ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 1 Report No: 83889 - TR ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY Prepared by: Dan Finn (CESI), Rebekka Grun (World Bank), Katia Herrera-Sosa (World Bank), Herwig Immervoll (OECD), Cristobal Ridao-Cano (World Bank), Gökçe Uysal (BETAM), and Ahmet Levent Yener (World Bank) November 2013 Human Development Sector Unit Europe and Central Asia Region WORLD BANK 2 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY CURRENCY EQUIVALENTS (Exchange Rate Effective November, 2013) CURRENCY = TL U$ 1.00 = 2.03 TL WEIGHTS AND MEASURES: Metric System ACRONYMS AND ABBREVIATIONS ALMPs Active Labor Market Programs ALS Active Labor Services CCTs Conditional Cash Transfers CESI Centre for Economic and Social Inclusion DSP Department of Social Protection EJML European Job Mobility Laboratory EU European Union GDP Gross Domestic Product GDSA General Directorate of Social Assistance İŞKUR Turkish Employment Agency IT Information Technology JVCs Job and Vocational Counselors LAFOS Labor Force Service Centers LLL Lifelong Learning LM Labor Market MIS Management Information System MoFSP Ministry of Family and Social Policies MSD Minister for Social Development MW Minimum Wage NAV Norwegian integrated labor and welfare service NDLP New Deal for Lone Parents NGOs Non-Governmental Organizations OECD Organization for Economic Co-operation and Development PA Personal Adviser PES Public Employment Services PISA Programme for International Student Assessment PLD Persistent Labor Market Difficulties SA Social Assistance SABER Systems Approach for Better Education Results SAIS Social Assistance Information System SASF Social Assistance and Solidarity Foundations ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 3i SILC Survey on Income and Living Conditions SNAP Supplemental Nutrition Assistance Program SWC Social Welfare Centers TANF Temporary Assistance for Needy Families TFP Total Factor Productivity TIMSS Trends in International Mathematics and Science Study TUBITAK Scientific and Technical Research Council of Turkey TUIK Turkish Statistical Institute TVET Turkey’s Vocational Education and Training UK United Kingdom UR Unemployment Rate US United States VET Vocational Education Training WAP Working-Age Population WfD Workforce Development WFIs Work-Focused Interviews Vice President: Philippe H. Le Houerou, ECAVP Country Director: Martin Raiser, ECCU6 Sector Director: Ana Revenga, ECSHD Sector Manager: Roberta Gatti, ECSHD Task Team Leader: Rebekka Grun, ECSHD 4 ii ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 5 iii CONTENTS Acknowledgments vii Executive Summary ix 1. Conceptual Framework and Motivation 1 1.1 The Vulnerable 1 1.2 Activation Policies 4 2. Strengthening Policies to Activate Social Assistance Beneficiaries into Employment in Turkey 7 2.1 Background and Summary of Findings 7 2.2 Coordination and Delivery of PES Programs Targeted at SA Beneficiaries in Turkey 9 2.2.1 The Current Situation 9 2.2.2 Strategic Planning and Legal Arrangements 11 2.2.3 Activation Implementation 13 2.2.4 Next Steps 16 2.3 A Review of OECD Activation Reforms 18 2.3.1 Defining the Concept of Activation 18 2.3.2 Organizational Reforms to Improve Coordination and Cooperation between PES and SA Delivery Systems 19 2.3.3 Activating SA Recipients: Work-Focused Interviews, Incentives, and Sanctions 26 2.3.4 Eligibility for Employment Services and Assessing Employability 37 2.3.5 The Role of Frontline Advisers and Employment Counselors 38 2.3.6 Contracting with External Providers for the Delivery of Employment Services 43 2.3.7 Making the Turkish SA System More Pro-employment 44 Annex 1: Profiling Systems and Employability Assessments 47 References 50 3. Profiles of Vulnerable People In Turkey 55 3.1 Introduction 55 3.2 Extent and Types of Employment Difficulties 56 3.3 What Characteristics are Associated with Persistent Market Labor Difficulties? 59 3.4 What Are the Main Groups of Individuals Experiencing PLD? 64 3.5 Targeting Activation and Employment and Income Support Policies 68 3.5.1 Targeting Based on Living Standards 68 6 iv ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 3.5.2 Targeting Based on Labor Market Difficulties 71 Annex 1: Variables and Results from Cluster Analysis 73 References 75 4. Mobilizing Skills for Productive Employment in Turkey 77 4.1 The Role of Skills in the Turkish Labor Market 77 4.1.1 Workforce Skills Supply 77 4.1.2 Workforce Labor Market Profile 82 4.1.3 Evidence of a Skills Barrier to Employment 87 4.2 A Solution From the Education System 98 4.3 Policy Options to Overcome the Skills Barrier 100 References 102 List of Boxes Box 2.1: Information Cross-Checked by SAIS at Beneficiary Registration 12 Box 2.2: Japanese PES for Mothers Seeking Employment 22 Box 2.3: Employment and Work Requirements Applied to SA Benefits in the U.S. 28 Box 2.4: The British New Deal for Lone Parents 30 Box 2.5: Work-Focused Interviews (WFI) and Mandatory Work Preparation in the U.K. 31 Box 2.6: Lone Parent Employment and Child-Rearing Allowance in Japan 33 Box 2.7: Benefit Sanctions for SA Claimants in Australia, Germany, and New Zealand 36 Box 4.1: The Turkish Education System 78 List of Figures Figure 0.1: From vulnerability to “good” jobs xii Figure 0.2: Different types of labor market difficulties/vulnerability xvi Figure 0.3: Broad categories of persistent labor market difficulties (PLD), 2005-2008 xvii Figure 0.4: Groups with PLD: complete groupings for the period 2005-2008 xix Figure 0.5: Targeting based on living standards: some groups have much greater need for support xxi Figure 0.6 : Targeting based on distance: some groups have much better chances of overcoming PLD xxiii Figure 1.1 : A number of structural factors behind long-term employment challenges 3 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY v 7 Figure 1.2 : From vulnerability to ‘good’ jobs 4 Figure 2.1 : ALMP offered by İŞKUR 15 Figure 2.2 : İŞKUR beneficiaries and allocation of services 16 Figure 3.1 : Different types of labor market difficulties/vulnerability 57 Figure 3.2 : Broad categories of persistent labor market difficulties, 2005-2008 59 Figure 3.3 : Groups with PLD: complete groupings for the period 2005-2008 68 Figure 3.4 : Targeting based on living standards: Some groups have much greater need for support 70 Figure 3.5 : Targeting based on distance: some groups have much better chances of overcoming PLD 72 Figure 4.1 : Educational profile of Turkey’s WAP, 2010 79 Figure 4.2 : Education level by age, 2011 80 Figure 4.3 : Percent of labor force that has attained at least secondary education in OECD countries 81 Figure 4.4 : Breakdown of Turkey’s WAP by employment status 83 Figure 4.5 : Labor market status by education level, 2011 84 Figure 4.6 : Unemployment rates (%) by education level, 2000 and 2011 86 Figure 4.7 : Returns to education in formal and informal labor markets (controlling for industry and occupation) 89 Figure 4.8 : Evolution of standardized skills in the year of reference 92 Figure 4.9 : Trends in skills utilization 94 Figure 4.10 : Majors pursued by male and female higher education graduates in Turkey, 2010 96 Figure 4.11 : Top 10 business environment constraints 97 List of Tables Table 2.1: Social assistance figures, 2012 13 Table 2.2: Common tasks of employment counselors delivering services for the unemployed 39 Table 2.3: Employment adviser caseloads in selected OECD countries 41 Table 3.1: Factors associated with PLD risks, 2005-2008 63 Table 3.2: Groups with PLD: An illustration 65 Table 4.1: Summary of skill types 90 8 vi ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 9 vii ACKNOWLEDGEMENTS This is a joint study by the World Bank (WB) and the Turkish Ministry of Development (MoD). This report was prepared by a team comprising Rebekka Grun v. Jolk (WB), Cristobal Ridao-Cano (World Bank), Herwig Immervoll (OECD), Ahmet Levent Yener (WB), Katia Herrera-Sosa (WB), Gökhan Güder (MoD), Sinem Çapar (MoD), Dan Finn (CESI) and Gökçe Uysal (BETAM). Cristobal Ridao-Cano (WB) and Herwig Immervoll (OECD) initiated the dialogue to start this work and conceptualized the components and analytical focus of the report. Amy Gautam substantially improved the report through her professional editing, and Norosoa Andrianaivo. Overall guidance was provided by Martin Raiser (Country Director for Turkey, WB), Roberta Gatti (Sector Manager, Human Development Economics, Europe and Central Asia, WB), and Ana Revenga (Director, Human Development, Europe and Central Asia, WB). 10 viii ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY ix 11 Executive Summary Introduction for non-agricultural jobs. Most of them will be young and low-skilled. Poor Despite Turkey’s strong economic labor market integration of young performance and rich job creation in people creates very significant costs recent years, the Turkish labor market for individuals and for the productive continues to suffer persistent low capacity of the economy as a whole. activity rates and low labor productivity. About half of the Turkish working-age The diverse opportunities and population (WAP) does not enter the challenges of the Turkish labor market labor market and nearly 40 percent motivated a comprehensive and of workers are informal (holding jobs programmatic policy engagement without social security benefits). While between the Turkish government and the overall unemployment rate in 2012 the World Bank, broadly following the stands at a moderate 9.2 percent, this objective of “Creating Good Jobs.” rate does not reveal the particular Four analytical reports were prepared situation of women and youth. to support this heading, by defining: (i) what “good jobs” are in Turkey; (ii) how Only 29 percent of working-age good jobs are created in the economy; women work (the lowest among (iii) the kinds of workers and skills OECD countries) and over 50 percent required for good jobs; and (iv) how of female employment is informal. At policy can facilitate the matching of the the same time, there are large returns existing labor force with good jobs. to increasing female employment: According to a World Bank simulation, The fourth topic is the focus of this increasing the share of working women report, which is particularly interested by just 6 percentage points could in how women and youth can be increase income by 7 percent and assisted to find productive and secure reduce poverty by 15 percent. employment. The relevant policies are generally called “activation policies.” Above 30 percent of youth (ages 15-24) Activation policies have a role to play are neither working nor going to school both for first-time labor market entrants (the highest among OECD countries). (often women) and for young people Urbanization, agricultural shedding, who have already been facing labor and the predicted strong increase in the market difficulties for a while. Indeed, WAP until around 2020 will continue to international evidence shows that push up the number of people looking well-designed employment activation 12 x ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY policies can have a greater impact on demonstrate the initial steps of profiling youth than on older people. and draw some initial conclusions on the types of clients that might be Turkey recently significantly expanded covered by activation policies in Turkey, the scope of its activation policies, and the extent of their socio-economic which were far more modest than vulnerability as well as labor market those of many other OECD countries. employability. A key conclusion is The authorities have identified a very that policy makers will need to decide large group that could potentially be whether to put the emphasis on the targeted by activation policies: namely, former or on the latter so that the dual all those out of a regular formal sector objectives of protecting the vulnerable job, roughly two-thirds of the WAP . and helping them move out of transfer dependence are achieved. With such a large target population, there is a real need to find ways to The initial profiling exercise shows prioritize assistance. Other countries that several large subgroups of the have achieved this using profiling (i.e., vulnerable comprise inactive females, dividing the vulnerable population into often with limited or outdated skills. A relatively homogenous subgroups of priority might be to rethink the offer of decreasing order of need). Subgroups public services involved in upskilling or “profiles” receive a tailored mix of the workforce, such as lifelong social benefits and active labor market learning, in order to mobilize the largest assistance designed to maximize the identified segments. incentives of recipients of government support to join the labor market and work themselves out of dependence. This Executive Summary reflects on Turkey intends to follow this course and four connected background papers. has recently taken a number of steps in A Conceptual Framework first defines this direction. As the scope of activation vulnerability and activation policies. A policies grows, however, Turkey needs second background paper takes stock to develop more robust profiling of the progress of activation policies methods, as the initial experience in Turkey to date. A third background suggests that Turkey’s Employment paper profiles the large and diverse Agency (İŞKUR) is unable to process group of vulnerable people in Turkey the large wave of clients transferred into units of higher or lower priority, from social services. while the final background paper examines how the capacity and skills of This report presents the results of a the vulnerable, especially those in the highly simplified profiling exercise to high priority units, can be built. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY xi 13 Conceptual Framework and that İŞKUR’s role in this regard will grow. The vulnerable are working-age people who are able to work but are at high risk Activation in this report describes of not getting a “good job.” They are, in policies that help vulnerable people other words, individuals who are “stuck” break out of a low-productivity (or without a job or in low-productivity/low- even poverty) trap by building their wage jobs (notably informal workers). aspirations, human capital, and job As the group of vulnerable people is seeking efforts, with the ultimate aim of relatively large and heterogeneous achieving their productive participation in Turkey, it is important to consider in society and self-sufficiency. This carefully which individuals should be definition comprises a broad set of a high priority for activation. Among policy instruments. Work-incentive the vulnerable, the report focuses on measures seek to strengthen the two groups of particular significance motivation for making use of existing for overall labor market performance earnings opportunities. Training, and social outcomes in Turkey today: workforce development, and related women and youth. active labor market programs (ALMPs) This report’s focus on activation policies address specific employment barriers is motivated both by the size of the jobs on the labor supply side by improving challenge for the vulnerable in Turkey the capacity or skills of job seekers and by Turkey’s heretofore relative and other activation clients. Finally, under utilization of activation strategies employment services and ALMPs that compared to other OECD countries. The support labor demand expand the 10th Development Plan recognizes that set of earnings opportunities that are activation policies play a key role in the accessible to those with no or only agenda to reduce employment barriers, limited income from work. 14 xii ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY FIGURE 0.1 From vulnerability to “good” jobs job ‘good’ job Barri Lack of Lack of Opportunity Opportunity Other relevant Other ers releveant Activation policies policies Activation Macro Macro Barri policies Lack of Education policies Lack of Education institutions Labor institutions Labor Capacity Capacity ers Labor regulations … Barri Lack of Lack of Motivation Motivation ers Vulnerable Vulnerable e.g., ‘trapped’ e.g., inactivity, ininactivity, ‘trapped’ in unemployment, informality unemployment, informality Source: Authors. Well-designed and implemented they can have a positive impact on a activation policies can promote range of outcomes but that, in many growth (by increasing employment cases, they have failed to do so. Their and making labor markets more mixed success points to an important efficient) and family incomes (by role for customizing and coordinating strengthening earnings prospects activation policies. Coordination and helping to direct public support between activation measures is to those who need it most). There especially important when, as is often is, however, conflicting empirical the case, vulnerability is associated evidence on the effectiveness of with multiple barriers that all need to activation policies: Studies show that be addressed in turn (Figure 0.1). ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 15 xiii Strengthening Policies to Activate beneficiaries to İŞKUR for registry, Social Assistance Beneficiaries into but the data transfer had to be put on Employment in Turkey hold (in December 2012) due to this high volume; a screening/prioritization Implementing policies to enhance mechanism to identify the top priority productive employment in Turkey beneficiaries was thus initiated. through “good jobs” is a priority for the government as it seeks to raise İŞKUR’s newly recruited job and living standards, increase productivity, vocational counselors (JVCs) had and enhance social cohesion. A registered 330,000 beneficiaries into complementary objective of the İŞKUR’s database by the end of March government is to activate the vulnerable 2013. Of those, 111,000 were directed into good jobs to benefit from the to open positions and 26,000 were above mentioned multiple gains. The employed. Another 62,000 beneficiaries Turkish government is aware that better were directed to İŞKUR’s vocational targeting, by customizing activation education training (VET) programs, of efforts and providing coordinated which 17,000 participated in them. By services, might lead vulnerable people the end of 2013, İŞKUR plans to reach all into good jobs. 1.1 million SA beneficiaries transferred into the İŞKUR database, but timely The ‘Activation Action Plan’ and its delivery of services is still an issue as status the JVC system was not designed to handle this increased caseload. Towards this end, an Action Plan on Linking Social Support System to An initial diagnosis Employment and its Activation was accepted in April 2010, followed by a Better profiling and prioritization is protocol signed between the Ministry the key to restarting the data transfer of Family and Social Policies (MoFSP) process and helping MoFSP and and İŞKUR in February 2012, to İŞKUR reach their mutual targets. incentivize job search, job finding, and Countries with well-developed public productive participation in society for employment services (PES) make an social assistance (SA) beneficiaries to initial employability assessment of job become and remain self-sufficient. seekers that is then used to “profile” them into different groups receiving As of 2012, there were around 6.4 different support packages, with the million SA beneficiaries, 1.8 million bulk of resources going to the hard-to- of which were men in the WAP . In the employ. In the activation process of SA first year of implementation, MoFSP beneficiaries to employment, İŞKUR and sent information on 1.1 million SA MoFSP are working closely to develop 16 xiv ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY broader and more detailed definitions program. International best examples of “employability” and “proper jobs.” in the activation of SA beneficiaries that are applicable to Turkey may One weakness of the activation help İŞKUR and MoFSP improve their program is the reluctance of some current model and accelerate the SA beneficiaries to be transferred to process. İŞKUR’s database, as they are afraid of losing their SA. To address this, MoFSP Case studies illustrate some of the implemented pilot programs to test policy design and implementation mechanisms to increase participation. elements that can facilitate more Raising awareness and providing better effective coordination of Turkey’s information to SA beneficiaries about SA and PES systems the activation program may increase participation and could even create The first priority concerns the strong advocates. Once all households development of national priorities are informed about the program’s and local agreements outlining successful results, the attitudes of service protocols, working methods, reluctant SA beneficiaries may change. and respective organizational responsibilities. Such protocols In the middle to long term, İŞKUR typically contain agreement on the and MoFSP should evaluate the following elements: (i) the clients to introduction of compulsory registry be served and the operating model; measures for SA beneficiaries into (ii) management arrangements İŞKUR’s database, and evaluate some and supervision of operations; (iii) punitive actions to discontinue SA personnel to be allocated to deliver the to employable SA beneficiaries not services; (iv) budgeting and monitoring participating in the activation program. of expenditure; and (v) any services If provided for a long period of time, to be outsourced or purchased from SA may blunt the skills and motivation external service providers. of beneficiaries. The government has implemented successful programs with The experiences of OECD countries scarce resources, but the sustainability that have reformed their delivery of these rigorous programs is of employment services to clients questionable in the long term. To raise receiving SA benefits would be the living standards of SA beneficiaries beneficial for Turkey to examine. and enhance social cohesion, İŞKUR These experiences include (but are and MoFSP should consider options to not limited to: (i) organizational continue SA only to the “impossible-to- reforms designed to better connect employ” and to require employable SA the respective PES and SA delivery beneficiaries to take part in the activation systems, from cooperation to co- ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY xv 17 location to integration; (ii) incentives setting clear and traceable performance and conditionality regimes that link indicators for monitoring and evaluation the receipt of benefits to active job in line with their strategic targets and search and participation in labor market action plans, İŞKUR and MoFSP may programs; (iii) initial employability enhance their ability to fine tune the assessments and profiling systems system and their institutional objectives. used to determine who is able to work More efficient cooperation between and the degree of market readiness related agencies would also help to and special needs of people that are resolve problems in implementation able to work; and (iv) the role of job and help in the further development of a counselors and their work in assessing pro-employment SA system in Turkey. employability, assisting with job search, agreeing on individual action plans, and Very importantly, in developing their monitoring progress and compliance. approach, Turkish policy makers might be assisted by further knowledge on The capacity of the Turkish PES the characteristics, employability, system to work with SA clients could and circumstances of existing SA be developed by reallocating some beneficiaries. This knowledge could aid of the resources now invested in in the planning, design, and delivery relatively high-cost training programs of employment services, and could be to support targeted core job search used for future program evaluations. In assistance and counseling services. The the future, such data might be captured employment needs of the client group through the initial employability should be identified to broker access assessment undertaken by social to complementary services, such as workers and İŞKUR’s registration child care, and to develop links with process. likely employers. Many employment services could be delivered more cost- These data would also help to prioritize effectively on a group basis, or through the stock of SA beneficiaries so that external providers that complement it can be processed more effectively. İŞKUR skills and capacities, especially if One promising strategy for grouping a they have particular expertise working beneficiary population is profiling. with specific client groups. Profiles of Vulnerable People in The integrated information system Turkey already in place in Turkey can be used to evaluate the performance of Spells of unemployment are a the overall activation system and to necessary element of a market-based make improvements, as the system growth process. If job reallocation architecture collects data from a wide is efficient, in the sense that jobs range of government agencies. By move from less productive to more 18 xvi ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY productive firms or sectors, then it the circumstances and labor market leads to a more productive economy histories of the WAP are both already and higher incomes. This is especially more diverse and evolving more rapidly true during a prolonged growth than in other parts of the OECD. process, such as the one seen in Turkey. But it also holds for the Using the Survey of Income and Living aftermath of economic downturns, Conditions (SILC), which follows which tend to be associated with individuals over a four-year period, it significant restructuring and changes is possible to examine some of these in the sectoral composition of an trends, and the histories of people’s economy. labor market experiences, in some detail. Four different types of labor Different barriers can prevent or slow market difficulties or “vulnerability” are down efficient job reallocation, creating significant economic and social costs. distinguished (and shown schematically For some groups, a combination of in Figure 0.2): (i) not working at all; (ii) policy-related barriers and insufficient working only a few months during the skills or work experience means that year (“low work intensity”); (iii) low they remain “stuck” without a job or earnings; and (iv) informal work. Some in marginal employment for extended of those categories can overlap (e.g., periods of time. Due to Turkey’s rapid those working informally or sporadically growth process, with its associated will typically have a higher risk of low social and economic transformations, earnings). FIGURE 0.2 Different types of labor market difficulties/vulnerability potential clients for employment and income support low not working earnings at all informal low work work intensity ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 1 xvii Between 2005 and 2008, nearly two- persistently jobless (either unemployed thirds of working-age individuals were or labor market inactive). About half as either out of work or in marginal or many are in persistent informal, low- informal employment for an extended paying, or unstable work, meaning that period of time (Figure 0.3). While low they do have both some labor market employment rates in Turkey are a well- attachment and some work experience. established fact, the very large share of About the same number again move people who are persistently vulnerable between joblessness and marginal, highlights the challenges for activation sporadic, or informal work. This pattern policies. Among the persistently indicates that persistent labor market vulnerable, only about half (30.3 difficulties (PLD) are indeed far from percent of working-age individuals) are one-dimensional. FIGURE 0.3 Broad categories of persistent labor market difficulties (PLD), 2005-2008 Notes: Persistence is defined as experiencing the relevant status during at least one-half of the observed years using the following categories. “Not working”: not reported to have worked during any month of the year, or no labor income at all (either in cash or in kind). “Low earnings”: labor income less than 2/3 of the full-time, full-year minimum wage, and not informal. “Low work intensity”: employed or self-employed for at least one, but less than six months during the year, and not informal. “Informal”: positive labor income but no (employer) social security contributions, or labor income is mainly earned in kind, or the person reports being an unpaid family worker. “Cyclers”: those who are in more than one of these categories during the period. In all cases, those who are younger than 15 or older than 64 and those in education or military service during most of the year are not categorized as facing labor market difficulties (and are therefore not in any of the above categories). Source: World Bank staff calculations based on TUIK SILC data, waves 2006-2009 (with 2005 to 2008 as the reference year for incomes and activity calendars). 2 xviii ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY Designing and targeting employment female employment rates in Turkey, it and income support measures requires is not surprising that the largest groups knowledge about the combined consist mostly of women. characteristics of people affected by PLD. For instance, case workers Together, the “female only” groups 1, 2, at the employment office or benefit 3, 8, and 11 account for about half of administration need to have as full a the persistently vulnerable individuals picture as possible about their clients’ in Turkey. The summary in Figure 0.4 education, income, family situation, clearly shows the great diversity in the health status, and work experience. circumstances and, hence, the potential They also need to know if there are employment barriers, of women with larger groups of comparable clients with PLD. They include those with significant relatively similar barriers or needs. work experience but low-paid, informal, or unstable employment (groups 3, “Latent class analysis” is a statistical 5, and 8), as well as large numbers of approach used to search for suitable women who have been persistently groupings over a wide range of out of work with no or very short demographic, family, social, and labor employment histories (groups 1, 2, 10, market characteristics. The results of and 11). such analysis provide a set of groups characterized by similar characteristics Among men, the groups accounting for of members within each group and the largest shares of the persistently dissimilar characteristics between vulnerable are the “informal, low- groups. A latent class analysis earning” group 6 (9 percent of all conducted using SILC data in Turkey persistently vulnerable in Turkey), as partitioned the PLD population into 12 well as two groups of older men (the separate groups. “older, self-employed family men” group 9 and the “retired urban career The full set of groups is depicted worker” group 7). In addition, there is a in Figure 0.4, showing only those sizable group of younger and relatively characteristics that are particularly well-educated single men (group 4). But relevant for distinguishing each group unlike the “young and educated single from the others. Groups are ordered by women” group 10, who are mainly out their size, with the largest group shown of work, the young men do have some first. (The share of the group in the labor market attachment and significant overall PLD population is given by the work experience (40 percent of them percentage in the first line of each box; report four years or more). e.g., 13 percent are “housewives with working husband.”) Given the very low Interestingly, only a small proportion ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 3 xix of the persistently vulnerable can be budgetary resources, a fundamental described as typical unemployed job policy decision is how to prioritize seekers with no or very little earnings government support among the 12 and actively looking for a job (group groups. Two notable criteria that 12 of “blue-collar, struggling, job- might be used for deciding on suitable seeking men”). Given the small size of targeting strategies are: this group relative to the total number of persistently vulnerable and the • People’s living standards; and relatively large size of groups with • The likely extent of their labor market inactive females, as well as limited difficulties. FIGURE 0.4 Groups with PLD: complete groupings for the period 2005-2008 Urban housewives, Older urban Young urban Educated single men working husband housewifes mothers: Low-pay / no-pay - 13 % - 11 % Group - 10 % -9% - Young child - 45 + size - Young children - 29- - Never worked - Never worked - Unstable employment - Unstable employment - Higher earning husband - ≤ basic edu - Higher earning husband 1 2 3 4 Women in long-term Younger informal Retired urban Female unpaid marginal work family men career workers agriculture workers -9% -9% -9% -8% - 45 + - 25 – 39 - Very long experience - Rural - Long experience - Long experience - Informal - High or low edu - Spouse vulnerable - ≤ basic edu - Mostly MW or less - Health issues 5 - Spouse vulnerable 6 7 8 Older self-employed Educated urban Uneducated with Blue-collar struggling family men single women struggling husband job seeking men -7% -6% -6 % -4% - 45 + - 15 – 29 - Young child - 35 – 49 - Very long experience - Living with parents - No experience - experience - Mostly agriculture - ≤ basic edu - No/little experience - Husband informal & low - ≤ basic edu - Health issues 9 10 pay 11 - Wife vulnerable 12 Notes: Group sizes are the percent of all individuals in the PLD group. Source: World Bank staff calculations based on TUIK SILC data, waves 2006-2009 (with 2005 to 2008 as the reference year for incomes and activity calendars). 4 xx ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY Targeting based on living experiencing hardship is less than the standards 79 percent average for the WAP as a whole (i.e., including the “non-PLD” With labor income the primary income groups). The proportions of people source for working-age people and reporting poor sanitation are also their families, extended spells without significantly lower than the average (18 adequately paid employment leave percent for the entire WAP) in these families financially vulnerable and at a three groups. The largest proportion high risk of poverty. As a result, income of people with poor sanitation is found support is an essential element of an for the mostly rural “female unpaid overall support package that fosters agricultural workers.” Perhaps more stable employment and reduces surprisingly, the mostly urban group poverty risks. of “educated single women” also shows an above-average incidence Figure 0.5 illustrates differences of poor sanitation (as well as a high in living standards and material incidence of financial hardship). The deprivation across groups by showing highest incidence of financial hardship the proportions of people facing forms is reported for the group of “educated of “financial hardship” and “poor single men,” but it is also high for groups sanitation.”1 Both measures show where both spouses are persistently a considerable divergence of living standards between groups and, hence, vulnerable (“younger informal family the importance of careful targeting. men,” “uneducated women with Three of the mostly urban groups struggling husband,” and “blue-collar (“urban housewives,” “retired urban struggling job-seeking men”). The career workers,” and “young urban results underline the importance of the mothers”) show the smallest incidence family situation and of addressing labor of financial hardship. In fact, in all market difficulties at the family level, three groups, the proportion of people rather than as a purely individual issue. 1- Using information in the SILC data, a household considered to be experiencing financial hardship is defined as unable to afford at least three of the following four items: a one-week vacation away from home; unexpected expenses; replacement of old/broken furniture; or new clothes. A family is considered to have poor sanitation if it reports having either no bath or shower, no toilet, or no piped water in its household. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 5 xxi FIGURE 0.5 Targeting based on living standards: some groups have much greater need for support experiencing “financial hardship”, reported by 79% of all working-age household with “poor sanitation”, reported by 18% of all working-age Urban housewives, Older urban Young urban Educated single men working husband housewifes mothers: Low-pay / no-pay - 75 % - 83 % - 77 % - 92 % -5% -7% -8% - 18 % Women in long-term Younger informal Retired urban Female unpaid marginal work family men career workers agriculture workers - 89 % - 95 % - 62 % - 89 % - 23 % - 30 % -1% - 33 % Older self-employed Educated urban Uneducated with Blue-collar struggling family men single women struggling husband job seeking men - 80 % - 84 % - 96 % - 98 % - 19 % - 20 % - 27 % - 19 % Source: See Figure 0.4. Targeting based on distance from holding a stable job or to groups where the labor market policy interventions are likely to have the greatest probability of success. Those Targeting issues also arise with two criteria need not, and generally do activation measures, employment not, provide the same answers about the services, and other types of ALMPs, desirable targeting mechanisms. This such as training. A crucial but difficult is illustrated in Figure 0.6, which shows question is whether resources should be channeled to those who are, in the shares of people with predicted some sense, furthest from obtaining or “high” and “low” risk of PLD in each 6 xxii ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY group. higher. But at the same time, some of those “low-risk” individuals may well If the objective is to focus activation succeed in overcoming their PLD even policy efforts on those who are, in without active policy support. a sense, furthest from finding and holding a stable and adequately paid In practice, it is useful to consider a wide job, then policy should focus on those range of information when deciding with a high risk of PLD (shown in red on policy design and targeting. The in Figure 0.6). Using such a criterion, information in Figure 0.5 and Figure 0.6 groups 2, 5, 8, and 11 should be should arguably be read in combination priorities for activation measures. One to inform an integrated policy approach can think of these group members that seeks to tackle both employment as those with multiple or relatively barriers and PLD risks as well as major employment barriers – they economic hardship. therefore clearly need support, but the challenges of successful and continuing For example, the red percentages in labor market integration will often be Figure 0.6 signal the share of people in formidable. each profile that have a high risk of PLD. Among the three largest segments of A very different set of priorities would the PLD population, this percentage result if the objective is to focus efforts is at least 46 percent (“young urban on groups where a significant number mothers”) and as much as 99 percent of individuals have relatively low risks (“older urban housewives”); it is also of PLD. Groups 4, 6, 10, and 12 would above 90 percent for “women in long- need to be prioritized in this case. Such term marginal work,” and “female a strategy may be attractive, as policies unpaid agriculture workers.” While would have to “bridge” a smaller gap these groups may face barriers to work since people may already be relatively based on social norms or the availability well-equipped to find a good-quality of public services (for child or elder job. The probability of successful care), arguably the greatest barrier they activation would therefore likely be face is a lack of skills. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 7 xxiii FIGURE 0.6 Targeting based on distance: some groups have much better chances of overcoming PLD how many are far from /close to “good job”? “high” or “low” risk of persistent LM difficulties Urban housewives, Older urban Young urban Educated single men working husband housewifes mothers: Low-pay / no-pay -0% -0% -0% - 38 % - 68 % - 99 % - 46 % -4% Women in long-term Younger informal Retired urban Female unpaid marginal work family men career workers agriculture workers -0% - 32 % - 10 % -0% - 96 % -1% - 36 % - 94 % Older self-employed Educated urban Uneducated with Blue-collar struggling family men single women struggling husband job seeking men - 15 % - 24 % -0 % - 25 % - 31 % -7% - 90 % -5% Mobilizing Skills for Productive acquire and maintain skills relevant to Employment in Turkey Turkey’s emerging economy.2 Turkey’s overall population is young This skills transition is critical for Turkey, and its WAP has increased over the last as low educational attainment can decade. The country can benefit from constrain the movement of workers the growing number of young people from inactivity to more productive jobs. in the labor force if they are gainfully The report’s focus on skills is restricted employed, improving individual living to the segment of the WAP that has standards and contributing to national completed formal education, and it economic growth. To do this, they must assesses the extent of skill barriers for 2- World Bank. 2013. “SABER-Workforce Development. Barometer Report. Turkey 2012.” Washington, DC: World Bank. 8 xxiv ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY their productive employment.3 The still needs to increase the skills of its report: explores the supply of and population to catch up with other OECD demand for skills and their intersection countries. in the labor market in Turkey; discusses existing and proposed government A growing share of wage employment policies to help young people develop in total employment reflects the rise skills and find gainful employment; in skill-intensive sectors that were and proposes additional policy options able to absorb a substantive part of for skills development to enhance job the growing number of educated entry. youth. Higher levels of education are associated with greater formal wage The skills of the Turkish population employment. The skills gained by the and its employed workforce have younger generations are reflected increased at all levels, albeit in the higher share of educational slowly attainment among the population in all employment categories. Vocational The 1997 educational policy that high school graduates have increased increased the duration of compulsory their shares among the employed as education from five to eight years well. Increasingly, Turkey’s workforce contributed to an improved educational has the skills required for good jobs. profile of the population. The flow of young people entering the labor market Nonetheless, unemployment and has more and higher quality skills than inactivity are important, generally those already in it. The gender gap is more so for the less educated closing fast, but women are still at a disadvantage when compared to men, The educational profile of the as their educational levels are much unemployed is mixed. Individuals with lower. The education system in Turkey below primary education have very low has shown substantive improvements levels of unemployment although their in education quality, as measured by unemployment rates increase with age. the 2003 and 2009 PISA assessments. Across all age groups, unemployment The scores of Turkish students in the is highest among individuals with three PISA disciplines (reading, math, primary education only. The high and science) increased the equivalent unemployment among youth with rising of a half year of additional schooling in skill levels suggests a slow school-to- each of the tested subjects, but Turkey work transition. Possible explanations 3- Although education is a barrier for women to participate in the labor force (particularly since almost 70 percent of women with tertiary education degrees participate in the labor force), the most frequently reason that women cite for not participating is “household work”. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 9 xxv include information failures and unclear time, the occupational profile of the pathways to link school graduates with labor force has changed over the last the world of work. decade. A demand for different types of skills (e.g., an increased demand for Among young people, inactivity rates routine cognitive skills and a decline are similar for those with both lower and in demand for routine manual skills) higher levels of education. However, as has emerged and is reflected in the age increases, inactivity rates rise for composition of the employed, while the those with lower levels of education and type and number of vacancies available decrease for those with higher levels of has also changed (favoring high school education. Inactivity also captures the graduates). discouraged workers, those who have given up looking for jobs. Almost every Current policies generally favor the second person in Turkey (48 percent of skills transition to better match the population) is either unemployed or labor market demand inactive, suggesting that while Turkey can take advantage of its increasingly Preliminary results of the Systems abundant labor, it is not fully utilizing Approach for Better Education it, and more policies are required to Results (SABER-WfD) instrument on activate people into employment. workforce development in Turkey Upgrading the skills of the “stock” of show good governmental and non- existing workers, particularly women, governmental championship at the could be a critical factor in these sets national and regional levels. Through of policies. different ministries, the Government of Turkey has taken concrete steps to The changing skills profile of the set a strategic direction to improve Turkish labor force is slowly starting the skills of the population and reduce to reflect employers’ needs the demand-supply skills mismatch. SABER reveals that there are relevant, A number of factors suggest that solid standards for quality programs in some mismatch between the demand workforce development. There is also for and supply of skills remains: (i) a well-developed skills testing process unemployment rates are relatively for major occupations. Private provision lower for high school graduates of training is well-recognized and and higher for primary and tertiary regulated and employers share their education graduates; and (ii) returns to demands and priorities for workforce education are high for vocational and development with the government and higher education levels. At the same the public. The Government of Turkey is 10 xxvi ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY working on six priority areas as part of no incentives for employers to upgrade its national Action Plan on Linking Social employees’ skills. More effort is needed Support System to Employment and its to improve the skills of people already in Activation to reduce unemployment those sectors and to induce high-skilled and inactivity and to improve the match individuals to move into them. Turkey’s between skills demand and supply. government is actively committed to improving the skills of its workforce in a Nonetheless, challenges remain in way that supports Turkey’s population the training system. First, the funding from inactivity to higher productivity mechanism for Lifelong Learning occupations. (LLL) needs to be strengthened, as low funding of Turkey’s Vocational There are further options to Education and Training (TVET) system’s overcome the skills barrier institutions often results in low-quality, supply-driven programs, removed from Additional policy options to overcome the private sector’s needs. Second, the skills barrier include: diversifying recognition of prior learning is not well the pathways for skills acquisition, organized. New jobs will require new complementing vocational programs skills and competencies; workers will with life skills training to facilitate need to maintain and upgrade their LLL and increase the impact of skills to be relevant in the market. LLL training; and developing information is supported when students are able dissemination mechanisms to reduce to transfer across courses and gain information asymmetries on the access to higher levels of training. demand for (vacancies) and supply of Third, there is limited awareness of the (graduates) skills between employers skills constraints in priority sectors and and jobseekers. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 1 1. Conceptual Framework and Motivation Abstract: The focus of this report is on groups. It is therefore a better basis for groups in a vulnerable labor market targeting and customizing activation position and how to activate them efforts than a static approach. into good jobs in Turkey. This requires defining who the vulnerable people The group of vulnerable people is are as well as understanding the relatively large and heterogeneous characteristics and uses of activation in Turkey, where about half of the policies. The conceptual framework working-age population (WAP) do not presented in this chapter addresses work and nearly 40 percent of workers both of these issues in turn before are informal (i.e., they hold jobs without outlining how the rest of the report social security benefits).4 Turkey has brings them together to provide relevant the biggest share of non-working and timely advice to policy makers in individuals in the OECD, and among Turkey. the highest informality rates, so it is important to carefully consider which 1.1 The Vulnerable individuals (or types of individuals) of this large group should be a high The vulnerable are defined as working- priority for activation. age people who are able to work but are at high risk of not getting a “good job.” Among the vulnerable, this report They are, in other words, individuals focuses on two groups that are of who are “stuck” without a job or in low- particular significance for overall productivity/low-wage jobs (notably labor market performance and social informal workers). The report therefore outcomes in Turkey today: evaluates vulnerability in dynamic terms rather than for a specific point • Women: Only 29 percent of in time. This dynamic perspective, working-age women in Turkey which considers both the frequency work (the lowest among OECD and duration of labor market problems, countries) and over 50 percent of provides a rich understanding of the female employment is informal. employment barriers faced by different Women account for the majority 4- There is a significant wage gap between informal and formal workers even after controlling for their characteristics. These differences are likely to be caused, at least in part, by productivity differentials between these sectors. There are also large differences in total factor productivity (TFP) between formal and informal firms even after controlling for firms’ and entrepreneurs’ characteristics (World Bank 2010, Country Economic Memorandum on Informality and Economic Growth). 2 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY of non-working youth (those 15-24 2020 will continue to push up the years old), although employment number of people looking for non- rates are even lower for prime- agricultural jobs. Most of them will age and older women. There are be young and low-skilled (Figure large returns to increasing female 1.1). Poor labor market integration employment: according to a World of young people creates very Bank simulation5, increasing the significant costs for individuals and share of working women by just 6 for the productive capacity of the percentage points could increase economy as a whole. Activation income by 7 percent and reduce policies have a role to play, both for poverty by 15 percent. first-time labor market entrants and for young people who have already • Youth (ages 15-24): above 30 been facing labor market difficulties percent of youth are neither working nor going to school for a while. Indeed, international (the highest rate among OECD evidence shows that well-designed countries). Urbanization, agricultural employment activation policies can shedding, and the strong increase have a greater impact on youth than predicted in the WAP until around on older adults. 5- World Bank 2011: “Female Labor Force Participation in Turkey: Trends, Determinants and Policy Framework” ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 3 FIGURE 1.1 Targeting based on living standards: some groups have much greater need for support Increasing working-age population until 2020 Growing urbanization 90 84 81 1.00 78 65+ 80 74 70 70 65 0.80 59 60 0.60 15-64 50 44 40 0.40 30 0.20 20 <15 10 0.00 0 2010 2020 2030 2040 2050 1980 1990 2000 2010 2020 2030 2040 2050 Agricultural shedding Large “stock” of low-skilled workers 50 60 56 52 45 50 40 43 35 40 30 30 25 22 23 19 19 20 20 16 16 15 10 12 10 8 10 5 0 0 Less than basic Basic Secondary College 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Total Men Women Source: TUIK (Labor Force Survey data), UN (urbanization, population projections), and authors’ calculations. 4 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 1.2 Activation Policies seek to strengthen the motivation for making use of existing earnings In this report, activation describes opportunities. Training, workforce policies that help vulnerable people development and related active labor break out of a low-productivity- or market programs (ALMPs) address even poverty trap, by building their specific employment barriers on the aspirations, human capital and labor supply side by improving the ultimately, job seeking effort; with capabilities or capacity of jobseekers the aim of achieving their productive and other activation clients. Finally, participation in society and self- employment services and ALMPs that sufficiency. This definition comprises a support labor demand (e.g., wage broad set of policy instruments. Work subsidies) expand the set of earnings incentive measures (including in-work opportunities that are accessible to support as well as obligations of benefit those with no or only limited income recipients, and associated sanctions) from work. FIGURE 1.2 From vulnerability to ‘good’ jobs job ‘good’ job Barri Lack of Lack of Opportunity Opportunity Other relevant Other ers arriers releveant Activation policies policies Activation Macro Macro B policies Lack of Education policies Lack of Education institutions Labor institutions Labor Capacity Capacity Labor regulations … Barri Lack of Lack of Motivation Motivation ers Vulnerable Vulnerable e.g., ‘trapped’ e.g., inactivity, ininactivity, ‘trapped’ in unemployment, informality unemployment, informality Source: Authors. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 5 Well-designed and implemented is also an important goal in Turkey, activation policies can promote growth and is part of the agenda of the 10th (by increasing employment and Development Plan, which provides a making labor markets more efficient) detailed roadmap for addressing, for and family incomes (by strengthening example, contractual issues (hiring earnings prospects and helping to and dismissal). As the Development direct public support to those who need Plan recognizes, a specific challenge it most). There is, however, conflicting in Turkey, more so than in other OECD empirical evidence on the effectiveness countries, is activating the large number of activation policies: Studies show of vulnerable youth and women who that they can have a positive impact on have no employment experience or a range of outcomes but that, in many whose experience is limited to low- case, they have failed to do so. Their quality jobs. mixed success points to an important role for customizing and coordinating This report works through the issues activation policies. Coordination identified in the conceptual framework between activation measures is as follows: Chapter 2 takes stock of especially important when, as is often the progress of activation policies in case, vulnerability is associated with Turkey to date, with specific regard multiple barriers that all need to be to the cooperation between İŞKUR addressed in turn (Figure 1.2). Likewise, and the General Directorate of Social activation policies are more likely to be Assistance. Chapter 3 presents a effective when embedded in a broader mechanism to divide the large group of growth and employment-friendly policy framework. vulnerable people into units of higher or lower priority for activation; in other In countries with higher activity rates, words, it defines “customer segments” activation policies often focus on for social sectors and services. Finally, facilitating job transitions and re- Chapter 4 focuses on the key issue of integration for those experiencing some priority segments and examines temporary labor-market problems. how the skills of the vulnerable can be Improving labor market dynamics built. 6 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 7 2. Strengthening Policies to Activate Social Assistance Beneficiaries into Employment in Turkey Abstract: As in many other countries, then enumerates some possible policy public employment services (PES) and options, derived from an assessment social assistance (SA) programs in of OECD countries that have sought Turkey are the responsibility of different to better integrate their PES and SA national ministries and are delivered system, that could inform Turkey’s through different institutions. Because strategy to do the same. the Turkish government aims to improve the ways in which both systems work 2.1 Background and Summary together, the Ministry of Development of Findings brokered the development of an Action Plan to establish more In the past few years, the Turkish effective connections between SA and government has introduced substantial employment services. The goal is that advancements in its labor market policy. people capable of working who have With the labor reform in 2008, non-wage applied for and receive SA will be labor costs were reduced significantly, “registered to the Turkish Employment to the OECD average, and Active Labor Institution (İŞKUR)’s system by related Market Programs (ALMP) were directed institutions.” Those people directed to all registered unemployed people. An from SA institutions to İŞKUR “will action plan to reduce informality was be contacted; their skill levels will be introduced in 2008. The 2011 Omnibus tested, they will be directed to training Law reduced disincentives to part-time programs they need and they will work and extended subsidies to new receive job seek support or will be hires (women and youth). Collectively, directed to vacancies and placed to the new government programs include jobs.” İŞKUR will also develop special measures to improve: labor market programs for these new clients. This flexibility, worker protection, ALMP, the report first describes the current relevance of education to market needs, situation in Turkey with respect to the and the employability of vulnerable effectiveness of SA and PES delivery. It groups, including youth and women. 8 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY Despite these latest efforts and formal wage employment in high Turkey’s strong economic performance productivity sectors and sectors with in recent years, the Turkish labor high productivity growth potential market continues to be characterized through incentives and supports. by persistently low activity rates and low labor productivity. Turkey has the A complementary objective of the lowest working-age population (WAP) government is to activate the vulnerable employment rate in the OECD, with less into good jobs to benefit from the than half of the WAP employed, of which multiple gains in terms of social equity only about a quarter are women. About and productive employment. The 35 percent of youth, mostly women, vulnerable are working-age people are neither working nor attending who are able to work but are “stuck” school. Informal employment has fallen with their current conditions (i.e., remarkably, but still affects 42 percent individuals without a job, in informal of workers, contributing to Turkey’s low-productivity/low-wage jobs, or lower labor productivity compared with not working as they don’t want to OECD and other peer countries. lose their social assistance (SA)). The Turkish government is aware that better Implementing policies to enhance targeting, by customizing activation productive employment in Turkey efforts and providing coordinated through “good jobs” is a priority for the services, might lead vulnerable people government as it seeks to raise living into good jobs. Towards this end, a new standards, increase productivity, and activation strategy is being developed enhance social cohesion. While the to incentivize job search, job finding, definition of what constitutes a “good and productive participation in society job” varies by country as well as over for SA beneficiaries to become and time for the same country, good jobs for remain self-sufficient. development in Turkey can be defined as formal wage sector jobs because: An Action Plan on Linking Social (i) these are typically associated with Support System to Employment and higher incomes, higher productivity, its Activation was accepted on April 1, and better protection and social 2010, with the aims of establishing a link inclusion of workers and their families between Turkey’s SA system and the than in informal wage employment public employment service (PES) and or self-employment; and (ii) a large supporting the economic activation of segment of the WAP does not have SA beneficiaries. The Ministry of Family access to these jobs (the vulnerable). and Social Policies (MoFSP) and İŞKUR Accordingly, the Turkish government (Turkish Employment Agency) signed a has channeled its efforts into promoting protocol on February 17, 2012, to direct ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 9 all SA applicants to İŞKUR’s database assisting with job search, of jobseekers and provide them with agreeing on individual action active labor force services. plans, and monitoring progress and compliance; The first part of this report provides an audit and assessment of the current • Five case studies of best practice Turkish programs and services targeted employment programs targeted at at SA beneficiaries and of how some particular groups of employable better coordinated PES and SA SA recipients. Information is given services are already being delivered on how these programs have been in Turkey. The second part compares designed, commissioned, delivered, the experience, impacts, and lessons monitored, and evaluated; and learned over the past two decades in selected OECD countries that have • The role, selection, and contracting reformed their delivery of employment of external and non-profit providers services to clients receiving SA for the delivery of employment benefits. This section includes an programs and services targeted at assessment of: hard-to-employ SA clients. • Organizational reforms designed 2.2 Coordination and Delivery to better connect the respective of PES Programs Targeted at PES and SA delivery systems, SA Beneficiaries in Turkey from cooperation to co-location to integration; 2.2.1 The Current Situation • Incentives and conditionality regimes İŞKUR has started to play a key role in that link the receipt of benefits to enhancing their PES. Since 2008, İŞKUR active job search and participation in has increased the coverage and quality labor market programs; of vocational training, introduced job and vocational counselors (JVCs), • Initial employability assessments and and linked SA receipts to registration profiling systems used to determine in İŞKUR. Participation in ALMP who is able to work and the degree of has been opened to all registered market readiness and special needs unemployed regardless of whether of people that are able to work; they are eligible for unemployment insurance (UI), and the number of • The role of case workers and/or vocational trainees increased from job counselors and their work 30,000 in 2008 to 464,000 in 2012, in assessing employability, representing almost one-fifth of the 10 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY registered unemployed.6 İŞKUR has option. Providers of general training been working to improve the quality are now also subject to more stringent of training through the development job placement requirements (at least of the national vocational qualification 50 percent of trainees must find system7 and, more recently, through employment for a period which cannot the selection of training providers be lower than the duration of the course on the basis of specific quality and or at least 120 days). The new regulation performance criteria. also links training to job search, and trainees who refuse job offers suitable A new regulation published in March to their characteristics are not allowed 2013 introduced a number of initiatives to benefit from an ALMP for 24 months. to improve the effectiveness of Finally, JVCs will be more active in the vocational training, including: (i) the selection of participants for training addition of new modules to training programs. programs to improve employability of participants (e.g., job search skills, İŞKUR’s efforts to transform into an interview techniques, and basic efficient employment agency have skills); (ii) rewards for job placement been remarkable, but the low levels performance and accreditation by the of labor force participation and formal Vocational Qualification Institution (VQI) employment have continued to cause when awarding contracts to providers; a gradual surge in the need for SA; and (iii) an increased job placement hence, both the content and magnitude rate requirement for job-guaranteed of SA in Turkey have increased rapidly courses. over the last decade. Social spending (including SA and services) increased In line with the new regulation, service from 0.5 percent of GDP in 2002 to providers with a low job placement 1.43 percent of GDP in 2012. Social performance face a more stringent expenditures include the SA provided supervision. ISKUR first legally warns by the Social Assistances General non-performing service providers. Directorate (mainly in the form of family If they still do not provide services support and conditional cash transfers in accordance with the contract (CCTs) for health and education), requirements then İŞKUR can collect an support to specific groups (i.e., upfront agreed performance security. handicapped citizens, poor citizens over In more severe cases, debarring is an 65, widowed women), health insurance 6- With the amendments to law no. 5763 in 2008 and to law no. 6111 in 2012, the ceiling of the annual amount transferred from unemployment insurance fund to active labor force programs was raised and determined to be 30% of government’s share transferred to this fund in a year. With this change, the expenditures on active labor force programs increased from 35 million TL to 1 billion TL. Accordingly, the number of persons benefitting from these programs has increased in these years. 7- This includes the development of 294 national occupational standards in 16 sectors, 127 of which are under implementation. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 11 coverage of poor people, and social SA system and employment and of services for children and youth. supporting the economic activation of SA beneficiaries. The Action Plan 2.2.2 Strategic Planning and Legal envisaged 13 actions to be carried out Arrangements in the short, medium and long term, including: registering SA beneficiaries Recognizing that the provision of social who are capable of working into benefits can create disincentives for İŞKUR’s database; developing formal employment and that programs suitable employment programs for SA that link the receipt of social benefits beneficiaries; adjusting the scoring to activation can have a large payoff, formula for other institutions that the government started developing execute SA activities; and maintaining strategies to strengthen the link the green card status of individuals between employment and social participating in İŞKUR’s vocational protection. The MoFSP , the Ministry of training programs and community Labor and Social Security, the Ministry work programs. MoFSP and İŞKUR of Development, and the Ministry signed a protocol on February 17, 2012, of National Education have been to direct all SA applicants to İŞKUR’s working on developing an effective database and provide them with active and integrated social protection system labor force services. İŞKUR’s JVCs that includes all parts of society in contact SA beneficiaries, administer line with economic, social, and fiscal skill level tests, and then direct clients policies, with the aims of both avoiding to open positions, vocational education needy citizens’ dependence on SA and training courses, or other labor and encouraging them to participate adaptation programs. in the labor force. The recent reforms address this by linking SA receipt to Thanks to significant improvements in registration in İŞKUR, opening up SA the infrastructure and administration to formal sector workers, and reducing of SA, activation of SA beneficiaries the disincentives to formal employment into employment is now much easier. for green card holders. Since the establishment of MoFSP in 2011 and placement of government After the development of the Action agencies dealing with social services Plan on Linking Social Support under it, SA and services have been System to Employment and its planned and implemented in a more Activation, MoFSP and İŞKUR began coordinated fashion. A new, integrated to register SA beneficiaries into the Social Assistance Information System İŞKUR database. The Economic (SAIS) was developed, while common Coordination Council accepted the mechanisms were revisited to target Action Plan on April 1, 2010, with the benefits more effectively. The General aims of establishing a link between the Directorate of Social Assistance (GDSA) 12 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY of MoFSP is currently coordinating the beneficiaries into the system. Payments SA system, with the help of around for each household are calculated in 1,000 local Social Assistance and accordance with the eligible support Solidarity Foundations (SASF) at the schemes and transferred to the bank district level. Applications are checked accounts of actual beneficiaries (e.g., by the SASF through a need-poverty to the mother in the case of a CCT for inquiry via SAIS. SAIS is interlinked her children’s education). One of the with other government databases to household members can apply to a monitor the assistance and prevent duplication. SASFs make necessary SASF by filling an application form for checks and controls and register all members in his household. Box 2.1 Information Cross-Checked by SAIS at Beneficiary Registration SAIS makes the following checks and controls: 1. Applicant(s)’ address, copies of birth/family/personal registrations and incident information (from MERNIS database of Ministry of Interior General Directorate of Civil Registration and Nationality). 2. Whether the applicant has an İŞKUR record, and s/he receives Unemployment Insurance, Short- time Working Allowance or Job Loss Compensation. 3. Whether the applicant has received a Poverty Allowance (from the database of General Directorate of Foundations). 4. Whether the applicant has received a Home Care Allowance (from the database of General Directorate of Social Services and Children Protection) 5. Whether the applicant has benefited from a Conditional Cash Transfer, Income-Generating Projects, or Other Assistance (from the database of General Directorate of Social Assistance). 6. Whether the applicant has Social Security; s/he benefits from health services and receives Allowance (assistance) according to the Law No. 2022 (from the database of Social Security Institution). 7. Whether the applicant has a Tax Liability and has registered a motor vehicle (from the database of Revenue Administration). 8. Whether or not the person has a green card (from the database of Ministry of Health). 9. Whether the applicant has registered real estate (from the General Directorate of Land Registry and Cadaster). 10. Whether the applicant receives scholarship or credit (from the General Directorate of Credit and Dormitories Agency). 11. The amount of cultivated/fallow land, registered wet/dry land and information about agricultural products cultivated on these lands (from Farmer Register System of the Ministry of Agriculture, Food and Livestock). 12. Temporary village guard salary (from the Ministry of Interior). 13. The student’s attendance and school success (from the Ministry of Education’s e-school database). Source: MoFSP 2011. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 13 In 2012, the Social Assistance and home-based care, etc.). Solidarity Law was amended to clarify There is still room to direct SA the eligibility of SA beneficiaries. beneficiaries to productive Under the amendment (in Article 2 of employment, as 28 percent of all the law), eligibility is defined by taking SA beneficiaries are working-age the total income of all household males. The development of SAIS members into consideration; an makes valuable data available to all employed beneficiary is eligible for SA institutions that deal with employment if the total income of all households is and SA. Almost one-third of the entire less than one-third of the net minimum Turkish population is registered in wage. Under such circumstances, SAIS (Table 2.1). Efficient use of this all households may be eligible for impressive information system should different kinds of SA (education enable MoFSP and İŞKUR to track support for children, coal support to beneficiaries and provide them with the family for heating, provision of appropriate services. TABLE 2.1 Social assistance figures, 2012 SA registration and beneficiaries Number Population 75,627,384 People registered in SAIS 23,668,942 Households registered in SAIS 6,768,126 SA beneficiaries (people) 6,370,100 SA beneficiaries (households) 2,101,611 SA beneficiaries (regular transfers) 1,657,144 SA beneficiaries (irregular transfers) 1,994,470 Source: TUIK database and MoFSP 2013. 2.2.3 Activation Implementation or objective tests of beneficiaries’ work capacity are used during this process. The beneficiaries of all SA programs8 The built-in control mechanisms of are covered by the new activation the SA payment system are being strategy; however, no screening tools improved, and the SASF staff make 8- SA can be provided regularly (i.e., monthly payments to widowed women), temporarily (i.e., housing benefits), or provisionally (i.e., CCTs). 14 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY necessary checks and controls at each old, and willing to be registered into the payment interval (generally bi-monthly). İŞKUR database, relevant information In addition to these controls, MoFSP is provided to İŞKUR. İŞKUR registry inspectors perform home visits to takes place in line with the provisions verify the eligibility of SA beneficiaries of the protocol dated March 17, 2012, by using a Household Visit Information and is completed by İŞKUR staff Form, developed with the Scientific and upon the receipt of information from Technical Research Council of Turkey MoFSP . Starting in January 2013, each (TUBITAK) under the “Defining a Scoring SA beneficiary directed to İŞKUR is Formula to Identify the Beneficiaries for assigned to one of İŞKUR’s 4,000 JVCs. Social Assistance Project.” During both A JVC working in the closest İŞKUR of these processes, beneficiaries are unit to the transferred SA applicant informed about the activation program monitors the applicant individually. and asked whether they would like The eligibility of a SA beneficiary may to be directed to İŞKUR to receive change over time; an SA beneficiary employment services. Information that receives support is called “active.” about those who volunteer and who JVCs track the information of both are in the WAP (those 15-64 years active and passive SA beneficiaries old) is sent to İŞKUR. No priority under a separate section of their IT criteria or classification exist during system and they can see whether a this information collection phase, and newly registered person receives SA. there is no other control related to the employability of the beneficiary. JVCs provide services in accordance MoFSP and İŞKUR are working closely with the new Active Labor Services to develop criteria for employability of (ALS) Regulation, which states that a household members and to improve SA beneficiary has to receive active the Household Visit Information Form. labor force services within two weeks of being transferred into İŞKUR’s İŞKUR registration of “able” database. Thus JVCs contact and give beneficiaries applies to the entire stock appointments to the transferred SA of SA beneficiaries; however, registry into İŞKUR’s database is not mandatory. applicants to administer skill level tests Any new applicant or existing SA to them, and to direct them to vocational beneficiary may inform the SASF education programs and labor personnel that s/he would like to be adaptation programs. Beneficiaries can registered into İŞKUR’s database during participate in all programs but the social the regular controls that are performed benefits of beneficiaries participating in by the SASF personnel at the time of public works programs are eliminated. each transfer (usually bi-monthly) or Still, SA beneficiaries do not have to anytime s/he wishes. If the SA applicant accept job offers or enroll in vocational is unemployed, between 15-64 years training. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 15 FIGURE 2.1 ALMP offered by İŞKUR Source: Authors. The activation program is still in its any sort of prioritization by the MoFSP first year of implementation. Although and their contact information is not MoFSP sent information on 1.1 million always correct, yet İŞKUR’s regulation SA beneficiaries to İŞKUR for registry, dictates that JVCs must contact the data transfer had to be put on hold SA beneficiaries within two weeks in December 2012 to deal with this to provide employment services. high volume and to initiate a screening/ Despite these problems, İŞKUR staff prioritization mechanism to identify had registered 330,000 beneficiaries the top priority beneficiaries. Though into İŞKUR’s database by the end of the situation differs from one province March 2013. Of those, 111,000 were to another, the workload of JVCs directed to open positions and 26,000 has increased due to the additional were employed. Another 62,000 number of SA beneficiaries assigned beneficiaries were directed to İŞKUR to them. Problems have arisen as these vocational education and training additional people are not subject to (VET) programs, of which 17,000 16 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY participated in them. It is not clear do not want further contact. The JVC when the data transfer from MoFSP to system was not designed to handle the İŞKUR will resume, but İŞKUR plans to caseload of SA beneficiaries, and there reach all 1.1 million SA beneficiaries currently exists no clear time frame for by the end of 2013. The JVCs will increasing their number (from 4,000). contact all of these SA beneficiaries Therefore, timely delivery of services at least once and record which ones is still an issue to be addressed. FIGURE 2.2 İŞKUR beneficiaries and allocation of services SA Beneficiaries Transferred Services provided by İŞKUR to İŞKUR 100000 80000 Registered 30% 60000 40000 Open Positions Not registered 20000 70% 0 VET Courses Not employed Enrolled Not enrolled Employed Source: Own calculations (based on data from MoFSP and İŞKUR database). 2.2.4 Next Steps them to assess the employability of job seekers and thus their need for Better profiling and prioritization is training or other services. Countries the key to restarting the data transfer with well-developed PES do make an process and helping MoFSP and initial employability assessment of İŞKUR reach their mutual targets. jobseekers that is then used to “profile” Using İŞKUR’s JVCs will most certainly them into different groups receiving result in a better match of trainees different support packages, with the to employment services, but their bulk of resources going to the hard- current workload makes it hard for to-employ. In the activation process ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 17 of SA beneficiaries to employment, before being implemented nationwide. İŞKUR and MoFSP are working closely MoFSP has also initiated a pilot study in to develop broader and more detailed Karabuk and Kirikkale, where all public definitions of “employability” and administration offices (including the “proper jobs.” It is expected that with municipality) will establish integrated these new definitions, SA beneficiaries cooperation for social services. can be better grouped and prioritized by the SASF staff, and thus receive Raising awareness and providing better employment services once they better information to SA beneficiaries are registered in İŞKUR’s database. about the activation program may increase participation and could even One weakness of the activation create strong advocates. All jobseekers program is the reluctance of some transferred to İŞKUR should receive SA beneficiaries to be transferred some services, but some will still to İŞKUR’s database, as they are be among the hard-to-employ. afraid of losing their SA. To address Once İŞKUR’s JVCs have informed this, MoFSP has been implementing households about the program’s pilot programs to test mechanisms successful results, the attitude of to increase participation. Most SA reluctant SA beneficiaries may change. beneficiaries are aware that they lack JVCs may work with these groups on necessary basic and technical skills for an action plan to find employment and good jobs, for which open positions are to monitor their progress. limited, and that if they are employed, their wages may not be sufficient In the middle to long term, İŞKUR and to cover the additional costs of MoFSP should evaluate the introduction employment (e.g., transport, clothing). of compulsory registry measures for To overcome this, MoFSP initiated SA beneficiaries into İŞKUR’s database, a pilot project in Denizli province to and evaluate some punitive actions continue making SA transfers to SA to discontinue SA to employable SA beneficiaries for a transitory period beneficiaries not participating in the after they are employed through İŞKUR activation program. If provided for a services. This pilot project not only long period of time, SA may blunt the covers the beneficiary’s compulsory skills and motivation of beneficiaries. costs in the first few months of work The government has implemented but also enables better cooperation successful programs with scarce between the İŞKUR JVC and the SASF resources, but the sustainability of these social worker assigned to the SA rigorous programs is questionable in the beneficiary. The pilot will be expanded long term. To raise the living standards to Kocaeli, Gaziantep, and Şanlıurfa of SA beneficiaries and enhance social 18 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY cohesion, İŞKUR and MoFSP should child-care responsibilities who are consider options to continue SA only receiving SA. to the “impossible to employ” and to require employable SA beneficiaries 2.3.1 Defining the Concept of to take part in the activation program. Activation International best examples in the activation of SA beneficiaries that are Research findings in the 1980s, applicable to Turkey may help İŞKUR especially in Europe, suggested that and MoFSP improve their current in many countries, the poor design model and accelerate the process. of benefit entitlements and the weak These examples are discussed next. organization of UI and SA systems weakened work incentives and 2.3 A Review of OECD induced welfare dependency. Since Activation Reforms then, OECD and EU policy makers have encouraged member countries Section 3 considers several factors to implement activation reforms for that have shaped the design and the unemployed, and increasingly in implementation of pro-employment the 2000s, argued for the extension SA programs in other OECD countries. of such policies to employable people of working age in receipt of disability, This includes findings on different early retirement, and lone parent or organizational reforms designed to other “inactive” SA benefits. Activation better connect the respective PES policies are intended to bring long-term and SA delivery systems. The section unemployed and inactive people into reviews the design of activation the effective labor supply and enhance incentives and requirements targeted their employability. They are regarded at SA recipients, the sanctions used to as a response to the challenge of aging enforce requirements, and the different populations and an effective way to ways in which the relative employability reduce poverty and social exclusion of claimants is assessed. It considers while containing the costs of social findings on the role and tasks of frontline protection systems. case workers in such delivery systems and the ways in which PES services It is useful to distinguish between have been supplemented for SA clients particular types of ALMP , such as job by contracting with external providers. search assistance, skills training, or The section also includes brief case employment subsidies, and wider studies of coordination reforms and “activation” strategies that typically employment programs from other encompass a combination of different countries, including examples of those ALMP within a reformed service delivery targeted at employable women with system. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 19 The precise nature of national activation often performed poorly in placing the strategies varies, but an important long-term unemployed into work, and variation across the OECD is whether although employable SA claimants individual countries implement a “work often had to register with the PES, first” strategy that requires unemployed they were viewed as low priority and people to enter employment as swiftly consequently received little attention. as possible, or a “human capital development” strategy that emphasizes OECD governments have introduced the importance of improving benefit various reforms that seek to improve the recipients’ skills so that they may gain coordination of service delivery; these access to higher quality employment. developments have been reviewed The former approach is most closely by the OECD and the European associated with “welfare-to-work” Commission. In some countries, such reforms in English-speaking countries, as Japan and Slovenia, reforms have such as the U.S., the U.K., and Australia. involved greater cooperation between The latter approach has been associated the social welfare system and the more with welfare states in Western PES. In others, such as Finland and and Northern Europe, especially the Norway, reforms have resulted in the Scandinavian countries. co-location and coordination of PES and SA services; and in others, such 2.3.2 Organizational Reforms as Britain and Ireland, benefit delivery to Improve Coordination and and employment services have been Cooperation between PES and SA fully integrated. In Ireland, current Delivery Systems reforms are being implemented despite high unemployment and deep In many OECD countries, responsibility public expenditure cuts because the for the delivery of benefits and government and international agencies employment services was often see such reforms as vital to future fragmented between different recovery. A further variant has involved institutions and agencies comprising more targeted reforms, as in countries national PES, benefits agencies, like Switzerland and Australia, where municipal welfare departments, and policy makers have sought to improve social insurance bodies. These were cross-institutional cooperation for generally accountable to different particular disadvantaged client groups ministries and levels of government, or localities. and social partners exercised a great deal of influence and control, especially The different approaches reflect the in social insurance systems. A major particular circumstances of each criticism of such systems was that they country, but five examples give insight 20 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY into design and delivery factors that support from both services, and there Turkish policy makers may want to is a joint assessment of clients who consider as they implement the current lose (temporarily or permanently) reform and consider possible future their work ability while they are options. unemployed. Linked databases were established in 2006 and data sharing Case study 1: Slovenia protocols enable daily transfer of data between the systems. This means that Slovenia’s government has sought SWC case workers have access to PES over time to ensure that SA recipients records, including where a client starts who are able to work become a part employment and if sanctions have been of the employment system (OECD imposed, allowing them to decide on 2009). Since 2000, 62 Social Welfare further payment of SA. PES counselors Centers (SWC), which assess eligibility do not, however, have access to benefit and distribute family and SA benefits, records. have been required to refer employable claimants to the PES. After an initial An OECD review found that while there assessment of employability at the had been progress, the strategy was SWC, the PES registers a person as less effective than it could be and that unemployed if s/he meets the criteria while the elements of an activation defined in legislation (“younger than system were in place, the impact 60 years, not homeless, not drug- or was limited “because the institutional alcohol addicted, not in the process of incentives to provide the services detoxification, not ill”). In 2007, nearly and controls necessary to make 40 percent of the people registered with activation work are not in line with the the PES received SA. Many of these responsibilities of actors” (OECD 2009: claimants were young people under 25 p. 109). In particular, as the PES was who did not qualify for unemployment not directly responsible for paying SA, benefits and the level of SA payment these clients, who were more difficult they received was below subsistence. to place, were given fewer services; PES advisers prioritized their work The PES provides activation and with UI recipients; and PES staff had placement services for those registered excessive administrative functions and reports to the SWC any change which reduced their capacity to deliver in the status of the claimant. There is job counseling. structured formal cooperation between SWC and PES advisors. Joint case Case study 2: Japan management is provided in the case of clients who need more intensive In Japan, there have been recent efforts ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 21 to improve coordination between the the client in a joint interview. The client PES (“Hello Work”) and prefectures and is registered for work at the PES office municipalities, which are responsible and is referred to vacancies once his/ for paying SA (Duell et al. 2010). Few her job preparation is completed. people who are capable of work qualify for SA because of strict eligibility rules, For those with special difficulties, but recently the number has increased the welfare office may appoint a somewhat, especially of lone parents. “navigator” who works from the Hello Work office to give support on a one- Welfare offices do not oblige to-one basis. Navigators are supposed employable clients to register with to increase their clients’ work focus and the PES, but they do generally advise propose appropriate activities, such as them to. There is no computer interface work experience or public or private with the PES, but to better understand vocational training. Some Hello Work recipients’ job search activity, the offices have set up a support team welfare office sometimes reviews how “corner” or access point in collaboration clients are using PES services and asks with municipal welfare institutions. the PES for information. Such specialist service points are also targeted at other mothers seeking Since 2005, welfare offices and employment (see Box 2.2). Hello Work have developed a special Employment Support Program for In 2009, an OECD review reported there welfare recipients and recipients were some 300 welfare office and Hello of the Child Rearing Allowance Work navigators targeting welfare and (see Box 2.6) that aims to enhance child allowance recipients, still less clients’ independence and self- than one per PES office on average. support capability. First, employment In 2007, 12,422 persons (about 2 consultants or other case workers in percent of the number of non-waged welfare offices identify suitable clients households in receipt of welfare) for the program and approach the were referred to the PES from welfare local Hello Work office for cooperation. offices. It was reported that 6,741 found Next, employment and welfare officers employment, a success rate of 54 develop an individual action plan for percent. 22 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY Box 2.2 Japanese PES for Mothers Seeking Employment In Japan, specialized “centers, salons and corners,” often located in PES offices, target mothers who are bringing up children and seeking employment or who wish to change jobs. Compared with the main Hello Work offices, these centers or corners offer more tailored services, such as information on child care (in collaboration with local government), more intensive jobsearch assistance and career counseling, special seminars, and some training (e.g., IT). They make special efforts to acquire vacancies from companies that offer conditions suitable for mothers (e.g., in terms of work-life balance). Free childcare is provided for visitors. Mothers seeking employment can either register directly at the Mothers’ Hello Work center or with the main Hello Work office. In the latter case, at the first counseling session, the employment officer can refer appropriate clients to the Mothers’ Hello Work center, where jobseekers are assigned to a single counselor throughout the jobsearch period. Counselors aim to place jobseekers within three months after drawing up an Employment Realization Plan. This involves identifying the specific employment barriers faced by each mother, the type of preparatory activities she needs to carry out (e.g., training for job interviews), and enterprises that may offer her suitable work. Case study 3: Finland of long-term benefit recipients. One reason is because LAFOS centers are In Finland, organizational reform established only in densely populated has been targeted at the long-term areas where they may serve several unemployed and others receiving municipalities. In 2010, it was reported minimum income benefits, which are that of 9,149 people completing the delivered through municipalities (Duell, service, about 10 percent were in open Grubb and Singh 2009). Between 2004 employment and nearly 12 percent were and 2007, the PES and municipalities participating in ALMP (EJML 2011). created 39 specialized “Labor Force Service Centers” (LAFOS). Clients are LAFOS centers are jointly run by referred to the LAFOS center by the PES municipalities and local PES offices or the municipality based on a needs and are based on local, rather informal, assessment. Participation can last for cooperation contracts between the two to three years, after which clients partners, and act under management without another outcome return to the jointly defined by them. Organizational PES or municipality. models vary, with the lead managerial position being taken either by the PES The 2007 LAFOS caseload of 23,500 or a municipality or sometimes shared represented about half of the number between them in a rotating system. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 23 The core LAFOS personnel comprise unemployed, to people on long-term PES counselors and municipal sick leave or with disabilities, to those on social workers. In addition, health vocational rehabilitation benefits, and professionals, such as nurses, doctors, to employable SA recipients. The units and psychologists, may be on site or and divisions comprise employees part of multi-professional teams. The from all three of the former services. sizes of the LAFOS centers vary, with the largest offices offering a wide range The reorganization included some of professional services. Clients can 14,000 staff under government control be referred to the full range of ALMP , and 4,000 municipal employees. The but one possible type of placement frontline offices had 6,000-7,000 staff is into subsidized work in social when the network was finalized. enterprises, which provide realistic In 2008, services were provided to work experiences designed to prepare an average of 150,000 unemployment disadvantaged groups for employment benefit, SA, and vocational in the open labor market. rehabilitation recipients and about 100,000 job seekers who were Case study 4: Norway not benefit recipients (e.g., people registered for a potential change of Norway’s reform involved the job). integration of its PES and its social insurance agency to form NAV, the Local NAV offices were established combined Labor and Welfare Service, through agreements between NAV which has now been co-located with at the regional or national level with the delivery of municipal SA benefits the municipalities. These agreements (Duell, Singh and Tergeist 2009). The related to the design and operation of objective was to bring separate frontline the office and the interaction between offices under one roof to create a single the two organizations and could further contact point per client to deal with all determine that, apart from SA, other the needs of the individual, ensuring municipal social services may be that the office is experienced by users provided. This has led to variation in the as a single unit. character of the agreements and in the services provided in local NAV offices. The co-located offices typically have two departments: reception and long-term In practice, local offices are steered by follow-up. The reception department two different “owners” (municipalities offers self-service and limited guidance and the central government). This to jobseekers and to employers with means there is no single chain of jobs to offer. The follow-up department command, staff are on different gives more intensive assistance to the salary scales, and at the time of the 24 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 2009 OECD review, IT systems were review; see Grubb, Singh and Tergeist not integrated to create a joint client 2009). The new approach is primarily database. Preliminary evaluations focused on those claiming benefits and of the merger process showed that the target is to get 75,000 currently long- services became more integrated in term unemployed people back into the that there was increased cooperation workforce and to reduce the average across previous agency borders. NAV time spent on the live register from 21 offices still differed, however, in how months to less than 12 months by the each local area defined the approach end of 2015. to client treatment, with some case workers in some offices handling the A key element of the Pathways to Work whole spectrum of clients while other approach is the transformation of social offices basically maintained the pre- welfare offices and the new service NAV division of work. was officially launched as Intreo in four offices in October 2012; a full network of Case study 5: Ireland 70 offices is to be established by the end of 2014. The service delivery approach In 2011, Ireland’s newly created combines a number of key elements. Department of Social Protection (DSP) These include the development of a was given responsibility for developing personal progression plan and a “social an integrated one-stop system to contract” whereby clients commit to administer all working-age benefits and engage with the DSP’s employment employment services. This involved services as a pre-condition for the the integration of some 1,700 PES staff receipt of benefit. In addition to job with existing benefit payments staff search and availability for work, clients and the development of systems and will be required to attend meetings and procedures to deliver the new service. participate in employment programs. The system is underpinned by new In February 2012, the Irish government sanctions which mean that failure to launched its wider Pathways to Work engage can lead to a reduction in, strategy, which combines reforms and ultimately a cessation of, income to the benefit system, employment support payments. programs, and services for jobseekers and employers (Government of On entry to the system, unemployed Ireland 2012). The strategy aims to people are asked to complete a prevent increased unemployment from profile questionnaire so that their case becoming entrenched in Ireland by worker can assess their “probability of transforming the comparatively passive exit” from unemployment during the system (as revealed in an OECD country subsequent 12 months. The results are ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 25 used to determine the level and timing of operating model; employment support to be given. Clients • Management arrangements and selected for support should benefit supervision of operations; from at least four intensive case review • Personnel to be allocated to deliver interviews over 18 months. For most the services; people, the first appointment consists • Budget and monitoring of of a group engagement where options expenditure; and and obligations are explained to groups • Any services to be outsourced or of up to 20 jobseekers with similar purchased from external service characteristics. The group engagement providers. is followed by more intensive “one- on-one” counseling interviews. If In this respect, the steps taken by MoFSP jobseekers are still on the live register and İŞKUR after the development at 18 months, they are mandated to of the Action Plan on Linking Social attend a personal development and/or Support System to Employment and employment program. its Activation are to the point. To direct all SA applicants to İŞKUR’s database Summary and implications of case and provide them active labor force studies services, MoFSP and İŞKUR signed a protocol that covers the operational The client groups, budgets, benefit model, implementation arrangements, entitlements, and institutions in and allocated personnel. With the these five countries differ from those efficient use of budget and monitoring involved in the Turkish reforms, but the of expenditures, more effective findings illustrate some of the policy coordination of SA and PES can be design and implementation elements attained. that can facilitate more effective coordination of SA and PES services, A further key factor has been the even when institutions are not as development of shared information well resourced. The most significant systems. A particular challenge for factor concerns the development of implementing reforms has been national priorities and local agreements creating effective linkages between outlining service protocols, working the separate established (legacy) methods, and respective organizational data systems of PES and SA offices, responsibilities. Such protocols especially in coordinating information typically contain agreement on the flows and the “handover” of service following elements: users between the respective parts of the delivery chain. Some countries, • The clients to be served and the such as Great Britain, merged the 26 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY existing benefit and employment income programs for unemployed service delivery organizations and working-age people and their families. created a new client management IT The terminology used to describe system. Other countries, such as the such income-transfer programs Netherlands, steered more integrated varies across countries, but they are activity between the PES, municipalities, often characterized as SA benefits and social insurance agency through (Immervoll 2009). Such benefits are data sharing protocols, service level typically means-tested and recipients agreements, and joint performance qualify only if they satisfy reasonably indicators. Turkey’s integrated SAIS, stringent tests of the income and assets which is interlinked with the databases of individuals and their households. of all other government agencies, will Such benefits may be generally be a very useful piece of infrastructure available to poor people, including to further enhance the linkages between long-term unemployed people who PES and SA services. have exhausted their UI benefits, or they may be categorical and targeted Such developments have been at specific population groups, such as facilitated by the existence or lone parents, widows, or families with development of reliable registers of school-age children. In most OECD clients, structured databases, and countries, SA benefits comprise regular improvements in IT infrastructure cash payments supplemented by in- that allow for the analysis of delivery kind support. In many countries, SA and performance based on sound benefit entitlement continues for as administrative data. At its best, long as the person shows that s/he is in investment in such information need, though other countries put a time systems has enabled service providers limit on eligibility.9 to coordinate separate administrative data on clients, establish eligibility The primary purpose of SA benefits and referral mechanisms, track is to alleviate poverty but a critical interventions, and monitor subsequent design issue concerns the extent to progress. which such income transfers may reduce work incentives and encourage 2.3.3 Activating SA Recipients: welfare dependency. This is a particular Work-Focused Interviews, concern in those countries where SA Incentives, and Sanctions benefits are relatively generous and without time limits. There are several Most OECD countries have fairly ways in which policy makers can generous means-tested SA or minimum design their benefit entitlements to 9- In the U.S., for example, individual states now impose work requirements and lifetime entitlement limits of between two and five years on those working-age adults who are eligible for cash assistance from “Temporary Assistance for Needy Families” (most of whom are single parents with dependent children). ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 27 minimize such work disincentives and of job search activity to be undertaken, make their systems pro-employment. the types of employment programs in Some important factors concern which they must participate, and when the employment-related behavioral exactly such participation should begin requirements that claimants are (OECD 2007). expected to meet; the avoidance of unemployment and poverty In some OECD countries, registration traps; and the use of sanctions and with employment services is now incentives that encourage claimants required before any cash benefits or in- to take up employment opportunities kind assistance are provided. In some or participate in services that would U.S. states, for example, “Temporary enable them to do so. Assistance for Needy Families” (TANF) applicants must show they have actively Behavioral requirements and looked for employment before they are benefit entitlement considered eligible to apply for welfare benefits. Unemployment-related benefits, whether insurance-based or means- In many countries, job search and tested, have traditionally been subject availability for work rules have been to a series of rules or conditions that extended and adapted to cover a require recipients to be available wider range of claimants, including for and actively seeking work, to be those claiming SA benefits. In the U.K., capable of doing the work for which Germany, and the U.S., for example, they say they are available, and to employment requirements now apply be willing to accept offers of suitable to single parents, spouses with child- employment. In OECD countries, such care responsibility, and employable unemployed claimants are typically claimants with work-limiting health required to register with the PES, report or disability conditions. Since 2005 on their job search, and attend periodic in Germany, a person claiming the interviews. The PES is expected to “basic allowance”has been considered assist individuals as they look for work, capable of working if s/he can work monitor their job search, and ensure for at least three hours per day under that claimants meet their obligations. normal labor market conditions or is Recent activation reforms targeted at only prevented from doing so by health unemployed benefit claimants have problems for a “foreseeable period.” In typically widened who is defined as most systems, there is often flexibility capable of work, the definition of what within legal regulations and such constitutes suitable and appropriate claimants may only be required to seek work for the person to seek, the level employment that fits with their child- 28 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY care or other caring responsibilities or In the U.S., mandatory job search and that matches their reduced capacity registration with employment services to work. Nevertheless, such claimants have been further extended and now are required to register with the PES apply to the primary means-tested cash and to show they are actively seeking and in-kind benefits available for poor employment within their agreed families, including food stamps (see constraints. Box 2.3). Box 2.3 Employment and Work Requirements Applied to SA Benefits in the U.S. The main SA benefits in the U.S. comprise ‘Temporary Assistance for Needy Families” (TANF) and the Supplemental Nutrition Assistance Program (SNAP), previously known as “Food Stamps.” TANF is a work-based cash assistance program targeted at poor families with dependent children. TANF budgets are devolved to individual U.S. states, which have much discretion in the local design of programs, but federal “work requirements” apply nationally and stipulate that: • Recipients (with few exceptions) must work as soon as they are job-ready or no later than two years after starting assistance. • Single parents are required to participate in work activities for at least 30 hours per week. Two- parent families must participate in work activities for 35 or 55 hours a week, depending upon circumstances. • Failure to participate in work requirements can result in a reduction or termination of benefits to the family. • States cannot penalize single parents with a child under six for failing to meet work requirements if they cannot find adequate child care.In addition, those with children under six are only required to complete a total of 20 hours of work activity per week. TANF rules stipulate that states must make an initial assessment of a recipient’s skills and may develop personal responsibility plans for each recipient to identify the education, training, and job placement services needed to move him/her into the workforce. Poor able-bodied adults without children are not eligible for TANF but may be eligible for SNAP benefits. However, work requirements and time limits also apply. With some exceptions, able- bodied adults between 16 and 60 years old claiming SNAP must register for work, accept suitable employment, and take part in an employment and training program to which they are referred by the local office. Failure to comply with these requirements can result in disqualification from the program. Additionally, claimants who do not have any dependent children can get SNAP benefits for only three months in a 36-month period if they do not work or participate in a workfare or employment and training program other than job search. This requirement is waived in some locations. Sources: http://www.acf.hhs.gov/programs/ofa/programs/tanf/about, accessed 29 April, 2013 http://www.fns.usda.gov/snap/applicant_recipients/eligibility.htm, accessed 29 April, 2013 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 29 The transition from voluntary to integrate employment assistance programs to employment with other related services including, requirements – the case of for example, access to supplementary mandatory work-focused provision, such as child care for lone interviews parents or “condition management programs” for people with health The extension of job search or disability problems.10 Such a requirements to SA benefits in different developmental process characterized countries has taken some time. Of reforms in the U.K. and the design and relevance to Turkish policy makers delivery mechanisms that contributed is that in several countries, such to the creation of an employment- activation reforms often commenced focused delivery system there may be with the introduction and testing of of relevance in Turkey. voluntary employment programs targeted at clearly defined groups of In the first phase of the British reform SA recipients. in the late 1990s, voluntary, relatively small-scale employment programs were This initial voluntary phase enabled introduced for lone parents, spouses, policy makers to build system and and people receiving disability benefits. delivery capacity and to accumulate Evaluations found that employment knowledge about what works in impacts in each program were high for assisting such claimants and about those who participated but because the relative financial costs and benefits the interventions were voluntary, of such interventions. For example, they did not reach the majority of in several systems, policy makers claimants. Nevertheless, during this were able to test new referral and period, experimentation with program attendance procedures concurrent design and funding helped build the with the development of effective operational knowledge and capacity of programs. As activation requirements the PES system to cater to more diverse were extended to more disadvantaged client groups, especially single parents clients, this also involved testing how (see Box 2.4). 10- Condition management programs are not intended to act as curative medical treatments but involve rehabilitation support designed to enable an individual to return to work and manage his/her health conditions. They were developed in the U.K. to complement job placement services targeted at clients with significant health barriers (Lindsay and Dutton 2013). 30 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY Box 2.4 The British New Deal for Lone Parents The NDLP was a voluntary program targeted at lone parents on means-tested income support whose youngest child was under 16. Its aim was to provide the advice and practical help lone parent claimants needed to enter or return to employment. The program was launched in eight areas as a prototype in July and August 1997, and rolled out nationally in April 1998to lone parents making new and repeat claims for SA benefits. The program was extended to all such lone parents in October 1998. At an initial interview, lone parents who wished to participate developed an individual action plan with their PES personal adviser (PA). The PA provided an integrated service, covering job search, help finding child care, advice on benefits, and help with claiming benefits. Participating lone parents were also eligible from the outset for the full range of programs for the unemployed administered by the PES and support for participating lone parents extended beyond the benefit claim period into the early weeks of employment. Provision within NDLP was developed and augmented continuously with improvements, including, for example, the introduction of basic skills screening, the introduction of a self- employment option, an increased training allowance and additional support with child-care payments for the first 12 months of work. The PA also had flexibility to provide small cash payments that helped overcome barriers arising at the transition from benefit to work. Evidence suggests that the advice, guidance, and support provided by NDLP was effective. Evaluations conducted by Dolton, Azevedo and Smith (2006) and Lessofet al. (2003) found that the proportion of lone parents who exited social security benefits within nine months of participation was 20-26 percentage points higher for those who participated in NDLP than for non-participants. The estimates for employment entry differed, however. Whereas the 2003 study estimated that NDLP increased the proportion of lone parents entering employment by 24 percentage points within nine months of participation, the 2006 study estimated it to be10 percentage points. Despite the variation in results, largely attributable to methodological differences, the positive outcomes recorded in both studies underpinned the confidence of policy makers to subsequently require other lone parents to engage with such services. After the voluntary programs had legislation introduced compulsory been implemented, the next phase “work-focused interviews” (WFIs). These in British welfare reform concerned interviews did not require lone parents the introduction of an intermediate or disabled claimants to take a job but behavioral requirement in the Welfare did require them to attend an interview Reform and Pensions Act (1999). The with the PES and subsequently to ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 31 engage in reasonable steps that would In countries like Australia and the U.S., improve their employability (see Box as well as the U.K., it was only 2.5). Initially, WFIs applied only at the after several years of testing such start of a benefit claim but Ministers intermediate activation requirements, and developing related pro- subsequently extended the frequency employment services, that formal job with which high-priority groups of search and availability for work rules previously inactive claimants had to were extended to lone parents with attend these interviews and participate school-age children and wider groups in related activities. of SA claimants. Box 2.5 Work-Focused Interviews (WFI) and Mandatory Work Preparation in the U.K. In the U.K., all working-age claimants are required to attend a face-to-face WFI with a Jobcentre Plus Personal Adviser (PA) at the start of their benefit claim. The PA has discretion to defer the WFI and there are some limited exemptions for prescribed groups, such as those with terminal illness. At the WFI, a claimant must be prepared to answer questions (if asked) about such matters as: • Educational qualifications/vocational training • Employment history and employment-related skills • Any current paid/unpaid employment • Caring responsibilities • Any medical condition that puts the person at a disadvantage in getting a job After the first WFI, different groups of claimants are subject to various interview requirements and further WFIs develop from an initial “one off” engagement at the start of a benefit claim into a flexible activation instrument targeted at lone parents, spouses, and people on disability benefits. Since October 2005, most claimants who attend a WFI are also required to complete an action plan agreed upon with a PA that might include referral to an employment program. PAs now have discretion to encourage and require such claimants to participate in an unspecified range of work-related activities but may not require a person to apply for a job, undertake work, or undergo medical treatment. This work preparation regime is also underpinned by a differentiated sanctions system. It is not as harsh as the one that applies to the unemployed, however, and the penalties involved reflect the nature of the rule breached, the conditionality group of the claimant, and any hardship that might be caused to children. 32 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY Work incentives: making work pay work benefit entitlements have been through earnings disregards and designed to encourage part-time or tax credits irregular work because they contribute to self-sufficiency and provide a route In many OECD countries, there is a to a better job. The design issue for concern that the availability of relatively Turkey is whether the SA system generous out-of-work benefits for encourages such employment and unemployed working-age adults if so, how payments from any work undermines employment incentives. undertaken should be offset against There has been particular concern benefit entitlement. about the unemployment and poverty traps that may be created when the In many of the EU8 countries (the most combination of work costs, income recent EU members), where SA levels tax, social insurance contributions, and are far lower than in other EU countries, benefits withdrawal mean claimants policy makers have gradually created have no financial incentive to work or “earnings disregard” rules and/or to increase their earnings in the formal introduced benefit tapers and in-work labor market. Variations in wage levels benefits or tax credits that can ease the across regions may also create such transition to employment and ensure traps in particular geographic areas. that claimants are better off in work. Estonia, for example, has introduced Where cash transfers are of low value, in-work tax credits for families with as in Turkey, their impact on work children, and Hungary and Slovakia incentives is usually negligible but have benefits that taper off gradually as there is a design issue concerning the claimants enter employment (Ringold treatment of earnings of SA benefit and Kasek 2007). households. A “dollar-for-dollar” reduction rule may simply encourage The design of such in-work benefits and dependency, undeclared earnings, and tax credits have been found to have a informal employment. Disincentives powerful impact on employment rates, may also be compounded if such especially in the U.S. and the U.K. (Meyer reduction rules mean that a person 2007; Gregg, Harkness and Smith might lose entitlement to other benefits. 2009). One particular example that This may be relevant in Turkey, where may be of relevance to Turkish policy households may receive more than one makers concerns the design and impact means-tested benefit. of Japan’s “Child Rearing Allowance,” associated with the highest rate of lone In many OECD systems, special parent employment amongst OECD “earnings disregard rules” and in- countries (see Box 2.6). ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 33 Box 2.6 Lone Parent Employment and Child-Rearing Allowance in Japan The exceptionally high employment rate of Japanese lone parents, at 85 percent, is related to their differential access to benefits in and out of work. Estimates vary, but there are at least 600,000 and possibly up to one million single-mother households in Japan, of which only about 93,000 received SA each month in 2006. In addition to the social stigma mothers face in claiming the benefit, municipal welfare offices evaluate the lone parent’s capacity to work rigorously and often suggest that other family members support them. In contrast, some 956,000 single-mother households in Japan received the in-work “Child Rearing Allowance” in 2006. This benefit can only be claimed by parents in paid work and it is paid until the youngest child is aged 18. The benefit amount is set well below subsistence level, which enables the benefit withdrawal rate in relation to earnings to be set at a low level. This creates a strong financial incentive to work long hours, a benefit absent in most other OECD countries. When combined with preferential access to daycare centers at heavily subsidized rates for mothers on low incomes, the allowance makes it possible even for mothers with rather low earnings capacity to achieve net incomes similar to – although probably still lower than, in some cases – the out-of-work SA rates. These factors facilitate the high employment rate of single mothers and help explain why a large proportion of this group works full-time. Unfortunately, the high lone parent employment rate does not translate into lower levels of child poverty and many single mothers report that their lives, working full-time with still relatively low net incomes, are difficult (Duell et al. 2010). Participation incentives and PES registration and participation in sanctions programs can be reinforced and made more attractive through the use of cash The effective referral and attachment or in-kind incentives that help meet the of SA claimants to the PES and direct costs of engagement. In several subsequent engagement with and countries, frontline case workers are participation in services can be able to provide discretionary support enhanced through the use of incentives, that might include transportation sanctions, or a combination of both. subsidies or vouchers for attending the PES, attending job interviews, A voluntary or mandatory approach to assistance with child-care expenses 34 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY and provision, and even assistance with sanctions comprise financial penalties work clothes and tools or equipment. that typically escalate in severity Many PES services also provide modest if a service user continues to miss participation or activity allowances or appointments or fails to undertake bonuses paid in addition to existing a required activity (see Box 2.7). benefits for targeted individuals who Compliance activities might start with engage in and complete longer duration a warning, as in Japan or in some employment or training programs. As cases in Australia and the U.K. Failure mentioned in section 4, MoFSP has to attend scheduled appointments with already started piloting these kinds of the benefit administration often results incentives, the results of which will be in the suspension of benefits until the taken into consideration in nationwide client complies and a failure to attend expansion. a mandatory employment program is likely to result in a fixed period of non- In many OECD countries, mandatory payment or be construed as making participation in services and related an individual unavailable for work and sanctions comprise an important therefore ineligible for benefits. component of activation systems. As Immervoll (2009) observes, imposing Where the sanction is designed to more demanding behavioral conditions change behavior, financial penalties for benefit receipt makes work relatively may often be suspended or withdrawn more attractive and limits opportunities if the individual attends the interview, for benefit claims that might be reengages with the service, and/or considered “undeserving” (e.g., from undertakes specified actions. In some those with incomes from undeclared U.S. states’ TANF systems, individuals employment or a strong preference may be required to undertake for leisure). At the same time, work- employment-related activities before related behavioral requirements they are allowed to qualify or re-qualify seek to improve employability. Both for cash payments, especially for effects may help reduce the number second or third penalties. of beneficiaries, and this effect can be further strengthened by providing When sanctions are imposed, there job search assistance and other are often safeguards designed to stop employment-oriented support. A family incomes from falling below a possible negative effect is that stringent given subsistence level or to minimize requirements may deter poor families the impact of financial sanctions on eligible for assistance from claiming it. particular groups, most commonly by ameliorating their impact on families In most systems, activation-related with children or on otherwise vulnerable ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 35 clients (such as those with mental health of sanctions may be less important issues).11 than the effective communication of their existence and the speed of their It is difficult to assess the contribution implementation. sanctions make to overall compliance or employment transitions because At this time in Turkey, where benefit many people may respond to the risk coverage and generosity are limited or possibility of sanctions without and many employment programs ever experiencing them. There is, and services are over subscribed however, evidence from a range of by volunteers, the development of OECD countries that welfare exit and a regular sanctions regime may be employment transition rates increase inappropriate. However, in time it might prior to clients being required to attend be necessary to test different ways of interviews and engage in mandatory imposing modest financial penalties to work programs or more intensive ensure that employable SA claimants assistance, so the threat of sanctions do attend formal interviews with may lead some to leave welfare and İŞKUR. Best OECD practice suggests move into work before they ever such modest requirements might best experience a sanction (Finn and be tested first in the Turkish UI system, Casebourne 2012; Griggs and Evans one of the few in OECD countries where 2010). such sanctions do not currently apply. Experimental evaluations of welfare reforms in the U.S. have found that Targeting activation requirements and states that communicated, monitored, sanctions at SA claimants who face and enforced sanctions secured significant employment barriers, such as greater compliance with employment lone parents or those with health issues requirements and incurred lower or disabilities, needs careful design and welfare expenditure than those that consideration, especially with regard did not (Kauff et al. 2007; Pavetti, Derr to when such requirements should be and Hesketh 2003). Studies have found imposed. Equally, the sanction regime that enforcing work requirements is to underpin such requirements needs important, but it is not clear whether calibration to find the best balance complete or partial termination of between properly enforcing obligations benefits is more effective. Some and encouraging engagement with the findings suggest that the severity services available. 11- It should be noted that in nearly all OECD countries, benefit entitlements have a legal, and sometimes constitutional, basis, which ensures fairness but adds to the administrative complexity of benefit sanction regimes. Claimants who feel they have been treated unfairly are normally entitled to ask for their case to be reviewed by an independent person within the public agency imposing the sanction. If claimants remain dissatisfied, they are often then entitled to appeal to administrative tribunals. Where legal entitlements have been breached, claimants may also be able to seek redress through the formal judicial system or have the right to take certain cases to an independent ombudsman. 36 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY Box 2.7 Benefit Sanctions for SA Claimants in Australia, Germany, and New Zealand Australia introduced a revised compliance and sanctions regime in 2009. The improved liaison between Centrelink, the public benefits agency, and contracted employment service providers and sanctions was designed to achieve or restore engagement with services. The system comprises “connection failures,” which apply to missed appointments, and a “no show, no pay” sanction that applies to non-engagement with activities, whereby a jobseeker who fails to attend without reasonable cause loses a day’s benefit payment for every day missed (Disney, Buduls and Grant 2010). Persistent non-compliance results in a “comprehensive case assessment” and if a recipient is found to have been “‘willfully and persistently” non-compliant or has unreasonably refused a job, s/he may be subject to an eight-week non-payment penalty. The non-payment period is not applied to some “vulnerable” jobseekers and can end earlier if a claimant undertakes a “compliance activity” (generally 25 hours per week for eight weeks in a “Work for the Dole” project). In Germany, welfare claimants are informed about the possibility of sanctions in their integration contract and each time they are assigned to a program. If a case manager, employed by a municipality or the PES, detects non-compliance, s/he may impose a sanction, which usually lasts for three months. For minor non-compliance, such as a failure to report to the welfare agency, benefits are cut by 10 percent. More severe infringements (lack of job search effort, refusal to accept a suitable job offer, refusal to participate in a program) lead to a benefit reduction of 30 percent. Where severe infringements are repeated within a year, a second sanction of a 60 percent reduction or a third sanction of a 100 percent reduction can be imposed. Sanctions are stricter for those aged under 25 (Schneider 2008). New Zealand implemented a revised sanction regime in 2010 alongside the extension of work obligations (of up to 20 hours per week) to lone parents and those on sickness benefit. The new regime introduced a 50 percent reduction in the rate of main benefit, followed by a 100 percent suspension and cancellation for second and third failures, respectively. There is a safety net for lone parents and couples with dependent children, who remain eligible for 50 percent of benefits for second and third failures. Those sanctioned are given “opportunities to comply and restore benefit payment” (MSD 2010: p.1). ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 37 2.3.4 Eligibility for Employment payment agencies now employ more Services and Assessing focused profiling instruments using Employability administrative data and/or answers to a short questionnaire to gather In most OECD countries, access to more additional information about clients’ expensive publicly funded employment characteristics. The results are then services and ALMP is targeted at priority used to assign a client to a particular groups (with access to wider job- category based upon a regression broking and information for jobseekers model, which predicts the client’s and employers increasingly available probability of becoming long-term only through self-service channels). unemployed as a function of his/her Eligibility rules will typically target characteristics. The categories range services and programs at people who from easiest-to-place to hardest-to- are receiving different cash benefits and place, with different categories referred may be directed at particular categories to different employment services and of service users who are perceived programs. to face common barriers, such as disadvantaged youth, single parents, Such profiling methods are used married women returning to the extensively in Europe, Australia, labor market, people with disabilities, and the U.S. to score and weight the minority ethnic groups, and so on. It barriers facing individual service users is important that policy makers define and to assign them to unsupported program eligibility rules in ways that job search, particular programs, and/ reflect the resources available and the or other forms of support (Konle-Seidl capacity of the system to deliver the 2011). In addition, there is a wide services envisaged. range of other diagnostic and usually more time-intensive assessment tools Simple eligibility rules, determined used to assess skills and employability by factors such as age, education, that help shape the services given to skills, work experience, or caring users, especially those with significant responsibilities, have obvious employment barriers (see Annex 1 for advantages in terms of their ease of a brief review of such profiling and operation and equality of treatment. assessment tools). Such criteria are, however, a “rather blunt instrument” and may be “inefficient In the Turkish context, where there are in terms of achieving a good match widely differing groups of poor people between individual client needs and receiving various SA benefits, it is provision of support” (Hasluck 2004: p. important that policy makers establish 5). In many systems, the PES or benefit clear eligibility criteria about those 38 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY claimants they consider will benefit played by frontline case workers and from İŞKUR services and the priority employment counselors in securing they should be given in the interview these core objectives. A primary and referral process. It should be instrument for such advisers has been noted that both MoFSP and İŞKUR are the development and use of action aware of this need and are working plans or reintegration agreements, in collaboration to define eligibility signed by service users, which spell criteria. The eligibility criteria chosen out responsibilities and rights and the should be such that the target priority agreed employment-related steps to be groups can be easily identified by both taken. SA and/or İŞKUR information systems and case workers, and that they are In OECD countries, there is significant commensurate with the resources variation in the resources and status available. Choices could then be of the frontline advisers and case made about the sequence in which workers who deliver employment individuals within these groups could services, ranging from the ways in be referred to İŞKUR services. This which they are employed, their job would help ensure that İŞKUR services titles, and the autonomy they have are not overwhelmed by inappropriate with their caseloads to the frequency referrals and that administrative and and nature of contact they have with service delivery capacity is developed unemployed people.12 Despite these in tandem with the roll-out of activation differences, such case workers have an requirements. In this way, resources important dual role in assisting clients can be allocated more effectively. with their search for employment 2.3.5 The Role of Frontline Advisers while communicating and enforcing and Employment Counselors the requirements clients must meet to remain eligible for benefits. Two core objectives of activation strategies are to encourage benefit In a review of the role of such claimants to become more active in counselors in the PES in European their efforts to find work and to help countries, Sienkiewicz (2012) identified reduce or better manage barriers core generic tasks associated with that diminish their employability and this role and reviewed how such staff capacity to take jobs. Over time, OECD are recruited and the forms of training and other comparative evidence available in different countries (see reviews have pointed to the key role Table 2.2). 12- In the U.K., for example, they are referred to as “personal advisers”; in Holland, they are “reintegration coaches”; and in Australia, they are called “recruitment consultants.” ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 39 TABLE 2.2 Common tasks of employment counselors delivering services for the unemployed Most common tasks Common tasks Less common tasks • Information about ALMP measures • Providing • Registering the telephone/online unemployed • Initial interviewing of jobseekers job information and (formal registration • Assessment of jobseekers’ assistance process) strengths/weaknesses and the • Group sessions for • Use of specialized adequate need for support jobseekers IT and other tools • Providing information on available for assessment • Providing and up- job offers keeping on site and • Guidance on • Job search assistance online information career planning/ • Preparation of individual action on available development plans training/ALMP • Planning of ALMP • Referral to appropriate ALMP • Monitoring of • Networking with measure/ provider implementation of other stakeholders, individual action • Maintaining contact with registered providers plans jobseekers through face-to-face contact, telephone, and/or email • Monitoring in case of referral to • Providing and maintaining onsite another provider/ and online information on available training program job openings and ALMP availability Studies of employment counselors and self-confidence. Such staff must show wide variation in the level of also be able to plan the range of support specialization, caseloads, and job or assistance needed, link clients roles of individual case workers, but with necessary external programs or the combined findings indicate that services, and then monitor clients’ such staff should have the expertise progress through an agreed series to assess clients’ circumstances and of steps or services, through to their employment constraints, develop placement into employment. action plans, check on job searches, refer clients to vacancies, raise In many countries, there is now a awareness of job search techniques, much closer working relationship and help improve clients’ motivation between PES counselors and social 40 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY workers. In countries where SA suggest caseload size is an important recipients are expected to prepare variable in the delivery of effective for and/or look for work (as in the services and that relatively smaller earlier case studies), social workers caseloads enable counselors to undertake an initial employability deliver more effective job-broking assessment and refer clients to the and activation services (see, for PES and liaise with employment example, Hainmueller et al. 2009). The counselors. In other systems, as in the effectiveness of advisers and the size Netherlands, municipal social workers of caseloads are shaped by the clients are themselves expected to perform with which they are expected to work, employment-focused roles and directly the tasks expected of them, and the help their clients prepare for and enter frequency with which they are expected employment, including referral to ALMP . to meet clients. In many activation In many U.S. states, social services systems, advisers are expected to case workers, whose primary function interview, assess, and agree on an action had previously been to assess benefit plan shortly after clients are referred eligibility, had to adapt to delivering to them and to subsequently follow employment-focused services as their this up with interviews that can vary jobs were redesigned to enable them in frequency from fortnightly to every to deliver welfare-to-work programs. three months. Findings from a number In some countries, it has been difficult of robust studies show that caseloads for managers to build a common in activation programs for hard-to-help service culture and overcome divisions groups in several OECD countries are in between the approach of social workers the range of 80 to 120 participants per and PES counselors. adviser, though the range can vary from 40 to 300, sometimes within the same Findings from evaluation studies country (see Table 2.3). ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 41 TABLE 2.3 Employment adviser caseloads in selected OECD countries Caseload per adviser % of adviser time Country Program (approx.) spent with clients U.K. (2008) Contracted employment 80 60% services Australia (2008) Job Network 100 45% U.K. (2012) Work Programme 117 N/A Australia (2012) Job Services Australia 114 N/A Netherlands (2008) Public, private, and non- 81 50% profit providers Germany (2009) Federal Employment 100 N/A Agency Ireland FAS (2006) Public employment 200 N/A service Ireland LES (2012) Local employment 211 N/A service Other impact evaluations give further administration and employment insight into the particular effects of assistance were given separately, and different case management styles used integrated case management, where by advisers in improving employment both functions were brought together outcomes and the frontline practices (Scrivener et al. 2001). The results that have contributed to their found that integrated case managers effectiveness. provided more personalized attention, engaged more people in welfare- U.S. evaluations of welfare-to-work to-work activities, and more closely programs that were largely targeted monitored participation in program at single parents investigated the activities. Both approaches reduced differential impacts of traditional welfare receipt and payments, but the case management, where benefit effects of the integrated program were 42 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY somewhat larger. The authors tried to establish whether those unemployed registered with a In a subsequent analysis, summarizing specific type of employment office and findings from experimental studies on advised by case managers with specific service strategies in 59 different welfare attributes had higher or lower chances offices across the U.S., researchers of finding a job than those registered reported higher employment and with other office types and followed earnings impacts in those offices where by other types of case managers. They case managers delivered a personalized found job-finding probabilities to be service and placed an emphasis on positively correlated with: quick job entry. There was also some evidence that high caseloads reduced • Good staff relationships with effectiveness (Bloom, Hill and Riccio employers, particularly knowledge 2003). Another experimental study of employer needs and careful use in the U.S. found that clients’ average of direct referrals; rapid reaction earnings were higher over a two-year to vacancies; and careful pre- period in offices that primarily used selection; both unified case management and • The extent of cooperation with employed a specialist staff member who private placement agencies; sourced job opportunities specifically • “Tough” rather than more cooperative for welfare clients (Hill 2005). attitudes of case workers towards their clients; Switzerland, by contrast, is one of the • The use of work-first strategies that few OECD countries where there have give priority to job placement over been high-quality evaluations into the training measures; and relative performance of different PES • The organizational separation of office processes and adviser placement counseling and sanctions. strategies, rather than evaluations of particular programs. Frölichet al. (2007) Egger and Lenz (2006a and 2006b) and Behncke, Frölich and Lechner (2007) analyzed the relationship between carried out a micro-economic analysis organizational and incentive structures of the influences of employment office and office leadership types on the one characteristics and counselor strategies hand, and performance results on the on the re-integration probabilities of other, searching for success factors jobseekers within 24 to 36 months responsible for rapid and durable after the beginning of job search. The jobseeker integration. Their study study followed the careers of all new identified four major success factors: jobseekers registered in 2003, and included a standardized survey of all • The early activation of registering job counselors and office managers. jobseekers, with a rapid start of ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 43 the re-integration process and to complement existing PES services, strong guidance by competent case in particular by utilizing the skills and workers; capacities of specialist providers • Personal contacts with employers by including, for example, organizations all job counselors; working with lone parents, • Recruitment of motivated and highly disadvantaged youth, or disabled trained personnel; and people, or those delivering specific skills • Good staff/client ratios. training programs. Public purchasers also may use subcontracting to 2.3.6 Contracting with External increase capacity to meet the needs of Providers for the Delivery of new client groups and/or respond to Employment Services increased cyclical demand, as witnessed recently when employment services in In Turkey, care must be taken not many OECD countries were expanded to overload İŞKUR with new client quickly in response to higher levels groups when extending coverage to of unemployment. Another reason SA recipients. Such a surge in demand to contract out services is to increase may weaken İŞKUR’s capacity to competition, either by requiring deliver its normal services and at the external providers of programs same time, the new client groups are to compete for contracts or, less likely to have circumstances different commonly, by requiring the PES itself from the more employable clients with to compete with external providers for which İŞKUR traditionally works, and the delivery of case management and may need access to different forms job search assistance services. of employment-focused provision. In these circumstances, it is worthwhile In individual OECD countries, the considering how to develop commissioning and contracting of appropriate delivery capacity by employment services is typically contracting with external providers, as complex and, in many countries, small- İŞKUR now does with its conventional scale, with a wide variety of procurement training programs. The use of external practices (Finn 2011). Contracted out employment services providers is activities typically include the delivery already a trend in many of the other of conventional labor market programs OECD countries that have introduced and more intensive forms of support wider activation and PES service targeted at disadvantaged groups, delivery reforms. including special programs for those with disabilities. Australia is the only Ministries, the PES, or other public OECD country to wholly outsource the agencies may use contracted services delivery of publicly funded employment for several reasons. They may do so services. 44 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY While systematic information is delivery costs, stimulate innovation in unavailable, it appears that in many service delivery, and give clients access OECD countries, purchasers procure to skilled staff and services unavailable specific services, typically specifying in the public sector. Such developments the detailed design of the particular also may spur improved performance employment intervention or training in the PES, both through competition course to be delivered. The public body and best practice transfer. also determines the price to be paid and the terms of the contract. Contracts 2.3.7 Making the Turkish SA System are often shortterm, with durations of More Pro-employment one year or less. Payment systems also vary from recurrent public funding to The findings from welfare reforms grants to staged payments or fees paid in other OECD countries provide a for services delivered. rich comparative source of ideas and evidence but the implications of the In several other countries, however, policy insights assessed in this report there has been extensive reform of have to be considered in the context procurement practices (Finn 2011). In of the distinctive social policy and these countries, procurement is now implementation challenges with which characterized by competitive tendering, Turkey is grappling. the selection of employment service providers on the basis of price and An immediate issue for the Turkish quality, and payment of providers based system concerns the development of on their performance in delivering simple eligibility criteria that will enable services and securing employment either İŞKUR counselors or the officials outcomes. delivering SA benefits to determine who should be regarded as employable Findings from evaluations and reviews and at which point such clients should of these contracting systems show be contacted and interviewed. There mixed results, but suggest that should be feedback to the SA system on contracting arrangements allow public the take-up of such referrals to ensure authorities to expand or reduce PES that registration has taken place and to service delivery capacity without consider any impact of non-registration assuming the long-term commitments or non-attendance on benefit eligibility. involved in public sector employment – although some stability in contracting The mandate of İŞKUR JVCs is clear, arrangements is desirable to build up but their workload is remarkable. private sector capacity. The introduction Theoretically, an equal number of SA of competition and tendering for beneficiaries are placed in each JVC’s contracts can also potentially reduce portfolio in accordance with their ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 45 place of residence, and the JVCs and service delivery capacity. They are supposed to perform at least 10 could be extended later, or not at all, to interviews with these beneficiaries other areas or to people receiving SA (in addition to their current work benefits prior to the change in policy. load). In reality, however, the number Alternatively, activation requirements of beneficiaries mapped to a JVC could be targeted at points of particular differs from one province/district to change, for example, when the youngest another. İŞKUR counselors expect clear child in a family starts school or when guidance about which SA clients they a claim for a particular assistance should register and what services (i.e., payment is reviewed. occupational guidance, transfer to VET programs, job placement, and labor The capacity of the Turkish PES adjustment services) they should offer. system to work with SA clients could As part of the registration process, be developed in a number of ways. İŞKUR counselors could develop a Currently, İŞKUR simply offers SA “Back to Work Plan” and then assist clients access to its existing services. clients with finding employment or Another option is to reallocate some of accessing appropriate ALMP . It may the resources now invested in relatively be feasible for İŞKUR to test the use high-cost training programs to support of specialist counselors who would targeted core job search assistance provide more intensive support to a and counseling services. Such job case load of targeted clients, such as search support might comprise advice younger female parents, and focus and guidance on how to seek and on placing them in female- and child- apply for jobs with facilities to enable friendly workplaces. participants to actively seek a number of vacancies during the time they In developing and further improving are supported. The key competence the system, Turkish policy makers must required would be for counselors to take care to deal with İŞKUR’s current understand the employment needs of problem of overloading JVCs with new the client group, to broker access to client groups, risking a poorly targeted complementary services, such as child diversion of resources from more care, and to develop links with likely employable to less employable clients. employers. Such provision might also Policymakers might, for example, want include access to a discretionary fund to consider phasing the implementation that would enable counselors to give of any activation requirements aimed some initial support with commuting at SA clients. For example, such expenses, work clothes, and child care. requirements could be targeted first It may also be worth while to consider geographically and/or at new claimants the use of short-term employment to test and refine referral mechanisms subsidies to encourage employers 46 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY to recruit individuals they would not it may be useful to undertake a normally hire. representative survey of possible future clients. This could capture information Many such employment services could on attributes that affect employability, be delivered more cost-effectively on previous work experience (if any), and a group basis, including, for example, employment aspirations, and provide group sessions on job search skills and an evidence base upon which to inform information about job opportunities the development of future policy. and available services. They might also be delivered through external The integrated information system in providers that complement İŞKUR place can also be used to evaluate the skills and capacities, especially if they performance of the overall system and have particular expertise working with to make improvements, as the SAIS specific client groups, such as mothers architecture allows data collection with child-care responsibilities. from a wide range of government Such providers should have clear agencies. Though merging the İŞKUR performance indicators focused on Management Information System effective employment assistance that (MIS) with SAIS seems difficult for places participants into jobs. The administrative and technical purposes, development of such an employment- establishing fully functional interfaces focused approach would provide insight between these two systems (as into the capacity of external providers well as other government systems) and what does or does not work with to develop evaluation reports for the client groups referred. managerial decision making can be achieved. By setting clear and In developing their approach, traceable performance indicators for Turkish policy makers might also be monitoring and evaluation in line with assisted by further knowledge on the their strategic targets and action plans, characteristics, employability, and İŞKUR and MoFSP may enhance their circumstances of existing SA claimants. ability to fine tune the system and their This could aid in the planning, design, institutional objectives. and delivery of employment services, and could be used for future evaluations. Finally, efficient cooperation between Existing information systems may senior officials and managers from not record the relevant information MoFSP and İŞKUR would help now, but in the future, such data identify and swiftly resolve problems might be captured through the initial in implementation and help them in employability assessment undertaken the further development of a pro- by social workers and İŞKUR’s employment SA system in Turkey. registration process. In the meantime, ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 47 Annex 1: Profiling Systems and Employability Assessments The PES and employment service educational attainment, work history, providers in OECD countries utilize a and previous type of work, as well wide range of approaches to profiling as local unemployment statistics). In and the assessment of employability. Australia, the Jobseekers Classification The instruments and tools vary in terms Instrument (JSCI) is administered by of the type of information they collect, staff as part of a client’s initial interview how they are administered, and the (after starting a benefit claim). This ways in which results are used. approach allows a greater number of issues to be included in the model Statistical modeling is an approach (e.g., stability of residence, access to that utilizes statistical systems to transport); however, like the models identify clients who require support based on administrative data only, the or the type of support to be provided. JSCI focuses on characteristics, work This “profiling” approach “allows the experience, and skills (rather than employment service to calculate the attitudes). risks faced by individual clients ... and then, by combining the risk scores In the Netherlands, the “Kansmeter” with a decision-making rule, to allocate (chance meter), later replaced by “AB clients to services” (Hasluck 2004). The Routering,” is also administered by statistical model can either produce a a member of staff at a client’s initial single “risk” score (e.g., predicting the interview, with questions on personal risk of long-term unemployment) or can characteristics, skills, and occupational place individuals into a particular group profile. This interview also examines distinguished by their need for different the client’s capacity for independent job levels or types of support. search. However, this approach differs from the other examples in that it does In the U.S., Denmark, and Germany, the not use a statistical model; the scoring model is based on administrative data system is based on the subjective already held about jobseekers (covering assessment of the member of staff demographic characteristics, prior administering the questionnaire.13 13- The details of these assessment tools are taken from Rudolph and Konle-Seidl (2005) and O’Connell et al. (2009). 48 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY A key question is how effective such work. This category includes a large tools are in predicting long-term number of tools assessing individuals’ unemployment or level of need. As skills, including several that have been noted by Bryson and Kasparova (2003), commercially produced in the U.S. As this type of profiling is controversial, discussed by Bimrose et al. (2007), these although some existing examples have typically ask individuals to assess their been quite successful; for example, own confidence in their ability to perform a model developed, tested, and now various occupational activities or skills, being implemented in Ireland reported and also to assess their occupational success rates of 83 percent (for men) interests. These can help advisers or and 85 percent (for women) in predicting counselors explore suitable career unemployment at 12 months (O’Connell and training options. Similarly, the et al. 2009). The model developed by French “Bilans de Compétences” (skills Bryson and Kasparova in the U.K. was check) is an appraisal service for career also found to “work reasonably well,” development. This model has also been although success varied by client adopted in other countries, including group (Bryson and Kasparova 2003). Switzerland, Belgium, and the Czech However, the Kansmeter used in the Republic. In the U.K., a skills check for Netherlands was accurate in predicting benefit claimants consists of an initial long-term unemployment in only “Entry Review” interview followed by three out of five cases, and in the U.S., a “Skills Diagnostic” assessment of an profiling generally performs poorly individual’s employability (Hasluck et in ranking benefit recipients based on al. 2006). their subsequent experiences. Studies have also acknowledged the limitations While the above examples focus on of these types of assessment tools assessments of clients’ characteristics (e.g., they are developed for particular and skills, an alternative approach is to client groups in particular economic focus on clients’ attitudes. Attitudinal conditions, and are often expensive diagnostic tools aim to identify to set up and run). In particular, this jobseekers whose attitudes act as a type of approach depends critically on constraint to obtaining employment and the quality of data available to predict are used to inform activities designed outcomes, and on how easily the results to change jobseeking behavior. can be used by advisers. Based on psychological models of “employability,” they can exist in Alternatively, a range of assessment the form of checklists for advisers to tools focus on specific characteristics, use during interviews with clients. such as strengths and job aspirations Examples of this approach can be found of jobseekers, and key barriers to in France (Copilote Insertion), Germany ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 49 (Placement Characteristics), Portugal modeling are often just the starting (Forecast Guide to the Difficulties point of an adviser’s assessment. In the of Insertion), and Denmark (Job models used in Germany and Denmark, Barometer), as reported by Bimrose et for example, advisers use the results of al. (2007). the model as part of the judgment they make to tailor personalized support. A key issue common to these different At the same time, an assessment tool types of approaches is the role of can provide useful objective evidence, the assessment tool alongside the which can be particularly helpful when judgment and discretion of advisers. decisions about the allocation of limited Even the tools based on statistical resources need to be made. 50 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY References Behncke, S., M. Frölich, and M. Lechner. dwp.gov.uk/asd/asd5/rports2005-2006/ 2007. “Unemployed and their rrep356.pdf Caseworkers: Should they be Friends Duell, N., D. Grubb, and S. Singh. 2009. or Foes?” IZA Discussion Papers No. “Activation Policies in Finland.” OECD 3149, Bonn. Social Employment and Migration Bimrose, J., S. Barnes, A. Brown, C. Hasluck, Working Papers, No. 98, OECD and H. Behle. 2007. “Skills diagnostics Publishing, at http://www.oecd-ilibrary. and screening tools: a literature review.” org/docserver/download/5ks6wdz7jhvj. DWP Research Report 459, London, at pdf?expires=1368202686&id=id&accn http://research.dwp.gov.uk/asd/asd5/ ame=guest&checksum=85C97DAE31 rports2007-2008/rrep459.pdf AB4C17152756DAD200D4FB Bloom, H., Hill, C. and Riccio, J. Duell, N., S. Singh, and P . Tergeist. 2009. “Linking Program Implementation “Activation Policies in Norway.” and Effectiveness: Lessonsfrom a OECD Social, Employment and Pooled Sample of Welfare-to-Work Migration Working Papers, No. Experiments”, Journal of Policy Analysis 78, OECD Publishing, at http:// and Management, Vol. 22, No. 4, 551– www.oecd-ilibrary.org/docserver/ 575 (2003) download/5ksmdtd6bw7j.pdf?expires= 1368202652&id=id&accname=guest&c Bryson, A., and D. Kasparova. 2003. “Profiling hecksum=90BA931D58E48AC5FB5CD benefit claimants in Britain: A feasibility 7CE9C52CDE0 study.” DWP Research Report 196, London, at http://statistics.dwp.gov.uk/ Duell, N., D. Grubb, S. Singh, and P . asd/asd5/rports2003-2004/rrep196.pdf Tergeist. 2010. “Activation Policies in Japan.” OECD Social, Employment Disney, J., A. Buduls, and P . Grant. 2010. and Migration Working Papers, No. “Independent Review of the Jobseeker 113, OECD Publishing, at http:// Compliance Framework: A Report to the www.oecd-ilibrary.org/docserver/ Parliament of Australia.” Department of download/5km35m63qqvc.pdf?expires Education, Employment and Workplace =1368202503&id=id&accname=guest Relations, Canberra, at http://foi. &checksum=253BD81C2B5F79B531B0 deewr.gov.au/system/files/doc/other/ 5A1DFEA8E429 impacts_of_the_new_job_seeker_ compliance_framework_report_of_the_ Egger, M., and C. Lenz. 2006a. “Évaluation independent_review.pdf des résultats du service public de l’emploi.” Politique du marché du Dolton, P., J. Azevedo, and J. Smith. travail No. 18. Study commissioned 2006.“The econometric evaluation of the by the Commission de surveillance du New Deal for Lone Parents.”Research Fonds de compensation de l’assurance- report 356, London: Department for chômage, SECO, Bern, 31 May. Work and Pensions, at http://statistics. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 51 Egger, M., and C. Lenz. 2006b. “Évaluation de Gregg, P., S. Harkness, and S. Smith. 2009. l’impact du service public de l’emploi.” “Welfare reform and lone parents in the La Vie économique No. 10: 26-29. UK.” The Economic Journal 119 (535): F38-F65. EJML. 2011. “Partnerships among Employment Services.” European Griggs, J., and M. Evans. 2010. “Sanctions Job Mobility Laboratory, European within conditional benefit systems: A Commission, atec. europa.eu/social/Blo review of evidence.” Joseph Rpwntree bServlet?docId=7122&langId=en Foundation, York, at http://www.jrf.org. uk/sites/files/jrf/conditional-benefit- Finn, D. 2011. “Subcontracting in Public systems-full.pdf Employment Services: Review of research findings and literature on Grubb, D., S. Singh, and P . Tergeist. 2009. recent trends and business models.” “Activation Policies in Ireland.” OECD The European Commission Mutual Social, Employment and Migration Learning Programme for Public Working Papers No. 75. OECD Publishing, Employment Services, GHK Limited / at http://www.oecd-ilibrary.org/ Budapest Institute, DG Employment, docserver/download/5ksq8s3tx4q8.pdf Social Affairs and Inclusion, Brussels, at ?expires=1369307189&id=id&accname ec.europa.eu/social/BlobServlet?docId =guest&checksum=10BC71C867ACF9 =6964&langId=en C7873AD8605F3985A2 Finn D., and J. Casebourne. 2012. Hainmueller, J., B. Hofmann, G. Krug, and K. “Lone parents anctions: a review of Wolf. 2009. “Do more placement officers international evidence.” Policy Paper, lead to lower unemployment? Evidence Centre for Economic and Social from Germany.” IAB Discussion Paper Inclusion, London, at http://www.cesi. 13/2009, Nuremberg, at http://doku.iab. org.uk/sites/default/files/publications/ de/discussionpapers/2009/dp1309.pdf L P % 2 0 S a n c t i o n s % 2 0 Po l i c y % 2 0 Hasluck, C. 2004. “Targeting Services in Paper_0.pdf the Individual Customer Strategy: The Frölich, M., M. Lechner, S. Behncke, S. Role of Profiling; a Review of Research Hammer, N. Schmidt, S. Menegale, Evidence.” DWP Research Report 192, A. Lehmann, and R. Iten. 2007. London, at http://research.dwp.gov.uk/ “Influence des ORP sur la réinsertion asd/asd5/working_age/wa2004/192rep. des demandeurs d”emploi.” Politique pdf du marché du travail No. 20. Study Hasluck, C., J. Bimrose, S. Barnes, J. Brown, commissioned by the Commission de L. Marris, G. McGivern, M. Orton, and surveillance du Fonds de compensation R. White. 2006.“Evaluation of Skills de l’assurance-chômage, SECO, St- Coaching trials and Skills Passports: a Gallen and Zurich. synthesis of qualitative and quantitative Government of Ireland. 2012.“Pathways to evidence.” DWP Research Report 391, Work: Government Policy Statement on London, at http://research.dwp.gov.uk/ Labour Market Activation.” Dublin. asd/asd5/rports2005-2006/rrep391.pdf 52 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY Hill, C. 2005. “Casework Job design agenda.” Policy & Politics Vol. 41(2): and Client Outcomes in Welfare- 183–200. Policy Press, Bristol. to-work Offices.” Journal of Public Meyer B.D. 2007.“The U.S. Earned Income Administration Research and Theory Tax Credit, its Effects, and Possible Vol. 16(2), at http://jpart.oxfordjournals. Reforms.” Harris School of Public Policy org/content/16/2/263.full.pdf Studies, 07.20, University of Chicago and Immervoll, H. 2009.“Minimum-Income NBER, at http://harrisschool.uchicago. Benefits in OECD Countries: Policy edu/research/working-papers-series/ Design Effectiveness and Challenges.” papers/us-earned-income-tax-credit- IZA Discussion Paper No. 4627, Bonn, its-effects-and-possible-reforms at http://ftp.iza.org/dp4627.pdf MoFSP. 2011.“Social Assistance Information Kauff, J., M. Derr, L. Pavetti, and M. E. Sama. System.” General Directorate of Social 2007. “Using work-oriented sanctions to Assistance and Solidarity, at http:// increase TANF program participation.” www.sosyalyardimlar.gov.tr/upload/ Washington D.C.: Administration sosyalyardimlar.gov.tr/mce/2012/ for Children and Families, Office of genel_sencer/soybis_eng.pdf Planning, Research, and Evaluation, MoFSP . 2013. “Social Assistances Bulletin, U.S. Department of Health and Human March.” at http://www.sosyalyardimlar. Services. gov.tr/upload/Node/11873/files/MART_ Konle-Seidl, R. 2011. “Profiling Systems istatistik_.pdf for Effective Labour Market Integration: MSD. 2010.“Future Focus: Overview of Use of profiling for resource allocation, Package.” Paper A, Cabinet Social Policy action planning and matching.” The Committee, Office of the Minister for European Commission Mutual Learning Social Development and Employment, Programme for Public Employment Wellington, at http://www.msd.govt.nz/ Services, GHK Limited / Budapest documents/about-msd-and-our-work/ Institute, DG Employment, Social Affairs newsroom/media-releases/news/2010/ and Inclusion, Brussels. paper-a-overview.pdf Lessof, C., M. Miller, M. Phillips, K. Pickering, O’Connell, P ., S. McGuinness, E. Kelly, S. Purdon, and J. Hales. 2003. “New and J. Walsh. 2009. “National profiling Deal for Lone Parents evaluation: of the Unemployed in Ireland.” findings from the quantitative survey.” Economic and Social Research Working Age and Employment Report Institute Research Series Number 10, 147, London: Department for Work and Dublin, at http://www.esri.ie/UserFiles/ Pensions, at http://research.dwp.gov. publications/20090825142343/RS010. uk/asd/asd5/rports2005-2006/rrep356. pdf pdf OECD. 2007. “Activating the Unemployed: Lindsay, C., and M. Dutton. 2013. “Promoting What Countries Do.” Chapter 5 in OECD healthy pathways to employability: Employment Outlook, OECD Publishing, lessons for the UK’s welfare-to-work Paris, at http://www.oecd.org/els/ ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 53 emp/40777063.pdf Nuremberg, at http://www.upjohninst. org/fdss/euroseminar.pdf OECD. 2009.“Slovenia: OECD Reviews of Labour Market and Social Policies.” Schneider, J. 2008. “The Effect of OECD Publishing, Paris. Unemployment Benefit II Sanctions on Pavetti, L., M. Derr, and H. Hesketh. Reservation Wages.” IAB Discussion 2003.“Review of sanction policies Paper 2008 19, Institute for Employment and research studies: Final Literature Research, Nuremberg, at http://doku. Review.” Washington D.C.: Office of the iab.de/discussionpapers/2008/dp1908. Assistant Secretary for Planning and pdf Evaluation, Department of Health and Scrivener, S., J. Walter, T. Brock, and G. Human Services. Hamilton. 2001. “National Evaluation of Ringold, D., and L. Kasek. 2007. “Social Welfare-to-Work Strategies: Evaluating Assistance in the New EU Member Two Approaches to Case Management: States: Strengthening Performance Implementation, Participation Patterns, and Labor Market Incentives.” Working Costs, and Three-Year Impacts of the Paper No. 117, World Bank, at http:// Columbus Welfare-to-Work Program.” www-wds.worldbank.org/external/ Manpower Development Research default/WDSContentServer/IW3P/IB/20 Corporation, at http://www.mdrc.org/ 07/08/28/000020953_20070828145422/ publications/85/execsum.pdf Rendered/PDF/405410ECA0Soci1LIC0d Sienkiewicz, L. 2012. “Job profiles and isclosed0Aug271.pdf training for employment counselors.” Rudolph, H., and R. Konle-Seidl. 2005. The European Commission Mutual “Profiling for Better Services, Report Learning Programme for Public on the European Profiling Seminar Employment Services, GHK Limited / Nuremberg.” January 12-14, 2005, Budapest Institute, DG Employment, Institute for Employment Research, Social Affairs and Inclusion, Brussels. 54 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 55 3. Profiles of Vulnerable People In Turkey 3.1 Introduction Recognizing that those with limited or no attachment to the labor market A common insight from policy are a highly heterogeneous group, evaluations of both employment and it seeks to contribute policy-relevant social support measures is that careful information on the types of social and targeting is crucial for activation and economic risks that different groups poverty reduction strategies to be face and on the barriers hindering their successful. This is especially the case labor market integration. The resulting in Turkey, where the increasing labor information can inform more effective supplied by women combined with targeting of policies to alleviate these the current demographic “dividend” barriers. For instance, information on mean that the group of potential policy the characteristics of inactive working- “clients” is growing. In Turkey, about age individuals can be a basis for: (i) half of the working-age population identifying which groups are served (WAP) (those aged 15-64 years) do not most effectively by existing activation work and nearly 40 percent of workers and support policies; and (ii) channeling are in informal jobs (i.e., they are not in policy efforts towards specific priority the social insurance system). As Turkey groups. has the biggest share of non-working individuals in the OECD and among the Those outside or on the margin of the highest informality rates, it is important labor market frequently move between to carefully consider which individuals non-employment and different states should be a high priority for activation of “precarious” employment. As a efforts. result, looking at only some of these states (e.g., unemployment) would This chapter aims to strengthen not capture the true extent of labor the empirical basis for designing market difficulties or the need for and targeting income support and policy intervention. To identify the size activation policies. It uses survey data and characteristics of groups that are to identify relevant groupings and facing severe labor market difficulties, characteristics of individuals with no section 2 develops a broad concept or limited labor market attachment. of weak labor market attachment, or 56 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY labor market “vulnerability,” including process. If job reallocation is efficient in unemployment, inactivity, and informal that jobs move from less productive to work as well as sporadic or low-paid more productive firms or sectors, then work. it leads to a more productive economy and higher incomes. This is especially Sections 3 and 4 then use household true during a prolonged growth process, survey data to characterize groups of such as the one seen in Turkey. But it also people showing weak labor market holds for the aftermath of economic attachment over extended periods downturns, which tend to be associated of time (“persistent vulnerability”). with significant restructuring and Two different but complementary changes in the sectoral composition of approaches are used. First, a simple an economy. regression model seeks to identify the factors associated with persistent labor Different barriers can prevent or slow market difficulties. Second, a statistical down an efficient job reallocation, clustering approach is used to identify creating significant economic and groups of vulnerable individuals who social costs.14 For some groups, are likely to share similar employment a combination of policy-related barriers and can therefore be expected to barriers and insufficient skills or work benefit from similar policy interventions. experience means that they remain Section 5 discusses the possible policy “stuck” without a job or in marginal implications of the findings. In particular, employment for extended periods of it asks whether some groups are, or time. Such long-term labor market should be, a priority for employment marginalization or detachment is known and income support. to erode human capital and reduce both current income and future earnings 3.2 Extent and Types of prospects. Addressing the causes of Employment Difficulties long-term labor market difficulties and alleviating their consequences are Spells of unemployment are a necessary therefore crucial challenges for labor element of a market-based growth market and social policy. 14- For instance, inadequately resourced or poorly targeted re-employment services reduce the quality of matches between jobseekers and job vacancies. Ineffective income support for jobseekers can have a similar effect. On the one hand, insufficient support can prevent jobseekers from engaging in a thorough search for suitable vacancies, or force them to accept low-productivity or informal jobs that do not adequately use or remunerate their skills.On the other hand, overly generous or unconditional out-of-work support can, in principle, delay or weaken job search activities.The relative importance of the different barriers is likely to vary with economic conditions.For instance, evidence for OECD countries summarized in Immervoll (2012) shows that adverse work incentives are a less relevant determinant of employment outcomes if labor markets are weak (e.g., because of a country-wide downturn, or due to high or deteriorating unemployment in a specific region). ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 57 Given that Turkey has the lowest Using the Survey of Income and Living employment rates in the OECD, Conditions (SILC), which follows customizing policy measures to the individuals over a four-year period, it needs and barriers of those with long- is possible to examine some of these term (or “persistent”) labor market trends, and the histories of people’s difficulties is arguably particularly labor market experiences, in some important. But at the same time, detail. Four different types of labor the challenges of designing and market difficulties or “vulnerability” implementing an integrated activation are distinguished (and shown and employment support strategy are schematically in Figure 3.1): also daunting. Due to Turkey’s rapid growth process, with its associated • Not working at all social and economic transformations, • Working only a few months during the circumstances and labor market the year (“low work intensity”) histories of the WAP are both already • Low earnings, and more diverse and evolving more • Informal work rapidly than in other parts of the OECD. The temporary downturn in 2008/09 Some of those categories can overlap further contributed to the diversity of (e.g., those working informally or labor market experiences and barriers sporadically will typically have a higher in Turkey (World Bank 2013). risk of low earnings). FIGURE 3.1 Different types of labor market difficulties/vulnerability potential clients for employment and income support low not working earnings at all informal low work work intensity 58 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY Between 2005 and 2008, nearly two- percent of working-age individuals) are thirds of working-age individuals were persistently jobless (either unemployed either out of work or in marginal or or labor market inactive). About half as informal employment for an extended many are in persistent informal, low- period of time (Figure 3.2). Note that paying, or unstable work, meaning that individuals whose main employment they do have both some labor market status during the year is full-time attachment and some work experience. education or military service are About the same number again move excluded from the analysis and, hence, between joblessness and marginal, are not classified as “vulnerable” sporadic, or informal work. This pattern regardless of their earning levels or indicates that persistent labor market employment patterns. difficulties (PLD) are indeed far from While low employment rates in Turkey one-dimensional. In particular, they are a well-established fact, the very large go beyond long-term unemployment share of people who are persistently or inactivity. The experiences and vulnerable highlights the challenges employment barriers of the people for activation and employment-support concerned can be captured only when policies. Among the persistently a broad range of out-of-work and in- vulnerable, only about half of them (30.3 work situations is considered. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 59 FIGURE 3.2 Broad categories of persistent labor market difficulties, 2005-2008 Notes: Persistence is defined as experiencing the relevant status during at least one-half of the observed years using the following categories. “Not working”: not reported to have worked during any month of the year, or no labor income at all (either in cash or in kind). “Low earnings”: labor income less than 2/3 of the full-time, full-year minimum wage, and not informal. “Low work intensity”: employed or self-employed for at least one, but less than six months during the year, and not informal. “Informal”: positive labor income but no (employer) social security contributions, or labor income is mainly earned in kind, or the person reports being an unpaid family worker. “Cyclers”: those who are in more than one of these categories during the period. In all cases, those who are younger than 15 or older than 64 and those in education or military service during most of the year are not categorized as facing labor market difficulties (and are therefore not in any of the above categories). Source: World Bank staff calculations based on TUIK SILC data, waves 2006-2009 (with 2005 to 2008 as the reference year for incomes and activity calendars). 3.3 What Characteristics are market statistics in several ways: Associated with Persistent • It accounts for people’s labor market Market Labor Difficulties? experience over a longer period, rather than at a specific point in This section takes a closer look at time; individuals categorized as experiencing PLD. The information presented • It is multi-dimensional, meaning complements commonly used labor that it accounts for a whole range of 60 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY potentially relevant characteristics are typically correlated (such as sex, (e.g., age, gender, number of family status, and work experience) – children, education, and work something that cannot be done using experience), rather than only one simple cross-tabulations. For simplicity, dimension at a time. This allows Table 3.1 shows regressors grouped for a more detailed examination into four clusters (demographics, of the factors that are positively family situation, factors likely related or negatively associated with the to productivity, and labor market risks of persistent joblessness or situation). precarious employment; and The averages in the first column give • It considers both individual and a useful description of the WAP in family characteristics. Family Turkey.15 The average age among this circumstances are central for group (i.e., without counting those designing and targeting employment aged below 15 or above 64) is about 37 and income support measures. years. The averages for 0/1 “dummy” Understanding them is arguably variables (denoted ‘d’) represent especially important during and frequencies: 52 percent of working- after a severe downturn, as families age individuals who are neither in can provide essential income education nor in military service are stabilization following the loss of women and just over one-fourth report one family member’s earnings. living in a “rural” area. The table also reports interactions between some of Table 3.1 reports the main results of a these characteristics. For instance, 7 simple statistical model, which relates percent are “young” (defined as under PLD status during a four-year period to 30) and report living in a “rural” area. a broad range of potentially relevant The next block of variables gives a individual and family characteristics. first impression of the quantitative This format is convenient for importance of family-related factors investigating which of a large number often thought to be closely related of factors are associated with a higher to employment barriers. A large risk of PLD. The regression approach number of working-age people live is particularly helpful for disentangling in households where a young child the importance of different factors that (36 percent) or an elderly person (8 15- Only individuals who are observed multiple times during the four-year period are included. For each individual, characteristics, including the values of dummy variables, are obtained by averaging over the period. For instance, someone who reports being married during at least half of the years is classified as “married.” ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 61 percent) potentially requires care by a effect in any direction. working-age household member. Three in ten (29 percent) are not married, With this caveat in mind, the results with only 2 percent heading a single- do provide evidence consistent with parent household. theoretical models and common presumptions of the risk factors leading On average, people of working age not to labor market exclusion. An upper in education or military service report secondary or higher degree is the single- about 10 years of work experience. most important factor for reducing Over half (55 percent) of them have less the risk of PLD (by 16 percentage than a basic education, and 29 percent points), while having less than a basic have a higher secondary or tertiary education is associated with a higher degree. Perhaps surprisingly, one-third risk. Interestingly, however, lower of the WAP reports a significant health education appears to be much more problem (defined as either a chronic damaging to the labor market prospects illness, a health problem that restricts of women and individuals under 30 (as daily activity, or a health status that is the marginal effects are sizeable for either “bad” or “very bad”). “lowEdu” as well as for the interacted “femXlowEdu” and “youngXlowEdu”). The estimated model coefficients in the Experiencing a health issue is also next column identify several factors that associated with a much higher risk of are correlated with PLD. To make these PLD. associations easier to interpret, the table also shows the so-called marginal Consistent with the “income effect” effects. The marginal effects show, for predicted by economic theory, the each variable, the predicted effect of a income of a spouse reduces the one-unit change on the probability of probability of no or limited labor PLD (while keeping all other variables market attachment. Although highly unchanged at their means). For instance, significant statistically, the effect an additional child in the household is small (it implies that doubling a is associated with a 2.5 percentage spouse’s income increases PLD risks point higher probability of PLD, while by less than 1 percentage point). an additional year of work experience Irrespective of spousal income, married (at a given age) is associated with a men are less likely to experience PLD reduction of 1.4 percentage points. In than unmarried men. However, young all cases, it is important to keep in mind married individuals are at much greater that these are statistical associations risk, especially women. A gender and, technically, do not imply a causal difference can also be seen for parents 62 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY of young children (“isYchildinhh” and Indeed, although the age/risk profile is “femXychild”): women with young clearly non-linear (the variables age 2 children are significantly more likely and age 3 are both highly significant), to experience PLD, which is consistent it can be shown that PLD risks increase with the primary caregiving role of monotonously with age, which is women, and the very limited availability consistent with the important role of of child-care facilities in Turkey. No family-related employment barriers significant effect can be discerned for as well as Turkey’s very low effective retirement age. the presence of an elderly person in the household. An interesting finding is that the regional unemployment rate (UR) is The results for the demographic not a statistically significant factor for variables show that women and those higher PLD risk. One interpretation of living in rural areas have a 10 and 8 this result is that with low labor force percentage point higher probability, participation in Turkey, the UR is a very respectively, of PLD. Age is also a incomplete indicator of labor market highly significant factor. While being conditions. This suggests that a broad married and having low education is concept of labor market difficulties shown to be a more powerful risk factor indeed provides a more useful starting for young people, young age per se is point for discussing activation and not associated with a higher risk of PLD. employment support strategies. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 63 TABLE 3.1 Factors associated with PLD risks, 2005-2008 Notes: Logistic regression, including a full set of occupation/former occupation, region, and year dummies as control variables (not reported). The sample consists of individuals aged 15-64. Periods where the main status was education or military service are excluded. For each individual, variable values are obtained by averaging over their four-year observation period. Squared age and experience are divided by 100; cubed age is divided by 1000. See Annex 1 Table A1 for variable definitions. Source: World Bank staff calculations based on TUIK SILC data, waves 2006-2009 (with 2005 to 2008 as the reference year for incomes and activity calendars). 64 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 3.4 What Are the Main Groups both meaningful statistically and useful for policy purposes. The result of this of Individuals Experiencing exercise is a set of groups displaying PLD? similar characteristics of members within each group and dissimilar The results in section 3 provide pointers characteristics between groups. This for understanding the risk factors means that: associated with a higher probability of persistent joblessness or marginal • Group members should be similar to employment. However, the regression each other; approach looks at one factor at a time (e.g., it identifies the role of education • Members of different groups should while holding constant family situation, be dissimilar; and age, and work experience). Designing • The characteristics used to define and targeting employment and income group membership should be support measures also requires observable by policy makers, knowledge about the combined administrators, or caseworkers. characteristics of people affected by PLD. For instance, case workers As an illustration, Table 3.2 shows two at the employment office or benefit examples of the resulting groups of administration need to have as full a PLD individuals. To focus on the main picture as possible about their clients’ variables that characterize these groups, education, income, family situation, the example shows only a selection health status, and work experience. of characteristics that are especially They also need to know if there are important as distinguishing features larger groups of comparable clients with (the subset of characteristics shown relatively similar barriers or needs. The therefore differs between the two regression approach does not provide groups). The example shows that the this information. clustering method is able to separate groups very sharply along some of To fill this gap, this section identifies the the characteristics. For instance, all size and characteristics of different PLD individuals in Group A are women but groups using a variant of a statistical all individuals in Group B are men. clustering approach called “latent class Likewise, Group A members have analysis”. This approach searches for been engaged in housework and have suitable groupings over a wide range of no employment experience, whereas demographic, family, social, and labor members of Group B have worked market characteristics. The basic idea informally and have a 97 percent is to cluster people into groups that are probability of having work experience ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 65 of six years or more. Members of both separates both groups from a number groups are married and have children, of other groups comprising almost so this is not a distinguishing feature. exclusively unmarried or childless But as will be shown below, it sharply individuals. TABLE 3.2 Groups with PLD: An illustration Other characteristics vary more widely The statistical clustering tool provides between group members. While the probabilities of group membership husbands of most women in Group A but not, of course, specific labels for earn more than the (annual equivalent these groups. Results can nevertheless of the) minimum wage (MW), this is not provide a basis for thinking about universally the case. And although a group labels. There are a large number majority of Group A are aged under 35, of characteristics, so care needs to be a significant number of group members taken to resist over simplification and to are older. In Group B, a large majority keep arbitrary judgments to a minimum. have less than a basic education. At With this in mind, careful labeling can the same time, a significant number be useful as a basis for discussing of them have finished basic education, suitable policies for each of the groups. but are nevertheless grouped together Table 5 illustrates an attempt to find because their other characteristics suitable labels that capture relevant make them “sufficiently similar” in a group characteristics (e.g., “informal statistical sense. family men”). 66 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY The full set of groups is depicted in may require quite different types of Figure 3.3, again showing only those employment and support than men. characteristics that are particularly relevant for distinguishing each group Together, the “female only” groups 1, 2, from the others. (Annex 1 Table A1 3, 8, and 11 account for about half of the provides a list of all characteristics that persistently vulnerable individuals in were used as inputs in the clustering Turkey. Groups 5 and 10 are also mostly analysis, along with their definitions. women, but there are also a significant Annex 1 Table A2 shows detailed number of men with otherwise similar results for all groups and the entire set characteristics in these groups. The of characteristics). Labels were derived total proportion of women among the using the procedure illustrated in Table persistently vulnerable is just over two- 3.2, and groups were ordered by their thirds (68 percent, obtained by adding size, with the largest group shown first. up the products of “female” percentages and respective group sizes in Annex 1 The clustering approach partitioned the Table A2). PLD population into 12 separate groups. Given the very low female employment The summary in Figure 3.3 clearly shows rates in Turkey, it is not surprising that the great diversity in the circumstances, the largest groups consist mostly of and hence the potential employment women. In fact, the clustering algorithm barriers, of women with PLD. They distinguished several groups very include those with significant work sharply by gender; in each of the three experience, but in low-paid, informal, or largest groups, as well as in groups unstable employment (groups 3, 5, and 8 and 11, the percentage of women 8), as well as large numbers of women is exactly 100 percent (see Annex 1 who have been persistently out of work Table A2 for detailed distributions of with no or very short employment all characteristics of each group). The histories (groups 1, 2, 10, and 11). interpretation is that the combination Women in the “no work experience” of characteristics in groups 1, 2, and groups (1, 2, and 10) mostly live in 3 is only observed for women or, in urban areas. They include women in other words, it is not only their gender early retirement (group 2) as well as that separates them from persistently young mothers (group 1) and better- vulnerable men, but they also differ educated single women.16 This stands in other relevant respects. A possible in sharp contrast with women in rural policy implication is that these groups areas (the majority of group 8 and just 16- Recall that people who are in education or military service are not counted here as they are not included in the “vulnerable” group. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 67 under half in group 5), most of whom work experience (40 percent of them are either working or have significant report four years or more). previous work experience. Interestingly, only a small proportion Among men, the groups accounting for of the persistently vulnerable can the largest shares of the persistently be described as the “archetypical” vulnerable are the “informal, low- unemployed jobseekers with no or very earning” group 6 (9 percent of all little earnings and actively looking for a persistently vulnerable in Turkey), as job (group 12). Given the small size of well as two groups of older men (the this group relative to the total number “older, self-employed family men” of persistently vulnerable, policies group 9 and the “retired urban career seeking to strengthen labor market worker” group 7). In addition, there is a sizable group of younger and relatively attachment among a broader group of well-educated single men (group 4). But working-age individuals need to find unlike the “young and educated single ways to reach out beyond the group of women” group 10, who are mainly out core jobseekers who have traditionally of work, the young men do have some been the main focus of activation and labor market attachment and significant employment-support measures. 68 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY FIGURE 3.3 Groups with PLD: complete groupings for the period 2005-2008 Urban housewives, Older urban Young urban Educated single men working husband housewifes mothers: Low-pay / no-pay - 13 % - 11 % Group - 10 % -9% - Young child - 45 + size - Young children - 29- - Never worked - Never worked - Unstable employment - Unstable employment - Higher earning husband - ≤ basic edu - Higher earning husband 1 2 3 4 Women in long-term Younger informal Retired urban Female unpaid marginal work family men career workers agriculture workers -9% -9% -9% -8% - 45 + - 25 – 39 - Very long experience - Rural - Long experience - Long experience - Informal - High or low edu - Spouse vulnerable - ≤ basic edu - Mostly MW or less - Health issues 5 - Spouse vulnerable 6 7 8 Older self-employed Educated urban Uneducated with Blue-collar struggling family men single women struggling husband job seeking men -7% -6% -6 % -4% - 45 + - 15 – 29 - Young child - 35 – 49 - Very long experience - Living with parents - No experience - experience - Mostly agriculture - ≤ basic edu - No/little experience - Husband informal & low - ≤ basic edu - Health issues 9 10 pay 11 - Wife vulnerable 12 Notes: Group sizes are the percent of all individuals in the PLD group. See Annex 1 Table A2 for full results showing probabilities/incidence for the full set of characteristics. Source: World Bank staff calculations based on TUIK SILC data, waves 2006-2009 (with 2005 to 2008 as the reference year for incomes and activity calendars). 3.5 Targeting Activation and might be used for deciding on suitable targeting strategies are: (i) people’s Employment and Income living standards; and (ii) the likely Support Policies extent of their labor market difficulties. With limited resources and very large 3.5.1 Targeting Based on Living numbers of people facing PLD, a Standards relevant question is whether some groups are, or should be, a particular Long-term labor market difficulties priority for employment and income can lead to economic hardship for the support. Two notable criteria that individuals and families concerned. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 69 With labor income the primary income bath or shower, no toilet, or no piped source for working-age people and water in its household. their families, extended spells without adequately paid employment leave Both measures show a considerable families financially vulnerable and at a divergence of living standards between high risk of poverty. As a result, income groups, reinforcing the importance of support is an essential element of an careful targeting. Three of the mostly overall support package that fosters urban groups (“urban housewives,” stable employment and reduces “retired urban career workers,” poverty risks. However, even among and “young urban mothers”) show people with PLD, some groups are the smallest incidence of financial better able than others to cope with hardship. In fact, in all three groups, periods of low or no labor income. the proportion of people experiencing They may have access to other income hardship is less than the 79 percent sources (including benefits in cash or average for the WAP as a whole (i.e., in kind), they may receive support from including the “non-PLD” groups). The other household members or from proportions of people reporting poor extended family, or they may have sanitation are also significantly lower savings on which they can draw. As than the average (18 percent for the factors with a direct impact on people’s entire WAP) in these three groups. The living standards, these circumstances largest proportion of people with poor are important when tailoring support sanitation is found for the mostly rural measures and targeting mechanisms. “female unpaid agricultural workers.” Perhaps more surprisingly, the mostly Figure 3.4 illustrates differences urban group of “educated single in living standards and material women” also shows an above-average deprivation across groups by showing incidence of poor sanitation (as well as the proportions of people facing forms a high incidence of financial hardship). of “financial hardship” and “poor The highest incidence of financial sanitation.” Using information in the hardship is reported for the group of SILC data, a household considered “educated single men,” but it is also to be experiencing financial hardship high for groups where both spouses is defined as unable to afford at least are persistently vulnerable (“younger three of the following four items: a informal family men,” “uneducated one-week vacation away from home; women with struggling husband,” and unexpected expenses; replacement of “blue-collar struggling job-seeking old/broken furniture; or new clothes. men”). The results underscore the A family is considered to have poor importance of the family situation and sanitation if it reports having either no of addressing labor market difficulties at 70 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY the family level, rather than as a purely more advanced) measures of material individual issue. deprivation, as well as monetary poverty measures (including poverty The measures of “need” shown here headcounts based on region-specific are informative because they are simple poverty thresholds) can bring additional and provide some sense of absolute insights. Most importantly, however, living standards, which is especially future work should examine whether useful in a country like Turkey, where existing income support measures are large regional differences in incomes and prices can make it difficult to well targeted, in the sense that they compare income levels. However, reach those with the greatest need for future analyses using other (perhaps such support. FIGURE 3.4 Targeting based on living standards: Some groups have much greater need for support experiencing “financial hardship”, reported by 79% of all working-age household with “poor sanitation”, reported by 18% of all working-age Urban housewives, Older urban Young urban Educated single men working husband housewifes mothers: Low-pay / no-pay - 75 % - 83 % - 77 % - 92 % -5% -7% -8% - 18 % Women in long-term Younger informal Retired urban Female unpaid marginal work family men career workers agriculture workers - 89 % - 95 % - 62 % - 89 % - 23 % - 30 % -1% - 33 % Older self-employed Educated urban Uneducated with Blue-collar struggling family men single women struggling husband job seeking men - 80 % - 84 % - 96 % - 98 % - 19 % - 20 % - 27 % - 19 % Notes: See text for definitions. Source: See Figure 3.3. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 71 3.5.2 Targeting Based on Labor a sense, furthest from finding and Market Difficulties holding a stable and adequately paid job, then policy should focus on those Targeting issues also arise with with a high risk of PLD (shown in red activation measures, employment in Figure 3.5). Using such a criterion, services, and other types of active groups 2, 5, 8, and 11 should be labor market programs (ALMP), such priorities for activation measures. One as training. A crucial but difficult can think of these group members question is whether resources should as those with multiple or relatively be channeled to groups who are, in major employment barriers – they some sense, furthest from obtaining therefore clearly need support, but the or holding a stable job or to groups for challenges of successful and continuing whom policy interventions are likely labor market integration will often be to have the greatest probability of formidable. success. A very different set of priorities would Those two criteria need not, and result if the objective is to focus efforts generally do not, provide the same on groups where a significant number answers about the desirable targeting of individuals have relatively low risks mechanisms. This is illustrated in Figure of PLD. Groups 4, 6, 10, and 12 would be 3.5, which shows the shares of people prioritized in this case. Such a strategy with predicted “high” and “low” risk of may be attractive, as policies would PLD in each group. The predictions use have to “bridge” a smaller gap since the estimated regression model results these people may already be relatively shown earlier in Table 4 (see figure well-equipped to find a good-quality notes for more details). Combining job. Their probability of successful the model with the clusters is useful activation would therefore likely be because although members of each higher. But at the same time, some of cluster group all face PLD and share these “low-risk” individuals may well many other characteristics, they are not succeed in overcoming their PLD even identical. Some will face greater labor without active policy support. market difficulties than others. Using the regression model, it is possible In practice, it is useful to consider to quantify how likely it is for each a wide range of information when member to experience PLD given his or deciding on policy design and her characteristics. targeting. The information in Figure 14 and Figure 3.5 should arguably be If the objective is to focus activation read in combination as an integrated policy efforts on those who are, in policy approach that seeks to tackle 72 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY both employment barriers and PLD employment services and ALMP . Low risks as well as economic hardship. benefit coverage among those with PLD The different perspectives on group is not only a concern from an equality characteristics and employment and poverty-reduction point of view; it barriers also highlight the need to link also makes it harder to implement and income support and activation policies. deliver effective activation strategies As part of an employment-oriented and employment services, as those policy framework, benefits in cash or outside the scope of benefits tend to find in kind provide a principal instrument accessing these services significantly for linking unemployed people to more difficult. FIGURE 3.5 Targeting based on distance: some groups have much better chances of overcoming PLD how many are far from /close to “good job”? “high” or “low” risk of persistent LM difficulties Urban housewives, Older urban Young urban Educated single men working husband housewifes mothers: Low-pay / no-pay -0% -0% -0% - 38 % - 68 % - 99 % - 46 % -4% Women in long-term Younger informal Retired urban Female unpaid marginal work family men career workers agriculture workers -0% - 32 % - 10 % -0% - 96 % -1% - 36 % - 94 % Older self-employed Educated urban Uneducated with Blue-collar struggling family men single women struggling husband job seeking men - 15 % - 24 % -0 % - 25 % - 31 % -7% - 90 % -5% Notes: PLD “risk” scores were calculated using the statistical model of PLD presented in Table 1 (the estimated coefficients were used to calculate a probability of experiencing PLD for each individual). Those with a score in the bottom/top third of the whole WAP (i.e., the 33 percent with the lowest/highest risk) were then classified as “low-risk”/”high-risk.” Source: See Table 3.1 and Figure 3.3. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 73 Annex 1: Variables and Results from Cluster Analysis TABLE A1 Description of variables used in the regression and/or cluster analysis 74 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY TABLE A2 Detailed results from the cluster analysis ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 75 References Immervoll, H. 2012. “Reforming the World Bank.2013. “Turkey: Managing Benefit System to ‘Make Work Pay’: Labor Markets Through the Cycle.” Options and Priorities in a Weak Ministry of Development, Turkey Labour Market.” IZA Policy Paper and World Bank, Washington, DC. No. 50. Institute for the Study of Labor: Bonn. 76 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 77 4. Mobilizing Skills for Productive Employment in Turkey Abstract: Turkey has a historic less skilled individuals is not straight opportunity to reap an economic forward. Thus, promoting mechanisms dividend from its growing number of that link graduates and employers youth by employing them in quality and expanding programs that help jobs. The skill levels of Turkey’s the WAP increase the relevance of its population are increasing, particularly skills will aid the transition to more for youth and women, and labor market productive employment. Whether outcomes in Turkey are better for the Turkey can successfully facilitate the most educated. Over the long run, move from inactivity to productivity individuals with less education have occupations that benefit society and worse employment and activity rates enhance economic growth will partly than their more educated counterparts. depend on how the Government Youth and women face challenges as of Turkey supports individuals in well: youth face difficulties breaking responding to the demand for skills. into the labor market and women Several promising activities have been must continue to acquire skills. This is proposed already. especially critical as the nature of the skills required in jobs in Turkey has 4.1 The Role of Skills in the changed over the last decade. The need Turkish Labor Market for routine cognitive skills is on the rise, while jobs that require non-routine 4.1.1 Workforce Skills Supply manual skills are on the decline. The changing demand for skills translates The working working-age population into a higher demand for graduates (WAP) in Turkey is young, with with secondary (high school) and approximately 48.4 million people higher education, reflecting Turkey’s between the ages of 15 and 64 (out ongoing sectoral transformation of a total population of 73.6 million towards more productive activities people).17 It’s increased 12 percent and towards the services sector. At the over the last decade (TUIK, Household same time, finding employment for Labor Force Survey 2011). This 17- The WAP definition, individuals between 15–64 years old, used in this chapter was used to match the OECD definition. 78 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY demographic transition in the labor productive employment by exploring force offers both opportunities and the supply of and demand for skills challenges: on one hand, Turkey can and their interaction in the labor market benefit from the growing number of in Turkey19; (ii) to discuss existing young people if they are gainfully and proposed government policies employed, improving individual living to help young people develop skills standards and contributing to national and find gainful employment; and (iii) economic growth; on the other, in the to propose additional policy options absence of adequate job creation, high for skills development that enhances unemployment and inactivity rates can job entry.The report focuses on the constrain present and future incomes segment of the WAP that has already as well as social cohesion (World Bank completed formal education. For this 2013a). reason, the recommendations do not address education policies that This Section has three objectives:(i) to target individuals still attending formal assess the extent of skill18 barriers for schooling. Box 4.1 The Turkish Education System The Turkish education system follows a 4-4-4 structure. Primary education comprises grades 1-8 (first and second levels), followed by four years of secondary education, equivalent to high school education. Until 1997, children in Turkey were obliged to take five years of education, while the reforms of 1997 introduced compulsory education for eight years. New legislation introduced in March 2012 prolonged compulsory education to 12 years. The number of high school years increased from three to four in 2004. Universities provide two to four years of education. Vocational and higher education is offered by multiple providers: teacher training colleges, technical and vocational training schools, and universities. Vocational tracks can start in secondary or higher education. A completed secondary education is required to pursue higher education. 18- The classification of skills broadly follows the distinction introduced by the World Bank 2012: “cognitive,” “technical,” and “soft.” These categories may not necessarily be mutually exclusive. Cognitive skills refer to those demonstrated by intellectual ability and command of academic subjects. Technical skills refer to those tasks that are required by specific occupations. Soft skills refer to a person’s attitudes and behaviors, like timeliness, accuracy, and the like (World Bank 2012a). Later in the chapter, we use the skills definition of a well-cited study by Autor, Levy, and Murnane (2003). 19- The conceptual framework used to analyze employment and skills in Turkey relies on the economic concepts of demand and supply. Supply refers to the skills that young people (individuals) have and make available to employers. Demand refers to the skills that employers (firms) seek in the labor market. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 79 The education of the Turkish population benefits of the reform have outweighed has increased at all levels, albeit its costs. The share of high school slowly. The 1997 educational policy graduates increased 1 percentage that increased the duration of primary point during the last 10 years, but this is education from five to eight years averaged across general and vocational contributed to an improved educational high school tracks. The share of general profile of the population. Roughly high school graduates decreased from 62 percent of the WAP held at most 9.5 percent in 2000 to 8.3 percent in a primary education degree in 2011 2011, while the share of vocational high (Figure 4.1), compared to 72 percent in school graduates increased from 4.4 2000. The change in education policy percent to 6.9 percent between 2000 resulted in a slight increase in the and 2011. These dynamics are also number of primary education dropouts, linked to the government’s policy to and women were affected more than increase access to and enrollment in men. The share of women without any vocational training. The share of tertiary degree increased more than twofold education graduates almost doubled in between 2000 and 2011 (from 3.9 the last 10 years, from 5.7 percent to percent to 8.5 percent). The impact 10.3 percent (TUIK, Household Labor was much less for men. However, the Force Survey 2011). FIGURE 4.1 Educational profile of Turkey’s WAP, 2011 Source: Turkey’s Household Labor Force Survey 2011. 80 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY The skill transition is critical for Turkey, age. For instance, the share of youth as low educational attainment can with at most eight years of education constrain the movement of workers is 58 percent for individuals between into employment. The transition of 20 and 24, compared to 78 percent for workers from inactivity to productive individuals between 50 and 54 years and higher earning jobs will require old. Despite these changes, the skills a larger share of the workforce to of the WAP are still very basic and the acquire higher-level skills. This is situation is worse for women. Turkey already happening. Younger cohorts needs to continue increasing the skills are more educated than the previous of its population to catch up with other generation. This can best be seen when OECD countries (OECD, 2012). Currently, education information is disaggregated by age group (Figure 4.2). The share it is ranked almost last in secondary of individuals with at most eight years education attainment compared with of education increases with age, while other OECD countries (Figure 4.3). the share of individuals with at least a Further, educational challenges remain high school education decreases with a concern for women particularly. FIGURE 4.2 Education level by age, 201120 Note: Excludes those still in school. Source: Turkey’s Household Labor Force Survey 2011. 20- The education distribution presented in Figure 4.2 excludes those still in school. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 81 FIGURE 4.3 Percent of labor force that has attained at least secondary education in OECD countries 25-34 year-olds 55-64 year-olds Notes: Year of reference 2000/2002 and 2009. Source: OECD 2012. Women have improved their skills dropped out of the education system acquisition over the last 10 years, but before completing primary education, they are still at a disadvantage when while only 4 percent of men did so. compared to men. In 2000, 21 percent Overall, women are less likely to of women had at most eight years of reach high school and pursue higher education; by 2011, this share had education degrees than men. increased to 53 percent. Despite the gains, women’s educational levels are The gender gap is closing fast, however. much lower than those of men. First, as The growth rate for women’s education of 2011, a higher proportion of illiterate completion is faster than that of men. women participated in the labor force For instance, the share of females with than illiterate men (13 percent and 2 a primary education degree almost percent, respectively). Second, 8.5 doubled between 2000 and 2011, while percent of women in the labor force the share of males with a primary 82 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY education degree increased only by half with a low quality and quantity of skills, during the same period. Similarly, the an impediment that must be overcome share of female high school graduates if their labor market conditions are to increased 14 percent compared to 7 be improved. percent for men. Equally impressive is the jump in the share of female 4.1.2 Workforce Labor Market tertiary education graduates, which Profile almost doubled, from 4.4 percent to 8.6 percent. Finally, the share of females It is important to know the economic still in school has almost caught up with activities in which the Turkish population that of males due to a sizeable increase is engaged. Jobs are important at the in female enrollment (74.2 percent). The individual level for living standards and share of women between the ages of 20 at the country level for social cohesion and 24 with tertiary education degrees and productivity (World Bank 2013a). is 11.3, while the share of men is 9.1 Working people can be divided into percent. The higher levels skills that four main categories: wage earners women are acquiring is an important in the formal and informal sector; development for activation as women entrepreneurs; the self-employed; and with higher education have much higher unpaid family workers. Most unpaid rates of labor force participation. family workers and informal sector workers are employed in agriculture. In addition to the improvements in Entrepreneurs are those who own educational outcomes for women, the a business and employ at least one education system in Turkey has shown person. The self-employed are those substantive progress in the quality who own a business without employing of education, as measured by PISA anyone else. Two other groups that assessments between 2003 and 2009. classify the WAP include: (i) individuals The scores of Turkish students in the who are inactive, not in education or three PISA disciplines (reading, math, training,or not looking for jobs; and and science) increased the equivalent (ii) the unemployed. The employment ofa half year of additional schooling categorization excludes individuals who in each of the tested subjects (Aedo, are still in school. An overall picture of Hassan, and Cahu 2013). In sum, young Turkey’s WAP is given in Figure 4.4. people entering the labor market (the While the bulk of the population is “flow”) have more and higher quality engaged in either formal or informal skills than those already in the labor wage employment, self-employment, market (the “stock”). A serious concern or entrepreneurship activities, over four is how to upgrade the skills of the large in ten (18.1 million) in the Turkish WAP stock of the existing WAP endowed are inactive. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 83 FIGURE 4.4 Breakdown of Turkey’s WAP by employment status21 Source: Turkey’s Household Labor Force Survey 2011. Turkey’s population has moved more population in formal wage employment towards wage employment over the increased 6 percentage points over the last decade. The proportion of people last 10 years, compared to an increase in formal and informal employment of less than 1 percentage point in increased from 19 percent and 7 informal employment. Rates of wage percent, respectively, in 2000 to 24 employment among men and women percent and 8 percent in 2011. At the are very similar, although during this same time, the share of the population time period, more women held formal in self-employment and unpaid family wage jobs than men.This growing work in 2000 decreased from 12 percent and 11 percent, respectively, to share of wage employment reflects the 9 percent and 7 percent in 2011.Most rise in skill-intensive sectors that were people in wage employment moved into able to absorb a substantive part of the formal employment. The share of the growing number of educated youth. 21- This refers to individuals between 15 and 64 years of age, excluding those in school. 84 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY FIGURE 4.5 Labor market status by education level, 2011 TOTAL 100 % 90 % 80 % 70 % 60 % 50 % 40 % 30 % 20 % 10 % 0% Over all Formal wage Informal Entrepreneur Self- Unpaid family Unemployed Inactive earner wage earner employed worker Illiterate Not a grad Primary school Primary educ or junior high High school Vocational high school Tertiary education Source: Turkey’s Household Labor Force Survey 2011. Higher levels of education are associated tertiary degrees; this share increased to with formal wage employment. A 30 percent in 2011. Similarly, the share third of formal wage earners have of individuals with at least eight years tertiary education degrees, another 26 of education, or primary completed, percent are high school graduates, and increased from 23 percent to 43 percent 42 percent have completed primary among informal wage workers in the school. There are very few illiterate same period. Vocational high school individuals or those with incomplete graduates increased their shares in primary education in formal wage all market groups as well. The share employment. The skills gained by the of entrepreneurs with a vocational newer generations are reflected in the high school degree increased from higher share of educational attainment 6.1 percent in 2000 to 10.5 percent in among the population in all employment 2010. The skills of workers varied by categories, suggesting that, increasingly, employment status. Self-employed the workforce has the skills required workers, unpaid family workers, and for good jobs. For instance, 21 percent informal wage workers had, on average, of formal wage earners in 2000 had lower levels of education than formal ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 85 wage workers, entrepreneurs, and seem to promote early retirement unemployed individuals. This suggests among workers, yet a significant that highly educated professionals may proportion of retirees continue to work prefer to be unemployed and wait for informally. Overall, informality may job opportunities rather than joining provide a school-to-work transition those in informal wage employment, as for young educated individuals, while the occupations of each of these groups concurrently offering post-retirement differ. individuals possibilities for continued employment. Women working in formal wage employment were more likely to have The educational profile of the a tertiary education degree compared unemployed is mixed (Figure 4.6). to men (50 percent versus 24 percent). Individuals with below primary This is linked to the high labor force education have very low levels of unemployment although their participation rate of women with college unemployment rates increase with degrees (71 percent). Along the same age. People with low skill levels usually lines, unemployed females were better belong to poorer households and educated than their male unemployed cannot afford to be unemployed; they counterparts. The lower female have to work to make ends meet. Across educational attainment can be seen in all age groups, unemployment was the groups of women in unpaid family highest among individuals with primary work and in inactivity; they had much education only. Unemployment was lower education levels than men in the lowest among those individuals with same groups. Working women were high school completed (general and also younger than men on average, vocational). Unemployment rates for suggesting that women drop out of the vocational graduates were lower than labor force earlier, for family reasons or those of high school graduates across earlier retirement. all age groups. Over the last 10 years, unemployment rates amongst higher The relationship between age and the education graduates almost doubled. probability of being an informal wage The unemployment rates decreased earner is U-shaped: younger and older for those individuals who had general people were more likely to be employed high school completed, and remained in the informal sector than those middle- the same for those with vocational high aged. Turkey’s retirement age is one of school completed. This suggests an the lowest among OECD countries.22 increased demand for skills provided by Labor regulations in Turkey, along with high school and a decrease in demand very generous retirement packages, for higher-level skills. 22- The average retirement age in OECD countries is 64 for men and 63 for women (OECD 2012). 86 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY FIGURE 4.6 Unemployment rates (%) by education level, 2000 and 2011 Source: Turkey’s Household Labor Force Survey 2011. In addition to unemployment, the this. Possible reasons could vary from inactivity of Turkey’s WAP is also family responsibilities, cultural and important. Around 40 percent of social norms, topossibly economic Turkey’s WAP is inactive, and inactivity motivations like low wages and low rates are higher among youth and employment quality for those women women. The large number of inactive who do not reach higher levels of people in Turkey partly reflects the education.23 low level of women’s labor force participation. At 29 percent, Turkey has Among young people, inactivity rates by far the lowest female participation are similar for those with both lower and rate of all OECD countries. More higher levels of education. However, as research is required to understand the age increases, inactivity rates rise for different contributor factors behind those with lower levels of education and 23- Girls in Turkey have solid cognitive skills; they perform equally well as or even better than boys on standardized tests like PISA and TIMSS (PISA 2009 and TIMSS 2011). ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 87 decrease for those with higher levels of Bank 2013a). In particular, the high education. For instance, inactivity rates unemployment among youth and the for 20- to 24-year-olds with at most relatively low unemployment among eight years of education are 67 percent high school and higher education and 66 percent for those with at least graduates suggest a slow school-to- eight years of education.In contrast, for work transition. Possible explanations 40- to 44-year-olds, the inactivity rate is include information failures and unclear 87 percent for those with at most eight pathways to link school graduates with years of education, and only 21 percent the world of work. for those with at least eight years of education. Inactivity also captures the 4.1.3 Evidence of a Skills Barrier to discouraged workers, those that have Employment given up looking for jobs. Almost every second person in Turkey (48 percent of The changing skills profile of the the population) is either unemployed or Turkish labor force is slowly starting inactive, suggesting that while Turkey to reflect employers’ needs. While the can take advantage of its increasingly evidence is hard to quantify, a number abundant labor, it is not fully utilizing of factors suggest a mismatch between it and more policies are required to the demand for and supply of skills: activate people into employment. (i) unemployment rates are relatively Upgrading the skills of the “stock” of lower for high school graduates and workers could be a critical factor in higher for primary and tertiary education these sets of policies. graduates; (ii) returns to education are high for vocational and higher education There may be a number of reasons levels; (iii) the occupational profile of the for the inactivity and unemployment labor force has changed over the last situation: (i) a mismatch between the decade; (iv) demand for different types skills required by employers and those of skills (e.g., an increased demand for offered by the population; (ii) lack routine cognitive skills and a decline in of jobs and a weakened or specific demand for routine manual skills) has demand for skills as a result of the recent emerged over the last decade; and economic crisis;(iii) market distortions (v) the type and number of available (e.g., high-paying career jobs in the vacancies has changed. A discussion of public sector); (iv) lack of incentives to each of these five factors follows. invest in education and training; and (v) market failures represented by lack of Returns to education or poor information about employment opportunities, transportation, housing Rates of return can provide evidence of market challenges, and the like (World the extent to which education pays off in 88 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY terms of higher earnings. Private rates of returns to education decreased for return were estimated using data from both high school and tertiary education the Household Labor Force Survey 2011 graduates. Data indicate that there was and applying Mincerian regressions.24 not a significant change in returns to Among those with wage jobs, the rate education in elementary occupations of return to education increases with between 2003 and 2011. higher education degrees. In Turkey, tertiary education graduates earn Returns to education changed higher wages than secondary education dramatically by formality status. graduates, and secondary education Figure 4.7 provides data on returns graduates earn higher salaries than to education for formal and informal primary school graduates. In the formal workers after controlling for occupation labor market, returns to a high school and industry of employment. Returns to degree decreased, while returns to a education were estimated in reference tertiary degree increased significantly to a wage earner with at most a primary between 2004 and 2011. When education degree. High school and controlling for occupation and industry tertiary education graduates earned a of employment, there was no change in premium regardless of their formality the returns to tertiary and high school status. In 2011, tertiary education degrees. graduates employed in the formal market enjoyed a premium of 44.9 Looking at occupations, returns to percent over individuals who had at education are higher for individuals most a primary education degree. working as legislators, senior officials, In the informal sector, the higher and managers (occupations usually education premium decreased to 18.1 associated with non-routine cognitive percent.25 Formally employed high skills), although this pattern shows a school graduates earned a premium of decreasing trend between 2003 and 15.3 percent, whereas the informally 2011. Further, for wage earners who employed earned 6.1 percent more than work as clerks (an occupation associated individuals with only a primary degree. with routine cognitive skills), the returns Returns to education in the informal to high school and tertiary education sector remained similar between 2004 graduates increased significantly and 2011. The high returns to tertiary between 2004 and 2011, suggesting an and high school degrees reflect the increase in demand for their skills. For relative demand for these skills in the technicians and associate professionals, labor market. 24- Mincerian regressions were estimated using educational attainment categories of“at most primary education” (illiterate, not a graduate, primary school, junior high and primary education), “high school” (secondary education), and “at least tertiary education.” 25- These premiums were 76.8 percent and 29.1 percent when industry and occupation were not controlled for. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 89 FIGURE 4.7 Returns to education in formal and informal labor markets (controlling for industry and occupation) 60.0 % 50.0 % 40.0 % 30.0 % 20.0 % 10.0 % 0.0 % 2003 2004 2005 2006 2007 2008 2009 2010 2011 Informal HS Informal Tert Formal HS Formal Tert Source: Turkey’s Household Labor Force Survey 2011. Changes in the demand for skills been on the rise in the U.S. since 1960. types The changes from lower to higher productivity economic activities are In a well-cited study, Autor, Levy and Murnane (2003) showed that the reflected in changes in occupations and demand for non-routine cognitive skills. (See Table 4.1 for a summary of (analytical and interpersonal) skills has skills types.) 90 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY TABLE 4.1 Summary of skill types Source: Aedo et al.2013. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 91 Turkey’s occupational profile reflects levels. a slow rise in the job’s content of routine cognitive skills and non-routine Rise of routine cognitive skills cognitive (analytical and interpersonal) skills26 (see Figure 4.8). The content of Rates of return to non-routine cognitive routine manual and non-routine manual skills are higher than those to routine physical skills in jobs have been on cognitive, routine manual, and non- the decline. Surprisingly, for younger routine manual skills. The non-routine cohorts, the content of non-routine cognitive skills earned very high rates manual physical and routine manual of return; interpersonal skills earned skills in jobs substantially increased relatively high rates of return; routine in 2008, partly due to the economic cognitive skills earned an average rate crisis and the increasing number of of return; and routine manual skills tertiary education graduates, which earned a relatively low rate of return. almost doubled in the last 10 years,27 This pattern was consistent across pushing young people into lower-skilled different age cohorts and levels of occupations. By 2010, the occupational education. The analysis suggests that content of cognitive skills had declined an increasingly more educated youth and non-routine manual physical skills, are at a disadvantage when entering the routine manual, and routine cognitive labor market because they go into low- skills job content is only now starting earning and low-skilled occupations, to increase again. Data suggest that mostly requiring non-routine manual younger generations are more likely to physical and routine manual skills, break into the labor market by working which affects their future jobs prospects in occupations that require lower skill and earnings. 26- Skills are classified as routine-manual, non-routine manual physical, routine cognitive, non-routine cognitive interpersonal, and non-routine cognitive analytical.The five skills measures by Acemoglu and Autor (2010) and Autor, Levy and Murnane (2003) are constructed by aggregating skill measures defined as follows:Routine- manual skills are characterized by repetitive movements requiring physical abilities used in labor-intensive agricultural production (manual harvesting), specific trades (brick-layers), construction workers, and for specific machine operators (e.g., textile workers), or for assembly lines (e.g., electronic equipment). Non- routine manual physical skills are characterized by the ability to vary and react to changing circumstances on a continuous basis. These skills are required by operators of heavy equipment in agriculture, industry, or construction and electricians. Routine cognitive skills are characterized by the ability to carryout repetitive, non-physical tasks – call center operators or book keepers would use these intensively. Non-routine cognitive analytical skills consist of thought processes requiring the absorption, processing, and decisionmaking of abstract information. Occupations requiring these skills include computer programmers, teachers, lawyers, doctors and nurses, and the like. Non-routine cognitive interpersonal skills are characterized by personality traits that underlie behaviors such as teamwork, reliability, discipline, and work effort. These are important for professional occupations as well as services which establish direct client contact (Aedo et al. 2013). 27- From 5.7 percent to 10.3 percent. 92 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY FIGURE 4.8 Evolution of standardized skills in the year of reference Source: Aedo and Moreno, 201328 28- Data used in the analysis was Turkey’s Household Labor Force Surveys 2004–2010. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 93 The analysis tracks the skills content of some of these skills like non-routine current, actual jobs in Turkey. It shows cognitive analytical have increased in that overall, and on average, the skills of Turkey faster than in other countries. the different tasks are starting to have a higher proportion of routine cognitive, Changes in occupational profile non-routine cognitive analytical, and non-routine interpersonal components Employers’ demands for skills reflect than before—higher level skills usually the ongoing sectoral transition in Turkey, associated with value-added. The which now favors secondary (high analysis does not say if more of these school) graduates over those with lower skills are being demanded but not degrees as compared to the situation a offered or if these skills are in excess few years ago. Overall, a slow but steady supply. decline of underqualified individuals is observed in all employment categories. This change in the labor market can The fastest decline of underqualified also be seen in the graphs below where individuals was in the entrepreneur the higher level skills content of jobs category, whose share fell by 10 (non-routine cognitive analytical, for percentage points over the decade. example) have increase in Turkey and The share of overqualified individuals in many other economies. The trends increased among technicians, clerks, in skills utilization (Figure 4.9) show that and skilled agricultural and fishery many other countries with higher levels workers, while the share of overqualified of growth (like the U.S.) and some individuals declined among service countries with high growth rates (like workers (shop and market sales Brazil) have a much higher demand for workers). A steady decline in the share non-routine cognitive analytical and of underqualified individuals over non-routine cognitive interpersonal time was observed in manufacturing skills than Turkey does (Aedo et al. and transportation, communication, 2013). However this is deceptive as the and storage, as well as in community, base and final year substantially vary social, and personal services. However, for the countries. Taking the number of the share of under qualified individuals years in consideration, we can see that in utilities (except telecom), finance 94 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY FIGURE 4.9 Trends in skills utilization29 Source: Aedo et al. 2013. 29- Base and final years are as follow: USA: 1980–2008; Chile: 1992–2009; Costa Rica: 2001–2008; Brazil: 1981– 2009; Poland: 2002–2010; Turkey: 2004–2008; Sri Lanka: 2000–2008; India: 1994–2010. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 95 industries, and business services percent of 215,243 positions), followed increased.30 These latter sectors have by construction (12.2 percent), and experienced a substantial increase in wholesale and retail trade sectors labor productivity in recent years and (11.6 percent). The lowest job vacancy may therefore be demanding higher rates were in the water supply sector, skill profiles which are not immediately sewage waste management and supplied, resulting in a detected under improvement activities, and other basic qualification of workers. utilities. The three occupations with the most vacancies were craftsmanship, These observations are consistent facility and machinery operators, and with the rise in the demand for routine assemblers, while the two occupations cognitive skills and the number of with the least number of vacancies were vacancies (discussed below). In all agriculture, forestry, and aquaculture sectors combined, a growing share of workers and managers. The vacancy higher qualified individuals aregoing profile is consistent with the picture into occupations that seem to already described: few openings exist require less skill. Recent university for university graduates, since most graduates may expect to be employed vacancies require only vocational or in occupations where they do more general high school level qualifications. routine cognitive work rather than non- routine cognitive work. The overall The skills most required for vacant increase in the share of over qualified jobs were: occupational knowledge, individuals may reflect the rising physical qualifications, teamwork, education levels of the WAP as well communication skills, and problem as the need for more dynamism in the solving. Knowledge of a foreign skill-intensive sectors that can absorb language was the least requested more highly educated youth. qualification. The high demand for occupational and physical qualifications Vacancies reflects the need for routine cognitive and manual skills. Educational levels are Vacancies are mainly recorded in not relevant for 36 percent of the vacant occupations that require skills that can jobs, as 27.5 percent of all vacancies be provided by high school graduates. require less than a high school degree, In 2012, the bulk of vacancies were 14.6 percent require vocational training in the manufacturing sector (36.2 (9.5 for vocational high school and 30- The analysis was done by choosing the median years of schooling of the formal wage earners to be the educational norm, as they are the ones that hold “good jobs” and constitute the majority of those that are employed. The median formal wage earner is a high school graduate, a situation that has not changed over the last 10 years. Individuals are underqualified if they are not high school graduates and overqualified if they have more than a high school degree. 96 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 5.1 percent for a vocational school through university education, probably of higher education), 10.7 percent explaining the high unemployment require general high school, and rate for those with a tertiary education. only 0.3 percent require a bachelor’s Further, about half of Turkey’s degree (Government of Turkey 2012). university graduates major in social Interestingly, there is an important sciences, as opposed to engineering segment of vacancies that require only and other applied sciences. Majors a high school or vocational degree, are, for the most part, gender neutral, as well as some that only require equally pursued by men and women, graduates to have a primary education. with the exception of engineering and As there are unemployed individuals manufacturing, where most graduates with these qualifications looking for job are men. Combined, the information opportunities, information asymmetries on vacancies, demand for new skills, may be limiting jobseekers’ mobility from unemployment and inactivity to and unemployment rates suggests employment. that high school graduates (general and vocational) are more likely to find The reported vacancies exhibit little employment quickly than are university- demand for higher-level skills obtained educated graduates. FIGURE 4.10 Majors pursued by male and female higher education graduates in Turkey, 2010 Source: UNESCO 200031 31- Accessed at http://stats.uis.unesco.org/unesco/TableViewer/document.aspx?ReportId=136 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 97 Employers report not finding the use of online websites, newspapers, employees with the skills they need. In and private employment agencies. This a survey administered to 1,152 firms in situation is problematic, as finding a Turkey in 2008, business owners and job is based more on who one knows top managers cited an inadequately than on a meritocratic process that educated workforce as one of the top rewards skills. The mismatch between five constraints to business (Figure vacancies and unemployment rates 4.11). Further, employers mentioned suggests a combination of a lack of that occupational groups that require skills in people with primary education, qualifications and skills were hard to as well as a combination of lack of recruit. The most common reason relevant skills in high school and higher for not filling vacancies was “lack of education graduates and information employees with requested occupational failures between employers and skills/qualifications,” followed by “lack employees. The good news is that of employees in the related occupation” and “lack of employees with necessary many unemployed people could experience. ”The relatives-friends move to wage employment if these channel wasthe most common way to information channels were improved. recruit—over half of employers tried to Turkey’s government is actively working hire through this channel—followed by to address this. FIGURE 4.11 Top 10 business environment constraints Access tı finance 25.9 Tax rates 18.2 Political instability 17.5 Practices of the informal sector 14.7 Inadequately educated workforce 9.1 Electricity 3.0 Business licenses and permits 2.5 Customs and trade regulations 2.0 Corruption 2.0 Labor regulations 2.0 0 5 10 15 20 25 30 Percent of Firms Source: Enterprise Survey: Turkey32 32- Retrieved June 4, 2013 from http://www.enterprisesurveys.org/Data/ExploreEconomies/2008/turkey#workforce 98 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 4.2 A Solution From the programs in workforce development. Education System There is also a well-developed skills testing process for major occupations. There is strong awareness of the Private provision of training is well- importance of training in Turkey. recognized and regulated, and Preliminary results of the Systems employers share their demands and Approach for Better Education Results priorities for WfD with the government (SABER-WfD) instrument on workforce and public. development (WfD) in Turkey show good governmental and non-governmental The Government of Turkey is working championship for the cause at the on six priority areas as part of its national and regional levels. Through national Action Plan on Linking Social different ministries, the Government Support System to Employment and its of Turkey has taken concrete steps to Activation33 to reduce unemployment set a strategic direction to improve the and inactivity and to improve the match skills of the population and reduce the between skills demand and supply. The demand-supply skills mismatch. First, in government identified three critical 2009, the government issued an action constraints for skills development: (i) plan that articulates the link between lack of relevance of training programs; employment and the vocational (ii) insufficient on-the-job training education and training (VET) system. activities; and (iii) limited chances for Second, a national employment strategy students to meet with experts in their being prepared has special emphasis on field. To address these, the Government workforce development. Third, İŞKUR of Turkey is working on six priority areas offers learning programs focused to: on literacy and skill development for upskilling Turkey’s WAP . Fourth, with 1) Create a national competency support from a European Union grant, framework. Turkey has instituted the Ministry of Education implemented the framework for occupational a project to improve the quality of the standards and qualifications. VET system. Fifth, İŞKUR initiated a However, certification mechanisms new program called Job Consultants in and appropriate regulations are still 2011, whereby 4,000 job and vocational being developed. counselors (JVCs) will provide support to employers and the unemployed to 2) Ensure quality in vocational and help match supply and demand in the technical education through job market. SABER reveals that there introduction of national are relevant, solid standards for quality professional standards for 33- The purpose of the Activation plan was to connect Turkey’s social assistance (SA) system and the public employment serviceto support the employment activation of SA beneficiaries. ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 99 education programs. National pathways for skills acquisitions have competencies are being developed been developed, the certifications to measure graduates’ knowledge from VET are not aligned with the and skills and an accreditation National Qualifications Framework. process is being initiated for VET institutions. However, there are 6) Create open lifelong learning few measures in place for quality environments and improve the control of the training available in effectiveness of active labor market the education market. policies (ALMP). This implies increasing the number of people 3) Develop a labor force suitable who benefit from entrepreneurship for the current labor market. training and distance education. In The government has introduced addition, a newly created database several labor market needs surveys, with relevant active employment conducted in coordination with labor programs will facilitate access to and employer unions, professional information. Some of these goals organizations, and NGOs, that have already been achieved, as periodically assess the demand some jobseekers are redirected to for skills. Nonetheless an incentive the İŞKUR database, however more system for engaging in skills- information on programs is required upgrading for workers is lacking in as well as analysis of the impact of the system. these policies (World Bank 2013b). 4) Improve the environment for VET The SABER-WfD diagnostic tool by working on the qualifications demonstrates that Turkey lags behind of educators and designing a new other countries such as Ireland, school management system that Singapore, and Chile, which have will increase the administrative more established systems. The SABER capacities and management of framework for the analysis of WfD higher education institutions. There ranks a country in three dimensions: is a wide mix of training providers, (i) Strategic framework assesses and more incentives are required the advocacy, championship, and for the private sector to increase coordination of the objective of WfD students’ access to on-the-job and with national priorities; (ii) System workplace training. oversight refers to provisions regarding funding, quality, and learning pathways 5) Harmonize vocational guidance that affect individual, employers, and and consultancy services and training providers; and (iii) Service put into practice the tools for delivery encompasses the diversity, mobility proficiency and vocational organization, and management of education. Although some of these training provision (World Bank 2013b). 100 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY Turkey is ranked as emerging—showing those sectors and to induce high-skilled some instances of good practice—on individuals to move into them. Turkey’s the three dimensions; in 2010, Chile was government is actively committed to ranked established, showing systemic improving the skills of its workforce in good practice, while Ireland was ranked a way that supports Turkey’s transition advanced in strategic framework from low-productivity activities to high- and service delivery and established productivity occupations. in system oversight in 2000 (World Bank 2013c and 2012b). The SABER 4.3 Policy Options to diagnostic identifies several challenges Overcome the Skills Barrier that the Turkish system needs to address to improve the elements of its This report presented a brief profile of WfD system. Turkey’s WAP , as well as an analysis of the evolving demand for and supply of First, the funding mechanism for skills in Turkey over the last 10 years. On Lifelong Learning (LLL) needs to one hand, there seems to be a growing be strengthened, as low funding of demand for high school vocational and Turkey’s Vocational Education and general graduates and, in the long run, Training (TVET) system institutions for tertiary education graduates. On (which receive 2.7 percent of education the other, over 60 percent of Turkey’s expenditure) often results in low-quality, population has completed only a supply-driven programs, distant from primary education degree; these the private sector’s needs (World Bank people need more skills to successfully 2013b). Second, recognition of prior join the labor market. Youth, women, learning could be better organized. and those with only a primary New jobs will require new skills and education are much more likely to individuals currently inactive will need face high unemployment and inactivity to upgrade theirs to be relevant in the rates. Education is a key determinant market. LLL works well when students of individuals’ future labor market are able to transfer across courses and outcomes. While the government is to access higher levels of education actively working on improving the and training (World Bank 2013b). Third, relevance and quality of VET in Turkey, there is limited awareness of the skills more work is required to reduce existing constraints in priority sectors, and no information asymmetries and to induce incentives for employers to upgrade individuals to shift from inactivity employees’ skills. Education, research to work. Many individuals with the and development, IT, health, and required skills seem to be waiting to find transportation are some of Turkey’s opportunities to join the labor market, key sectors. More effort is needed to while at the same time, employers are improve the skills of people already in looking for these individuals. Three ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY 101 optionsto scale up the skills of the WAP and entrepreneurial skills. This is could be considered by policy makers: important as the private sector is increasingly looking for these types 1) Continue working on diversifying of soft skills; additionally, many the pathways for skills acquisition. more highly educated people are Increasingly, the global economy becoming entrepreneurs due to lack will require individuals to adopt of formal opportunities. Examples of and learn new skills as new jobs are success of such training programs created and others are destroyed. include the Empowering Adolescent Diversifying pathways for skills Girls in Uganda (Bandiera et al. acquisition include vocational 2013). The idea is to find ways that degrees to be recognized by the increase the value-added of the National Qualifications Framework, courses provided by the different but also to invest in general high government agencies. Improving the school. The current situation in relevance of vocational programs Turkey seems to reflect a tradeoff will allow a smoother transition into between pursuing vocational the labor force for youth, and for the education and further general stock of the population a retooling studies. Using data for 18 countries to better meet current and emerging from the International Adult Literacy job market needs. Survey, Hanushek, Woessmann and Zhang (2011) found that gains in 3) Develop incentives for the private youth employment from graduates sector to get further involved in of vocational programs are offset by vocational training to make it more limited employability later in life, as demand-driven. The example of graduates from vocational programs Ireland (World Bank 2012b) shows have limited skills. Promoting that the functions of the industry linkages between TVET and general and business stakeholders were university programs can help lower much more than just consultations. the barriers to building skills over a Industry leaders were key players lifetime. in the training review process and had executive authority in setting 2) Complement vocational programs and implementing aspects of WfD. with life skills training to facilitate LLL Incentives could be placed so that and increase the impact of training. the private sector in Turkey gets In addition to building technical further involved in the development programs that respond to the needs of curricula of training programs. of the private sector, vocational Other incentives could target programs could be complemented mechanisms for employers to invest with training on behavioral, life, in upgrading employees’ skills. 102 ACTIVATING VULNERABLE PEOPLE INTO GOOD JOBS IN TURKEY References Acemoglu D. and D. Autor. 2010. “Skills, Hanushek E., Woessmann L., and L. Zhang. Tasks and Technologies: Implications 2011. “General Education, Vocational for Employment and Earnings.” NBER Education, and Labor-Market Outcomes Working Paper No. 16082. Cambridge, over the Life-Cycle.” NBER Working MA. Paper No. 17504. Cambridge, MA. Aedo, C., N. Hassan, and P. Cahu. 2013. OECD. 2012. Education at a Glance 2012. Promoting Excellence in Turkey’s OECD Indictors. Retrieved from: http:// Schools. Washington, DC: World Bank. www.oecd.org/edu/eag2012. Aedo, C., Hentschel J., Luque J., and M. PISA. 2009. Key Findings. Retrieved Moreno. 2013. From Occupations to June 8, 2013 from http://www.oecd. Embedded Skills: A Cross-Country org/pisa/pisaproducts/pisa2009/ Comparison. Washington, DC: World pisa2009keyfindings.htm Bank. TIMSS. 2011. Retrieved June 8 2013 from Aedo, C., and Moreno, M. 2013. Skills http://timss.bc.edu/timss2011/ Analysis in Turkey 2004–2010. Turkish Statistical Institute, TUIK, Turkstat. Washington, DC: World Bank. Household Labor Force Survey 2000– Autor D., Levy F., and R. J. Murnane. 2013. 2003. The Skill Content of Recent World Bank. 2012a. “Stepping Up Skills for Technological Change: An Empirical More Jobs and Higher Productivity.” Exploration. Quarterly Journal of Washington, DC: World Bank. Economics 118(4), November: 1279- 1334. World Bank. 2012b. “SABER-Workforce Development. Ireland 2012.” Bandiera O., Buehren N. , Burgess R., Washington, DC: World Bank. Goldstein M., Gulesci S., Rasul I., and M. Sulaiman. 2013. Empowering World Bank. 2013a. World Development Adolescent Girls: Evidence from a Report. Jobs. Washington, DC: World Randomized Control Trial in Uganda. Bank. Available at: http://econ.lse.ac.uk/staff/ World Bank. 2013b. “SABER-Workforce rburgess/wp/ELA.pdf Development. Barometer Report. Enterprise Survey. 2008. Enterprise Survey: Turkey 2012.” Washington, DC: World Turkey. Retrieved June 1, 2913 from Bank. http://www.enterprisesurveys.org/ World Bank. 2013c. “SABER-Workforce Government of Turkey. 2012. İŞKUR Labor Development. Chile.” Washington DC: Market Analysis. World Bank.