows firms to progress up the industries’ value chain. By contrast, the inadequacy of such skills negatively affects the firm’s performance. As shown in Figure 88, firms in large labor markets such as Istanbul, Ankara, and Izmir find the inadequacy of education in the workforce as a top constraint. These provinces have relatively diversified economies, indicating that this obstacle is encountered by firms across sectors. Firms in surrounding subregions (Manisa-Afyonkarahisar-Kutahya-Usak and Denizli-Aydin-Mugla) where wholesale and retail, and services in general, dominate the economy, this obstacle is also listed as a top constraint. Figure 88 Inadequately educated workforce, by region, 2015–2016 (percentage of firms reporting it as a top obstacle) Source: Authors’ elaborations based on WB Enterprise Survey 2015–2016 Page 79 of 114 4. RECOMMENDATIONS TO FACILITATE JOB CREATION IN TURKEY The Turkish labor market remains dynamic—creating new and quality jobs but faces increasing challenges to absorb large subsets of workers. As shown through this diagnostic report, it faces various challenges that will need to be addressed in the near term to facilitate new job creation, job retention and labor productivity increases. There is also a regional disparity regarding ease of job creation and economic opportunities more broadly. Fortunately, the government has recently undertaken important reforms, and is planning to undertake more to ease hiring, incentivize job creation, strengthen workforce skills and improve the business climate. This section of the report reflects on the analysis presented in previous sections and outlines a comprehensive set of proposed policies and actions. The actions proposed in this section aim to address key challenges identified to catalyse job creation, job retention, productive inclusion and labor productivity. Recommendations in this section respond to key challenges in the Turkish labor market, as identified through this diagnostic (Table 3). Seventeen distinct but interrelated challenges have been identified; they fall under five broad categories: Growth and productivity, formalization, job creation, labor force participation, and workforce skills/talent. Each challenge affects how the labor market functions and formal employment can come about. For instance, decreasing growth rates and declining productivity affect the ability firms to grow or new ones to be created and as a result it affects job retention and new job creation. Informality and the risk of further informality in the economy and the labor market limit fair competition among firms, and quality employment opportunities and wage growth for workers. Limited job creation and double-digit unemployment stems from structural constraints firms face; high employment costs and low access to resources are two examples that firms in Turkey often cite as factors hindering their growth. Low labor force participation of women and youth pose risks to the country’s ability to harness its most important resource, human capital. Lastly, inadequacy of workforce skills not only limit productivity and growth but also limit people’s own ability to embark in productive economic activity (e.g. formal employment, formal self-employment, entrepreneurship). Page 80 of 114 Primary Challenges (and Outcomes Resulting from Challenges) Declines in economic growth, and flat productivity growth Growth and High net job creation since 2009, but clear declines in net job creation since 2016. productivity Similarly, job quality rose since 2009 but began to decline since 2016 Regional divergence in labor market opportunities and performance Competitiveness gaps relative to high performing comparator countries in labor market, human capital and innovation Increasing informality and unfair competition disincentivize abidance with regulation High agglomeration of formal workers at minimum wage levels. Such clustering Formalization indicates concerns (e.g. wage under reporting, vulnerability to wage increases) Increased influx of refugee labor, most working informally, exacerbate informality in the labor market, especially in less developed regions Persistent long-term unemployment and underemployment Large segments of firms don’t grow, and they offer limited new job creation Job creation High tax and high labor costs disincentivize new job creation Limited access to finance limit firm growth and job creation Low labor force participation, especially women, youth, and recent university graduates. This is illustrated by the high shares of the population who are neither in Labor force education nor in employment (NEET) participation Difficulty for new labor market entrants (recent graduates) and women returning to the labor force face to access formal work. This leads to discouragement and exit from the labor market Inadequate skills in the workforce limit firm's ability to innovate and increase productivity Workforce Declining returns to higher education since 2016 skills and Gender parity in education levels not achieved; women's education levels remain talent lower than that of males Labor-technology substitution emphasizes why employers want workers with more cognitive (technical and analytical) and soft skills Source: Author’s summary The government’s own development objectives highlight the importance of the same categories. In the 10th National Development Plan published in 2013 and the National Employment Strategy published in 2014, the Government highlighted that it is important to focus on growth, productivity, formalization, employment, labor force participation and fostering talent to deepen Turkey’s development. In those documents, the government identified similar challenges shown throughout this diagnostic report and summarized in this chapter. In those documents the government also put forward a set of actions planned to tackle them. Unfortunately, despite time passing, many of the challenges persist and some new challenges have arisen, making labor market conditions ever more complex. A summary of the action plan that the government proposed in its’ 10th National Development Plan is provided here (Table 4). In the government’s plan, it is shown that to address growth and productivity constraints, it is important to boost industrialization and expand export-led activities. To reduce informality, the government proposed to strengthen the social security system. To promote Page 81 of 114 employment the actions proposed revolve around upskilling the workforce and improving labor conditions more broadly. To promote labor force participation actions were directed to incentivizing productive work and supporting migration management to ensure foreign workers contribute to the economy. Even though many of these actions are sensible, even in today’s context, they fall short from concretely outlining what needs to be done. They also do not address many of the structural challenges that prevent job creation and firm expansion. And of course, there are no actions that address new challenges or challenges that have emerged since those documents were drafted. Table 3 Employment and job creation, Turkey’s 10th National Development Plan Growth and Develop the export-oriented, private sector-led competitive production productivity structure via increasing productivity and accelerating industrialization. Reduce informality by equipping the social security system with a structure Formalization that is fair, of high quality and financially sustainable. Form a labor market in which decent job opportunities are provided to all segments of the society, skills of the labor force are upgraded and utilized Job creation effectively, gender equality and occupational health and safety conditions are ameliorated and flexicurity is embraced. Labor force Benefit from demographic window of opportunity created by the young participation, population at the maximum level, to increase people’s life quality conserving workforce skills and productive and dynamic population structure and to develop an effective talent migration management system which supports the economy. Knowing the challenges is an important step in addressing them. The main contribution of this diagnostic report is to review, validate and update the status of previous challenges, identify new ones and provide recommendations to tackle them. This part of the chapter focuses on what actions should be taken to address these challenges. It also identifies the priorities that the Government of Turkey can consider addressing its challenges and facilitate future improvements of employment opportunities and conditions. Many actions are likely to yield positive results when implemented simultaneously. In fact, there are actions that can be applied to more than one challenge. In one case an action is repeated in two sections because it is relevant to both, but in many cases, actions could be applied across sections but were not repeated. In all cases, there should be an emphasis on monitoring and evaluation to ensure the costs of programs merit the benefits and to measure impacts and make necessary adjustments to programs. Table 5 provides a summary of the activities recommended. The following section provides details on each of the activities. Page 82 of 114 Table 4 Summary of the actions proposed to address Turkey’s labor market challenges Actions to Address Labor Market Challenges 1 Strengthen the investment climate through policy predictability and transparency; lower administrative burden on businesses; an efficient tax system; access to longer-term finance; and agglomeration economies to accelerate longer-term productive investments including in R&D. 1.1 2 Strengthen the efficiency of fiscal support for businesses, including through better targeting of Growth and support to firms with high growth and/or large job creation potential, winding down fiscal productivity support that keep inefficient firms in business. 3 Support firms to adopt technology by reforming labor market policies that disincentivize them from doing so and increasing worker support measures to obtain complementary skills. 1 Reductions in excessive employment costs, elimination of employment barriers, and increase the capacity for monitoring compliance with (enforcement of) labor regulations. 2 Increase flexibility of types of formal employment so firms can adjust to changing markets while increasing worker protections through a reformed unemployment insurance system, social 1.2 assistance with a strong graduation emphasis, and re-employment support services. Formalization 3 Streamline new business setup procedures and exit procedures, including targeting support (financial and non-financial) to new firms entering the economy while reducing financial support to inefficient firms that rely on subsidies to stay in business. 4 Promote the positive perception and practice of formality in society and strengthen informality prevention measures. 1 Rationalize existing employment support programs by targeting them to support disadvantaged workers (youth, women, disabled, refugees), ensure their effectiveness in the long term by accompanying them with structural reforms that improve employability (e.g. education and skills, regulations that reduce youth and women's competitiveness). 2 Improve employment related services, including operational efficiency and using technology and data-based tools to improve service delivery. Continue to monitor and address occupations 1.3 and skills requirements and consider matching jobseekers to jobs based on skills not occupations. Job creation 3 Review and reform financial support programs to firms, to increase their impact on productivity and sustain jobs by targeting their support to economically sound firms with high growth potential and plans for longer-term productive related investments. Shift support from (old and stagnant) SMEs and toward younger start-ups and vibrant SMEs that bring innovation and create new jobs. 4 Strengthen system of incentives (monetary and non-monetary) to support fast growing firms and sectors and those that have high export potential. 1 Facilitate improvements in the provision of quality affordable childcare, especially among the most vulnerable women, for whom it should be strongly subsidized. Reform policies to allow flexible working times and improve parental benefits to prevent gender discrimination. 2 Rationalize employment support programs by targeting them to support disadvantaged workers 1.4 in firms that are productive and economically sound and can sustain employment once the subsidy Labor force is done. (Follows recommendation 1.3.1) Introduce accompanying non-financial measures to promote participation economic activity. 3 Improve employment related services to increase their effectiveness placing disadvantaged groups into sustainable jobs. Customize service delivery and strengthen referral system for expert services (beyond employment) to address distinct needs that disadvantaged groups may have. 1 Modernize vocational and lifelong learning systems, provide flexible adult learning options which offers modular courses, which lead to certifications and improvements in qualifications. 1.5 2 To improve the skills of existing workers and promote upskilling of the workforce, consider Workforce financial support to employers for allowing (lower-skill and at-risk of dismissal) staff to take time skills and off to train in recognition of the costs they will incur and to promote lifelong learning. talent 3 Facilitate the establishment of employer-led sector skills councils that focus on skills development that address the needs in the economy. Councils should ensure occupational standards, learning provision and qualifications meet sector standards. Page 83 of 114 1. Details of actions recommended 1.1. Growth and productivity 1.1.1 Drivers of growth. Strengthen macroeconomic stability and the investment climate through policy predictability, much needed reforms to positively affect productivity and employment growth. A recent World Bank report finds that macroeconomic imbalances in Turkey, which intensified economic stress starting in mid-2018, has led to contraction of investment and prolonged unemployment. Key actions that can improve stability and predictability are outlined in the recently released Turkey Economic Monitor. For instance, sustaining tight monetary policy, enacting financial sector policies that enable gradual deleveraging and enhance financial risk monitoring and management. Critical to deleveraging is a strong corporate debt restructuring framework. Apart from strengthening the macro side of the economy, it will also be important to strengthen the investment climate by improving policy predictability and transparency, lowering the administrative burden on businesses, improving the efficiency of the tax system, and improving access to longer-term finance for firms to invest. The recently launched New Economic Program (NEP) sets out measures to address immediate vulnerabilities and increase longer-term productivity. Thus, enactment of the actions listed in that plan will likely result in a strengthening of the macro and financial environments as well as the investment climate. Strengthen climate for foreign direct investment and domestic firms’ linkages to local and global value chains. By strengthening the linkages between domestic firms and foreign firms one could expect to also see improvements in technology transfer, product and process upgrading, management practices, and improvements in the supply and demand for more skilled and educated workers. On the other hand, low foreign direct investment and weak linkages to international markets and technology can hinder productivity improvements and limit firm capacities to produce and remain competitive. Strengthening of linkages with FDI is of importance in sectors that invest in R&D and longer-term productive investments because these are the sectors most likely to drive growth and create permanent formal jobs. There are sub-sectors in the economy that grew quickly and have potential to continue doing so; these are sub-sectors employ large numbers of workers, formally, and have large potential to grow (e.g. motor vehicles, textiles are two examples). There are other more sophisticated sectors that contribute immensely to growth and productivity, but are relatively small direct employers (e.g. pharmaceuticals, and chemicals); however, those sub-sectors can create positive spillovers on the economy (agglomeration economy) and can contribute to increasing the absorption capacity of higher skilled workers. 1.1.2 Streamline support measures. Strengthen the efficiency of fiscal support for businesses, including through better targeting of support to firms with high growth and/or large job creation potential. Targeted and finite countercyclical fiscal policies are important to help the economy tide over a predicted difficult period ahead. Such fiscal policies include incentive and support measures to help firms and workers stay economically active. But incentive and support measures pose fiscal expenditures, so selection of the right measures will be of utmost importance. There is a diversity of fiscal incentives provided to firms in Turkey. These incentives are often meant to boost demand and sustain supply and demand and Page 84 of 114 keep the economy functioning. There are incentives to innovate, incentives to employ workers, incentives to export, support measures to access credit, incentives to locate in certain areas, support to access government contracts, tax incentives and discounts, among others. Some of these measures are short-term support measures that help offset the impacts of difficult economic conditions, market or institutional failures of some sort. Fortunately, there are also support measures that focus on longer term productive investments that can help increase productivity and economic performance in the longer term. Incentivize formality in the economy to help broaden the tax base. This is done by offering support to firms operating informally to register and enter the formal economy. Amplifying the tax base by adding new formal firms can also help increase fiscal revenues. Many of these measures have positive effects and meet their objective. So even in periods where fiscal discipline is needed, support measures that help productive firms, high-tech and skill-intensive firms, and informal firms become formal should be continued. For instance, there are firms that require support at the start because they involve greater risk in terms of high start-up costs and long gestation periods. Given their potential for productive contribution to the economy and job creation, public subsidies and incentives can be targeted to these sub-sectors to overcome market failures that lower initial investments. Support can also prevent these firms from operating informally or engaging in informal activities (e.g. hiring, not registering their products). Wind down fiscal support that keep inefficient firms in business and prevent reallocation of resources to more productive firms. Fiscal discipline is essential to improve and sustain economic performance, maintain macroeconomic stability, and reduce vulnerabilities in the Turkish economy. Periods of expansionary policies have led to a concerted effort to rein in fiscal expenses. But to do that in a manner that supports economic growth, it is imperative to know which expenses need to be wound down and/or eliminated. For instance, support programs aimed to help firms pose immense fiscal costs. Some of these measures effectively support firms with high potential for growth and economic success in the long term while others do not. In fact, some of these short-term measures offset inefficiencies in eligible firms which prevent them from adjusting to changing market conditions. In the case of employment incentives, these short-term measures help offset labor market rigidities that would otherwise precipitate layoffs, especially among the least productive workers. There are also support measures such as government guarantees for credit which help firms (mostly micro, small and medium) access working capital during downturn cycles. Without such guarantees, banks would have been unlikely to extend credit to these firms due to financial market frictions. Such limited access to credit would have likely prompted many businesses to close; unfortunately, many of those businesses are unproductive and remain reliant on subsidies and guarantees to subsist. Thus, before cutting programs, it is extremely important for the government to assess the effectiveness and efficiency of all its firm incentives to maintain only those that can support economically sound firms with high-growth potential and with a clear potential for catalyzing economic activity (multiplier effects). 1.1.3 Support firms to adopt technology by reforming labor market policies that disincentivize them from doing so and increasing worker support measures. Address constraints for firm growth and technological adoption. Labor-saving technologies have played a role across countries in advancements in industry, and economies. As new labor-saving technologies emerge employers consider the cost-benefit of acquiring new technology vis a vis hiring or retaining workers. This can lead to decreases in new employment of workers with technologically Page 85 of 114 substitutable skills and increase firm-level labor productivity. This is arguably a good outcome in the long term given that higher labor productivity is good for the economy. However, many firms with large numbers of potentially substitutable workers are unable to acquire such technology because dismissal costs are too high. And small firms, which are the majority in the Turkish labor market, have difficulty accessing new technologies because of high upfront costs which make it difficult for them to adopt new technologies or compete for skilled workers to work the equipment. Moreover, many smaller formal firms find it cost prohibitive to dismiss existing workers and make the necessary changes to their production process. The main lessons emerging from sectoral studies that measure the short- and longer-term impacts of technology adoption on employment find that in the short-term impacts on employment can be negative because workers are dismissed. However, in the longer term, as productivity improvements help firms be more competitive new more sustainable jobs emerge. Having adequate support measures for workers dismissed due to technological upgrades is necessary to incentivize technological uptake, catalyze productivity and ease labor market adjustments. The ultimate impact of technological adoption on workers depends on having a support system that helps workers retrain and upskill, provides them temporary income support, and helps them reenter employment. A recent study finds that there is an optimal amount of labor protection and labor flexibility that helps firms to make optimal technology-labor investment decisions. Policies should facilitate worker transition from one job to another—through income support and retraining— rather than trapping them, through costly dismissal polices, in jobs where their productivity is low, and their career outlook in the medium and longer term is dire. A key lesson from various studies is that employers will continuously look to acquire technological options to substitute relatively costly workers, especially when labor costs are high (and increasing) and technological options to replace workers are available. Thus, instead of keeping labor costs high as an artificial means to prevent worker dismissal it is best to adapt labor costs for the labor market to be competitive, facilitate technology adoption for firms to remain competitive, and strengthen worker support measures. Doing so will help Turkey to strike the optimal balance between worker protection and labor flexibility and will not prevent firms from adopting productive technologies that will benefit them in the longer term. (See subsequent recommendations in 1.2.1 and 1.2.2 on reforms on labor market policies and recommendations in 1.4 and 1.5 discuss support measures for workers). 1.2. Formalization 1.2.1 Address labor costs and barriers to formal employment. Labor regulation needs to be updated continuously to ensure policies remain in line with the reality of the labor market. Job creation, especially in medium- and high-value-added sectors, can be re-energized by improvements in the regulatory environment, including reductions to the costs of employment and doing business. One immediate action is to undertake a full review of the labor regulation code to identify overlaps in regulation, excessive costs and constraints to formal job creation and retention placed on employers. There have been several regulations and adopted and reforms made in the recent years which improved the regulatory environment in creating jobs. For instance, regulations concerning temporary work and the legal use of temporary agencies have been adopted; the turquoise card, which allows foreign skilled workers to work in Turkey legally, was introduced. However, there is still room for improvement particularly to tackle rigid and expensive labor laws; namely old-style severance payment regime and a minimum wage system that does not systematically consider the socio-economic realities of Turkey. Page 86 of 114 Reduce excessive employment costs and eliminate employment barriers by reforming the severance pay system. Various studies, including government studies, find that the severance pay policy is too costly and preventing natural churning and worker protection. In the current severance pay system, employers pay for severance payments if the worker is dismissed and is eligible to receive benefits. Dismissed eligible workers are entitled to receive pay equal to the last 30 days’ gross wages for each full year of employment (or prorated for a portion of a year). The structure of the system puts a large economic burden on employers, limits portability of benefits and leads to negative incentives all around. A reformed system should look to decrease costs to employers, increases coverage for workers, and reduce non-compliance. To do this, consider a model that creates individual severance accounts which are funded by the employer and the government and made available to each worker upon separation from a job, for any reason. In such model funds are managed by private pension companies, which allow workers better access to information related to their own account and increases trust among workers and employers that the resources will be available when they are needed. On minimum wage, as the economy continues to face difficulty creating formal jobs, it is imperative to align minimum wage levels with socio-economic conditions. The current minimum wage level in Turkey is not aligned with the socio-economic conditions; such a misalignment can have negative consequences on the most disadvantaged segments of the workforce. One way of better aligning minimum wage increases to the real socio-economic picture of the country is by designing and using a quantitative mechanism and technical tool that considers multiple socioeconomic factors to revise the level. Such a mechanism can consider different criteria, such as living costs, the development of workers’ wages, employment and informality, increases in labor productivity, among other factors. Results can be more in line with the context, and the level derived by the mechanism can serve as a reference in the tri-partite consultation process. In other words, the technical tool can provide a basis or starting point for the process of deliberation among the parties, and the basis would reflect the socio-economic reality of the context. Consider including a group of independent experts and academics to participate in the minimum wage consultation process and to monitor the effects of the new level of the economy and workers. Doing so can increase transparency and objectivity in the level deliberation process, and representation of workers not represented by traditional stakeholders. Increase the capacity for monitoring compliance of labor regulation. Consider improving the capacity of labor inspectors to monitoring compliance with labor law and support (not punish) firms to remain compliant. To increase fair competition in the business sector and labor markets, strengthen formalization efforts by enhancing labor tax compliance and improving auditing functions, especially risk-based audits for labor regulation compliance. About this, strengthen the capacity of tax and social security audits, share databases among agencies involved in overseeing employment and industry, move toward a risk-based labor inspection systems and audits. Regarding risk-based audits, introduce a national wage index (NWI) which provides appropriate wage level for employees that are in accordance to their qualifications, skills, and productivity. Labor inspectors can employ NWI to undertake risk-based audits based on wages. Labor inspectors should also be equipped to support firms to stay compliant or become compliant; for instance, once they identify a firm at risk of non- compliance labor inspectors can support them through referrals to various services to remain formal. Page 87 of 114 1.2.2 Increase flexibility of contracts while increasing worker protection. Flexibility of contracting workers and security to protect workers must go hand in hand. Flexicurity means shifting from the protection of jobs to protection of workers, by improving unemployment insurance, self-insurance, enhancing the capacity of employment related services to more effectively place people back into work, and improving the effectiveness of the social assistance system to provide better support while encouraging beneficiaries to seek work. Evidence shows that facilitating the use of diverse forms of formal work such as part-time work and temporary formal work can help create more jobs, particularly for disadvantaged populations. However, these types of contracts can also increase layoffs during downturns which increases uncertainty for workers. Therefore, flexible employment systems should be accompanied with (at least partial) guarantees of income protection for workers between jobs. Thus, providing income support during periods of transition from one job to another, ensuring portability of social benefits to support internal migration of workers, introducing safety nets for those that cannot transition easily, and improving access and quality of reemployment services (including training) to help facilitate transition are essential ingredients to any flexicurity system. Continue to implement recently adopted reforms aiming to diversify the types of labor engagement contracts in the formal sector. These reforms are meant to promote job creation in formal work by easing entry and re-entry into work. Part-time work has become easier thanks to recent regulations (2016) that help mobilize people who are difficult to place in full-time employment. Diverse forms of employment such as renewable fixed term contract, outsourcing, working from home, part-time working have been shown to facilitate job creation in the formal sector, particularly for women and youth. Recent reforms on temporary work agencies and part-time work were made in recognition of the need to increase flexibility; but their effectiveness in facilitating entry to formal employment and retention should be assessed to minimize imbalance between flexibility and worker security. Reforms in the diversification of work contracts should be accompanied with reforms in the unemployment insurance system and social assistance system to ensure temporary coverage during unemployment spells and incentivize returning to work in the short-term. A recent reform made to the social assistance system was a good start; in that reform, beneficiaries are immediately registered with ISKUR to receive employment services and opportunities. However, implementation has been very slow. Thus, strengthening the implementation will be critical to helping people transition to work, and preventing social assistance costs from rising. Reform the unemployment insurance system by broadening access to income support during unemployment periods. Reforms to labor contracts and the links between social assistance and employment have already been made (though they both need further implementation support). And a previous section discussed the need to reform the current severance payment system with ‘portable’ severance saving accounts that helps workers move easily across jobs. But no flexible labor system can be effective without ensuring that the unemployment insurance system offers adequate coverage of the workforce. In Turkey, the unemployment insurance system is too narrow; it offers coverage to few workers because the eligibility criteria are too strict. Thus, consider changing the criteria to be more inclusive, and link this reform to severance pay reform so that the costs imposed on employers for both policies are lower than the currently are but the benefits to workers are not only higher but also more accessible than with the current system. Also, improve the social safety net system that helps displaced workers have access to income support. With respect to social assistance, strengthen the social assistance system to ensure coverage of displaced workers and adequate benefit levels and ensure that there is a link between social assistance and employment and that benefits do not prevent Page 88 of 114 work disincentives. See World Bank’s social assistance review (2019) for more details on the recommendations to improve the social assistance system in Turkey. See next section for how to improve the effectiveness of employment related services to help displaced workers. 1.2.3 Streamline new business entry procedures and exit. Improvements in the regulatory environment that lower the cost of starting a business and facilitate access to services (and finance) to grow and create jobs. While providing financial support to new businesses can help incentivize innovation and entrepreneurship, it is also important to reduce financial support to existing firms that operate inefficiently (many of them staying small) and rely on subsidies to stay in business. There are many lessons learned from past efforts to promote formality in Turkey. The most effective measures should be enhanced. For instance, Turkey has embarked in simplifying new formal firm registration procedures, which makes setting up a new business more possible for a broader set of people. It takes less than a week to set up a new business in Turkey; fast compared to other peer countries. Ensure that there are continual reductions of administrative steps, and simplification of financial and legal obligations of enterprises, especially those faced by micro and SMEs, and those looking to restructure. Support firms that are not performing to exit the market in a planned and managed manner. Bankruptcy and insolvency frameworks need to be in place. There should be an effective bankruptcy settlement law to ensure that entrepreneurs and investors are able to close a failing entrepreneurial experience and move on to new challenges instead of holding on by relying on subsidies. Unfortunately, Turkey, compared to other countries, fares badly on the time and cost required for firms to resolve insolvency. As recently reported in the Turkey Country Economic Memorandum (2019), it takes a creditor up to five years to recover only 15 cents on the dollar from an insolvent firm, compared to a process that requires less than two years, with a recovery rate of over 70 percent for creditors in high-income OECD economies.17 The process is also expensive in Turkey, where it costs the equivalent 14.5 percent of the creditor’s estate to recovery owed assets, including a tax of 4.6 percent and legal fees of nearly 5 percent. To address these challenges, a new ‘concordat’ procedure was recently introduced, which enables authorities to set timelines for the procedure, and puts a heavy focus on business continuation rather than its liquidation through new financing, confirmation of contracts and sale of essential assets in bankruptcy (Law No. 7101 Amending to Code of Enforcement and Bankruptcy, 2018). 1.2.4 Promote the positive perception and practice of formality in society and strengthen informality prevention measures. Promote changes in cultural norms and attitudes towards informality through awareness raising campaigns. Increase public information and education to raise awareness of the benefits of formality among firms and worker by raising awareness on labor rights and tying access to public services with formality. Evidence shows that strategies that are educational and persuasive, while also being participatory and transparent can be effective. For instance, consultative councils where employers, social partners and government (namely labor inspectorate) work together to find avenues for reducing non-compliance through education of workers, business and society at large. 17 Based on the 2018 Doing Business Report Page 89 of 114 Enhance formalization efforts by strengthening prevention initiatives and enforcement measures. Setting up or enhancing mechanisms for joint monitoring between various governmental authorities—such as social security, tax authorities, labor and health inspectorate—using interlinked databases to preempt non-compliance, undertake risk-based audits, as well as embark in outreach efforts targeting workers and SMEs are shown to be effective. Turkey has a coordinated online system that helps target firms vulnerable to informality; the system can serve as a platform for identifying vulnerable sectors and workers to support them to remain compliant and to educate the workers on the risks and negative consequences of accepting informal forms of payment. There are various online tools that can facilitate inspection as well; for instance, e-inspection in Albania or hotlines and online portals, as used in many countries, to report non-compliance can also play a role. 1.3. Job creation 1.3.1 Rationalize existing employment—wage and hiring—support programs (see Section 1.1.2 for recommendation details on the need to streamline incentive and support programs) Employment support programs can act as a significant countercyclical policy tool to stabilize the labor market. Employment support programs can also support economic recovery in economic downturns. Unemployed and underemployed people as well as people at risk of unemployment need distinct forms of support to become and/or remain engaged in the formal economy. Long-term unemployment (and even underemployment) can lead to discouragement, skill attrition and have detrimental effects on future employment prospects of people. This is especially true in and after periods of economic crises, when employment growth is slow, and the economy cannot absorb high levels of workers. Addressing this problem requires the use of effective, well designed, active labor market policies. Among the most salient policies are wage and hiring subsidies. These support measures can provide temporary incentives for firms to hire unemployed workers. Evidence from OECD countries shows that private-sector hiring subsidies can generally be more effective than public education, training measures, or public works programs. In fact, employment support programs were successfully used in Turkey to counter the negative effects of the crisis in 2009. They acted as countercyclical policy tools that helped to stabilize the labor market and support economic recovery. But many of these policies were kept in place, after the crisis was over, and new ones were added instead of tightening them after the economy recovered. The success of employment support programs is highly dependent on the design of these policies. Begin by strengthening the targeting of existing employment support programs by focusing them on new job creation. Apart from better targeting, review whether these programs are cost- effective means of reducing unemployment during both periods of economic stability and recovery. A sensible design focuses the targeting to balance direct impacts and indirect effects, encourages labor market persistence, and enables a more equitable distribution of employment. A critical aspect in the design is to ensure that while targeting unemployed workers, the policy does not lead to a displacement or crowding out of regularly employed workers. Targeting that is too loose may create high potential deadweight and displacement costs. Targeting that is too tight may create exclusion errors, can be administratively expensive to monitor, and result in low employer take-up rates. Targeting the long- term unemployed redistributes employment incentives to the disadvantaged and strengthens their labor market attachment. Programs should not only be well targeted to support new hires among disadvantaged workers (hard to place skilled youth, disabled, refugees and subsets of low-skilled Page 90 of 114 women), but they should also be provided to firms that are productive and economically sound, so they can sustain the workers once the subsidy is done. Hiring subsidies can be more effective than full wage subsidies. Both hiring and wage subsidies provide financial incentives to employers by temporarily reducing their labor costs. The magnitude of support and eventual dependence on the support differs between the two subsidy types. Compared with wage subsidies, which are usually targeted at specific groups of workers (e.g. youth, women, refugees), irrespectively whether they are new hires or already employed workers, hiring subsidies are usually offered to employers for new hires and for a specific period. They tend to be more cost- effective as they can have lower deadweight costs than wage subsidies. ISKUR currently provides many programs to help workers access jobs through on-the-job training support (wage subsidies) and other subsidized forms of employment, including hiring subsidies (covering social security premiums and other non-wage costs for new workers). Wage and non-wage support programs can both be effective when they are well targeted. Both typs of subsidies, can enable employers to take a risk on unemployed workers by giving them an opportunity to gain or regain work experience and strengthen their future labor market options. But to make them sustainable, it is best to pay them for a limited period and cover new hires, so that fewer workers are covered. Of course, there may be a need to also have programs that target soon to be unemployed workers as well as programs that prevent detachment in down cycles; those programs need to be well targeted and discontinued as the economy recovers to prevent subsidizing labor costs in unproductive firms and trapping workers in those firms. Increase effective use of public resources on employment support programs by continuously evaluating their effectiveness. Accompany employment support programs with structural changes that prevent workers from being employable. A good design that includes effective targeting can address deadweight issues and substitution effects; both which lower the cost- effectiveness of subsidies. But even a good design needs to be continuously evaluated to ensure their benefit outweighs their cost. Evaluation can indicate that implementation is consistent with the design, ensures the programs continues to yield the results needed and the programs remain relevant to the context. Also, even though hiring subsidies can serve as automatic countercyclical stabilizers that provide direct incentives for job creation it is also important to accompany them with measures that address constraints that prevent workers from being competitive and from entering the labor market. For instance, regulatory constraints (see 1.2.1 and 1.2.2), education and skills constraints (see 1.5.1), household, cultural or personal constraints (see 1.4.1), information and employment service constraints (see 1.3.2), among others. 1.3.2 Improve employment related services. Improve the effectiveness and efficiency of labor market intermediation mechanisms to maximize the quality of the match between firms and workers and make best use of employment support resources. The public employment services agency (ISKUR) is the main public entity charged with placing workers into employment in Turkey. ISKUR services range from basic guidance to worker placement. Given its centrality in job placement, and resources spent to support job seekers, it is important to consistently improve the effectiveness of how ISKUR delivers services. Increasing the efficiency and effectiveness of ISKUR is most salient in a downturn economy when services and resources required are likely to exceed existing capacities for quality service delivery. One way to improve service delivery is to introduce tools that aid job counselors in their work; some Page 91 of 114 of these tools are listed below. Improvements should be made from outreach to job placement and including adding a strong emphasis on monitoring and evaluation. Consider establishing a profiling system that uses data-driven information to support decisions and facilitates a path to formal employment. The introduction of this tool in ISKUR is currently being considered under an ongoing project (funded by the EU and World Bank). It is important to fast-track its adoption so that it can help job vocational counsellors (JVCs) to place an increasing number of job seekers into jobs and training programs. This tool will help JVCs to differentiate the types of job seekers to be able to customize the services offered. Job seekers can be classified from easy to place and in need of low intensity services to, less employable and in need of high intensity services. Such a system can help ISKUR to support its JVCs with more evidence base (objective) information to make decisions and limit the use of fiscally expensive services (such as wage subsidies) to those job seekers that can benefit from them most effectively. Introduce a labor market information system (or a workforce dashboard) to provide guidance to job seekers on the jobs, skills and occupations in demand. Evidence shows that the problem of mismatches of skills is exacerbated by inadequate information both in the search of employment and in the recruitment of workers. To reduce mismatches in the labor market, facilitate access to labor market information through the creation of a Labor Market Information System (LMIS). The LMIS should provide real-time labor data, including jobs and skills in demand, wage information, career paths, relevant trainings available, and region-specific labor market demand data. It should also contain up-to-date information on vacancies, as well as current information on the education and competencies required for specific occupations. Such an online information system can provide present and future labor market prospects and other relevant dimensions to help in the job-search effort and making decisions on career paths. Guide job-seekers and prospective job-seekers (in the education system) to use the LMIS to make training and career related decisions. In addition to the creation of an LMIS, it is important to undertake various information dissemination campaigns to increase the utilization of such an information platform to make education and career decisions. Such campaigns can be distinctly targeted to potential users, counselors (job vocational counselors as well as school-university counselors), job seekers, parents, industry, educators and others. Continuously monitor occupations and skills in demand so that people can be trained and retrained according to current market needs. The Turkish labor market, much like many dynamic economies, is constantly changing. This report has shown that the labor force is distributed across sectors and occupations and how that distribution is constantly evolving. To ensure the education and training sector can respond to labor market needs, it is important for Turkey to invest in establishing a system that continuously identifies and monitors the occupations and skills in demand from employers. The World Bank, under an ongoing project with the Ministry of Family Labor and Social Service and ISKUR, recently developed a methodology that draws uses various data bases (administrative and survey firm and worker level data), live online data (from recruitment portals) and employer qualitative assessments to identify the occupations and skills in demand. Such an effort yielded a set of lists of occupations and skills in demand, by distinct geographic regions, and determined an estimated labor market absorption. This labor market assessment analysis can also be used for the education system and training providers to reform their curricula and ensure their services deliver the skills required by the labor market. Such a tool should be adopted by relevant agencies of government and institutionalized for continuous monitoring. Page 92 of 114 1.3.3 Review and reform financial support programs to firms to support firms most likely to grow and be productive. Consider other non-financial measures of support. There are arguably too many incentive programs offered by the government, imposing a heavy fiscal burden, and only some are effective. There are currently several special loan programs, interest rate subsidies and credits, and grants of many kinds offered by the Turkish banking system, with support from the government. Many government subsidized programs are targeted to firms under some criteria. They are implemented through various Development Banks (subsidized lending), and institutions such as KOSGEB and TUBITAK (direct grants). Apart from subsidized loans and grants, there are other types of public incentives, such as tax breaks, preferential credit, and public procurement advantages. These types of support to firms are often meant to be for investments in technology, R&D, and innovation; but also, to support employment creation/retention. In all cases, these programs should support firms that are economically sound, with high growth potential and explicit plans to use resources to make longer-term productive related investments. However, many of the programs do not always target appropriately or impose their criteria for selection loosely. In fact, these programs sometimes help to sustain otherwise failing firms to stay afloat. Going forward, before enlarging these programs or continuing to finance them, they should be rigorously evaluated to assess their effectiveness; results from such evaluations should help to make relevant reforms to them to increase their impact on innovation, productivity and catalyze their ability to (indirectly) create sustainable jobs. In cases where programs are ineffective or duplicative, they should be closed, and resources reallocated to programs that deliver measurable results. Incentive programs targeted to SMEs should support newer firms rather than support older and stagnant SMEs. A previous World Bank study (2011) finds that the two priorities for SMEs in Turkey are access to finance and the ability to adopt and use technology and to innovate. The recent Investment Climate Assessment 2015 identifies access to finance as one of the most important areas for SME job creation and competitiveness (2017). As such, despite the many finance related programs, addressing financial ‘access’ constraints, especially by SMEs, remains a priority. However, to prevent financial resources (from government supported programs) from distorting market and leading to allocative inefficiencies, these programs should shift away from supporting (older and stagnant) SMEs that rely on these resources to operate, and toward younger firms (such as start-ups) that bring innovation and create new jobs or existing vibrant SMEs with a credible restructuring plan. These resources should be targeted to help overcome market failures for productivity enhancing investments and to facilitate long-term productivity enhancing investments (including investments in retraining the workforce and purchasing productive assets). By refocusing efforts to make credit available to growing firms rather than subsistence firms, the government can expect these firms to be the motors of growth that improve the economy sustainably. Apart from financial support to eligible firms, it is also important to stimulate the development of firms through other (non-financial and less costly) support instruments. There are other ways in which SMEs can be supported which do not promote reliance on subsidized financing. For instance, credit guarantee schemes are one option. But also connecting SMEs to regional and global value chains or connecting them to commercial support services and export channels to increase their client base. Many SMEs face limited access to important information and training; thus, another way of support is by providing information and advisory support for better management. Two examples are the UK’s Growth Voucher Scheme and Growth Accelerator Schemes. Helping SMEs obtain external financing by easing conditions for private investors to invest in firms can also help SMEs access capital. To promote smart investments, tax breaks for R&D and Page 93 of 114 growth-oriented costs can also be supported. This section (and report) provides only a partial list of effective support measures that can be considered. 1.3.4 Strengthen system of incentives (monetary and non-monetary) to support sectors that drive growth and have high export potential. Apart from financing, support can be provided to promote entering global value chains and economic (sectoral) clusters. Various firms in the Turkish economy invest in R&D and longer- term productive investments. These firms have the potential to make productive investments and increase their productive capacity and create sustainable jobs. Apart from helping these firms access resources to make such investments for continued growth, it is also important to help them become part of global value chains and participate in sectoral clusters. For instance, programs (not necessarily financial) can target co-located firms (usually SMEs) in the same sub-sector to provide them with services that help them improve their productive capacity, be better linked to larger firms and broader markets, and take part in regional and global value chains. Many European countries such as Spain, Italy and Germany have good cluster programs. In the immediate term, enhance support toward firms to use more technology in the production process, move-up the value production chain, and promote partnership between SMEs and organizations overseas to develop or transfer technology or processes. Though technological adoption and changes in the production process of these firms may decrease demand for labor in the short term, it is likely that the development of these firms and clusters will lead to increases in more sustainable job creation in the longer term. Promoting innovation, which leads to new products and markets, in the economy can help create new employment opportunities that are sustainable and resilient to deteriorating economic conditions. Support for innovation need not be monetary only. In Turkey, apart from financial support measures, there are also programs that encourage links between enterprises and research universities in activities such as training, skills exchange programs, and joint ventures. But to intensify private R&D (rather than publicly supported R&D) and encourage research to be developed in private sector economic activities, government may need to introduce new (or revise existing) support measures. To start, a thorough mapping of the current policy mix, followed by an assessment of the overlaps, efficiency, and effectiveness of existing programs should be considered. Thereafter, adjustments to the mix of programs can be made to ensure a strong focus on incentivizing excellence in publicly funded research, facilitating technology transfer and commercialization of that research to the private sector, and boosting R&D undertaken by firms. This process could result in a more strategic, targeted approach to R&D policy that supports more frontier innovation. 1.4. Labor force participation 1.4.1 Address critical barriers that prevent (many) women from working. Policy efforts to attract (and retain) women into economic activity must encompass several fronts—from childcare to barriers to achieve gender equality in the labor market. Efforts to attract women back into work should start by improving the provision of quality and affordable childcare. Support for childcare is shown to ease women’s access to the labor market; especially low- income women with marketable skills for whom it should be strongly subsidized. Countries with more conservative views on men and women’s roles in society tend to spend less on early childhood education and child care and are less likely to accommodate flexible working arrangements. Such underinvestment limits women’s ability to return to work and/or work full-time. Thus, better Page 94 of 114 childcare options constitute a fundamental policy to support parents, mainly women in Turkey, combine work and the care of young children. In low-income households, access to free or subsidized public childcare increases women’s labor force participation. But the availability of child-care services needs to increase to ensure broader coverage. Thus, one important action is to increase the availability and access to adequate child care services by increasing the offering of crèches. Two aspects of childcare are fundamental to successful programs: quality and convenience. Building up the quality and availability of kindergartens to meet the demand for preschool enrollments is therefore an important policy priority and one that can not only expand women’s opportunities but simultaneously address equality and efficiency challenges. In this respect, implement previously agreed upon measures to compel municipalities to open nurseries that can be accessed by local women and provide individual firms incentives to offer child care options on-site. For instance, larger firms may access discounted loans to help them set up the creches. Similarly, government may consider a voucher system targeted to low-income parents to take their children to childcare. Reform policies to allow flexible working times for women and men to be able to balance work with competing responsibilities. Flexible work includes part time work, or if full time, it includes reduced hours, selection of shifts or the ability to work at home during the working week. Flexible work is not a women’s only issue; in fact, emerging studies from developed countries show that both men and women (especially younger workers) prefer flexible arrangements to balance competing responsibilities. But even though flexible work is not a women’s only the lack of flexibility is still a constraint to work. Continuing to support flexible work can help increase women’s labor force participation. Flexible work time can help women balance family and work. It can help workers who want to balance career and family responsibilities, this is especially appealing to Turkish women who generally shoulder the greater part of caring responsibilities. Improving parental benefits can minimize potential discriminatory practices against women and positively affect social norms. Implementing parental leave benefits rather than maternal leave alone, which hold women back from career progression, can help women be more competitive in the labor market. But when companies encourage parental leave, where men can also take leave, the negative impact on women’s career advancement is significantly reduced. Evidence shows that by changing from maternal leave to parental leave rights and offering family-related subsidies rather than mother only support, changes in a country’s social norms and attitudes towards gender roles in the home and the market can come about. Improve the skills of low-skilled women (and other disadvantaged groups) to be better equipped to work and be less vulnerable to unemployment. Job and business training programs can fill the skill gap for people who are out of the education system. Evidence shows that training programs can be effective in boosting female employment and earnings. A lack of skills is one of the main obstacles to low-skilled women’s participation in the labor market in Turkey. Training programs for wage employment, as currently offered through various channels (e.g. ISKUR, MoNE), should be combined with on the job training (employment support) to ensure skills are applied and further developed. But the success of training program depends on how well training institutions match the skills transmitted in the classroom to the needs of firms. Programs offering business training for self- employment can also be effective means of providing skills and economic opportunities to women, especially when coupled with financial support (e.g. cash grants) to start a business; however, there is evidence that only a subset of the businesses supported through entrepreneurship training and cash grants increase profitability in the longer term and grow beyond subsistence businesses. Therefore, it Page 95 of 114 is important to selectively support these businesses though financial support (selecting those that are economically viable and have growth potential) and accompanying them with advisory services. Improving the working environment by identifying and eliminating potential discriminatory practices against women. One of the most important factors in job quality is the working environment. Identifying work-related factors associated with increased risk of harm (e.g. physical or mental harassment, discrimination, mobbing) in the workplace and eliminating them can help to attract women (and other groups) vulnerable to harm. This is especially critical in sectors and occupations where women are a small minority and where tasks are traditionally male-oriented. Negative working conditions are prominent across sectors and occupations. In fact, a study in 2018 by the US national academy of sciences found that there are high levels of harassment of women faculty and staff in academia in science, engineering and medicine, with women in medicine reporting a higher incidence of gender harassment compared to academics in engineering and science. Studies in Turkey find similar results for women, not only in occupations in industry and services,18 but also in high-skilled professions in academia and medicine. Women in Turkey find it hard to make it to top positions in academia, unless in disciplines that are traditionally considered feminine, such as humanities.19 Gender expectations thus play a role when women select a specialization, not only in academia but also in medicine.20 Such difficult working environments in high skilled professions showcase the importance of recognizing the problem and addressing it through various means, including regulatory and cultural changes. Not doing so not only discourages women (or targeted workers) from working but it also costs firms substantial resources (e.g. high turnover, low productivity, absenteeism). 1.4.2 Rationalize employment support programs and introduce accompanying non- financial measures to promote economic activity. (see Section 1.3.1 for employment support programs and see Section 1.3.2 for recommendation details on broader improvements to employment related services and) Reinforce the involvement of disadvantaged groups in entrepreneurial activities by reinforcing the teaching of entrepreneurship and management in the education system. Productive entrepreneurship should become an integral component in the educational system by embedding courses that foster creativity and innovation and the positive perception of entrepreneurship earlier on, and throughout the educational cycle of a person. Strengthen efforts to support entrepreneurship in the higher education system. Do this by training and developing the capacities of instructors to deliver these concepts, linking students with private sector and business networks, creating institutions that support entrepreneurship through experimentation (incubators), and facilitating access to start-up capital and innovative ways of financing can be helpful in instilling and promoting productive self-employment and entrepreneurship. Increase the availability of business-oriented education and financial inclusion courses in the formal and non-formal education and training system to reinforce productive entrepreneurship as a viable means for income earning 18 see, e.g., Celik and Akar Sahingoz (2018) 19 Ince Yenilmez (2015) 20 Koyun et al. (2013) Page 96 of 114 opportunities. Consider promoting productive entrepreneurship of youth and women through income tax exemptions for the first three years, access to grants and loan guarantees. Continue to develop apprenticeship programs so that young and new job seekers can learn on the job and have career experience early on. Turkey is currently implementing an apprenticeship program for students in higher years of formal schooling and technical vocational education. Such programs need to be assessed to ensure the costs merit the benefits in the longer term. For instance, youth-guarantee programs in various EU countries provide good practices that can serve as a starting point for Turkey. For instance, those programs have been reformed to be youth-centred or new-worker-centred and to emphasize technical and vocational skilling, and enhancement of socio-emotional skills. Turkey should continue with its own model, given its economic and social reality is different than that of European countries, however, it should candidly evaluate the programs and continue to adapt them to account for lessons learned through the evaluations. Target youth-at-risk of vulnerability in the long term through adaptive learning. To prevent having large cohorts of young people unskilled, consider reforming second chance education programs (such as distance learning) that target youth at risk of low educational attainment (namely drop outs) by improving its appeal/attraction (through various incentives), the quality of the content taught, broadening the pathways of participants, and enhancing its school-to-work focus. In the case of refugee youth, add the relevant services (e.g. language support, psycho-social emotional support, specialized outreach) to entice them to join the skills training and/or education system, and/or stay in the education and training system. 1.4.3 Improve employment related services. (see Section 1.3.2 for recommendation details on broader improvements to employment related services) Empower employment service agencies (private and public) to customize their support to provide services to disadvantaged (e.g. women, youth, disabled and refugee) job seekers and strengthen referral system, to services beyond employment. Customization should focus from the outreach campaigns, to attract disadvantaged groups using means that relate to them all the way through the way services and information are provided, and monetary incentives are targeted. Employment services should strengthen its own provision of complementary services (when the service is within the mandate and competency of the public employment service provider) or strengthen its referral service capacities to ensure job seekers are delegated to expert services offered by other institutions to access complementary services that hinder them from entering or re-entering the labor market. For instance, many jobseekers need targeted interventions such as psycho-socio emotional support (mental health), life-skills, transport assistance, translation, extra guidance, etc. Some of these services fall within the mandate of ISKUR and should be provided in the centers or affiliated locations while others require more in-depth or specialized support from other institutions (e.g. ministry of health, family services, education, non-government institutions). In all cases, employment services will be more effectively delivered if the capacity of job vocational counselors is strengthened (e.g. more trainings, additional support staff, better materials and equipment). Page 97 of 114 1.5. Workforce skills and talent 1.5.1 Modernize the workforce development system. Despite great improvements in the level of education of the Turkish workforce, a large segment of the workforce does not have the job-related or technical skills necessary to meet Turkey’s aspirational goals. Many Turkish workers are trapped in occupations that entail using routine cognitive skills rather than non-routine cognitive analytical or interpersonal skills, which are typical of a more mature economy. TVET, the lifelong learning system and employment training through short-term courses are all part of the workforce development system21. Critical improvements are needed in the quality of the skills of the workforce to positively affect their productivity levels. An adequate, modernized, workforce development system is responsive to the dynamism of the economy and fast changing labor demands. Strengthen the workforce development system by improving its alignment with industry. Enterprises and industry requirements should be at the center of any workforce development system to increase the relevance of skills imparted on adults. Evaluations show that successful workforce development systems coordinate content delivery (learning) with vocational guidance and encourage close cooperation between enterprises and training bodies. By putting enterprises at the center and combining learning with guidance, the workforce development system not only increases employability by increasing the match between the supply of skills and demands from employers but also rationalizes financial resources spent on adult training and placement. Action 1.5.3 provides further details on how industry can be brought in closer into the workforce development system. Create flexible adult learning options, with pathways to skill up and by offering a system that is modular, with portable and stackable modules that offer credentials that respond to labor market needs. Mechanisms should be created to facilitate worker access to further training via modular training. Many countries recognize that for employed adults the opportunity cost of learning is very high, so courses need to be delivered in a way that does not take workers away from work for long or at all. Thus, to increase the use of lifelong learning among the workforce, redesign the system to offer modular courses, delivered in small dosages, which lead to certifications and improvements in qualifications. Modular training must make it possible to construct a learning pathway leading to a properly recognized and certified vocation/profession. Portability of credentials is critical; but to ensure they are recognized across the system each module must be verified and accredited under a unified qualification standard. Stackability can also be an effective model to meet the demands of students in need of basic skills and those seeking more advanced skills and lifelong learning. Stackable credentials allow students to earn shorter term credentials with clear labor market value that builds on into more advanced credentials that easily translate into higher wages. These stackable degrees or certificates ease the entrance into the workforce, and function as a “gateway” for attainment of more advanced skills. Delivery of lifelong learning should also be flexible to fit distinct learners and learning needs. One way to widen access and improve the provision of lifelong learning is by making better use of technology in the delivery of content. Various OECD countries such as Portugal, Canada, US, UK provide distinct examples. 21Primary education is not included in the definition of workforce development system and not the focus of this report; however, it is important to note that investments in skills in early years can have higher returns on investments than in later years. Page 98 of 114 Integrate a system of recognition of skills and experience acquired through non-formal education. People are consistently learning, through formal and non-formal education, work and life. Unfortunately, learning that takes place outside the formal education system is often undervalued. For instance, on the job training, informal apprenticeships, and engaging in various non-formal work activities are some ways in which people learn and obtain new skills. However, these forms of learning are not captured through a certification system and the knowledge and skills are not formally recognized. Establishing a system that allows people to have those skills recognize can be beneficial for people, employers and society at large. People are more motivated to continue learning when their skills are recognized, and the labor market can reward the worker for it through better wages. Countries like Canada, Mexico, Norway, New Zealand and South Africa have set up systems for recognizing prior learning (obtained through various means) by measuring learning outcomes against standards used in the formal education and training systems. They have done so to provide an opportunity for people to acquire qualifications (and accrue credits towards qualifications) without having to learn it again formally (ILO, 2018). To ensure young adults’ transition successfully from school to work, there is a need to strengthen the formal TVET taught in the secondary and post-secondary education systems. The vocational education system can help improve the quality of skills imparted on future workers. But the TVET system is underutilized. To improve its utilization and ensure students transition into work after completing their education it will be important to modernize the curricula, set up a system to regularly update it to reflect industry demands, and adopt qualification standards aligned with local needs and international standards. The TVET system should continue providing practical training through workplace learning. It should also place a strong emphasis on career guidance to ensure students understand the returns on their educational investment, make best use of their newly acquired skills and are guided through the system and into work. Consider also enhancing business-oriented education and financial education to promote entrepreneurship as another path into the economy. 1.5.2 Facilitate access to financial support for firms to enable workers to upskill. Firms that invest in worker training often make net gains from training through increases in workers’ productivity and worker retention. But training is costly and often needs financial support. Training benefits employers as well as trainees in the long term, but staff taking time off to train imposes a cost on firms. The potential costs to firms are: the wage costs of staff taking time off to train; and the frictional costs of staff taking time off to train (for example the recruitment costs of finding temporary replacement staff. To improve the skills of existing workers and promote upskilling of the workforce, provide financial support to employers for allowing (lower-skill and at-risk of dismissal) staff to take time off to train in recognition of the costs they will incur and to promote lifelong learning. Consider an employer training investment program that assists interested businesses in training their employees. Financial support provided to employers whose staff take time off, is in recognition of the costs they will incur. Government can support employers willing to allow workers to take time off to train by allowing them to claim subsidies towards the wage costs of eligible workers registered with an approved training provider to work towards acquiring basic skills or relevant qualifications that will allow them to remain employed. Employers should be able to claim the support if the employee is paid for the period that he or she is training. To prevent abuse, program participation can be made contingent on criteria that address productivity concerns (such as having clear job creation/retention goals) and workers targeting can be done by setting an eligibility that Page 99 of 114 targets workers in occupations where wages are below expected levels, and whose need to upskill is most salient. To prevent that these programs are only benefiting medium and large firms, special concessions can be made for growth oriented small firms willing to allow their workers to train. Smaller firms tend to provide less training for their staff than larger firms because of cost concerns and limited cash flow. Many of these firms are productive and have growth potential but their limited staff limits their ability to spare any workers for training. Most firms in Turkey are small, and many have higher proportion of staff with low skill levels than larger firms, which makes them more likely to have a larger share of people at risk of retrenchment. Given that small firms face higher costs associated with training for low skilled people because they likely face higher frictional costs when staff take time off work and have smaller internal labor markets and more specific jobs in many cases, these firms require support paid at a higher rate than larger firms. As in the UK, small firms willing to allow their workers to access training receive up to 150 percent of the average wage costs of low-skilled staff for the required period off. The higher amount covers costs that the firm may face beyond wage costs and to ensure full support. In Turkey, the amount need not be that high, but further analysis to determine the right amount should be considered. Administration of this program can be established through existing bodies (e.g. ISKUR, KOSGEB) and accessing it should not be cumbersome for firms to apply for the program to increase its’ take up rate. 1.5.3 Facilitate the establishment of employer-led sector skills councils. Strengthening the links between training providers and employers requires reforming the way these links are harnessed into the actual training. Many countries around the world, struggling with skills mismatches and unemployment, have established national skills councils as well as local and sectoral skills councils to improve learning provision and qualifications imparted on students and adult trainees meet sector needs and standards. Such councils are most effective when they work with the education and training system, government bodies, trade unions and professional bodies and focus on skills development that address the needs of industries and public service sectors. These councils serve as a platform to regularly obtain industry demands which are immediately fed back into the formal and informal education and training systems. Councils can be employer-led, they are usually sectoral, and work with professional bodies and trade unions to guide development needs of specific economic sectors or cluster of sectors. Skills councils are not new, many OECD countries have some form of national or sector skills council in place. The UK uses sector skills councils (SSCs) as strategic bodies to identify the skills and productivity needs of sectors and provide inputs for skills providers on sectoral needs. The SSCs in the UK are part of a system and work in partnership with various government institutions (e.g. PES, development agencies, education and training bodies) to improve the delivery of skills policy and learning provision. They support the effective operation of sector labor markets by ensuring occupational standards, learning provision and qualifications meet sector needs. Government usually (financially) supports the formation and administration of SSCs in exchange for improvements in sector skills and productivity performance, and employer investment in workforce development. To conclude this chapter, to ensure efficient and effective use of resources (mainly when using incentive programs) it is important to evaluate programs regularly. Thus, keeping track of the costs and benefits of these programs is important to adapt as needed. Also, support for these Page 100 of 114 programs should be made contingent on results. This kind of diagnostic should be done regularly to status of the labor market, identify new challenges, assess the impact of local policies/incentives and identify policies that may hinder the creation of jobs across the economy. Page 101 of 114 APPENDIX App 1 Nuts Classification of Turkey NUTS-1 NUTS-2 NUTS-3 Istanbul Region (TR1) Istanbul Subregion (TR10) Istanbul Province (TR100) Tekirdağ Province (TR211) Edirne Province (TR212) Tekirdağ Subregion (TR21) Kırklareli Province (TR213) Balıkesir Province (TR221) West Marmara Region (TR2) Balıkesir Subregion (TR22) Çanakkale Province (TR222) Izmir Subregion (TR31) İzmir Province (TR310) Aydın Province (TR321) Denizli Province (TR322) Aydın Subregion (TR32) Muğla Province (TR323) Manisa Province (TR331) Afyonkarahisar Province (TR332) Kütahya Province (TR333) Aegean Region (TR3) Manisa Subregion (TR33) Uşak Province (TR334) Bursa Province (TR411) Eskişehir Province (TR412) Bursa Subregion (TR41) Bilecik Province (TR413) Kocaeli Province (TR421) Sakarya Province (TR422) Düzce Province (TR423) Bolu Province (TR424) East Marmara Region (TR4) Kocaeli Subregion (TR42) Yalova Province (TR425) Ankara Subregion (TR51) Ankara Province (TR511) Konya Province (TR521) West Anatolia Region (TR5) Konya Subregion (TR52) Karaman Province (TR522) Antalya Province (TR611) Isparta Province (TR612) Antalya Subregion (TR61) Burdur Province (TR613) Adana Province (TR621) Adana Subregion (TR62) Mersin Province (TR622) Hatay Province (TR631) Kahramanmaraş Province (TR632) Mediterranean Region (TR6) Hatay Subregion (TR63) Osmaniye Province (TR633) Kırıkkale Province (TR711) Aksaray Province (TR712) Niğde Province (TR713) Nevşehir Province (TR714) Kırıkkale Subregion (TR71) Kırşehir Province (TR715) Kayseri Province (TR721) Sivas Province (TR722) Central Anatolia Region (TR7) Kayseri Subregion (TR72) Yozgat Province (TR723) Page 102 of 114 Zonguldak Province (TR811) Karabük Province (TR812) Zonguldak Subregion (TR81) Bartın Province (TR813) Kastamonu Province (TR821) Çankırı Province (TR822) Kastamonu Subregion (TR82) Sinop Province (TR823) Samsun Province (TR831) Tokat Province (TR832) Çorum Province (TR833) West Black Sea Region (TR8) Samsun Subregion (TR83) Amasya Province (TR834) Trabzon Province (TR901) Ordu Province (TR902) Giresun Province (TR903) Rize Province (TR904) Artvin Province (TR905) East Black Sea Region (TR9) Trabzon Subregion (TR90) Gümüşhane Province (TR906) Erzurum Province (TRA11) Erzincan Province (TRA12) Erzurum Subregion (TRA1) Bayburt Province (TRA13) Ağrı Province (TRA21) Kars Province (TRA22) Iğdır Province (TRA23) Northeast Anatolia Region (TRA) Ağrı Subregion (TRA2) Ardahan Province (TRA24) Malatya Province (TRB11) Elazığ Province (TRB12) Bingöl Province (TRB13) Malatya Subregion (TRB1) Tunceli Province (TRB14) Van Province (TRB21) Muş Province (TRB22) Bitlis Province (TRB23) Central East Anatolia Region (TRB) Van Subregion (TRB2) Hakkâri Province (TRB24) Gaziantep Province (TRC11) Adıyaman Province (TRC12) Gaziantep Subregion (TRC1) Kilis Province (TRC13) Şanlıurfa Province (TRC21) Şanlıurfa Subregion (TRC2) Diyarbakır Province (TRC22) Mardin Province (TRC31) Batman Province (TRC32) Şırnak Province (TRC33) Southeast Anatolia Region (TRC) Mardin Subregion (TRC3) Siirt Province (TRC34) App 2 Manufacturing NACE Rev.2 2digit Subsectors 10 Manufacture of food products 11 Manufacture of beverages 12 Manufacture of tobacco products 13 Manufacture of textiles Page 103 of 114 14 Manufacture of wearing apparel 15 Manufacture of leather and related products Manufacture of wood and of products of wood and cork, except furniture; manufacture of articles of straw and 16 plaiting materials 17 Manufacture of paper and paper products 18 Printing and reproduction of recorded media 19 Manufacture of coke and refined petroleum products 20 Manufacture of chemicals and chemical products 21 Manufacture of basic pharmaceutical products and pharmaceutical preparations 22 Manufacture of rubber and plastic products 23 Manufacture of other non-metallic mineral products 24 Manufacture of basic metals 25 Manufacture of fabricated metal products, except machinery and equipment 26 Manufacture of computer, electronic and optical products 27 Manufacture of electrical equipment 28 Manufacture of machinery and equipment n.e.c. 29 Manufacture of motor vehicles, trailers and semi-trailers 30 Manufacture of other transport equipment 31 Manufacture of furniture 32 Other manufacturing App 3 Construction NACE Rev.2 2digit Subsectors 41 Construction of buildings 42 Civil engineering 43 Specialised construction activities App 4 Service Nace Rev.2 2-digit Subsectors and Sector Classifications Wholesale and retail trade and repair of motor vehicles 45 and motorcycles Retail and Wholesale Wholesale trade, except of motor vehicles and 46 motorcycles 47 Retail trade, except of motor vehicles and motorcycles 49 Land transport and transport via pipelines 50 Water transport Transportation and Storage 51 Air transport 52 Warehousing and support activities for transportation 53 Postal and courier activities 55 Accommodation Accommodation and Food 56 Food and beverage service activities Information and Communication Technologies 58 Publishing activities Page 104 of 114 Motion picture, video and television programme 59 production, sound recording and music publishing activities 60 Programming and broadcasting activities 61 Telecommunications Computer programming, consultancy and related 62 activities 63 Information service activities Financial service activities, except insurance and 64 pension funding Insurance, reinsurance and pension funding, except Finance 65 compulsory social security Activities auxiliary to financial services and insurance 66 activities Real Estate 68 Real estate activities 69 Legal and accounting activities Activities of head offices; management consultancy 70 activities Architectural and engineering activities; technical testing Scientific, Occupational and Technical 71 and analysis Activities 72 Scientific research and development 73 Advertising and market research 74 Other professional, scientific and technical activities 75 Veterinary activities 77 Rental and leasing activities 78 Employment activities Travel agency, tour operator and other reservation 79 service and related activities Administrative and Support Services 80 Security and investigation activities 81 Services to buildings and landscape activities Office administrative, office support and other business 82 support activities Public administration and defence; compulsory social Public Administration 84 security Education 85 Education 86 Human health activities Health 87 Residential care activities 88 Social work activities without accommodation 90 Creative, arts and entertainment activities Libraries, archives, museums and other cultural 91 activities Culture 92 Gambling and betting activities Sports activities and amusement and recreation 93 activities Page 105 of 114 94 Activities of membership organisations 95 Repair of computers and personal and household goods 96 Other personal service activities Other Services Activities of households as employers of domestic 97 personnel Undifferentiated goods- and services-producing 98 activities of private households for own use 99 Activities of extraterritorial organisations and bodies App 5 Variables, definitions and formulas used in Labor demand chapter Variable Definition Employment Simple average of number of employees in each quarter Total Wages Total wages paid to all employees within a year Capital Stock Tangible Assets + Intangible Assets (both depreciated values) Real Capital Stock Capital stock deflated with PPI for capital goods Output Net Sales + Change in Stocks of Finished and Semi-Finished Goods Real Output Output deflated with sectoral deflators Total Operating Profits + Total Wages + Depreciation of Tangible and Value Added Intangible Assets Real Value Added Value added deflated with sectoral deflators Material Inputs Output - Value Added Real Material Inputs Material inputs deflated with PPI for intermediate inputs ( + + ) ⁄ , = ( + ℎ ) ⁄ , = Page 106 of 114 App 6 The percentage of the number workers at around minimum wage in sectoral employment, Accommodation sector, by province, 2016 Source: Authors’ elaborations based on EIS data App 7 The percentage of the number workers at around minimum wage in sectoral employment, Construction and Specialised construction activities sector, by province, 2016 Source: Authors’ elaborations based on EIS data App 8 The percentage of the number workers at around minimum wage in sectoral employment, Education sector, by province, 2016 Source: Authors’ elaborations based on EIS data Page 107 of 114 App 9 The percentage of the number workers at around minimum wage in sectoral employment, Architectural and engineering activities, technical testing and analysis sector, by province, 2016 Source: Authors’ elaborations based on EIS data App 10 The percentage of the number workers at around minimum wage in sectoral employment, Entertainment and recorded media sector, by province, 2016 Source: Authors’ elaborations based on EIS data App 11 The percentage of the number workers at around minimum wage in sectoral employment, Food and beverage service activities sector, by province, 2016 Source: Authors’ elaborations based on EIS data Page 108 of 114 App 12 The percentage of the number workers at around minimum wage in sectoral employment, Forestry and fishing, by province, 2016 Source: Authors’ elaborations based on EIS data App 13 The percentage of the number workers at around minimum wage in sectoral employment, Legal, accounting, financial and insurance sector, by province, 2016 Source: Authors’ elaborations based on EIS data App 14 The percentage of the number workers at around minimum wage in sectoral employment, Manufacture of food products sector, by province, 2016 Source: Authors’ elaborations based on EIS data Page 109 of 114 App 15 The percentage of the number workers at around minimum wage in sectoral employment, Manufacture of furniture sector, by province, 2016 Source: Authors’ elaborations based on EIS data App 16 The percentage of the number workers at around minimum wage in sectoral employment, Manufacture of leather sector, by province, 2016 Source: Authors’ elaborations based on EIS data App 17 The percentage of the number workers at around minimum wage in sectoral employment, Manufacture of textiles sector, by province, 2016 Source: Authors’ elaborations based on EIS data Page 110 of 114 App 18 The percentage of the number workers at around minimum wage in sectoral employment, Manufacture of wearing apparels sector, by province, 2016 Source: Authors’ elaborations based on EIS data App 19 The percentage of the number workers at around minimum wage in sectoral employment, Manufacture of wood sector, by province, 2016 Source: Authors’ elaborations based on EIS data App 20 The percentage of the number workers at around minimum wage in sectoral employment, Transportation sector, by province, 2016 Page 111 of 114 Source: Authors’ elaborations based on EIS data App 21 The percentage of the number workers at around minimum wage in sectoral employment, Wholesale and retail sector, by province, 2016 Source: Authors’ elaborations based on EIS data Page 112 of 114 BIBLIOGRAPHY Achor, S. 2012. “Positive Intelligence.” Harvard Business Review. Arias, Omar S., and Carolina Sanchez-Paramo. 2014. “Back to Work: Growing with Jobs in Europe and Central Asia”. Washington, DC: World Bank. Bagir, Yusuf Kenan. 2018. "Impact of the Syrian refugee influx on Turkish native workers : An ethnic enclave approach," Central Bank Review, Research and Monetary Policy Department, Central Bank of the Republic of Turkey, vol. 18(4), pages 129-147. Blalock, G., and P. J. Gertler. 2004. “Learning from Exporting Revisited in a Less Developed Setting.” Journal of Development Economics 75 (2): 397–416. Ceritoglu, E., H. B. Gurcihan Yunculer, H. Torun, and S. Tumen. 2015. “The Impact of Syrian Refugees on Natives' Labor Market Outcomes in Turkey: Evidence from a Quasi- Experimental Design." Unpublished manuscript, Central Bank of the Republic of Turkey. Del Carpio, X., and M. Wagner. 2016. “The Impact of Syrian Refugees on the Turkish Labor Market.” Journal of Labor Economics. Dollar, D., M. Hallward-Dreimeier, and T. Mengistae. 2005. “Investment Climate and Firm Performance in Developing Economies.” Economic Development and Cultural Change 54 (1): 1–31. Doms, M., and E. J. Bartelsman. 2000. “Understanding Productivity: Lessons from Longitudinal Microdata.” Journal of Economic Literature 38 (3): 569–594. Fafchamps, M. 2009. “Human Capital, Exports, and Earnings.” Economic Development and Cultural Change 58 (1): 111–141. Flabbi, L., M. Macis, M. Andrea, and S. Fabiano. 2014. “Do Female Executives Make a Difference? The Impact of Female Leadership on Gender Gaps and Firm Performance.” IZA Discussion Papers 8602. Gagliarducci, S., and M. D. Paserman. 2014. “The Effect of Female Leadership on Establishment and Employee Outcomes: Evidence from Linked Employer-Employee Data.” IZA Discussion Papers 8647. Guadalupe, M., O. Kuzmina, and C. Thomas. 2010. “Innovation and Foreign Ownership.” NBER Working Paper Series 16573. Heyman, F., H. Svaleryd, and J. Vlachos. 2013. “Competition, Takeovers and Gender Discrimination.” Industrial and Labor Relations Review 66 (2). Lyubomirsky, King, and Diener. 2005. “The Benefits of Frequent Positive Affect: Does Happiness Lead to Success?” Psychological Bulletin 131 (6). Page 113 of 114 OECD (Organisation for Economic Co-operation and Development). 2014. “Education at a Glance2014, OECD Indicators.” http://www.oecd.org/education/eag.htm. Olcott, G., and N. Oliver. 2014. “Social Capital, Sense-Making and Recovery from Disaster: Japanese Companies and the March 2011 Earthquake.” California Management Review 56 (2): 5–22. Oswald, et al. 2015. “Happiness and Productivity.” Journal of Labor Economics 33 (4). Paula, Aero, and J. A. Scheinkman. 2007. “The Informal Sector.” PIER Working Paper 08-018. Scarpetta, S., et al. 2016. “Measuring the Quality of Jobs in OECD Countries.” http://voxeu.org/article/measuring-quality-jobs-oecd-countries. Sekkat, K., A. Szafarz, and I. Tojerow. 2015. “Women at the Top in Developing Countries: Evidence from Firm-Level Data.” IZA Discussion Papers 9537. Sener, S., M. Savrul, and O. Aydın. 2014. “Structure of Small and Medium-Sized Enterprises in Turkey and Global Competitiveness Strategies.” Social and Behavioral Sciences 150: 212–221. Siegel, J., L. Pyun, and B. Y. Cheon. 2010. “Multinational Firms, Labor Market Discrimination, and the Capture of Competitive Advantage by Exploiting the Social Divide.” Harvard Business School Working Paper 11-011. Srour, I., Erol Taymaz, and Marco Vivarelli. 2013. “Skill-Biased Technological Change and Skill- Enhancing Trade in Turkey: Evidence from Longitudinal Microdata.” IZA DP No. 7320. Syverson, C. 2011. “What Determines Productivity?” Journal of Economic Literature 49 (2): 326–65. Tuncay, A. C. 2013. “Brief History and Flexibilisation Efforts of Turkish Labour Law.” Dokuz Eylül Üniversitesi Hukuk Fakültesi Dergisi, Cilt 15, Özel S: 341–367. Verhoogen, E. A. 2008. “Trade, Quality Upgrading, and Wage Inequality in the Mexican Manufacturing Sector.” The Quarterly Journal of Economics 123 (2): 489–530. World Bank. 2011. World Bank Enterprise Survey Database. www.worldbanksurveys.org. ———. 2018. “Doing Business 2018: Reforming to Create Jobs”. Flagship Report, Washington DC. Yildiz, O., O. Cetinkaya-Bozkurt, A. Kalkan, and A. Avci. 2013. “The Relationships between Technological Investment, Firm Size, Firm Age and the Growth Rate of Innovational Performance.” Procedia - Social and Behavioral Sciences 99: 590–599. Page 114 of 114