KOSOVO COUNTRY REPORT: FINDINGS FROM THE SKILLS TOWARDS EMPLOYMENT AND PRODUCTIVITY SURVEY March 2019 Report No: AUS0000290 Document of the World Bank Education Global Practice Social Protection and Jobs Global Practice Europe and Central Asia Region 2 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 3 Contents ACKNOWLEDGEMENTS.................................................................................................. 8 OVERVIEW...................................................................................................................... 9 Skills matter for Kosovo......................................................................................................9 Key findings from the STEP Household and Employer Surveys......................................10 Going forward: policy reforms to reduce skill gaps in Kosovo........................................12 Addressing inequity and quality problems in skill development..............................13 Connecting supply and demand: firms, students, and training institutions............14 Reducing information gaps........................................................................................16 1 SKILLS MATTER FOR KOSOVO................................................................................... 17 1.1 Employment challenges in Kosovo...........................................................................18 What types of jobs are available?.............................................................................18 © 2019 The World Bank What kind of workers?...............................................................................................19 1818 H Street NW, Washington DC 20433 A strategy to improve jobs.........................................................................................22 Telephone: 202-473-1000; Internet: www.worldbank.org 1.2 The role of skills..........................................................................................................22 Some rights reserved Skills influence growth, productivity and job creation..............................................22 A broad and flexible skill set is important.................................................................24 This work is a product of the staff of The World Bank. The findings, interpretations, and Measuring skills - the basics of the Kosovo STEP surveys......................................26 conclusions expressed in this work do not necessarily reflect the views of the Executive Directors 1.3 Organization of the report...........................................................................................29 of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other 2. SKILL GAPS FROM THE FIRM PERSPECTIVE........................................................... 30 information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. 2.1 Lack of skills are a constraint on growth but not the primary concern of firms....30 2.2 Skill constraints are holding back job creation.........................................................32 This publication was produced with the financial support of the European Union. Its contents 2.3 Skill gaps restrain more dynamic firms....................................................................36 are the sole responsibility of World Bank staff and do not necessarily reflect the views of the European Union. For the European Union, the designation of Kosovo is without prejudice to 3. THE SKILLS THAT MATTER IN KOSOVO.................................................................... 39 positions on status and is in line with UNSCR 1244/1999 and the ICJ Opinion on the Kosovo Declaration of Independence. 3.1 Skills needed on jobs..................................................................................................39 Rights and Permissions 3.2 Skills valued by employers........................................................................................41 3.3 There are no perceived skill gaps in the employed workforce.................................43 The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for 4. WHAT ARE THE CONSEQUENCES OF SKILL noncommercial purposes as long as full attribution to this work is given. GAPS FOR KOSOVO’S POPULATION?............................................................................ 46 Attribution—Please cite the work as follows: World Bank. 2019. Kosovo Country Report: Findings from the Skills towards Employment and Productivity Survey. World Bank, Washington, DC. 4.1 Skill gaps limit access to good jobs..........................................................................46 4.2 Skill gaps reinforce social and economic exclusion.................................................50 All queries on rights and licenses, including subsidiary rights, should be addressed to World Bank 4.3 Skill and gender gaps.................................................................................................57 Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202- 522-2625; e-mail: pubrights@worldbank.org. 4 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 5 5. WHAT CAUSES SKILL GAPS IN KOSOVO?................................................................ 62 Figures 5.1 Potential causes behind skill gaps............................................................................62 5.2 Access to quality education and training..................................................................63 FIGURE 1: The Kosovo population has moderate levels of Early skill formation...................................................................................................63 education with significant gender differences, 2016..................................20 Formal education.......................................................................................................64 Training opportunities for job seekers.......................................................................71 FIGURE 2: Access to employment and access to education are linked, 5.3 Firms involvement in education and training............................................................72 especially for women...................................................................................21 5.4 Information gaps........................................................................................................75 FIGURE 3: Low levels of skills are a constraint for firms in Kosovo............................23 6. GOING FORWARD: POLICY REFORMS TO REDUCE SKILL GAPS IN KOSOVO........... 77 FIGURE 4: Obstacles to operation and growth..............................................................31 Addressing inequity and quality problems in skill development.....................................78 Connecting supply and demand: firms, students, and training institutions....................79 FIGURE 5: Share of firms that tried to hire in each occupation...................................33 Reducing information gaps...............................................................................................81 FIGURE 6: Firms that tried to hire experienced difficulties for a variety of reasons..............................................................35 FIGURE 7: Recruitment difficulties due to skills, by key characteristics.....................37 FIGURE 8: Skill use by occupational group (worker reported).....................................40 FIGURE 9: Ranking of skills by employers.....................................................................42 FIGURE 10: Firms reporting skill gaps.............................................................................43 FIGURE 11: Skill gaps from the workers’ perspectives..................................................45 FIGURE 12: Decomposition of the contribution to earnings by different characteristics.......................................................49 FIGURE 13: Core literacy score by different characteristics..........................................51 FIGURE 14: Score on full literacy test by different characteristics................................52 FIGURE 15: Ethnicity and education are correlated with intensity of skills use at work..............................................................53 FIGURE 16: Education and ethnicity are correlated with intensity of skill use outside of work...................................................55 FIGURE 17: Education and ethnicity are correlated with scores on socio-emotional skills and characteristics........................56 FIGURE 18: Gender differences in skills (on the job)......................................................57 6 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 7 FIGURE 19: Gender differences in cognitive, socio-emotional and technical skills, according to employers in Kosovo.. 59 Annexes FIGURE 20: Role of personal characteristics in hiring....................................................60 ANNEX 1: Skills in STEP Household and Employer Survey..........................................82 FIGURE 21: Employer reported issues when hiring women...........................................61 ANNEX 2: Socio-emotional characteristics in the Household Survey.........................84 FIGURE 22: Participation in ECD, especially for longer durations, is low is...................64 ANNEX 3: Full literacy test: Reading Proficiency Levels..............................................86 FIGURE 23: Distribution of population by level of education in Kosovo, ANNEX 4: STEP surveys in Kosovo: Background information......................................87 Turkey, and EU (selected).............................................................................66 ANNEX 5: Descriptive Statistics for STEP Employer and Household Surveys............90 FIGURE 24: Access to higher levels of education is low, especially for certain groups........................................................................67 ANNEX 6: Determinants of employment and earnings................................................93 FIGURE 25: Mean ETS Literacy Scores, ANNEX 7: Determinants of Skills..................................................................................99 total working age population and employed (15-64)..................................68 FIGURE 26: PISA performance, Kosovo and comparator economies............................69 FIGURE 27: Firms’ views on education and training systems........................................71 Annex Tables FIGURE 28: Beneficiaries in active labor market measures (incl. training)...................72 ANNEX TABLE 1: Target sample size...............................................................................88 FIGURE 29: Less than one out of five firms engage with education systems; when they do, it is not at a systemic level..................................................73 ANNEX TABLE 2: STEP Household Survey: Occupation of the currently employed (aged 15-64, urban) by gender (%)....................92 FIGURE 30: Incidence of on-the-job, internal, or external training among firms.................................................................75 ANNEX TABLE 3: Probability of employment (vs. non-employment, i.e. unemployed and inactive)..............................................................93 FIGURE 31: Firms use informal channels to recruit workers..........................................76 ANNEX TABLE 4: Probability of unemployment (vs employment).................................94 ANNEX TABLE 5: Determinants of Earnings (OLS)..........................................................96 ANNEX TABLE 6: Cognitive skills used (reading, numeracy, computer use): ordered probit with svy........................................................................99 Tables ANNEX TABLE 7: Socio-emotional skills: OLS with svy................................................100 TABLE 1: Skills Towards Employment and Productivity: The five STEPs..................26 ANNEX TABLE 8: Literacy levels....................................................................................101 8 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 9 Acknowledgements Overview Skills matter for Kosovo This report has been prepared by a World Bank team under the guidance of Stefanie Brodmann and Flora Kelmendi. Sara Johansson de Silva led the report writing and Olga Kupets carried out the data analysis. Carey Jett was responsible for final language editing. The report has benefitted from overall guidance from Cristian Aedo (Practice Manager, Education Global Practice), Cem 1. Kosovo faces severe labor market challenges. Less than one third of the adult population Mete (Practice Manager, Social Protection and Jobs Global Practice), and Marco Mantovanelli holds a job, almost nine out of ten women are not working, and over half of active youth (Country Manager, Kosovo). The team is grateful to Mohamed Ishan Ajwad, Kevin Hempel, Victo- are unemployed. Kosovo’s labor market is characterized by limited employment oppor- ria Levin, and Alicia Marguerie for helpful comments and suggestions, as well as the participants tunities and low quality of existing jobs which increases the risk of poverty, reduces labor of a consultation workshop with the Government of Kosovo and donor representatives held on productivity, and fosters discontent, especially among the many youth who enter the labor June 21, 2018. Flora Kelmendi initiated the task and led the extensive data collection efforts. market each year. Promoting job creation is therefore essential to increasing productive Maria Laura Sanchez Puerta and Alexandria Valerio led the data collection protocol of the Skills employment, improving well-being and reducing poverty in Kosovo. toward Employment and Productivity (STEP) surveys. 2. Skills are central to enhancing business climate, fostering job creation and increasing The team is grateful to the Ministry of Labor and Social Welfare, the Ministry of Education, Sci- well-being. Workers with more skills improve their own productivity as well as that of ence and Technology for supporting the overall task. In particular, the team is grateful to the other workers, increase capital productivity, and facilitate innovation and the adoption of Kosovo Agency of Statistics for their cooperation and support in the data collection for the house- new technology. There is solid empirical evidence that higher levels of skills are associat- hold and employer skills surveys. ed with better labor market outcomes for individuals. Cognitive (analytical) skills matter, but socio-emotional skills – often called soft, or behavioral skills – also have strong and The team acknowledges generous financial support from European Commission Trust Fund long-lasting effects on employment and earnings. Moreover, skills explain most of country “World Bank Partnership for Europe and Central Asia” which supported the full data collection, variations in long-term growth rates, and even small improvements in skills could raise from the Trust Fund “Promoting Gender Equality in the Western Balkans”, supported by the Swiss growth rates significantly. Increasing the level and relevance of skills of the workforce is Cooperation Agency via the World Bank’s Umbrella Facility for Gender Equality multi-donor trust consequently a central focus of the Employment and Social Welfare Strategy 2018-2022 fund, and from the Rapid Social Response Multi-Donor Trust Fund. of the Government of Kosovo. 3. What are labor market relevant skills in Kosovo? Labor market relevant skills represent the ability to do a job-related task well, in this context, the tasks necessary for firms to provide goods and services. Thus far, there has been no systematic attempt to measure skills directly in Kosovo and analysis has been limited to inexact approximations of skills such as education levels, years of schooling, or occupational categories. This report pro- vides new empirical insights into skill levels in Kosovo, based on two recent surveys. The Skills Towards Employment and Productivity (STEP) initiative of the World Bank has de- veloped household-level and firm-based surveys to provide policy-relevant information on skills from both the supply and demand side. Household-level surveys measure the supply and use of skills in the adult urban population in a comprehensive way, including cognitive and socio-emotional skills that affect work-place readiness and effectiveness. Firm-level (employer) surveys focus on identifying the cognitive and socio-emotional job-related skills that employers seek, skills that are difficult to find, and linkages be- tween firms and education systems. 10 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 11 7. Skill gaps negatively affect labor market outcomes. As in many other countries, higher Key findings from the STEP Household skill levels are correlated with better labor market outcomes in Kosovo although oth- and Employer Surveys er factors (gender, age, ethnicity, family background) also matter. Estimates from the Household Survey indicate that workers who are conscientious and work well under pressure are more likely to be employed. This is consistent with findings on the skills valued by employers in the Employer Survey. Thus, although earnings are primarily in- 4. Job-seekers lack the skills that firms need. A majority of recruiting firms in Kosovo fluenced by cognitive skills, not having the right mix of skills may represent a significant find hiring new workers challenging because applicants have neither the skills nor work handicap. experience they require. According to the Employer Survey, firms were more likely to recruit for higher skill occupations than medium to lower skill occupations. Three out 8. Skill gaps reinforce existing inequities and affect vulnerable groups. Individuals with of four firms that attempted to fill a higher skill position, and three out of five firms lower levels of education, from certain ethnic minorities (other than Serbian) and from that sought to fill a medium to lower skill position, encountered problems because of a more vulnerable socio-economic background, consistently perform worse on literacy applicants’ lack of skills and/or experience. At the same time, the vast majority of firms tests (even at basic levels), use fewer cognitive, information processing skills (reading, report that their current employees are not lacking in important job-related skills. From numeracy, computer use), and score lower on measures of socio-emotional skills that the firm perspective, skill deficits appear to be related more to new labor market en- are important for labor market success. Hence, differences in skill levels leave disad- trants and jobless youth, than to experienced workers. When recruiting, firms rely more vantaged groups even further behind. By contrast, gender gaps in job outcomes are not on informal channels such as personal networks and recommendations and poaching well explained by observed (as opposed to perceived) differences in skill levels. Social (i.e., approaching workers in other firms). This, along with firms’ relatively poor ratings norms surrounding family obligations and limited options for family care are a more likely of Kosovo’s education and training system, suggest that they find it difficult to identify explanation for women’s limited access to employment. skill levels based on education and training achievements alone. The ensuing recruitment approach that emphasizes work experience penalizes new labor market entrants without 9. Skill gaps are rooted in issues of quality and equity within the educational system. access to social or professional networks. Access to preschool is increasing but only one third of youth (15-24) attended early child- hood education, and children from poorer families are much less likely to do so. Access to 5. Conscientiousness, problem solving, and working well under pressure are skills that primary and secondary education is now near universal in Kosovo, but because of lower are most needed and valued by employers. The Employer Survey shows that employ- enrolment in the past, the proportion of adult workers with basic level of education (9 ers look for workers that are disciplined and can be trusted upon to complete tasks well years) or less is high compared to the majority of European countries. Ethnic minorities without supervision (conscientious), that are able to solve non-routine problems as they other than Serbian also have significantly lower levels of education than other ethnic occur in the work process, and that can work well under duress (emotionally stable). groups, reflecting limited access to education historically. Gender gaps persist and are For medium-to-lower skill occupations, typically service, sales, and low skill technical/ larger than in other Western Balkan countries. A higher share of young women has com- manual workers, it is also important to work well with others, be they co-workers, cli- pleted tertiary education compared with young men, but the share of young women with ents, or suppliers. Findings from the Household Survey indicate that a broad skill set is basic levels of education or less is also twice as high as that for young men. Differences needed across occupations. Although higher skill occupations (managers, professionals, in access are compounded by the poor quality of the educational system. Between 70 technicians) are more likely to require advanced cognitive skills, medium-to lower level and 80 percent of Kosovo students have not acquired basic proficiency in math, science skills occupations also require the ability to do non-routine tasks, solve problems and and reading, leaving Kosovo far behind Western Balkan peers and European countries. interact successfully with others in the work place context. These assessments are consistent with employers’ low appraisal of education systems, especially the ability to deliver graduates with practical and up-to-date knowledge in 6. Skill gaps have negative consequences for firm growth and job creation in Kosovo their fields. and impede productive employment in more dynamic firms. Although skills are not the most pressing constraint that firms face, skill gaps limit their ability to hire and grow. 10. Beyond school, training opportunities for job seekers in the context of active labor This is especially true for more productive firms, such as large firms, foreign-owned market programs are limited and the quality and cost effectiveness of programs re- firms, innovative firms, and firms investing in research and development (R&D). These main unclear. Vocational training is the most common intervention in active labor market firms are more likely to have actively recruited workers and have the potential to create programs in Kosovo, aimed at skilling or re-skilling job seekers who are out of school. more jobs and, importantly, more productive jobs than other firms. Yet, these dynamic However, these adult training programs are not based on surveys of labor demand, are firms are also more likely to identify skills as a major constraint on recruiting efforts: over disconnected from employers and social partners, and lack systematic quality assur- eighty percent of large firms, foreign firms, and firms investing in R&D report skill gaps ance, including monitoring and evaluation. as a major challenge in the recruitment process. 12 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 13 11. Firms are not active in skill development. Few firms provide further training to upskill Addressing inequity and quality their employees, and work-based training for students through internships or appren- problems in skill development ticeships is underdeveloped. Firm-provided training is largely a complement to and not a substitute for school-based training. Most firms, however, do not offer training to their 14. The early years are formative and have long-term effects on further skill acquisition staff, especially training that is external to the firm. Dynamic markets for products and and broader aspects of well-being. Participation in preschool education has a positive services are also likely to change demand for skills among those currently employed. impact on the development of foundational and advanced cognitive skills, but only a The limited opportunities for skill development, once employed, are hence likely to hold small number of children have access to early childhood education in Kosovo and the back both labor market flexibility and firm productivity and expansion. quality of preschool education is unknown. The immediate priority and challenge for the Government of Kosovo – as recognized in its sector strategy for education – is to expand 12. Kosovo firms are not well connected with education and training systems, especially access to preschool education, while ensuring the quality of services. In this context, not at a strategic level. The involvement of firms is critical to support the development priority should be given to including vulnerable groups that have traditionally had little of labor market relevant skills, but there is little interaction between educational institu- access to such systems. Tapping into the private sector for expanding access should tions and firms, especially at the level of design and evaluation of the content of training. be explored along with developing and implementing quality and monitoring standards Firms that do engage in regular contact with education systems (fewer than one in five) across providers. A significant upscaling of affordable preschool options, if matched with do so to provide internships or train their workers and not to address the larger issue of complementary activation policies, could have the additional and direct benefit of encour- the relevance of skills development systems. Only a small minority of firms that interact aging women’s labor force participation. with education systems provide feedback on curriculum or assist in the testing of stu- dents and few use educational institutions for recruitment. Instead, firms rely on informal 15. Education reform must focus on raising the quality of education and training, espe- networks, personal recommendations, and recruitment of workers from other firms to cially for vulnerable groups. The low learning outcomes in Kosovo, even for foundational identify suitable workers. skills like literacy, point to serious weaknesses in educational systems that can compro- mise Kosovo’s ability to become more integrated into international markets. The Kosovo education system needs to ensure that students develop a wide range of skills, including socio-emotional skills that have been shown to have lifelong impact on well-being. There Going forward: policy reforms to is significant evidence that socio-emotional skills, such as discipline, long-term goal setting and decision-making ability, are in fact malleable over time and can be nurtured reduce skill gaps in Kosovo and developed with adequate pedagogical methods. Increasing the quality of education and improving learning will require increasing investment in education, focusing efforts on targeting disadvantaged groups and investing in quality enhancing measures such as 13. The broad set of cognitive and socio-emotional skills that are valued in labor mar- teacher policies and training, curriculum reforms, and ongoing monitoring and evalua- kets must be nurtured from an early age and throughout schooling and beyond. Labor tion. Education expenditure per pupil is low in Kosovo compared to peer countries and market relevant skills include a set of transversal cognitive and socio-emotional skills investing in education early and for all will be essential to the development of a more that are necessary in modern, competitive firms: conscientiousness, independent and productive economy. Because access to education differs significantly among different creative work, ability to work under rapidly changing or stressful conditions, and effec- groups, it will be important to evaluate potential demand side constraints to education tive communication and collaboration with others, among others. These develop from as well. This includes norms around schooling and gaps in information on career paths birth and throughout the life course: from interventions in early childhood programs that and pay-offs to education for children, families, and communities. lay the foundation for basic cognitive and socio-emotional skills, through school-based academic or vocational learning, and after formal schooling, through training and re- 16. There is significant scope for improving vocational education and training, which ac- training programs for adults and ongoing skill development in the work place. Education count for half of all students at the higher secondary level (grades 10-12). This will re- and training systems cannot be expected to produce graduates fully prepared for the quire aligning vocational education training with assessments of technical skills needed workforce but, instead, graduates with the capacity for further on-the-job training and in the labor market, involving employers in the design of training curricula, fostering adapting to new challenges as they arise. Hence, stepping up skill development in Kosovo cooperation with businesses in the delivery of training, and strengthening quality assur- will require effort and collaboration among all stakeholders: education systems, firms, ance mechanisms. students, and policy makers. 17. It will also be important to increase the relevance of vocational training provided by the Employment Agency. In particular, there is a need to develop effective training pro- grams based on accurate measures of demand for skills and labor, and on evidence of 14 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 15 what works. It will be necessary to collect information on skills in demand as evidenced ments have been signed with companies for internships in Kosovo, but monitoring and through firm-level surveys and other relevant sources of labor market information. It will evaluation of these initiatives are weak.4 also be important to strengthen the monitoring and evaluation of existing initiatives and develop effective programs. Tracer surveys, employer interviews, internship reports, 20. To fulfil its core function to match jobseekers with employers, the Employment Agen- and other data collection tools, together with information on regional and international cy needs timely information on the number of available vacancies. According to the best practice, can help identify the strongest programs and how to continuously improve STEP Employer Survey, there remains significant potential to increase this matching be- them. cause only a minority of employers recruit through the Agency. To this end, strengthening outreach to employers may be needed, as well as providing high quality services to them (e.g., through effective preselection of candidates, and fast response times). In prepa- ration, the Employment Agency can also provide job seekers with critical information on Connecting supply and demand: firms, students, and workplace skills that matter. Socio-emotional skills like discipline and perseverance are training institutions developed over many years. However, the Employment Agency can emphasize the im- portance of various workplace-related competencies, including timeliness, independent 18. The involvement of firms in multiple dimensions of skill development systems is of work, and team work in order to influence priorities and incentives. utmost importance. Connecting employers, workers, education systems and students is essential to increase the relevance of school-based training. Collaboration can take 21. Capacity building would help firms, especially smaller ones, to identify and evaluate different forms, ranging from public-private sector partnerships involving shared financ- skill and training needs. Smaller firms often lack the in-house capacity to identify and ing and management responsibilities, to providing technology and equipment, job skills develop the skills of their staff. They are also less likely to use sophisticated tools like needs assessments, mentoring and career advice, partnerships around curricula reform, personality tests or other forms of assessments in the hiring process (hence the reliance and work-based learning.1 on informal channels). Industry associations can play an important role as partners in both identifying skill development needs and in representing firms in interactions with 19. Work based-training, in the form of internships and apprenticeships2, can increase education systems and government. Capacity building at the industry or firm level is also the relevance of skills and provide youth with both experience and references. Intern- useful to enhance the identification of skill needs as well as the use of additional and ship and apprenticeship systems are under developed in Kosovo. Well organized, these more sophisticated techniques to evaluate skills among applicants. In this context, there arrangements have several advantages: (i) they provide trainees with an opportunity to is a need to reinforce the message that skill levels are not different for women and men, participate in actual work place situations and solve concrete work-related problems; (ii) once education differences and participation in labor markets are accounted for. Advo- they contribute to building job specific skills as well as transversal labor market relevant cacy to firms of all sizes may also be needed to reinforce the importance of proactive skills that are highly valued by employers; (iii) they bring together firms, students and engagement by the private sector – through approaching other stakeholders, providing education systems and facilitate collaboration between stakeholders to improve educa- work-based training for students, and reducing prejudice to promote the employment of tion systems; (iv) they provide trainees with references from the private sector that can youth, women, and other disadvantaged groups. signal capabilities, which may be especially important given employers’ emphasis on work experience; and (v) they provide an entry into a professional network for students. In the EU, an estimated 60-70 percent of apprentices find employment immediately after graduation. To work well however, these arrangements need strong institutions and governance systems, including three party collaboration (e.g., firm-student-school, or firm-student-employment agency), formal learning arrangements or learning plans, monitoring and evaluation, and quality assurance systems.3 Recently, cooperation agree- 1  Dunbar, M. (2013). Engaging the private sector in skills development, Health & Education Advice and Resource Team, https://www.educationinnovations.org/ 2  In the EU, apprenticeships and internships are part of formal TVET in many countries. Apprenticeships lead to formal (VET) qualifications. Internships do not necessarily lead to formal TVET qualifications but can be an in- tegral and formal part of a broader TVET program. See Broek, S., T. Hogarth, L. Baltina, and A. Lombardi (2017), Skills Development and Employment: Apprenticeships, Internships and Volunteering, IP/A/EMPL/2016-04, Report prepared for European Parliament’s Committee on Employment and Social Affairs. 4  Kosovo Education and Employment Network (2017). Evaluation Report on Implementation of Kosovo Educa- 3  Ibid. tion Strategic Plan in 2017. http://www.keen-ks.net/site/assets/files/1345/raporti_i_vleresimit_psak_eng-1.pdf. 16 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 17 Reducing information gaps 22. Assisting parents, children and youth in making informed choices on schooling, train- ing, and jobs can have significant benefits for education and employment outcomes. The private sector must be involved in collaborative efforts to strengthen the availability of labor market information. Students and jobseekers need adequate and timely infor- mation on current opportunities, wages of different sectors and occupations, and quali- 1. Skills matter for Kosovo fications needed for various careers, as well as forecasts on future skill needs. Research indicates that in both developing and developed countries, students and their families sig- nificantly misjudge returns to different levels and forms of education, and that providing 23. Kosovo is a young country whose working population is expected to grow over the relevant information can change incentives and career paths.5 The STEP surveys focus foreseeable future. Unlike in most European countries, Kosovo will experience a “demo- on understanding the demand and supply of basic and higher-order transversal skills. graphic bump” that could contribute to increases in long-term economic growth rates, to This type of analysis must be complemented with a better understanding of the techni- the extent that the working age population can secure productive employment. cal vocations in demand. Strengthening Kosovo’s Labor Market Information System will require (i) establishing a systematic collection of information on growth sectors with high 24. Job opportunities are currently scarce in Kosovo, however. Inactivity and unemploy- employment potential, as well as wages across sectors and occupations, (ii) integrating ment rates are high and less than one third of the working-age population (those between labor market information from key stakeholders (employers, the Ministry of Education, 15 and 64 years of age) is employed.7 Women and youth are most at risk for labor market Ministry of Labor and Social Welfare, the Employment Agency, the Statistical office); and exclusion; Kosovo has one of the largest gender gaps in the world in labor outcomes. Low (iii) making labor market information accessible through relevant channels such as the access to employment is compounded by a lack of productive job opportunities reflected Employment Agency website, Busulla6, and others. Career guidance systems in school in the high informality of existing jobs.8 Poor labor market outcomes are at the center of (both general and vocational) can also assist students and their parents in choosing voca- poverty and exclusion, high emigration rates, and disenchantment in Kosovo.9 tions that match labor market demand, their own aptitudes, and interests, and in making the most appropriate educational choices to pursue a chosen career. 25. Those from poor and socially excluded groups, face a lack of employment oppor- tunities and low earnings. In 2015, some 11 percent of employed adults were poor, compared to 23 percent of the unemployed and 17 percent of the inactive. Poverty is also linked to quality of work. Households whose primary source of income was occasional work were more likely to be poor than those depending on household enterprises or steady waged work. Those in the poorest 40 percent of the population (the two poorest quintiles) were significantly less likely to work in professional occupations than the top 60 percent.10 26. Lack of jobs is cited as the most urgent problem facing the nation by Kosovars. In a re- cent survey of major issues affecting social well-being, respondents cited unemployment more than twice as frequently as corruption, and four times as often as poverty.11 The poor economic prospects and ensuing disenchantment fuel social unrest and emigration of high skilled workers to European countries. 7  Unless otherwise specified, data on labor market outcomes refer to 2016 and are from the SEE Jobs Gateway database, available at: https://www.seejobsgateway.net. 8  2012 data, reported in Cojocaru, A. (2017). Kosovo Jobs Diagnostic, the World Bank, Washington, DC. 5  Guallar Artal, Silvia, S. Johansson De Silva, V. Levin, A. Safir, and A-M Munoz Boudet (2016). From aspirations 9  Ibid. to occupations: the role of information in educational and labor market decisions in Moldova (English). Washing- 10  Based on a measure of absolute poverty – a basic needs poverty line of €1.82 per adult equivalent per day. ton, D.C.: World Bank Group. See Kosovo Agency of Statistics and the World Bank, 2017, Consumption Poverty in the Republic of Kosovo, 6  Busulla.com is an online platform developed by Ministry of Education for the purpose of helping students in 2012-2015, the World Bank, Washington, FC. career planning. 11  UNDP Kosovo, 2017, Public Pulse XIII, October 2017. 18 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 19 27. Fostering job creation and productivity growth, the basis for access to high quality jobs, manufacturing, utilities) saw the most job losses whereas job opportunities increased in should be a top priority for policy makers in Kosovo. As emphasized in the recent Jobs commerce (in particular) and services. Although start-ups account for most of net job cre- Diagnostics for Kosovo, this involves improving the conditions for business, including re- ation, the formal firm sector is relatively static by international comparison, with low entry moving infrastructure bottlenecks, strengthening business and regulatory frameworks and of new firms, and low exit of incumbents. As a result, a much-needed shift of jobs from less developing the skills necessary for new job opportunities.12 Little is known, however, about productive to more productive firms has not occurred. the skill levels of the working age population, the skills that matter in labor markets, and how skill gaps affect hiring and productivity, which is a major constraint on policy formu- 32. The informal sector is sizeable in Kosovo.14 About fifteen percent of wage workers do not lation. This report fills an important knowledge gap by presenting new evidence on the have written contracts and as such can be considered informally employed. However, if the demand and supply of skills in Kosovo based on the World Bank Skills Towards Employment definition of informality is broadened to include unpaid workers, employees in micro-firms, and Productivity (STEP) surveys. and those self-employed in non-professional occupations, more than two out of five work- ers are informal. Informality is highest in agriculture and construction, reflecting low skill requirements, casual labor, and unpaid family work. Informality is also higher in the tradable than the non-tradable sector, where informal jobs serve as entry to paid employment for 1.1 Employment challenges in Kosovo low educated youth. What types of jobs are available?13 What kind of workers? 28. Kosovo’s formal sector firms are small, even by regional comparison, and geographically concentrated in the capital. Productive job opportunities in the formal sector stem from 33. Kosovo’s population - and working age population - is young by European standards.15 In a dynamic and job creating private sector. Kosovo’s formal sector is dominated by mi- 2011 (the most recent population census), children under age 14 and elderly above age 65 cro-firms, which account for 91 percent of firms and 36 percent of formal jobs. The share of made up 28 and 7 percent of the population in Kosovo, compared to 16 and 19 percent in the large firms in formal employment is low by international comparison. Formal employment EU28. The working age population in Kosovo is also relatively young, with youth between is concentrated in the commerce and services sectors. There is a significant geographical the ages of 15 and 24 making up 30 percent of the working age population in Kosovo versus concentration of firms and productive job opportunities in Pristina, where over half of all 18 percent in EU28. jobs in formal firms are based. 34. The working age population has moderate levels of education: Twelve percent of the pop- 29. Firms in Kosovo are not well integrated into the global economy, which is a constraint ulation aged 15-64 has tertiary education, but 50 percent has no more than upper secondary on job creation. Only four percent of firms are exporters and less than one percent have education, and 38 percent has no more than lower secondary education or less (Figure any foreign direct investment. Given the small size of Kosovo’s economy, the lack of access 1). Women have significantly lower levels of education than men. Education levels have to international markets means that few firms are able to significantly expand production been increasing, however, and the share of tertiary educated is twice as high (24 percent) and workforce. Exporters are also more likely to survive long-term, generate jobs, and pay among the 25 to 29-year-old cohort as among the 30-64 cohort. Education gaps between higher wages than non-exporters. women and men have also shifted; young women have surpassed men in terms of tertiary education, but the share of females with less than upper secondary education is more than 30. In Kosovo, larger, foreign-owned and exporting firms contribute to productivity and job twice as high as that for men. creation. Firms that are more productive should, inter alia, provide better jobs (i.e., pay higher wages). In Kosovo, larger, foreign-owned and exporting firms are most productive. Across sectors, relative to manufacturing, productivity is lower in the service sector but higher in commerce, construction, and agriculture. Exporting and more productive firms are also more likely to create jobs in Kosovo. 31. Net job creation in the formal sector declined significantly between 2005 and 2014, from 7,000 to just over 1,600 per year, and the formal sector remains static. Industry (mining, 14  Ajwad, M. I., E. Vasquez, and H. Winkler (2016). European Informality: A study of Kosovo’s shadow economy, Background paper for the Kosovo Jobs Diagnostic (mimeo). Washington, DC: The World Bank. 12  Cojocaru (2017), op. cit. 15  Estimates for Kosovo are based on data from the Population, Households and Housing Census 2011, available 13  The section draws on analysis presented in the Kosovo Jobs Diagnostic (Cojocaru, 2017, op. cit.). at http://askdata.rks-gov.net; for the EU, data are from Eurostat (2017 data). 20 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 21 The Kosovo population has moderate levels of education with the public sector and, possibly, social norms that prevent them from accessing private FIGURE 1: significant gender differences, 2016 sector jobs.16 38. Kosovo has a sizable diaspora working abroad, primarily in Germany, Switzerland and Working age population: distribution by level of education, age group and gender Austria. Remittances, together with foreign aid, are key drivers of economic growth in recent years. The large stock of migrant workers and the remittances they send home also influence consumption and investment patterns in Kosovo as well as incentives to 12% 8% work. Individuals in Kosovo who live in households that receive remittances have slightly 13% 10% 22% 26% 14% higher reservation wages than others and are less likely to be registered with employ- 32% 50% 40% ment offices, although the differences are small.17 60% 61% 64% 47% 59% 38% 50% 27% 27% 25% Access to employment and access to education are linked, especially 13% FIGURE 2: for women. TOTAL Male Female Male Female Male Female 15-64 25-29 30-64 Employed vs. jobless: distribution by level of education, age group and gender LOWER SECONDARY OR BELOW UPPER SECONDARY* TERTIARY 5% 3% 23% 19% 36% 7% 8% 7% 24% 56% 22% 20% 21% 35% 60% 30% 46% 57% 39% 35. Access to employment is low in Kosovo. In 2017, 57 percent adults (aged 15-64) were SOURCE: SEE 63% 64% 48% 61% Jobs Gateway 59% 64% inactive and the level of inactivity was more than twice as high among women as among database. *Includes 43% men (80 vs. 35 percent). Low participation rate is compounded by high unemployment post-secondary non- 45% rates in the active population, at 30 percent on average and reaching 53 percent for youth tertiary education. (aged 15-24). Just 13 percent of women and 47 percent of men hold a job in Kosovo 39% 67% – among the lowest employment-to-population ratios in the world. Unemployment is 54% 46% structural rather than transitionary. Once unemployed, the chances of transitioning into 35% 31% 35% employment are low, and nearly three quarters of the unemployed had been without a 18% 18% 19% 18% 22% 12% 15% job for more than 12 months. 5% TOTAL Male Female TOTAL Male Female Male Female Male Female Male Female Male Female 36. Women with more education are more likely to work. Education is a strong correlate of access to employment, especially for women, and especially for young people. More 15-64 15-64 25-29 25-29 30-64 30-64 than half of employed women aged between 25 and 29 have completed tertiary levels Employed Jobless Employed Jobless Employed Jobless of education, compared to 20 percent of males (Figure 2). Differences in the share of tertiary educated between the employed and jobless are smaller but still substantial for SOURCE: LOWER SECONDARY OR BELOW UPPER SECONDARY* TERTIARY older workers (aged 30-64). SEE Jobs Gateway database. *Includes 37. Employment is concentrated in non-tradeable sectors. Public administration and social post-secondary services, trade and construction employ 50 percent of the work force. Gender profiles nontertiary vary significantly, with 42 percent of women working in the public sector, compared to education 14 percent of men. These substantial differences likely reflect working women’s higher levels of education compared to men, women’s preferences for working conditions in 16  Cocjocaru (2017), op. cit. 17  Rudi, J. “Remittances and Labor Supply: The Case of Kosovo”, unpublished mimeo, September 2014. Avail- able at: http://paa2015.princeton.edu/papers/150522. 22 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 23 A strategy to improve jobs 43. Skills matter for Kosovo’s private sector, especially for firms with higher potential for job creation and productivity growth. The World Bank’s enterprise surveys indicate 39. Kosovo needs more jobs, better jobs, and more inclusive jobs to foster growth, reduce that whereas Kosovo firms consider high levels of informality, lack of finance and the poverty and social exclusion, and encourage qualified workers to remain in Kosovo. The impact of corruption as the most significant obstacles to business growth, a quarter of focus of such a three-pillar strategy revolves around key policies to (i) accelerate job cre- firms consider skills a major obstacle to their growth. The share in Kosovo exceeds that ation in the formal sector, (ii) increase the quality of existing jobs by increasing productivity of Eastern Europe and Central Asia on average. Moreover, large firms – which contribute and earnings, extending social security, and improving working conditions, and (iii) help more to productivity and growth than smaller firms – are more likely to report that low vulnerable groups, including women and first-time labor market entrants, connect to jobs. levels of skills are a constraint on growth (Figure 3). Possible priorities for a jobs strategy in Kosovo entail business environment reforms to stimulate private sector development, skill development strategies that increase the level of foundational skills as well as those relevant to current and future employer needs, and FIGURE 3: Low levels of skills are a constraint for firms in Kosovo targeted activation policies focusing on the constraints faced by specific groups. 40. Skills are only one aspect of any jobs strategy, but they are especially important as they Skills as a major obstacle (% of firms) cut across all three pillars. Skills are critical to removing competence bottlenecks in the private sector for firms to grow, increase productivity in the production of goods and services, and increase access to jobs with higher wages and good working conditions. 43 33 26 19 21 15 1.2 The role of skills SMALL MEDIUM LARGE KOSOVO ALL GLOBAL ECA Skills influence growth, productivity and job creation (5-19) (20-99) (100+) 41. Skills are defined as the ability to perform a task well, owing to a combination of knowledge, practice, and aptitude. This includes technical skills as well as socio-emo- SOURCE: tional skills such as the ability to collaborate with others or communicate well. Skills are Estimates based 44. Skills matter for individual earnings. Evidence from the Programme for the Interna- not the same as “years of schooling”. The latter has traditionally been used to measure on World Bank tional Assessment of Adult Competencies (PIAAC) survey in OECD countries indicates human capital, but it does not capture a number of factors that affect learning and skill Enterprise Surveys, that stronger analytical skills—measured by numeracy, literacy, and problem-solving— 2013 survey for development, such as the quality of schooling, family background and location. Efforts Kosovo. are significantly related to higher wages for individuals, more so than additional years have been now made to measure skills (both supply and demand) directly, through tests of schooling.19 In other words, actual skills matter more than education, reflecting the and surveys. importance of the quality of skills development system as a whole, and not only access to schooling. A recent study of skills in several middle-income countries (Armenia, Bo- 42. Skills matter for nations’ long-term economic development. Over time, the productivity livia, Colombia, Georgia, Ghana, Kenya, Ukraine and Vietnam) also found that skills pay and growth of a country are directly linked to the level of “economically relevant skills” in off significantly in terms of access to jobs and higher earnings, even when education is the population. There is logical evidence to support this link, but empirical research has accounted for.20 only recently established a strong causal relationship, using new and direct measures of skills. An aggregate measure of cognitive (analytical) skills, developed using scores on international tests in mathematics and science, explains three quarters of cross-country variation in long-term growth rates, and even small differences in skills measures have strong impact on growth rate.18 19  Hanushek, E; G. Schwerdt; S. Wiederhold; and L. Woessmann, 2015, “Returns to Skills Around the World: Evidence from PIAAC”, European Economic Review, Vol 73 (1), pp. 103-130. 20  Valerio, A., Sanchez Puerta, M-L, N. Tognatta and S. Monroy-Taborda, 2015, “Are There Skills Payoffs in Low 18  Hanuschek, E., 2017. “For long-term economic development, only skills matter”. IZA World of Labor and Middle Income Countries? Empirical Evidence Using STEP Data Policy Research Working Paper; No. 7879. 2017:343, March 2017. World Bank, Washington, DC. 24 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 25 A broad and flexible skill set is important 48. Skills begin to form in early childhood and are honed throughout childhood, adoles- cence, and adulthood (Table 1). Transforming the composition and level of skills takes 45. Globally, technological and organizational changes are transforming jobs and the cor- time. Policy to promote and reinforce skill development must therefore span across responding skills in demand. Significant structural change, in particular the rising role of the life course—from interventions in early childhood programs to university education information and communications technology in the production of goods and services, has as well as training and re-training programs for adults. New findings on school-based led to a decline in the need for routine, manual skills and an increased demand for “new programs aimed at developing socio-emotional skills strongly emphasize the value of economy skills” that complement automated work.21 These shifts in skill demand have integrating such training into the core curriculum, and show that efforts are more suc- also been observed among developing countries that are more advanced on the reform cessful when specifically targeted at young children (as well as vulnerable groups).25 The agenda.22 In these economies, structured and repetitive tasks are giving way to tasks that time lag involved in building skills means that countries, even those less advanced on the require abstract thought, decision making, team work, and leadership. There is also an reform agenda, must address skill gaps long before it becomes a major barrier to growth increase in manual tasks that require adapting and reacting to changing circumstances and labor market success. This involves collaboration among policymakers, workers, with tools and manual dexterity, as well as social interaction and collaboration. Thus, the jobless, students, schools, and firms. To inform stakeholders, more information is workers must accumulate a large tool box of skills that form the basis for acquiring needed to identify the nature of the skill gap and where it occurs, as well as the children job-specific skills and adapting to changes in tasks over time. and adults most in need of skill development. 46. Socio-emotional skills influence labor market outcomes. Socio-emotional skills – such as discipline, perseverance and openness to experience and change – matter significantly for how individuals perform in school, in the labor market, and on other measures of success in life. As discussed, there is ample evidence to suggest that these skills are malleable and can be developed and honed during school. High quality early childhood Table 1: Skills Towards Employment and Productivity: The five STEPs and primary school programs have the potential to build and improve socio-emotional STEPS Preschool age School age Youth Working age skills in the short and long run.23 A recent study groups together skills that have been 5. Facilitating Apprenticeships, Intermediation services, shown to be valued by employers into eight major categories, including the cognitive skill labor mobility and skills certification, labor regulation, social of problem solving, and several socio-emotional skills such as resilience, achievement job matching counselling security portability and motivation, control, teamwork, initiative, confidence and ethics.24 4. Encouraging Fostering inquiry Universities, innovation clusters, basic entre- entrepreneurship preneurship training, risk management systems 47. Skill development involves a number of stakeholders: (i) children/youth, their parents and innovation and communities, and adult workers that identify preferred educational and vocational 3. Building Basic vocational train- Vocational training, Firm-provided training, paths (ii) education and training systems (including active labor market programs) that jobs-relevant ing, behavioral skills higher education, recertification, reskilling provide youth and job seekers with training and education (iii) firms that employ workers skills apprenticeships, with different skill levels and provide them with experience and in some cases training targeted programs (iv) policy makers. In a well functioning system that rewards skill development and where 2. Ensuring that Cognitive skills, so- Second chance labor market information is available, youth choose educational paths suited to their all students learn cialization, behavioral education, behavioral skills skills interests, abilities, and labor market prospects; firms communicate and interact with education systems to improve relevance and provide further skill development to their 1. Getting children Nutrition, psychologi- School health and workers; and education systems respond to labor market needs. off to the right cal and cognitive stim- remedial education start ulation, basic cognitive and social skills SOURCE:  Banerji, A., W. Cunningham, A. Fiszbein, E. King, H. Patrinos, D. Robalino, and J-P Tan, 2010, 21  Autor, D., F. Levy and R.J, Murnane, 2003, “The Skill Content of Recent Technological Change: An Empirical Stepping up skills for more jobs and higher productivity, World Bank, Washington, DC. Exploration,” The Quarterly Journal of Economics, vol. 118(4), pp 1279-1333. 22  Arias, O.; Sánchez-Páramo, C.; Dávalos, M; Santhos, I; Tiongson, E; Gruen, C; de Andrade Falcão, N., Saiovici, G., Cancho, C., 2014, Back to Work: Growing with Jobs in Europe and Central Asia, World Bank: Washington DC. 23  Heckman, J. and T. Kautz, 2012. “Hard evidence on soft skills,” Labour Economics, Elsevier, vol. 19(4), pp 451-464; and Heckman, J. and T. Kautz, 2014, “Fostering and Measuring Skills: Interventions that Improve Char- acter and Cognition”, in Heckman, J. J.E. Humphries and T. Kautz (editors), The Myth of Achievement Tests: The GED and the Role of Character in American Life, Chicago, IL: University of Chicago Press. pp. 341-430. 25  Sánchez Puerta, M-L, A. Valerio, and M. Gutiérrez Bernal (2016). Taking Stock of Programs to Develop Socio- 24  Guerra, N.; K. Modecki; and W. Cunningham, 2014, Developing Social-Emotional Skills for the Labor Market. emotional Skills: A Systematic Review of Program Evidence. Directions in Development. Washington, DC: World The PRACTICE Model. Policy Research Working Paper 7123. World Bank. Bank. 26 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 27 Measuring skills - the basics of the Kosovo STEP surveys • Job-specific skills include some transferrable cognitive and socio-emotional skills (as described above) related to employment in general, such as making presenta- 49. Our understanding of skills from the demand side (skills required by firms, presently and tions (i.e., a communication skill) or working in teams (an interpersonal skill). They in the future) and the supply side (potential work force) has been held back by a lack of also include task-specific skills that are less transferrable across occupations systematic information on the relevance and performance of education and training.26 such as the ability to use various methods, materials and tools to drive or operate Educational attainment is an imperfect approximation of skills: skills developed in childhood a specific type of machinery. and adolescence, through education and training systems, and on and off the job in adult life, are what matter. For these reasons, gaps between developing and more developed econo- 52. The Household Survey focuses on measuring skills in the urban population. It provides mies in actual skills are much larger than the (already significant) gaps in education levels information on self-reported cognitive skills used on the job and during spare time (reading, would suggest.27 Thus, measuring access to education in order to evaluate skill supply, or writing, numeracy, computer use), as well as (for those employed), on the job cognitive the effects of skills on labor market outcomes, is not sufficient. skills like problem solving and learning new things, socio-emotional skills such as interper- sonal skills, supervising staff, or the ability to work independently, and technical job specific 50. The STEP Household and Employer Surveys is an important step toward filling knowl- skills that can be described as job characteristics, such as operation of machinery, or driving edge gaps in the supply and demand for skills. The STEP28 Skills Measurement Program, vehicles. The Household Survey also provides an assessment of socio-emotional skills, launched in October 2010 by the World Bank, is a systematic attempt to fill knowledge gaps based on the so called “Big Five” inventory of personal traits: extroversion, agreeableness, related to skills that enhance productivity and increase earnings in developing countries. The conscientiousness, emotional stability, openness – as well as grit, and approaches to deci- program is designed to provide policy relevant information on skills, beyond basic informa- sion making. The questions related to identifying these seven different traits are presented tion on education levels and literacy, from both the supply and demand side. Household-level in Annex 2. Finally, the Household Survey also includes a literacy assessment which consists surveys focus on measuring skills and the use of skills in the population in a comprehensive of two levels, a basic (“core”) and a more advanced reading test module. Respondents are way, including cognitive, or analytical skills, socio-emotional skills that affect workplace first given the core literacy test consisting of eight rapid questions that helps to screen the readiness and effectiveness, and job-specific technical skills. Firm level (employer) surveys least literate from those with higher-level skills. Passing the core literacy test (level 3 and focus on understanding the skills that employers need, skills that are difficult to find, and above) is an indicator of basic literacy. Those that pass the test also undertake a more ad- different means to address the skill gap. Annex 1 provides a table mapping different skills in vanced (“full”) literacy test. The different levels (1-5) of literacy in the full test are explained the Employer Survey and Household Survey into a coherent framework and explaining how in Annex 3. they are defined and measured. 53. The STEP Employer Survey focuses on workplace-related skills. The Employer Survey 51. The STEP Employer and Household Surveys focus on a broad set of skills: asks employers about a wide range of skills (primarily cognitive and socio-emotional skills), including how important they are to the job, and whether they are used by a typical worker. • Cognitive skills are analytical skills. These include foundational skills such as lit- Skill gaps are addressed by asking employers about the nature and size of skill gaps among eracy, numeracy and, in some settings, basic proficiency with computers (ITC the current workforce, as well as how they affect business in general, and recruiting and literacy) – skills that are valued in their own right but also as preconditions for hiring in particular. The Employer Survey also includes some specific questions aimed at acquiring higher-order skills. Higher-order cognitive skills include logical, intuitive, understanding whether employers think that women experience larger skill gaps than men, and creative thinking and problem solving. or vice versa. • Socio-emotional skills (in different contexts referred to as behavioral skills or 54. The STEP Employer Surveys provide new and important information on skill gaps: they (i) characteristics, soft skills, life skills, workplace skills) include skills that are em- provide detailed information on skills as viewed from the demand side: skills that employers pirically linked to long-term labor market success, including perseverance/grit, consider important, and skills that are lacking, for both high- and medium-skilled workers; conscientiousness, flexibility/openness, communication skills, interpersonal (ii) focus on labor market-relevant skills, rather than on diplomas, certificates, or levels of skills, independence, and emotional stability. education; and (iii) focus on generic cognitive, socioemotional, not just job-specific technical skills associated with a particular occupation. 55. The STEP surveys focus on transversal skills. The STEP surveys focus on generic skills that are important to varying degrees across a variety of occupations, and some very aggregate 26  Sondergaard, L. and M. Murthi, 2012, Skills, Not Just Diplomas: Managing Education for Results in Eastern Europe and Central Asia, the World Bank, Washington, DC. job specific technical skills (driving, interacting with people). Their lack of detail makes them 27  Hanushek, E., and L. Woessmann, 2012, “Do better schools lead to more growth? Cognitive skills, economic unsuited for evaluating the supply and demand for specific job-technical skills related to outcomes, and causation,” Journal of Economic Growth, vol. 17(4), pp 267-321. different occupations (manpower planning). Understanding the technical skills in demand is 28  Skills Towards Employment and Productivity. 28 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 29 1.3 Organization of the report central to improving the matching of job seekers to vacancies in Kosovo’s labor markets but requires complementary survey instruments. Yet, there are strong grounds for focusing on a broad skill set as is done in the STEP approach. Basic cognitive and socio-emotional skills provide the foundation for further skill development, including technical skills. Additionally, 59. To understand whether there are skill gaps in Kosovo, what they look like, and how they can demand for products and services is changing over time and with it demand for specific be addressed, the report proceeds as follows. vocations. Transversal skills are needed to allow for flexibility and reskilling. 60. Section 2 investigates whether there are skill gaps in Kosovo’s labor markets. If firms 56. Skill levels, skill needs, and skill gaps differ for different types of workers. All STEP cannot find the workers they need because applicants lack the requisite skills and experi- Employer Surveys separate workers into two categories, under the assumption that the ence, then skill gaps can hold back job creation. At the same time, if current employees lack demand for the skill sets and the systems addressing the skill gaps may differ for the two the skills necessary to perform their jobs, then skill gaps can constrain productivity in the types. The nine different occupational categories (ISCO) are ordered into groups A and B. short run and growth and job creation in the long run. Type A includes managers, professionals, technicians and associate professionals. The “A” group corresponds to occupations with greater need for “new economy skills” (i.e., non-rou- 61. Section 3 focuses on the skills that matter in Kosovo. For example, if a majority of the tine, non-manual skills). Type B includes the remaining occupations: clerical support, service Kosovo population does not speak a foreign language other than English, but this is not re- and sales workers, skilled agriculture workers, construction, crafts and related trades work- quired on the job, or not highly valued by firms in general (and not likely to become so in the ers, drivers, plant and machine operators and assemblers, and elementary occupations. future), then a lack of foreign language skills does not translate to skill gaps. The section The “B” occupations require more traditional skills, based on more routine tasks and – for explores aspects of the demand for skills, both those that are employed by workers and some – more manual skills. For this group of workers, international evidence suggests that those that are valued by firms. the non-routine components of tasks have also increased.29 In the following, these two groups of occupations are referred to as higher skill (Type A) and lower/medium skill 62. Section 4 looks at the consequences of skill gaps on firms and individuals. Although labor (Type B) occupations. On a few occasions the analysis also separates the lower/medium market outcomes are driven by multiple factors, and not just skills, the inequitable distri- skill occupations into two groups: occupations in the service industry (clerks, sales and bution of skills in the population suggests that some individuals lack opportunities in labor service workers), and physical/technical occupations (crafts/construction, plants/machine markets and that some of the work force lack the basic cognitive and socio-emotional skills operators and drivers, and elementary occupations). to respond to rapid changes in labor demand. 57. For the Employer Survey, the analysis also highlights specific outcomes for more dynam- 63. Section 5 discusses the potential driving factors behind existing skill gaps. This includes ic firms: firms that could potentially contribute more to productive job creation in innovative, issues related to education and training systems, and the extent to which the private sector competitive sectors. This includes firms that are innovative (introduced new or significantly helps develop relevant labor market skills and supports lifelong learning. improved methods of manufacturing or producing goods or services in the past 3 years), are invested in research and development, or are large with more than 50 employees. 64. Section 6 summarizes the primary findings and makes suggestions for moving forward. 58. The Kosovo STEP Employer Survey was rolled out in 2015 and the Household Survey in 2016- 17. Details around the two surveys are described in more detail in Annex 4 and Annex 5. 29  Autor and others (2013), op. cit. 30 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 31 (Figure 4, b). Over half of firms find corruption and crime, as well as electricity, a major obstacle to business (55 and 52 percent respectively), whereas only 28 percent cite la- bor-related factors as a major problem. The Employer Survey presents a similar picture to that of the World Bank Enterprise Survey from 2013.30 2. Skill gaps from the firm perspective FIGURE 4: Obstacles to operation and growth a. Labor-related obstacles Key messages: • Skill gaps are a constraint on growth. Firms cite lack of skills and experience as a con- 36% straint on the growth of their business, albeit one of many obstacles. 22% 23% 21% • Skill gaps are a constraint on job creation. Recruiting firms find it difficult to hire because 18% 14% of problems in identifying sufficiently experienced and skilled workers. 11% 10% • Skill gaps are holding back more dynamic firms and hence productive employment Work Skills Skills (General Labor Avail- Worker Turn- Payroll & So- Labor Wage Levels growth. Skill constraints appear to be especially detrimental to firms with the greatest Experience (TVET) Education) ability over cial Security Regulations potential for creating more, more sustainable, and better paying jobs. Skills Availability Cost b. All obstacles 2.1 Lack of skills are a constraint on 55% growth but not the primary concern of 49% 45% 52% firms 29% 28% 28% 22% 23% 65. Firms report that a lack of experienced workers is the biggest labor-related obstacle 15% to operation and growth. The quality of education and training systems, and relevant 11% work experience are both related to labor market relevant skills. For Kosovo firms, the lack of workers with relevant experience is the primary labor-related concern, over and Finance, Political, Anti-com- Licens- Cor- Access & Custom & above the quality of formal education and training, and the availability and cost of work- Economic, ruption; petitive; Electricity Telecom & Labor Tax Rates Trade Reg- ing & Access to Financial Legal Transport Cost Factors & Admin Operating Land ers (Figure 4, a). Yet, skill gaps must be evaluated in the context of other barriers to job Instability Crime system ulations Permits creation and growth, as only some 35 percent of firms report workers’ lack of experience Labor as a major or severe problem. Governance Infrastructure Finance Taxes and regulations Land related 66. Skills are but one of several constraints on business development in Kosovo. According SOURCE: STEP Employer to the Employer Survey, difficulties related to weak governance – political and economic Survey instability and uncertainty, corruption and crime, and a lack of rule of law – and infra- structure are more detrimental to business than labor-related problems, including skills 30  See: http://www.enterprisesurveys.org/data/explore economies/2013/kosovo 32 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 33 FIGURE 5: Share of firms that tried to hire in each occupation 2.2 Skill constraints are holding back a. All firms job creation Tried to hire in past 3 years (% of firms) 67. A limited share of firms has attempted to recruit workers. In the last three years, for Elementary 11% each occupational category, between 15 and 30 percent of all firms in the Employer Survey Drivers, operators 16% tried to hire at least one person (Figure 5, a). Demand was greater for higher skill occupa- tions (managers, professionals, technicians) and medium skill service-related occupations Construction, 13% (clerks, sales workers and services workers) than for low skill occupations (construction craft and crafts, elementary occupations, drivers and operators). These low skill occupations are Skilled ag 1% also occupations where men traditionally have been overrepresented.31 Sales workers 25% 68. Efforts to recruit workers was more common among potentially dynamic firms. Firms that introduced new or significantly improved methods for manufacturing or producing Service workers 24% goods or services in the past 3 years were more likely to seek employees, especially for higher skill occupations (Figure 5, b and c). Larger firms also attempted to hire workers Clerks 14% more often than small firms or micro-firms.32 These firms sought workers for higher skill Work 22% occupations in particular, as well as for non-manual medium skill occupations. Further Experience disaggregation (not shown here) indicates that foreign owned firms, and firms that invested Professionals 27% in R&D were also more likely than other firms to have recruited workers for higher skills occupation. As seen, these findings on job creation are in line with the findings from the Managers 26% analysis of firm registry data in Kosovo, discussed in Chapter 1.33 69. Skills and experience were the primary constraints to recruitment, especially for occu- b. By innovation in new technology pations in demand by firms. In fact, a majority of firms attempting to hire workers reported Tried to hire in past 3 years (% of firms) having difficulty identifying candidates with requisite skills or work experience. Employers’ 13% perception of applicants’ lack of skills and work experience were more detrimental to hiring Elementary 8% than the low number of applications or the number of applicants who would not accept the 18% firms’ working conditions or wages. In fact, 77 percent of firms that attempted to fill a higher Drivers, operators 16% skill position, and 59 percent that tried to fill a medium to lower skill position encountered Construction, 16% problems because of either skill or experience. The significance of these skill gaps was craft 8% consistent across all major occupational categories (Figure 6, a and b).34 Among the medi- Skilled ag 1% um to lower skills positions, fewer firms experienced difficulties recruiting for elementary occupations or jobs requiring driving or operating machinery (lower skill occupations), than Sales workers 27% 22% for clerks, sales, services workers, or construction workers. In other words, skill constraints 24% were more severe for occupations for which there was higher demand (more firms attempt- Service workers 24% ing to recruit). 17% Clerks 8% Work 25% 31  Women were overrepresented among professional occupations (have a higher share of jobs in this sector than Experience 16% in all sectors taken together), and underrepresented among the construction and crafts, and well as plant and 33% machine operators and assemblers. Professionals 14% 32  Because larger firms experience higher worker turnover in absolute terms, they are more likely to attempt recruitment than smaller firms. This does not necessarily imply, however, that larger firms create more jobs. 33% Managers 33  Cojocaru, 2017, op. cit. 12% 34  There is not an absolute difference between work experience and skill levels in this context as work experience should build the skills necessary for the job. NEW TECHNOLOGY NO NEW TECHNOLOGY 34 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 35 c. By firm size Tried to hire in past 3 years (% of firms) FIGURE 6: Firms that tried to hire experienced difficulties for a variety of reasons 18% a. All firms Elementary 15% Reasons for difficulties in hiring (% of firms that experienced problems) 8% 59% 27% Lacked experience 63% Drivers, operators 25% 60% 9% 49% 14% Lacked skills 59% Construction, 17% 53% craft 12% 8% Did not like work- 12% 5% ing conditions Skilled ag 9% 1% 29% Wages too low 24% 56% 31% Sales workers 28% 19% 15% No or few 19% 60% applicants 19% Service workers 27% 17% TECHNICIANS PROFESSIONALS MANAGERS 45% Clerks 18% b. Medium-to lower skill occupations 8% Reasons for difficulties in hiring (% of firms that experienced problems) 25% 34% 32% Work 27% 53% Experience 18% Lacked experience 45% 40% 50% 42% 41% Professionals 33% 20% 22% 25% 48% 55% Lacked skills 42% Managers 33% 35% 20% 42% 43% LARGE (51+EMPLOYEES) SMALL (10-50 EMPLOYEES) MICRO (1-10 EMPLOYEES) 7% SOURCE: 8% STEP Employer 3% Did not like work- Survey 45% ing conditions 8% 12% 6.% 14% 17% 18% Wages too low 20% 23% 15% 19% 11% 10% No or few 49% applicants 11% 12% 7% ELEMENTARY DRIVERS, OPERATORS CONSTRUCTION, CRAFT SKILLED AG SOURCE: STEP Employer SALES WORKERS SERVICE WORKERS CLERKS Survey 36 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 37 FIGURE 7: Recruitment difficulties due to skills, by key characteristics 2.3 Skill gaps restrain a. Sector, size, and age more dynamic firms % of firms that tried to hire but encountered difficulties because of skill gaps 97% 70. Skill gaps are especially problematic for job creation and productivity if they affect 86% the more dynamic firms in the economy. Firms that are likely more competitive and 83% 83% 81% 82 81% productive should also be able to grow and provide more jobs and jobs of higher quality 77% 76% 73% 73% in terms of earnings and working conditions. As discussed above, particular firms (i.e., 70% 68% large firms, innovative firms, firms that invest in R&D) were more likely to have previ- 65% 61% 61% 61% 62% 61% ously attempted to hire workers than other firms. These firms were also more likely to 59% 58% 58% 56% be productive and competitive, and thus more likely to provide both more jobs and jobs 52% 53% that are more sustainable and of better quality. 45% 71. Modern and potentially dynamic sectors were affected to a greater extent by skill gaps. In all economic sectors, with the exception of construction and “other services” (including domestic household workers), a lack of applicant skills affected more than 80 TOTAL Up to 5 years 6 to 15 years 16+ years & Industry Agriculture Construction Trade, Transpt, Ac- comm, Entertnmnt Business Services Public Services Other Services Micro (1-10 empl) Small (11-50 empl) Large (51+ empl) percent of firms trying to recruit higher skill workers and more than 50 percent trying to hire medium to lower skill workers. Large firms were also more likely to have encoun- tered skill constraints in hiring than smaller firms for lower-to-medium skills occupations (Figure 7, a). Sector Size Age 72. Innovative firms were also more likely to have experienced skill gaps when recruiting. HIGHER SKILL MEDIUM TO LOWER SKILL Firms that introduced new or significantly improved methods of manufacturing or pro- ducing goods or services in the past 3 years were more likely to be skill constrained, at least for medium-to-lower level skills (Figure 7, b). Firms that invested in R&D were also b. Additional characteristics more likely to be skill constrained, but more so for higher level skills. These dynamic and % of firms that tried to hire but encountered difficulties because of skill gaps technology-oriented firms were also more likely to consider lack of experienced workers and low levels of skills emanating from general and vocational education systems as 88% 86% major constraints to conducting business. Finally, foreign owned firms experienced high 83% 83% 78% 78% levels of skill deficiency for both higher and medium-to-lower skill occupations. 77% 76% 76% 73% 73% 73% 69% 73. In all, this suggests that skill constraints, especially for higher skill occupations, are 65% 60% holding back the creation of better jobs in Kosovo. Whereas skill constraints affected 59% 57% 56% 56% all hiring firms, they had a stronger effect on those that attempted to hire in the past 42% three years and on dynamic firms. Moreover, higher level skill constraints are large even in the Western Balkan context. A majority of firms in Albania and Serbia also experienced skill-related problems in recruiting workers for higher skill occupations but the share of firms affected in Kosovo was even higher (Figure 7, c). TOTAL Private State Yes No Yes No Yes No Domestic Foreign Innovation International Busi- Ownership R&D Investments (technology) ness Contacts HIGHER SKILL MEDIUM TO LOWER SKILL 38 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 39 c. Skill constraints in Kosovo, Serbia and Albania % of firms that tried to hire but encountered difficulties because of skill gaps 63 68 55 54 77 59 3. The skills that matter in Kosovo SERBIA ALBANIA KOSOVO HIGHER SKILL MEDIUM TO LOWER SKILL Key messages: SOURCE: STEP Employer Surveys. The charts refer to • A range of cognitive and socio-emotional skills are needed for different types of jobs in the share of firms Kosovo, emphasizing the importance of fostering broad and transversal skill sets. that tried to hire but encountered difficulties because • Skills most valued by employers include the abilities to work well with others (e.g., of lack of skills or co-workers and clients), to accomplish tasks, either independently or in collaboration work experience. with colleagues, to handle stressful situations, and to solve non-routine problems. • Skill gaps affect outsiders – job seekers – more than insiders – those with a job. Firms re- port that important skills are not lacking among the current employees, in sharp contrast to the difficulties related to recruitment. A majority of employed workers also consider their education level useful and well matched to the requirements of their job. 3.1 Skills needed on jobs 74. Skills are lacking among job seekers, according to employers. What are the skills valued by employers and how do they vary across different types of jobs? 75. A broad set of skills, both cognitive and socio-emotional skills, is needed for jobs in Kosovo. The Household Survey shows that workers report using a variety of skills on their jobs. Cognitive skills, both foundational (numeracy, literacy) and higher order skills (learn- ing new tasks and problem solving), are used to a medium or high degree on the job by between 40 and 60 percent of the workers, depending on occupation. Among socio-emo- tional skills, interpersonal skills (when working with clients and team members) are an important part of the job for nearly 70 percent of workers. 76. Higher skill occupations require greater use of foundational and higher order cognitive skills on the job. Figure 8 presents averages for the degree to which different skills are used on the job as reported by workers. The figure categorizes occupations into three groups: higher-level skill professions (managerial, professional, technicians and associate 40 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 41 Job-specific technical and socio-emotional skills, (average, Yes=1, No=0) professionals), lower/medium skill occupations in the service industry (clerks, sales and service workers), and physical/technical professions requiring lower-medium skills (crafts/construction, plants/machine operators and drivers). Workers in higher-level skill 0.548 occupations report using cognitive skills intensively, especially computer work and read- 0.426 0.412 ing, but also problem solving and learning new tasks on a regular basis. They are also 0.361 0.356 more likely to be supervising other workers. For these occupations, the mix of non-rou- 0.335 0.299 0.281 0.213 tine higher order skills – communication skills, team skills, management skills, creativity 0.184 0.171 and adaptability to new tasks – likely matter for employers (Figure 8). Not surprisingly, 0.109 0.066 0.021 0.017 workers in higher skill occupations are less likely to undertake physical tasks or operate machinery. Physical/technical occupations (lower-medium skills) are the least likely to others Operating machinery Repairing equipment Driving a vehicle Making Supervising presentations use cognitive skills intensively on the job. 77. People skills and the ability to work independently is needed across occupations. The use of interpersonal skills and a relatively high degree of independent work characterize Technical Socio-emotional the three groups of occupations. Higher level communications skills (e.g., making formal HIGHER LEVEL (A) SERVICE (B) PHYSICAL/TECHNICAL (B) presentations to share information or persuade clients or colleagues) are not frequently SOURCE: STEP Household Survey. used, but they are equally important for higher skills and services skills occupations – For the left chart, reflecting the need to interact closely with clients in service and sales work. responses range from 0 (not used on the job) to 3 (high 78. Employers confirm that jobs overall are skill intensive, using a broad range of cogni- intensity of use). tive and socio-emotional skills. According to the Employer Survey, for all skills listed, For the left chart, at least 60 percent of firms report that the particular skill is used regularly by workers in responses are binary (skill is used, or not). different occupations. Employers also confirm the importance of interpersonal skills and cognitive skills for both higher skill and medium/lower skill occupations. FIGURE 8: Skill use by occupational group (worker reported) Skill use (0=not at all, 3=high intensity use) 3.2 Skills valued by employers 2.2 2.1 2.0 2.0 79. When recruiting, employers look for a broad set of skills across occupational categories. 1.9 1.9 1.8 These findings are consistent with the picture emerging from the Household Survey. When 1.8 1.8 1.7 employers in Kosovo are recruiting new workers for higher skill occupations, they place 1.7 1.5 1.5 1.4 importance on cognitive and socio-emotional skills that affect productivity: conscientious- 1.4 1.4 1.3 1.3 1.2 1.3 ness, perseverance, problem solving, numeracy, and the ability to work under stress (i.e., 1.1 1.1 stress resistance). Employers of medium and low skills occupations also value interperson- 0.4 0.6 0.5 al skills, as well as reliability, problem solving skills, and stress resistance. These general 0.4 0.3 skills are more important for hiring decisions than job-specific technical skills in Kosovo (Figure 9). The rankings are consistent across different categories of firms, such as those Physical Repetitive skills tasks Problem use solving Computer Numeracy Reading Independent Interpersonal Learning new that faced hiring constraints because of skills compared with other firms, and hiring firms versus non-hiring firms (not shown here). Technical Socio-emotional Cognitive HIGHER LEVEL (A) SERVICE (B) PHYSICAL/TECHNICAL (B) 42 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 43 FIGURE 9: Ranking of skills by employers 3.3 There are no perceived skill gaps Ranking of skills for higher skill occupations (index= 0-300) Conscientiousness 89 in the employed workforce Problem solving 73 80. A vast majority of Kosovo firms are satisfied with the skill levels of their workers. This Numeracy 55 finding is in sharp contrast to the skill gaps hiring firms encounter among applicants. For the five skills that are most highly rated for either higher skill or medium to lower skill Stress resistance 54 occupations (see above), fewer than ten percent of firms report a skill gap among their Perseverance 54 current workforce (Figure 10). The only skill for which more than 10 percent of firms Grit 39 reported skill gaps is fluency in foreign languages other than English. Job specific 31 technical 81. Even among skill constrained firms, and firms planning to hire, only a minority of Advanced computer 28 firms report skills gaps among their employees. Firms that experienced skill-related English 27 difficulties during recruitment (“skills constrained”), as well as firms planning to hire in the coming years, rank skills similarly to other firms. Even for these two categories, Interpersonal skills 27 the share of firms reporting a skills gap among their employees is below ten percent for Foreign language** 22 most skills evaluated, including those ranked most important. Only for language skills, Flexibility 19 between twenty and thirty percent of firms report that there is a gap. Among skill con- strained firms, language skills are not considered very important, reflecting the low share Official language* 13 of international integration and sophistication of Kosovo’s production of goods and ser- vices. Nevertheless, the lack of communication skills in foreign languages may become a constraint on the goal of increasing FDI and exports. Ranking of skills for lower and medium skill occupations (index= 0-300) Conscientiousness 99 FIGURE 10: Firms reporting skill gaps. Problem solving 81 Interpersonal skills 71 % firms reporting a skill gap (top five rated skills) Stress resistance 67 Conscientiousness 3.5 96.4 0.1 Grit 44 Problem solving 5.7 92.7 1.6 Higher skill Job specific 40 Numeracy 3.6 94.6 1.8 technical Flexibility 33 Stress resistance 4.5 95.4 0.1 Numeracy 23 Perseverance 4.0 95.8 0.1 Literacy (Serbian) 18 Conscientiousness 5.1 90.1 4.5 Lower/Medium skill English 14 Problem solving 7.6 85.7 6.4 Foreign language** 14 Interpersonal skills 4.6 91.7 3.4 Literacy (Albanian) 10 Stress resistance 7.7 87.4 4.6 ITC literacy 10 Grit 7.9 86.3 5.5 REPORTED GAP NO GAP SKILL NOT NEEDED SOURCE: STEP Employer Survey. Firms are asked to identify the three skills that are first, second, and third most important when deciding to retain a job seeker who is on probation. The index is calculated as 3 x (% of firms listing it as most important) +2 x (% of firms listing it second) +1 x (% of firms listing it third). If all firms rated the same skill as the most important, its index value would be 300. For conscientiousness for higher skill occupations, the index value (89) is calculated as 3 x 15.9 +2 x 14.8 +1 x 11.7. *Official language that is not one’s mother tongue. **Foreign language other than English. 44 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 45 % skill constrained firms reporting a skill gap (foreign languages) FIGURE 11: Skill gaps from the workers’ perspectives Higher skill Other 29.2 44.1 26.7 Education and jobs matching, Kosovo and comparators English 23.7 69.6 6.7 Tertiary 76.7 23.3 Post-sec, non-tert 59.8 19.1 21.1 Medium skill Other 24.9 26.9 48.2 Lower/ Upper Sec 10.6 61.8 27.5 English 29.4 31.8 38.8 Lower Sec 23.2 76.8 All 10.8 66.4 22.8 REPORTED GAP NO GAP SKILL NOT NEEDED SOURCE: STEP UNDER-EDUCATED WELL-MATCHED OVER-EDUCATED Employer Survey. Skill constrained firms are those that reported difficulties Education and jobs matching, Kosovo and comparators in hiring due to lack of skills or work Kosovo 10.8 66.4 22.8 experience among Serbia 5.3 71.4 23.3 applicants. Ukraine 5.3 65.6 29.1 82. A majority of workers report that their education level is matched with their jobs. In FYR MK 6.4 72.9 20.7 Kosovo, two-thirds of workers report a good match between their level of education and Georgia 6.3 60.6 33.1 job requirements, while 23 percent consider themselves overeducated, and 11 percent Armenia 7.4 63.5 29.1 undereducated (Figure 11). This is broadly in line with the pattern observed in comparator countries such as Serbia or FYR Macedonia. Across all levels of education, most workers are well matched, with the exception of post-secondary, non-tertiary education, where a UNDER-EDUCATED WELL-MATCHED OVER-EDUCATED majority consider their level of education too low compared to what is required by the job. SOURCE: STEP Household 83. Taken together these findings suggest that skill gaps, or perceptions of skill gaps, Survey. Over/ under education mostly affect labor market outsiders – job seekers - in Kosovo. This apparent dichot- level is measured omy – firms finding difficulties with recruiting new workers because of skill gaps, but as the difference few firms reporting skill gaps among their employees – is not unique to Kosovo. Similar between the level of education that patterns occur in Serbia and Albania. Assuming that this to some extent concerns youth, workers consider the reason could be that education systems have deteriorated in quality, or that skills adequate for their valued by employers are currently only developed on the job. job, and the worker’s level of education. 46 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 47 play a role in labor markets in Kosovo, although the influence on access to employment is stronger than that of earnings. The results discussed below are presented in detail in Annex 6. 4. What are the conse- 85. Socio-emotional skills and cognitive skills matter for access to employment in Kosovo (Annex 6, Annex Table 3). Scoring high on certain socio-emotional characteristics or skills is positively linked to the probability of being employed. Relative to the population at large, quences of skill gaps for those that score higher on conscientiousness and emotional stability (i.e., those that can be relied on to complete a job, even under stressful and difficult circumstances) are more likely to be employed. High scores on agreeableness indicate kindness, sympathy, Kosovo’s population? cooperation and consideration, whereas low scores point to ruthless, manipulative and highly competitive behavior. Somewhat surprisingly, individuals with higher scores on agreeableness have less access to employment. The use of cognitive skills outside of work (e.g., reading and computer work) are also related to job access.35 Literacy test scores are not systematically related to employment. Because literacy is a foundation- Key messages: al skill that underpins learning of other skills (e.g., computer use) and is essential for reading on and off the job, the independent effect of literacy may disappear when these • Skill gaps have economic and social consequences for individuals in Kosovo as having other skills are accounted for. When focusing on the active population only, comparing the right skills matter for access to jobs and higher earnings. unemployed with employed, the effect of conscientiousness on employment remains important. Those that score lower are more likely to be unemployed. Surprisingly, those • Socio-emotional skills, especially higher degrees of conscientiousness and emotional that are more likely to ask for advice and think about the long-term consequences of stability, are systematically related to access to employment. their actions (i.e., score higher on the decision making measure), are more likely to be unemployed (Annex Table 4). It is possible that this reflects a more selective approach to • Earnings are conditioned on access to good jobs (formal, higher paid occupations and employment, for example a preference for public sector work. sectors), jobs that require higher-level cognitive skills. 86. Women and youth are penalized in labor markets. Aside from skills, personal charac- • Skill gaps reinforce social and economic exclusion as those with lower skills tend to teristics unrelated to skills also have a strong association with access to jobs. Age is pos- belong to vulnerable groups, including those with lower levels of education, from ethnic itively correlated with employment and, as expected, the effect diminishes with age. This minorities other than Serbian, and from less favorable socio-economic backgrounds. pattern holds whether the analysis is undertaken for the entire population or is limited to the active population. Given the low participation rates of women in Kosovo, it is not sur- • There are no significant measurable skill gaps between men and women that could ex- prising that being female, and especially a married female, is negatively associated with plain worse labor market outcomes for women. Gender biased perceptions and norms job access, even after accounting for skills and education. Social norms assign family surrounding family obligations are likely more important factors. responsibilities to women in Kosovo. A combination of lack of affordable child and elderly care options, large family sizes, and disincentives such as extended maternity leave have several negative effects: (i) discourage women from participating in labor markets (ii) make it difficult for them to re-enter labor markets after extended absence (iii) reduce 4.1 Skill gaps limit access to good jobs incentives for employers to hire women. Available institutional child care centers are either not affordable or not sufficiently flexible to accommodate full-time employment.36 Family obligations are a factor behind female inactivity but not female unemployment 84. Skill levels influence both employment and earnings in Kosovo, even after accounting as women are more likely to be unemployed than men, but whether they are married for other important factors. Skills are expected to matter for labor market outcomes – does not matter for unemployment. Other factors, perhaps perceptions about women’s how well an individual can do job-related tasks should be rewarded by greater access to higher paying jobs. In economies that are not productivity driven or economies that have information gaps, this may not be the case. Diplomas, personal networks and personal characteristics may instead serve as the most important decisive factors for accessing 35  Note that because the analysis focuses on access to jobs, only skill use outside of jobs can be included in the estimates. good jobs. Multivariate analyses of the Household Survey data suggest that skills do 36  Cojocaru (2017), op. cit. 48 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 49 capacity or norms limiting what kind of jobs are available to women, may matter more FIGURE 12: Decomposition of the contribution to earnings by different characteristics. once women do enter labor markets. All age groups 87. Ethnicity and Family background matters. Ethnicity is also a significant correlate of job access. Serbians have greater access to employment than Albanians or those of other ethnicities. Also, a favorable family background in terms of having grown up with a more educated mother is positively related to job access. 37 YEARS OF EXPERIENCE DEMOGRAPHIC JOB CHARAC- SOCIO-EMO- COGNITIVE GENERIC 88. Earnings are linked to cognitive skills but not socio-emotional skills (Annex Table 5). EDUCATION (PROXIED BY CHARACTERIS- TERISTICS TIONAL SKILLS SKILLS SKILLS AT The impact of different types of skills differs for access to jobs and for earnings. Cogni- AGE) TICS WORK tive skills like reading, problem-solving and autonomous work are correlated with the 4.6 2.4 2.8 42.2 1.2 1.7 3.8 hourly earnings of wage workers. One explanation for this finding is that these particular 15-29 years cognitive skills, by design, overlap with job characteristics and may not fully capture an individual’s innate cognitive ability. The skill variable is measured by the intensity in which particular skills are used on the job, such that jobs requiring more cognitive skill use are also jobs that offer higher earnings. Unlike in the case of access to employment, socio-emotional skills are not related to earnings, with the exception of agreeableness which has a negative effect. 38 YEARS OF EXPERIENCE DEMOGRAPHIC JOB CHARAC- SOCIO-EMO- COGNITIVE GENERIC EDUCATION (PROXIED BY CHARACTERIS- TERISTICS TIONAL SKILLS SKILLS SKILLS AT 89. Job characteristics (sector, occupation, etc.) are by far the strongest correlates of AGE) TICS WORK earnings, even when controlling for education and skills. This is particularly true for -4.6 5.6 4.4 63 -0.5 -0.9 17.8 public sector work which pays off more than other sectors and jobs for people with similar qualifications and backgrounds. By contrast, personal characteristics, including 30-49 years gender, have no links with earnings. It appears that the biggest hurdles women face are participating in labor markets and finding employment in the first place, rather than discrimination in the work place once employed. It may also reflect the higher propensity for women to work in the relatively well-paid public sector. YEARS OF EXPERIENCE DEMOGRAPHIC JOB CHARAC- SOCIO-EMO- COGNITIVE GENERIC 90. A decomposition of the contribution to variation in hourly earnings of different character- EDUCATION (PROXIED BY CHARACTERIS- TERISTICS TIONAL SKILLS SKILLS SKILLS AT AGE) TICS WORK istics for different age groups reinforce the message on the importance of job character- istics such as sector, occupation, formality, etc. (Figure 12). In fact, these characteristics 7.9 0.1 0.6 41.7 2 2.3 1.5 are the main explanatory factor for earnings variability for every age group. For youth (15- 50-64 years 29), generic work skills (e.g., interpersonal skills, problem solving, learning new tasks and autonomy, all of which are necessary for modern jobs) also influence earnings, but much less than jobs characteristics. Additionally, socio-emotional skills appear to have a stronger effect on earnings for the age group 50-64 than for younger workers. YEARS OF EXPERIENCE DEMOGRAPHIC JOB CHARAC- SOCIO-EMO- COGNITIVE GENERIC EDUCATION (PROXIED BY CHARACTERIS- TERISTICS TIONAL SKILLS SKILLS SKILLS AT AGE) TICS WORK 7.2 2.5 2 48.2 8.5 3.9 4.2 37  This finding may indicate that educated females make better educational choices for their children, but the variable could also be picking up higher socio-economic status which in turn may be a proxy for access to higher quality education, access to professional networks, or simply – in the case of female children – the impact of a SOURCE: Estimates based on STEP Household Survey. NOTE: Results obtained using regression-based decomposition according to the Field’s role model in terms of career achievements. approach (Fields, 2004) where the dependent variable is the log of hourly wages trimmed at the 5th and 95th centiles. The height of each bar represents 38  Agreeableness has shown mixed results on earnings in other studies of low and middle-income countries the total R-squared for that regression. The sub-components of each bar show the contribution of each factor (or set of regressors) to the R-squared. using STEP data: there was a significant negative association in Armenia and Ukraine and a positive association Education=Years of education; Demographic Characteristics=Gender, Marital status, Ethnicity, Mother’s education; Job characteristics=Sector, in Colombia (Valerio and others, 2016, op. cit.). The explanation for a negative sign is that agreeableness could Occupation, Formality, Private/Public, Permanent/Temporary, Region; Socio-emotional skills=Big 5+Grit and Decision-making; Use of information- carry costs in terms of career advancement if individuals put others’ needs in front of their own. processing skills at work=Reading, Numeracy, Computer use; Generic skills at work=Interpersonal skills, Problem solving, Learning, Autonomous work. 50 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 51 FIGURE 13: Core literacy score by different characteristics 4.2 Skill gaps reinforce social and economic exclusion Distribution of core literacy score (demographics) 91. Skills are not the only determinant of labor market outcomes in Kosovo, but they 50% 51% 49% 55% 54% 46% 46% 39% 49% 83% 25% matter. Low literacy levels, lower use of cognitive skills, and lower levels of important socio-emotional skills among certain groups put them at significant disadvantage in the 12% labor market which reinforces social and economic exclusion. 25% 14% 92. Basic literacy is low in Kosovo overall. Literacy is a critical foundational skill that serves 31% 31% 30% 31% 29% 32% as the basis for further skill acquisition. The core literacy test in the Household Survey 32% 29% provides an indication of basic literacy. In Kosovo, nearly one in seven respondents (14 9% 50% percent) failed the core literacy test. For those that continued on to the full literacy 9% 8% 11% 7% 5% 8% 7% test, the average score remained low, barely reaching level 2 of 5, indicating that they 5% 5% 27% 1% 14% 13% 14% 11% 14% 15% 13% are capable of recognizing basic vocabulary, evaluating the meaning of sentences, and 8% 4% reading paragraph text. They are, however, not capable of understanding and evaluating TOTAL Male Female 15-24 25-34 35-44 45-54 55-64 Albanian Serbian Other dense and complicated texts that include relevant and irrelevant information or require background knowledge (see Annex 3 for more details). Gender Age group Ethnicity FAIL (0-2) PASS (3-5) PASS (6-7) PASS (8) 93. Low average literacy scores reflect significant differences in foundational skills across different groups. Results from the core literacy test suggest that a few groups are significantly more handicapped than others, even in terms of basic literacy (Figure 13). The handicap is particularly strong for individuals belonging to an ethnicity other Distribution of core literacy score (education and labor market status) than Albanian or Serbian (Turks and Roma are the biggest minorities after Serbs), among whom more than half of the respondents failed the basic core literacy test and thus were 37% 54% 62% 52% 46% 61% 41% never subjected to the full literacy test. Education also matters for basic literacy; twenty percent of those with less than upper secondary education failed the test compared to 8 percent of those with tertiary education. It is surprising, however, that only 60 percent of those with tertiary levels of education had a full score (8) on the core literacy test. 32% 28% 34% 30% 31% 27% 29% 10% 10% 6% 5% 8% 21% 3% 4% 21% 11% 8% 12% 11% 6% Lower seconadry Upper seconadry Teritary Employed Unemployed Students Other inactive or less Education Labor market status FAIL (0-2) PASS (3-5) PASS (6-7) PASS (8) SOURCE: STEP Household Survey, Core literacy assessment 52 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 53 94. Age and ethnicity remain important for scoring higher on the full literacy test. A similar picture There are also significant differences in the use of advanced cognitive skills on the emerges when looking at scores on the full literacy test (Figure 14). Literacy scores are higher job by ethnicity and education levels. Lower educated workers and those of non-Alba- for the younger generation, which is a sign that proficiency is improving. Not surprising, it is also nian or non-Serbian ethnicities are much less likely to use higher order cognitive skills higher for those with tertiary education, for formal sector workers and for workers in higher skills on the job than others. For education, the largest differences are between those that occupations (type A). Literacy rates are also significantly higher among those with Serbian eth- have tertiary education (about 17 percent of the total employed) and those with less nicity than among other ethnic groups. In fact, this is the only sub-group in Kosovo that is at par education (Figure 15, a and b). The differences are largest in the use of cognitive skills with average national levels in more developed European economies. Participation in early child- and physical tasks (which are used more intensively by low educated workers). There hood development (ECD) programs is also associated with higher literacy rates. Further analysis are surprisingly small differences in socio-emotional skills needed for jobs held by work- shows that literacy is also associated with greater wealth, higher levels of education, as well as ers with different levels of education, reflecting that interpersonal, communication, and participation in ECD, although the marginal effect of ECD is smaller for higher levels of literacy. supervision skills are needed in many different occupations. The significant differences in cognitive skills on the job can mean certain groups are excluded from jobs in which skills are used more intensively. These groups, therefore, are not developing important FIGURE 14: Score on full literacy test by different characteristics transversal skills necessary to acquire more productive jobs. Literacy Level 1 Literacy Level 2 Literacy Level 3 FIGURE 15: Ethnicity and education are correlated with intensity of skills use at work Total 228 a. By education (tertiary versus upper secondary) Gender Male 230 Skill use by level of education Female 227 Tech- nical 15-24 237 Physical tasks 0.7 1.1 1.8 25-34 232 Interpersonal Age group Socioemotional 2.0 1.9 skills 35-44 225 Repetitive 0.4 0.3 45-54 223 55-64 210 Independent 1.9 1.8 Albanian 227 Ethnicity Learning 2.0 1.5 0.5 Serbian 278 Problem 1.6 1.2 0.3 Other 189 solving Cognitive Attended Yes 246 Computer 2.0 1.2 0.7 ECD No 224 Numeracy 1.5 1.4 0.2 Lower secondary or less 210 Education Reading 2.1 1.0 1.1 Upper secondary 234 Tertiary 245 DIFF TERT_UPPER-SEC TERTIARY UPPER SECONDARY Employed 236 Job-specific technical and socio-emotional skills by level of education Labor market Unemployed 229 Operating status 0.1 0.1 machinery Technical Students 244 Repairing 0.1 0.1 equipment Other inactive 211 Driving 0.1 0.4 0.4 Higher skill 243 a vehicle Job character- Lower/Medium Skill 230 Supervising emotional istics 0.3 0.3 others Socio- Formal 237 Making 0.4 0.4 Informal 231 presentations 175 225 275 325 DIFF TERT_UPPER-SEC TERTIARY UPPER SECONDARY SOURCE: STEP Household Survey, full literacy assessment (Educational Testing Service, ETC). See Annex 3. Green bars denote “best performer” in the respective category. ECD: Early Childhood Development. 54 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 55 b. By ethnicity (Albanian vs. non-Serbian, non-Albanian) 96. There are also marked differences in individual scores on socio-emotional character- Skill use by ethnicity istics for different education levels and different ethnicities. Statistically significant dif- Tech- nical Physical tasks 2.0 1.7 ferences in means for men versus women, education levels, and ethnic groups (pairwise) are presented in Figure 17. Those who stick to a task, ensure that it is well done, and Interpersonal can handle stressful situations are more likely to have higher education levels. However, Socioemotional skills 1.7 1.9 the association is bi-directional because, as discussed in the introduction, these abilities Repetitive 0.4 0.3 are also developed during school. There are marked differences in some important so- cio-emotional skills between ethnic groups, especially for grit (the ability to pursue long Independent 0.4 1.4 1.8 term goals, even if they are difficult). Learning 0.8 0.9 1.7 97. Socio-economic background influences cognitive and socio-emotional skills out- Problem 0.8 0.5 1.3 comes independent of education and ethnicity. In multivariate analysis of the determi- solving nants of skills (presented in Annex 7), more education remains linked to higher levels of Cognitive Computer 0.9 1.4 cognitive, information processing skills such as reading, numeracy and computer use, even when other factors such as family background and gender are accounted for (Annex Numeracy 0.4 1.1 1.4 Table 6). Ethnicity and schooling are also positively associated with socio-emotional skills scores (Annex Table 7). However, personal characteristics like gender (women tend to Reading 0.7 0.5 1.3 score higher on several socio-emotional characteristics) and family background (level of education of one’s mother, household wealth, and ethnicity) matter more for cognitive DIFF ALB_OTHER OTHER ALBANIAN skill levels and for most socio-emotional skills, than an additional year of schooling. There Job-specific technical and socio-emotional skills is no effect of ECD attendance on socio-emotional skills. As discussed below, it is possible Operating that the unique influence of ethnicity and socio-economic background on socio-emotional machinery 0.1 0.1 skills reflects inequities in access to higher quality schools. Technical Repairing equipment 0.1 0.1 Driving a vehicle 0.4 0.4 FIGURE 16: Education and ethnicity are correlated with intensity of skill use outside of work Supervising emotional others 0.2 0.3 Differences in skill use outside of work Socio- Making 0.2 0.2 0.4 Computer 0.5 1.4 1.9 presentations DIFF ALB_OTHER OTHER ALBANIAN Numeracy 0.3 1.6 1.9 SOURCE: STEP Household Survey. For the top chart, responses range from 0 (not used on the job) to 3 (high intensity of use). For the bottom Reading 0.8 1.7 2.5 chart, responses are binary (skill is used, or not). The blue bar represents the gap – the difference in means between the two categories. Only statistically significant differences in means reported (5% level). DIFF TERT_UPPER-SEC UPPER SECONDARY TERTIARY 95. Those with low levels of education, and of ethnicities other than Serbian and Alba- Differences in skill use outside of work nian also use skills less outside of work. One advantage of measuring cognitive skills Computer 1.3 1.3 use outside of job is that this measure includes the entire population, not only those employed. Including also the inactive and unemployed, those with tertiary education Numeracy 0.4 1.2 1.6 are significantly more likely to use computer skills, numeracy skills, or to read outside Reading 0.8 0.9 1.6 of work. Similar differences, albeit smaller, are also present for Albanian versus other (non-Serbian) ethnicities (Figure 16).39 DIFF ALB_OTHER OTHER ALBANIAN SOURCE: STEP Household Survey. The blue bar represents the gap – the difference in means between the two categories. Responses range from 39  Differences between Serbian and Albanian nationalities are not shown here. The Serbian ethnic group tends 0 (not used on the job) to 3 (high intensity of use). Only statistically significant differences are reported (5% level). to display higher skill levels compared to Albanians. 56 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 57 Education and ethnicity are correlated with scores on socio-emotional FIGURE 17: skills and characteristics 4.3 Skill and gender gaps Differences in socio-emotional skills (original scale 1-5) 100. There are no strong differences in skill levels between women and men with respect to literacy or the skills that characterize their jobs. As shown above, the literacy test shows no differences between men and women in basic reading proficiency. Men use 0.5 numeracy and computer skills at home more than women, but this isn’t true for use of these skills on the job. Overall, differences in skill use on the job between men and wom- 0.3 0.3 en are small, except on-the-job skills that reflect occupational segregation, such as driv- 0.3 ing, or doing physical tasks (Figure 18). Women are also less likely than men to work on 0.3 0.3 0.3 0.2 0.2 independent tasks or to be in a supervisory position, although the differences, again, are 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.1 0.1 small. Working women are, however, more likely to read on the job than working men. 0.1 0.1 0.1 0.1 0.1 0.1 -0.1 101. Women score slightly higher on socio-emotional skills such as agreeableness, deci- -0.1 -0.1 sion-making, and grit. There are some statistically significant gender differences in the socio-emotional skills measured in the STEP (agreeableness, decision-making, and grit) Openness “Conscientious- Extraversion Agreeableness “Emotional Grit “Decision ness” stability” making” but these differences are small, especially when compared to difference across other, non-gender related characteristics, such as education or ethnicity (as seen in Figure 17 TERTIARY_UPPER-SEC TERTIARY_LOWER-SEC ALBANIAN_OTHER SERBIAN_ALBANIAN MALE_FEMALE above). As seen above, decision making is positively correlated with access to employ- ment, whereas agreeableness is negatively correlated. In all, differences in skills do not SOURCE: STEP Household Survey. Each bar represents the difference in means between group x and group y (X_Y). Only statistically significant appear to explain gender differences in employment and earnings outcomes. differences in means are presented (5% level). 102. Some employers nonetheless report that men’s skill levels are superior to those of women. Perception of skill level is what matters for recruitment, but these may be 98. Participation in early childhood development programs is linked to the development of cognitive skills. In multivariate analysis, ECD participation before age 7 had a positive, FIGURE 18: Gender differences in skills (on the job) significant impact on cognitive skills used on the job, especially numeracy and the use of computers (Annex 7). ECD participation was more important for cognitive skills than an- Skill use by gender other year of schooling, pointing to the importance of early education in getting children Tech- nical off to a good start. Cognitive skills were also higher for those with favorable family back- Physical tasks 0.4 1.4 1.8 grounds (having grown up with an educated mother and in households with higher wealth Interpersonal Socioemotional levels). As noted, ECD participation does not influence the levels of socio-emotional skills, skills 1.9 1.9 but education and family background are important for the development of both cognitive Repetitive 0.3 0.4 and socio-emotional skills. This is consistent with findings from the 2015 PISA survey in Kosovo in which those who had attended preschool performed significantly better than Independent 0.2 1.6 1.8 others in tests on reading, math and sciences.40 Learning 1.7 1.5 99. In summary, education (including preschool attendance), ethnicity and socio-economic Problem 1.3 1.3 solving background influences the level of skills of individuals, whether measured directly (lit- Cognitive eracy and socio-emotional skills) or in terms of how much these skills are used at work. Computer 1.4 1.3 The disadvantaged groups are consequently penalized in labor market outcomes which, Numeracy 1.4 1.4 in turn, reinforces inequities in wellbeing. Reading 1.4 1.2 0.3 DIFF MALE_FEMALE FEMALE MALE 40  Findings from “Kosovo, PISA 2015 Analysis Report (2017)”, prepared by the Kosovo Ministry of Education, Science and Technology under the Bank-financed Education System Improvement Project. 58 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 59 Job-specific technical and socio-emotional skills, by gender Gender differences in cognitive, socio-emotional and technical skills, FIGURE 19: Operating according to employers in Kosovo 0.1 0.1 machinery Technical Repairing a. All firms equipment 0.1 0.1 Who is better at skill x? (% of firms) Driving Conscientiousness 17 9 26 0.4 0.2 0.5 a vehicle Supervising Problem solving 14 8 22 emotional 0.1 0.2 0.3 others Socio- Interpersonal skills 10 9 20 Making 0.4 0.4 presentations Numeracy 1 9 10 Stress resistance 29 5 34 DIFF MALE_FEMALE FEMALE MALE Grit 31 5 35 SOURCE: STEP Household Survey. The blue bar represents the difference in means between group x and group y (X_Y). For the top chart, responses range from 0 (not used on the job) to 3 (high intensity of use). For bottom chart, responses Perseverance 33 3 37 are binary (skill is used, or not). Only statistically significant differences in means are presented (5% level). Job specific technical 27 4 32 Flexibility 13 10 23 Computer 5 6 11 English 14 8 7 Foreign language 14 5 9 biased. A majority of employers do not think that skills differ between men and women. GENDER PERCEPTION GAP FEMALE MALE Those that do (between ten and forty percent of firms) favor men, however (Figure 19, a). Between 35 and 40 percent of employers think that men are better at working towards b. Women dominated firms only long-term goals (grit) and not giving up, even with complex and difficult tasks (perse- Who is better at skill x? (% of firms) verance) and in stressful situations (stress resistance). Only a few percentages of firms Conscientiousness 19 8 27 think women are better at these skills. Moreover, this is the case even for firms that hire predominantly women (Figure 19 b).41 The perception bias sets Kosovo apart from other Problem solving 18 6 24 countries in the Western Balkans. In Serbia and Bosnia-Herzegovina, for example, most Interpersonal skills 13 8 22 firms report no gender difference in skills but, when they do, they tend to favor women. This said, in Kosovo, the biggest perceived gender differences in skills were not for the Numeracy 4 8 12 most highly valued skills like conscientiousness or problem solving. Stress resistance 31 4 35 Grit 33 4 37 Perseverance 35 3 38 Job specific technical 5 6 11 Flexibility 17 9 25 Computer 5 6 11 English 14 7 7 Foreign language 14 4 10 GENDER PERCEPTION GAP FEMALE MALE SOURCE: STEP Employer Surveys, estimates from Davalos, M., S. Johansson de Silva, A-M Munoz Boudet, and A. Silva (2018), “Are Skills a Constraint to Women’s Access to Jobs? Perspectives from the Western Balkans”, mimeo, forthcoming. Ranking of skills from top to bottom by 41  Given the low employment levels of women in Kosovo, the average for total firms reflect largely male-dom- importance. The gender perception gap is the difference between the shares of employers who found men versus women better at the particular inated firms. skill. Female-dominated firms are those that employ more than 50% women. 60 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 61 103. Some employers view male applicants more favorably. When asked to rank the im- FIGURE 21: Employer reported issues when recruiting women portance of personal characteristics in the recruitment processes, 42 and 45 percent of employers (hiring for higher skill occupations) report that personal characteristics do not matter in the hiring decision (Figure 20). For medium to lower skill occupations, age Recruitment of women is affected by.... and gender do weigh into hiring decisions. An applicant’s age is reported most important, … women’s family obligations 22 19 followed by gender, and then personal connection to the applicant. Among employers who do value personal characteristics in hiring, they overwhelmingly state a preference … the higher cost of employing 22 19 women for hiring males and applicants between the ages of 25 and 44. … women’s lack of experience 16 17 …women’s expectations of greater 13 18 FIGURE 20: Role of personal characteristics in hiring benefits … women’s lack of skills 10 14 Ranking of importance of personal characteristics for hiring (index = 0-300) % of firms agreeing with statement Ethnicity -7 3 STRONGLY AGREE SOMEWHAT AGREE SOURCE: STEP Family relations/ Employer Survey -33 35 personal ties Gender -96 80 Age -131 132 105. Skills constraints do not differ significantly between firms that employ mostly wom- MEDIUM-LOWER SKILL OCCUPATIONS HIGHER SKILL OCCUPATIONS en, or mostly men. These two groups of firms – defined as firms which employ a majority SOURCE: STEP Employer Survey. For of women versus men – tend to value similar skills (e.g., interpersonal skills, conscien- an explanation on index tiousness and grit). There is no evidence that female dominated firms experience great- calculation, see footnote er skill constraints when recruiting than male dominated firms do, except that female to Figure 9 dominated firms are more likely to report that applicants for managerial jobs lack skills or experience.42 106. In conclusion, lack of education and socio-cultural norms rather than skills emerge 104. Firms associate recruitment of women with difficulties, but for reasons other than as the two most important obstacles to women’s successful integration into labor skills. A majority of firms do not find any specific gender issues with respect to hiring. markets in Kosovo. Education is a stronger correlate of women’s access to employment However, over 40 percent of employers agree or strongly agree that the recruitment of than of men. Once on the job, there are no significant differences in skill level between women is affected by their competing demands due, in part, to prevailing norms that men and women’s access to employment. A majority of firms report that women and identify women as the primary care taker (Figure 21). Employers also report that women men have similar skill levels and that gender does not weigh into hiring decisions (al- are costlier to hire because of labor regulations for maternity leave and similar provisions. though when they do, men are favored). Social norms surrounding family obligations and These constraints could both affect women’s performance on the job, and discourage labor regulations that increase the cost of hiring women matter more. women from applying in the first place. 42  Davalos and others (2018), op. cit. 62 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 63 workers seeking to develop certain skills, their parents and communities that influence their choices, education systems that promote these skills, firms that search for skills to provide goods and services, and policy makers. From this perspective, insufficient levels of labor market relevant skills in Kosovo are potentially caused by: (i) insufficient oppor- 5. What causes skill gaps tunities for skill formation in early childhood; (ii) low quality and relevance of school- based education and vocational training as well as training programs delivered through active labor market programs; (iii) low involvement of the private sector in education and in Kosovo? training; and (iv) information gaps regarding skills and vocations that: limit the ability of parents to make informed decisions about children’s schooling; limit the ability of youth to choose relevant vocations and develop necessary skills; impede education and training systems from providing relevant and high quality education; and limit firms’ ability to evaluate job applicants based on education and training paths. Key messages: • Access to early childhood education is limited, especially for more vulnerable groups. • Low quality of education is a problem: firms complain about the quality of education 5.2 Access to quality education and training systems and Kosovo students perform poorly on cognitive skill tests compared to stu- dent from other countries. • Inequities in terms of access to quality education persist, and children from disadvan- taged socio-economic backgrounds do worse in school. Early skill formation • Training delivered through active labor market programs are not closely connected with labor market demand and not well monitored or evaluated. 109. Skills developed in early childhood (preschool aged children in this case) are instru- mental in shaping future education and labor market outcomes. Linguistic develop- • Firms use predominantly informal channels and poaching to recruit workers, which ment, the ability to ability to learn, cooperative behavior, are forming at this time, and penalizes new job seekers without references. failure to develop these critical skills can have long-term negative effects on education, health and earnings. The design and implementation of high quality early childhood inter- • Firms are not involved with education or training systems and do not provide opportu- ventions can have significant positive effects on overall well-being. Moreover, because nities to their staff. for further skill development. these early years are formative, interventions during early childhood are more cost ef- fective relative to those aimed at youth or adults.43 110. Access to preschool is low but increasing. Less than one out of five adults in Kosovo’s 5.1 Potential causes behind skill gaps urban population participated in any form of early childhood education, and only three percent of the population attended pre-school for two or more years. Encouragingly, participation is higher among the younger cohorts (15-24), but still only about a third of 107. The analysis above has shown that for potential employers and especially for more this age group attended preschool. Access to preschool is hampered by the limited num- dynamic firms, lack of skills and relevant work experience are major constraints to hiring ber of programs, and an unequal distribution of facilities in urban areas. The incidence of in Kosovo. By contrast, firms report that their current employees have the necessary children under six in some form of child care outside of the family is lower in Kosovo than skills. The analysis also indicates that different groups have different skill levels. Those would be expected given its GDP per capita.44 lacking certain skills and characteristics are less likely to be securing or working in jobs with higher earnings. 108. What accounts for these skill gaps? Skills development depends on the information, 43  Banerji et al., 2010, op. cit. resources and incentives available to the key stakeholders: children, youth and adult 44  Cojocaru, 2017, op. cit. 64 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 65 111. Socio-economic status is a large component of participation in ECD programs: those Western Balkans) on primary and secondary education, suggesting that expenditure on of Serbian nationality, those with highly educated mothers, and those from wealthy education may still be insufficient. households are the only groups in which more than twenty percent have participated in ECD. Evidence from the PISA 2015 survey corroborate these findings: students of the 113. Access to quality education is a necessary condition for skills development: foun- bottom quintile of PISA scores in reading, mathematics and science, were nearly twice dational skills like literacy and numeracy that should be developed during the first years as likely as students of the top quintile of not having attended preschool education.45 of school are the basis for further skill development, and both higher order analytical skills and socio-emotional skills are expected to develop in school. The global evidence suggests that, overall, each additional year of schooling raises earnings by 8-10 percent. FIGURE 22: Participation in ECD, especially for longer durations, is low This correlation may occur for two reasons: first, school develops skills that are valued by labor markets; and second, education credentials can signal high levels of both cognitive % participating in ECD before age 7, by length of program and socio-emotional skills, or access to networks, both of which could result in higher earnings irrespective of schooling. However, analysis shows that the effect on earnings is 21.7 largely driven by the skills acquired in school, as opposed to individual with higher levels of ability or stronger networks performing better in school.46 Education, hence, matters for skills, and equitable access to quality education is a primary objective to counter socio-economic exclusion. 8.7 5.6 11.3 114. By European standards, Kosovo’s population has low levels of education. Enrollment 17.1 15.9 is now near universal in basic education (primary and lower secondary) and around 90 percent for upper secondary. The stock of workers changes slowly, however. According 2.3 9.5 6.4 1.3 to workers evaluated in the STEP survey, however, just around 22 percent of the popu- 2.8 4.2 3.5 4.4 3.1 3.4 0.7 lation between ages 25-64 have tertiary levels of education, and over 30 percent have 4.6 2.9 5.7 4.5 0.3 5.3 7.0 4.9 5.0 4.5 13.4 1.0 14.5 0.6 12.5 not completed upper secondary school.47 In comparison with countries in the European 18.9 13.2 10.3 2.0 1.5 9.8 9.8 8.5 Union, that are also in Central and Eastern Europe, Kosovo has a lower share of tertiary 1.5 0.9 0.2 9.1 8.3 9.6 0.7 0.8 8.9 0.6 7.7 7.0 educated, and a higher share of lower educated (Figure 23). Moreover, the estimates for 4% 0.1 1.0 1.8 1.9 education levels above upper secondary in Kosovo are likely inflated given that the STEP survey by design focuses on urban areas only where we expect education levels to be Total Male Female Low 15-24 25-34 35-44 45-54 55-64 Other Medium High Lowest Albanian Second Third Serbian Forth Highest higher, whereas EU countries include both rural and urban areas. Mother’s Demographics education Household wealth SOURCE: STEP 1 YEAR 1-2 YEARS OVER 2 YEARS Household Survey Formal education 112. Public expenditures on education have grown in recent years and are catching up with regional averages. Public spending on education grew steadily from 3.3 percent of GDP in 2007 to 4.3 percent in 2013. Yet, Kosovo still spends less than the average ECA country (4.6 percent) and short of the OECD recommendation of 5 percent. Moreover, 46  World Bank (2018). World Development Report 2018: Learning to Realize Education’s Promise. Washington, owing partly to high enrollment and partly to a large youth population, Kosovo spends DC: World Bank. 47  In the 1990s, the education system for the Albanian speaking population was largely informal and under- less per pupil than most comparator countries (and less than any other country in the ground after the Serb-dominated government at that time took control of the formal system. Mass dismissals of ethnic Albanian teachers who refused to adhere to the central government curriculum and widespread boycotts of government schools by Albanian Kosovars led to the creation of informal Albanian-language schools. As a 45  Kosovo PISA 2015 Report, op.cit. result, a large share of Kosovars currently in the 25–45 age-group have limited formal education. 66 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 67 FIGURE 23: Distribution of population by level of education in Kosovo, Turkey, and EU (selected) FIGURE 24: Access to higher levels of education is low, especially for certain groups a. Kosovo – gender, age and ethnicity Population 25-64 by highest level of education completed (%) 18 19 22 23 24 24 24 28 30 31 33 34 40 40 Highest level of education achieved 17 19 15 25 16 17 22 4 60 28 18 48 68 51 47 48 60 70 56 51 60 48 64 55 62 46 64 55 57 68 49 55 33 38 37 32 24 28 31 14 22 23 16 16 17 12 11 9 10 6 8 Total Male Female 25-49 50-64 Albanian Serbian Other 5 Romania Turkey Kosovo Slovakia Croatia Czech Republic Hungary Bulgaria Poland EU 28 Slovenia Latvia Estonia Lithuania Gender Age Etnicity LOWER THAN UPPER SECONDARY UPPER SECONDARY AND TERTIARY SOURCE: STEP LOWER THAN UPPER SECONDARY UPPER SECONDARY AND TERTIARY POST-SECONDARY NON-TERTIARY Household Survey POST-SECONDARY NON-TERTIARY (Western Balkans), Eurostat (EU) b. Kosovo, Macedonia and Serbia, level of education, total and by gender Highest level of education achieved 115. Inequities in the education system remain, especially concerning gender and eth- nicity. On average, women have almost one year less of schooling than men. As shown 17 25 26 19 24 23 15 26 28 above, a higher share of young women (ages 25-29) than young men have completed 51 47 tertiary education. However, there are still fewer girls enrolled in upper secondary edu- 56 54 60 65 cation and, even among youth, the share of women with basic education or less is much 60 50 57 higher than for men. These gaps are likely to be even higher in rural areas. Within Kosovo, the Serbian and Albanian ethnic groups also have much higher levels of education than other ethnic groups (Figure 24 a). 38 116. In Kosovo, levels of education are lower and gender gaps are larger than in Serbia 33 and Macedonia. The low level of education of Kosovo women and the large gap between 21 24 17 24 14 12 15 men and women place Kosovo apart from EU countries and its West Balkan neighbors (Figure 24 b). The share of the population with education levels below upper secondary is considerably higher (33 percent compared to 21 and 14 percent in FYR Macedonia and Kosovo MK Serbia Kosovo MK Serbia Kosovo MK Serbia Serbia respectively. The gap between the share of women and men with education levels Total Male Female below upper secondary, at 14 percentage points, is also much wider than in Macedonia (7 percentage points) and Serbia (3 percentage points). SOURCE: STEP LOWER THAN UPPER SECONDARY UPPER SECONDARY AND TERTIARY Household Surveys POST-SECONDARY NON-TERTIARY 68 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 69 117. Although access has increased, the quality of education remains a problem. Literacy below basic proficiency in reading, and 78 percent are below basic proficiency in math- is a foundational skill that should begin to develop in the early years of schooling. As dis- ematics Kosovo’s PISA results are low even in relation to its average level of income.49 cussed in section 4.2, average literacy scores reach just above basic levels, prompting concerns that schools are not, in fact, successfully imparting those skills. Compared to 119. These low achievement levels reveal a wide performance gap between Kosovar stu- OECD countries, but also other Western Balkan countries (Serbia) and countries in the dents and their peers in the region. Kosovo’s outcomes in science places its students, on former Soviet Union, Kosovo’s population scores poorly on the full literacy assessment. average, 4.5 years of schooling behind OECD and EU averages, and 2.5 years of schooling With an average score of 235, on a scale of 0-500 Kosovo is at the bottom of the distri- behind ECA country averages. Reading and math outcomes are equally poor. Kosovar stu- bution for employed individuals, significantly below Serbia, where average scores are dents’ scores are equivalent to 5 years of schooling behind the OECD average in reading, similar to those of Central and Western Europe, and far below the OECD average (Figure and a little over 4 years behind in math. 25). This is a significant skill gap since reading literacy is key to the development of a large range of skills – and the flexibility to adapt those skills – in a changing labor market.48 FIGURE 26: PISA performance, Kosovo and comparator economies. FIGURE 25: Mean ETS Literacy Scores, total working age population and employed (15-64) 564 556 535 Mean ETS literacy score in Western Balkan, FSU and OECD countries 493 490 493 475 464 487 446 441 432 427 418 411 427 279 279 413 384 277 405 277 378 371 276 274 362 273 352 274 347 270 268 268 261 257 256 249 246 235 Kosovo FYROM Kosovo Kosovo FYROM Croatia Bulgaria Montenegro Albania Singapore OECD average Croatia Bulgaria Albania FYROM Montenegro Singapore OECD average Croatia Bulgaria Montenegro Albania Singapore OECD average 228 Czech Slovak OECD Kosovo Georgia Armenia Serbia Ukraine Estonia Republic Republic average SCIENCE MATH READING STEP PIAAC (OECD) SOURCE: Kosovo SOURCE: STEP PISA 2015 Report, TOTAL POPULATION EMPLOYED Household Surveys Ministry of (Kosovo, Serbia, Education, Science Armenia, Georgia, and Technology Ukraine), PIAAC (Others) 120. PISA results confirm inequities inherent in the Kosovo education system. In science, 118. Young students’ overall performance in science, math, and reading significantly lags students from rural areas are about 1 year of schooling behind their urban peers, and behind those of comparator economies (Figure 26). In the 2015 PISA study, a large students from low-income households (bottom quintile) are about 1.5 years of schooling majority of Kosovar students do not reach basic levels of proficiency. Some 68 percent behind those from high income households (top quintile). While this is the smallest gap of students are below the level required for basic proficiency in science, 77 percent are in ECA – where the average between top and bottom income groups is almost 3 years of schooling – it is indicative of education systems that reinforce rather than eradicate social inequities. Moreover, given the overall low level of performance by Kosovo youth, the lagging groups are far behind the OECD population on average. 48  Pierre G., M.L. Sanchez Puerta, A. Valerio, and T. Rajadel (2014). STEP Skills Measurement Surveys. Innovative Tools for Assessing Skills, Washington D.C.: World Bank; STEP Skills Measurement Snapshot 2014, Washington D.C.: World Bank. 49  Kosovo PISA 2015 Report, op.cit. 70 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 71 121. Vocational training is prominent in Kosovo, but not in technical fields. About half of all FIGURE 27: Firms’ views on education and training systems are not positive. students in secondary education are enrolled in a vocational program and almost half of these are in turn enrolled in business and law, or health programs. These areas have offered Education/Training systems DO NOT produce people that... few employment opportunities in recent years.50 46 … have soft workplace skills 41 122. Students enrolled in vocational training perform worse than those in general education, … have practical relevant expe- 59 and the gap is significantly larger in Kosovo than in OECD countries. For the PISA tests, rience 39 students enrolled in academic programs outperformed students in vocational schools by 49 55, 51 and 64 points in science, math and reading respectively. These differences are much … have up-to-date knowledge 38 larger than the OECD average. Even after accounting for the socio-economic profile of stu- 44 dents and schools, students in general/academic and modular programs score 22 points … have the highly ranked skills 41 higher in science than students in vocational schools.51 … meet the skill needs of the 42 business adequately 36 123. Employers’ views reaffirm the presence of quality problems in Kosovo’s education sys- % of firms agreeing with statement tem. A vast majority of firms are satisfied with the level of education of a typical higher skill SOURCE: STEP or medium to lower skill worker in their firm (92 and 89 percent respectively are satisfied Employer Survey GENERAL EDUCATION SYSTEM TVET with the level of education). However, when asked about the quality and specific outputs of the education system, a significant share of firms identify problem areas. Nearly 60 percent Training opportunities for job seekers of employers report that the general education system does not produce students that have practical relevant experience, and nearly 50 percent report a lack of up-to-date knowledge, 124. Beyond school-based education, training provided through active labor market programs or socio-emotional skills like discipline, timeliness, or interpersonal skills. Although they are can facilitate the transition to employment and increase earning potential. Typically, such somewhat more satisfied with the experience and skills from the TVET System, between 35 training programs are short-duration interventions (less than 6 months) that focus on im- and 42 percent of firms find that the training systems do not graduate people with necessary proving skills and sometimes offer practical work experience to job seekers. Globally, there skills. This is worrisome, given that the objective of TVET is to provide students with direct is mixed evidence on the effectiveness of training programs for employment and earnings. work relevant skills in a profession or vocation. That being said, well designed and comprehensive skill development interventions could have positive effects on employment outcomes in the long run.52 125. In Kosovo, the coverage of active labor market measures in terms of beneficiaries is limit- ed. Public funds for ALMPs, which include skill training programs, have been limited to around 2 million Euros per year, or 0.03% of GDP, which is very low compared to an OECD average of 0.42%. The total number of beneficiaries still only represents about one tenth of all registered jobseekers in the country. The limited funding also affects the overall functioning of basic employment services as reflected in the low counselor to jobseeker ratio (about 1:1000).53 126. Vocational training is the largest category of ALMPs. The Employment Agency manages 8 Vocational Training Centers (VTCs) and 5 mobile Centers, offering vocational training and retraining through modular short-term training. These centers account for 71 percent of all beneficiaries. Given significant budget constraints, all measures other than vocational training, including those seeking to provide work experience such as on-the-job training and internships, have limited coverage. 52  Kluve, J., S. Puerto, D. Robalino, J.M. Romero, F. Rother, J. Stöterau, F. Weidenkaff, and M. Witte (2016). Do Youth Employment Programs Improve Labor Market Outcomes? A Systematic Review. IZA Discussion Paper No. 10263, Card, D., J. Kluve and A. Webber (2015). What Works? A Meta Analysis of Recent Active Labor Market Program Evaluations. 50  Kosovo Education and Employment Network (2017), Op. cit. IZA Discussion Paper No. 9236. 51  Ibid. 53  Republic of Kosovo (2018). Ministry of Labour and Social Welfare: Sector Strategy 2018-2022. 72 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 73 FIGURE 28: Beneficiaries in active labor market measures are mostly in training development of curricula, provision of work place experience, and testing of students is important for increasing the relevance of education and training. 6736 Number of beneficiaries 129. Kosovo firms are generally not well connected with education systems, especially not for medium to lower skill occupations. In Kosovo, as in other countries in the West- ern Balkans, firms and education systems do not interact with one another (Figure 29, a). On average, fewer than one in five firms is in regular contact with education systems. 4055 Some firms – notably those with a potential for hiring – are more likely to engage with 3230 education systems, at least with respect to higher skill occupations: large firms, state 2650 firms, public services and business services, those that invested in R&D, and those that 1769 experienced problems in hiring. 951 819 483 474 434 377 281 188 153 130. Firms are not interacting with education systems on a strategic level. Firms that do 49 54 2 Vocational training On-the-job training Internship Self-employment Wage subsidy Public works engage in regular contact with education systems largely do so to provide internships or train their workers and not at a strategic level. The share of these firms that provide feedback on curricula or help in the testing of students is low (Figure 29, b). SOURCE: Republic 2012 2015 2016 of Kosovo (2018). Ministry of Labour and Social Welfare: Less than one out of five firms engage with education systems; when FIGURE 29: Sectoral Strategy they do, it is not at a systemic level 2018-2022 a. Firms in contact with the education system 127. The vocational training provided through ALMPs is often of low quality and lack- % of all firms ing in relevance. The main challenges affecting the relevance and effectiveness of skill 37 training include: (i) lack of regular assessment of labor market demand; (ii) insufficient involvement of the private sector and social partners in the design of training curricula to ensure their relevance; (iii) lack of cooperation with businesses in the delivery of training 52 53 (in 2016 for example, less than 100 trainees conducted part of their vocational training 43 in companies); (iv) lack of quality assurance mechanisms; and (v) lack of monitoring and evaluation (e.g., absence of systematic tracer studies).54 37 23 24 25 21 19 17 19 16 17 18 17 15 13 13 12 11 15 5.3 Firms involvement in 17 15 13 10 8 8 7 7 8 8 5 6 5 5 4 4 4 3 4 2 education and training TOTAL Private State Yes No Yes No yes No domestic Foreign Invested Agr and Ind Construction Trade, Transpt, Ac- comm, Entertnmnt Business Services Public Services Other Services Large (51+ empl) Didn’t invest Micro (1-10 empl) Small (11-50 empl) 128. Firms are important stakeholders in demand-led, labor market relevant skill devel- opment systems. They embody the essential knowledge of skill requirement and pick up rapidly changing skill needs driven by global changes in demand and supply of goods and services. The private sector can also be a source of further skill development by Inter- Prob- Size Innova- providing training for their workers or internships for students. Their involvement in the Sector Ownership national R&D lems in (employees) tive Contacts hiring 54  World Bank (2018). Kosovo Social Protection and Health Expenditure Note. 74 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 75 b. Firms in touch with education systems, by purpose of contact FIGURE 30: Incidence of on-the-job, internal, or external training among firms. (% of those in contact) % firms reporting that they provided training Internships 81 78 Training of workers 58 66 66 62 57 Recruitment 52 41 49 47 47 18 47 43 40 41 40 37 35 29 34 Feedback on 34 31 31 30 28 27 education (curriculum) 45 23 23 22 22 21 21 21 19 18 19 19 17 15 Student testing 13 15 24 11 11 8 7 7 21 5 HIGHER SKILL MEDIUM TO LOWER SKILL OJT Internal External OJT Internal External SOURCE: STEP Employer Survey Higher skill occupations Lower/Medium skill occupations 131. The incidence of workplace-based training is low. To remain competitive, firms need SOURCE: STEP BUSINESS SERVICES PUBLIC SERVICES LARGE FOREIGN to upgrade the skills of the work force as well as identify those new skills among job Employer Survey STATE INT. BUSINESS CONTACTS PROBLEMS IN HIRING seekers. A majority of Kosovo firms do not provide training to their staff, not even in the form of on-the-job training (Figure 30). The incidence of training is higher among firms that are more likely to be in contact with education systems: public services, large firms, 5.4 Information gaps foreign-owned firms, firms that experienced skill problems in hiring (for higher skill occupations), and state-owned firms (for medium to lower skill occupations). Although firms are more likely to find women than men lacking in skills, men are more likely to receive training (in 60 percent of firms). This emphasizes that work place training is not 132. If education systems are plagued with quality problems, or grades and diplomas are likely to serve as a substitute for prior training, but more of a complement, to strength- influenced by other factors than skills (e.g., socio-economic background), employers en/impart firm specific skills, and that firms choose workers they identify as better able may have little trust in education credentials. If so, young graduates entering the labor to acquire these new skills. market are significantly handicapped because they lack the means to prove their ability. Individuals with limited social capital – likely the ones who already belong to vulnerable groups – will then be more constrained in their choices. There could be significant gains to providing information that would help workers identify jobs that are appropriate for their skill levels, maximizing their productivity potential given investments in education.55 133. Firms in Kosovo overwhelmingly recruit through informal channels. Skills that are highly ranked by employers (conscientiousness, ability to work well under duress) are not necessarily easy to evaluate in face-to-face interviews. Lacking other means of evaluating skills, firms in Kosovo often resort to informal networks, personal contacts, recommen- dations from friends, as well as poaching (i.e., approaching workers who are employed in other firms). The emphasis on personal recommendations and poaching implicitly raises the value of prior work experience. 55  Banerji and others, 2010, op. cit. 76 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 77 134. A small share of firms uses the public employment agency for identifying potential employees. Just over one fifth of firms connect with the public employment agency for recruitment, whether for higher or medium to lower skill occupations. Even fewer report connecting with education systems to identify potential job applicants. 135. The importance of contacts and networks for securing employment could also reduce incentives for students. New labor market entrants, especially those lacking access to 6. Going forward: policy professional networks through family and friends, are highly disadvantaged. Ethnic mi- norities, those from rural areas or poorer households, are at a higher risk of becoming marginalized. Excelling in school is not necessarily an optimal strategy if it is not likely reforms to reduce skill to pay off without connections. In fact, two thirds of Kosovars believe that non-merit factors like family connections, bribes, and party alliances for example are most helpful for securing employment in the private sector (an even higher share, 79 percent, believe gaps in Kosovo non-merit factors to be of highest value for the public sector). Thus, only 34 percent of respondents believe education, training and work experience matter.56 136. An in-depth analysis of the STEP Household and Employer Surveys shows significant scope for improving skill development in Kosovo for the benefit of firms, individuals, job FIGURE 31: Firms use informal channels to recruit workers creation and growth. To help foster more relevant skills in the work force, as set out in the Government’s own strategy for employment and social welfare, it will be necessary to (i) improve the quality of education from early school years through tertiary education, % of firms using specific channel for recruitment and assist vulnerable groups in accessing quality education; (ii) connect labor demand Informal channels (personal 64 contacts, recommendations) 59 – firms – with students and educational institutions in policy formulation and training delivery; and (iii) reduce information gaps relating to labor market developments more 46 Poaching broadly and skill needs more particularly. 41 Internet 37 137. The broad set of cognitive and socio-emotional skills that are valued in labor markets 30 must be nurtured from an early age and throughout education systems and beyond. Private Employment Services 26 Labor market relevant skills include a set of transversal cognitive and socio-emotional 18 skills that are necessary for modern, competitive firms: conscientiousness, independent 23 and creative work, ability to work under rapidly changing or stressful conditions, collabo- Public Employment Services 22 ration with others, and communication. These are developed from birth and throughout 20 the life course: from interventions in early childhood programs that lay the foundation for Other media (not internet) 17 basic cognitive and socio-emotional skills, through school based academic or vocational 16 learning onto university and, after schooling, through training and retraining programs Education systems for adults, as well as further skill development in the work place. Education and training 13 systems cannot be expected to provide workers fully prepared for employment. Their 11 Job Fairs responsibility is to graduate students with the capacity for developing abilities though 8 on-the-job training and adapting to new challenges as they arise. Hence, stepping up skill 1 Other development in Kosovo will require effort and collaboration between all stakeholders: 4 education systems, firms, students, and policy makers. SOURCE: STEP HIGHER SKILL OCCUPATIONS MEDIUM TO LOWER SKILL OCCUPATIONS Employer Survey 56  UNDP Kosovo, 2017, op. cit. . 78 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 79 Addressing inequity and quality problems in skill programs based on actual labor demand and on evidence of what works. Again, it will be development necessary to both collect information on skills in demand through firm level surveys and other relevant sources of labor market information and strengthen monitoring and eval- 138. The early years are formative and have long-term effects on further skill acquisition uation of existing initiatives and develop effective programs. Tracer surveys, employer and broader aspects of well-being. Participation in preschool education has a positive interviews, internship reports, and other data collection tools, together with information impact on foundational and advanced cognitive skills, but only a small number of children on regional and international best practice, can help identify the strongest programs and have access to early childhood education in Kosovo and the quality of preschool educa- how to continually improve upon them. tion is unknown. The immediate priority and challenge for the Government of Kosovo – as recognized in its sector strategy for education – is to expand access to preschool education, while increasing the quality of services. In this context, priority should be given to including vulnerable groups that have traditionally had little access to such sys- Connecting supply and demand: firms, students, and tems. Tapping into the private sector for expanding access should be explored along training institutions with designing and implementing quality and monitoring standards across providers. A significant upscaling of affordable preschool options, if matched with complementary 142. The involvement of firms in multiple dimensions of skill development systems is of activation policies, could have the additional and direct benefit of encouraging women’s utmost importance. Connecting employers, workers, education systems and students is labor force participation. essential to increase the relevance of school-based training. Collaboration can take dif- ferent forms, ranging from public-private sector partnerships involving shared financing 139. Education reforms need to increase the quality of education and training, especially and management responsibilities, to providing technology/equipment, jobs skill needs for vulnerable groups. The very low learning outcomes in Kosovo, even for foundational assessments, mentoring and career advice, partnerships around curricula reform, and skills like literacy, point to serious weaknesses in educational systems that can compro- work-based learning.57 mise Kosovo’s ability to become more integrated in international markets. The Kosovo education system needs to ensure that students develop a wide range of skills, includ- 143. Work based-training, in the form of internships and apprenticeships58, can increase ing socio-emotional skills that have been shown to have lifelong impact on well-being. the relevance of skills and provide youth with both experience and references. Intern- There is substantial evidence that socio-emotional skills are malleable over time and ships and apprenticeships systems are under developed in Kosovo, as in other countries that personal characteristics like discipline, long-term goal setting and decision-making in Western Balkans. Well organized, these arrangements have several advantages: (i) abilities can be nurtured and developed with adequate pedagogical methods. Increas- they provide trainees with an opportunity to participate in real work place situations and ing the quality of education and improving learning will require raising investment in solve concrete work-related problems; (ii) they contribute to building job specific skills as education, focusing efforts on targeting disadvantaged groups and investing in quality well as transversal labor market relevant skills that are highly ranked by employers; (iii) enhancing measures such as teacher policies and training, curricula reform, and ongoing they bring together firms, students and education systems and facilitate collaboration monitoring and evaluation. Education expenditure per pupil is low in Kosovo compared between stakeholders to improve education systems; (iv) they provide trainees with to similar countries and investing in education early, including for poorer families, will references from the private sector that can signal capabilities, which may be especially be essential to the development of a more productive economy. Because access to ed- important given employers’ emphasis on work experience; and (v) they create an entry ucation differs significantly between groups, it will be important to evaluate possible into a professional network for students. In the EU, an estimated 60-70 percent of ap- demand side constraints to education as well. This includes norms around schooling and prentices find employment immediately after graduation. To work well however, these gaps in information around career paths and pay-offs to education for children, families, arrangements need strong institutions and governance systems, including three party and communities. collaboration (e.g., firm-student-school, or firm-student-employment agency), formal learning arrangements or learning plans, monitoring and evaluation, and quality assur- 140. There is significant scope for improving the performance of vocational education and training, which account for half of all students at the higher secondary level (grades 10-12). This will require aligning the supply of vocational education training with assess- ments of technical skill needs in the labor market, involving employers in the design of training curricula, fostering cooperation with businesses in the delivery of training, and 57  Dunbar, M. (2013). Engaging the private sector in skills development, Health & Education Advice and Re- strengthening quality assurance mechanisms. source Team, https://www.educationinnovations.org/ 58  In the EU, apprenticeships and internships are part of formal TVET in many countries. Apprenticeships lead to formal (VET) qualifications. Internships do not necessarily lead to formal TVET qualification but can be an 141. It will also be important to increase the relevance of vocational training provided integral and formal part of a broader TVET program. See Broek, S., T. Hogarth, L. Baltina, A. Lombardi (2017), by the Employment Agency. In particular, there is a need to develop effective training Skills Development and Employment: Apprenticeships, Internships and Volunteering, IP/A/EMPL/2016-04, Report prepared for European Parliament’s Committee on Employment and Social Affairs. 80 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 81 ance systems.59 Recently, cooperation agreements have been signed with companies Reducing information gaps for internships in Kosovo, but monitoring and evaluation of these initiatives are weak.60 146. Assisting parents, children and youth in making informed choices on schooling, train- 144. To fulfil its core function to match jobseekers with employers, the Employment ing, and jobs can have significant benefits for education and employment outcomes. Agency needs access to a critical number of available vacancies. According to the The private sector must be involved in collaborative efforts to strengthen the availability STEP Employer Survey, there remains significant potential to increase this number be- of labor market information. Students and jobseekers need adequate and timely infor- cause only a minority of employers recruit through the Agency. To this end, strengthening mation on current opportunities, wages of different sectors and occupations, and quali- outreach to employers may be needed, as well as ensuring high quality of the services fications needed for various careers, as well as forecasts on future skill needs. Research provided to them (e.g., through effective preselection of candidates, and fast response indicates that in both developing and developed countries, students and their families sig- times). In preparation, the Employment Agency can also provide job seekers with crit- nificantly misjudge returns to different levels and forms of education, and that providing ical information on work place skills that are important for finding and retaining a job. relevant information can change incentives and career paths.61 The STEP surveys focus Socio-emotional skills like discipline and perseverance are developed over many years. on understanding the demand and supply of basic and higher-order transversal skills. However, the Employment Agency can inform job-seekers of the importance of different This type of analysis must be complemented with a better understanding of the techni- work place related competencies, including the importance of timeliness, independent cal vocations in demand. Strengthening Kosovo’s Labor Market Information System will work, and team work in order to influence, if not skills, at least priorities and incentives. require (i) establishing a systematic collection of information on growth sectors with high employment potential, as well as wages across sectors and occupations, (ii) integrating 145. Capacity building would help firms, especially smaller ones, to identify and evaluate labor market information from key stakeholders (employers, the Ministry of Education, skill and training needs. Smaller firms are not likely to have the in-house capacity to Ministry of Labor and Social Welfare, the Employment Agency, the Statistical office); and know how to continuously develop the skills of their staff and are not likely to be using (iii) making labor market information accessible through relevant channels such as the more sophisticated tools like personality tests or other forms of assessments when hir- Employment Agency website, Busulla62, and others. Career guidance systems in school ing, hence the reliance on informal channels. Industry associations can play an important (both general and vocational) can also assist students and their parents in choosing voca- role as partners in both identifying skill development needs and in representing firms in tions that match labor market demand, their own aptitudes, and interests, and in making the interaction with education systems and government. Capacity building at the industry the most appropriate educational choices to pursue a chosen career. or firm level is also possible to enhance the identification of skill needs as well as the use of different techniques to evaluate skills among applicants. In this context, there is a need to reinforce the message that skill levels are not different for women and men, once education differences and participation in labor markets are accounted for. Advo- cacy to firms of all sizes may also be needed to reinforce the importance of proactive engagement by the private sector – through approaching other stakeholders, providing work based training for students, and opening up work places to youth, women, and other disadvantaged groups. 61  Guallar Artal, Silvia, S. Johansson De Silva, V. Levin, A. Safir, and A-M Munoz Boudet (2016). From aspi- rations to occupations: the role of information in educational and labor market decisions in Moldova (English). 59  Ibid. Washington, D.C.: World Bank Group. 60  Kosovo Education and Employment Network (2017). Evaluation Report on Implementation of Kosovo Educa- 62  Busulla.com is an online platform developed by Ministry of Education for the purpose of helping students in tion Strategic Plan in 2017. http://www.keen-ks.net/site/assets/files/1345/raporti_i_vleresimit_psak_eng-1.pdf career planning. 82 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 83 ANNEX 1: Skills in STEP Household and Employer Annex 1 cont. Survey Can be relied on to get things done Conscientiousness STEP Household Survey STEP Employer Survey (only job related) Can continue in the face of challenging situations at work Grit Generic skills Degree to which skill Degree to which skill Job-related skills is important for job is used by typical (employed only) Can finish long and difficult tasks Perseverance (Type A workers) worker Cognitive Self-reported skills Self reported skills • Literacy • Reading skills used outside of job used on the job: (Read and write) • Writing Can work well in very busy times or difficult situations Stress resistance (Emotional Stability) and on job • Problem • Numeracy • Numeracy (Foundational • Reading solving • Problem solving* • Problem solving* Can easily adapt to new tasks or changes in the workplace Flexibility and ad- • Numeracy • Learning • ITC literacy* • Use of computer vanced) • Writing new things • Advanced • Foreign language • Computer computer use* Can find new and better ways to do things Problem solving • English Direct assessment • Foreign Language Can read and write in a foreign language, not English Foreign Language • Literacy (not English) Socio-emo- Personality traits Self reported skills • Conscientious- • Making presenta- tional skills (Assessment): used on the job: ness* tions Can read and write in English English • Openness • Making • Grit* • Interacting with a • Conscientiousness presentations • Perseverance* team Can read and write in mother tongue Literacy • Extraversion • Supervising others • Stress resis- • Agreeableness • Interpersonal skills tance* • Emotional stability • Independent work • Flexibility* Can read and write in other official language (not mother tongue) Official Language • Grit • Repetitive • Interpersonal • Decision making skills* Can use a computer for basic word processing tasks, email and ITC literacy (Type B workers) Job-specific Self-reported • Technical skills   internet searches Technical characteristics of job: Can use a computer for making presentations and/or other ad- skills • Repair/mainte- vanced purposes like creating and managing databases or using Advanced computer use (Type A workers) nance of specialized computer programs, etc. machinery • Operation of Can work well with others and listens to others’ views interpersonal skills machinery • Driving vehicle • Physical * see next page. Source: Compiled by authors. 84 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 85 ANNEX 2: Socio-emotional characteristics in the Household Survey Emotional Stability Socio-emotional skills Items (Questions in the background questionnaire) Do you have sudden changes in your mood?* (Neuroticism) Openness Are you curious, interested in learning new things?   Do you feel sad, depressed?*   Are you inventive, and discover new ways of doing things?   Do you stay calm in tense or stressful situations?   Do you like to think a lot, and reflect about ideas?   Do you manage stress well?   Do you come up with original or new ideas?   Do you get nervous easily?*   Are you interested in nature, art or music?   Do you get easily upset?*   Do you prefer work that involves repetitive tasks and routines?*   Do you worry a lot?*   Conscientiousness Are you organized? Grit Do you try very hard even after making mistakes?   Do you make plans and stick to them?   Do you stick to your goals despite obstacles and setbacks?   Do you work hard to do things well and on time?   Do you go after a goal, even if it takes a year or more to reach?   Do you get to work and appointments on time?   Do you finish your tasks, even if they take a long time to complete?   Do you put off your duties in order to relax?    Do you keep working very hard even when you feel like quitting?   Do you get easily distracted?    Do you finish whatever you begin?   Do you complete your duties on time?  Decision-making Do you ask for help when you do not understand something? Extraversion Do you enjoy being with people?   Do you think carefully before you make an important decision?   Do you easily share your thoughts and feelings with other people?   Do you think about how the things you do will affect others?   Are you enthusiastic and full of energy?   Do you think about how the things you do will affect you in the future?   In social gatherings, do you like to talk to many people? NOTE: Response categories range from 1 “almost never” to 5 “almost always”. The aggregation process was based on a simple average across items. Are you talkative? Negatively scored items marked by “*” were recoded with a score of 5 assigned for “almost never” and so on prior to the aggregation.     Are you comfortable expressing your thoughts and opinions to others? Agreeableness Do you work well with other people?   Do you try to understand how other people feel and think?   Are you generally trusting of other people?   Do you tolerate faults in other people?   Are you helpful with others?   Do you forgive other people easily?   Do you tend to be rude to other people?* 86 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 87 ANNEX 3: ANNEX 4: Full literacy test: Reading Proficiency Levels STEP surveys in Kosovo: Literacy Level Below Level 1 Score 0 to 175 Background information The tasks at this level require the respondent to read brief texts on familiar topics to locate a single piece of specific information. There is seldom any competing information in the text and the requested information is identical in form to information in the question or directive. The respondent may be required to locate information in short continuous texts. a. Kosovo STEP Household Survey However, in this case, the information can be located as if the text were non-continuous in format. Only basic vocabulary knowledge is required, and the reader is not required to understand the structure of sentences or paragraphs or make use Target population: The target population for the Kosovo STEP survey comprises all non-institutionalized persons 15 of other text features. Tasks below Level 1 do not make use of any features specific to digital texts. to 64 years of age (inclusive) living in private dwellings in urban areas of the country at the time of data collection. Level 1 176-225 This includes all residents except foreign diplomats and non-nationals working for international organizations. Most of the tasks at this level require the respondent to read relatively short digital or print continuous, non-continuous, or mixed texts to locate a single piece of information that is identical to or synonymous with the information given in the question or directive. Some tasks, such as those involving noncontinuous texts, may require the respondent to enter Sample design: The sample frame of Primary Sample Units (PSUs) is based on a list of Census Enumeration Areas personal information onto a document. Little, if any, competing information is present. Some tasks may require simple (CEAs) from the Kosovo Agency of Statistics CEA Frame for 2011. To facilitate the selection of separate Albanian cycling through more than one piece of information. Knowledge and skill in recognizing basic vocabulary, determining and Serbian samples, the sample frame was partitioned into two separate sub-frames: one sample frame for Al- the meaning of sentences, and reading paragraphs of text is expected. banian-majority populated PSUs, and another sample frame for Serb-majority populated PSUs. The sample frame Level 2 226-275 for the target population is a list of 1153 Albanian PSUs and 60 Serbian PSUs from the 2011 Census of Population. At this level the medium of texts may be digital or printed, and texts may comprise continuous, non-continuous, or mixed types. Tasks in this level require respondents to make matches between the text and information, and may require paraphrasing or low-level inferences. Some competing pieces of information may be present. Some tasks require the The survey used a 3-stage stratified sample design. The sample was explicitly stratified by 2 ethnic groups, i.e., respondent to: • cycle through or integrate two or more pieces of information based on criteria • compare and contrast Albanian majority PSUs and Serbian majority PSUs, and by 36 municipality areas. or reason about information requested in the question • navigate within digital texts to access and identify information from various parts of a document. Random selection method was used at all stages of sample selection. At the first stage 215 Albanian PSUs and 37 Level 3 276-325 Serbian PSUs were selected with probability proportional to size from 36 strata with PSUs. At the second stage 15 Texts at this level are often dense or lengthy, and include continuous, non-continuous, mixed, or multiple pages of text. Understanding text and rhetorical structures become more central to successfully completing tasks, especially navigat- households were systematically selected as the target sample in each selected PSU; in addition, 15 households were ing of complex digital texts. Tasks require the respondent to identify, interpret, or evaluate one or more pieces of infor- systematically selected as the reserve sample in each selected PSU. At the third stage the main respondent was mation, and often require varying levels of inference. Many tasks require the respondent to construct meaning across randomly selected in each visited household from among all household members aged 15 to 64 years; the substitu- larger chunks of text or perform multi-step operations in order to identify and formulate responses. Often tasks also tion of the main respondent was not allowed. demand that the respondent disregard irrelevant or inappropriate content to answer accurately. Competing information is often present, but it is not more prominent than the correct information. Level 4 326-375 Fieldwork: 9 Sep 2016 - 29 Mar 2017 Tasks at this level often require respondents to perform multiple-step operations to integrate, interpret, or synthesize information from complex or lengthy continuous, non-continuous, mixed, or multiple type texts. Complex inferences Final sample: 3,511 individuals aged 15-64 years (inclusive) living in urban areas. and application of background knowledge may be needed to perform successfully. Many tasks require identifying and understanding one or more specific, non-central ideas in the text in order to interpret or evaluate subtle evidence-claim Weighting: The data weighting was undertaken at three main stages (PSU, household- and individual-levels) by the or persuasive discourse relationships. Conditional information is frequently present in tasks at this level and must be taken into consideration by the respondent. Competing information is present and sometimes seemingly as prominent STEP survey methodologist to ensure consistency across sampling strategies. Based on the 3,511 participating as correct information. selected persons, the sum of the final person weights is 511,999. Level 5 376-500 At this level, tasks may require the respondent to search for and integrate information across multiple, dense texts; con- Languages: The literacy assessment was carried out in two languages, Albanian and Serbian. struct syntheses of similar and contrasting ideas or points of view; or evaluate evidenced based arguments. Application and evaluation of logical and conceptual models of ideas may be required to accomplish tasks. Evaluating reliability of b. Kosovo STEP Employer Survey evidentiary sources and selecting key information is frequently a key requirement. Tasks often require respondents to be aware of subtle, rhetorical cues and to make high-level inferences or use specialized background knowledge. 88 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 89 Target population: Establishments, or workplaces, defined as distinct physical locations at which of potential replacement firms. Given the smaller number of firms in the strata with 51+ em- employers undertake economic activity. These are not an office, department, building or assem- ployees, all firms were selected for this first sampling phase: 154 firms for Prishtina, and 165 bly line within a larger, geographically contiguous ensemble. firms for the Other geographic domain. A separate frame of 39 Serb firms in Northern Kosovo was combined with this large sample of firms for the first phase. Sampling design: The sampling frame is based on a database of firms compiled by the Kosovo Agency of Statistics (ASK), and a complementary frame of 39 Serb firms in Northern Kosovo that At the second stage the target number of firms/branches for each stratum under 51 employees was obtained independently. was selected from the combined first phase frame using random systematic sampling with equal probabilities within the stratum. The firms in the frame for each stratum were sorted in The sample was stratified by two geographic domains (Prishtina and remaining regions of Koso- the following order: municipality, Activity ID and number of employees. In the case of the stratum vo) and four size strata in terms of the number of employees: 5-9, 10-15, 16-50 and 51+ employ- of 51+ employees in the Prishtina and Other geographic domains, the subsample of firms were ees. The target sample size for the selection of firms by stratum is shown in Annex Table 1 below: selected systematically with probability proportional to size (PPS), where the measure of size was based on the number of employees. The largest firms were selected with a probability of 1 At the first stage, the ASK selected a larger sample of firms from their frame to select a reserve in the second stage, and some of the firms with more the 1600 employees were allocated 2 to 4 sample branches depending on their size. ANNEX TABLE 1: Target sample size Majority of smaller firms had only one workplace, so it was not necessary to select a branch. In the case of firms with more than one branch, a branch was selected randomly. There were only Geographic domain 6 self-representing firms with an allocation of 2 or more branches; one branch is selected in each Prishtina of the remaining 484 sample firms. Fieldwork: November-December 2015 60 60 60 60 240 Final sample: 500 workplaces (484 firms and 16 branches) EMPLOYEES EMPLOYEES EMPLOYEES EMPLOYEES (5-9) (10-15) (16-50) (51+) TOTAL Weighting: The data weighting was undertaken by the STEP survey methodologist. Based on the 500 participating establishments, the sum of the final weights is 6,115. Other Languages: The base questionnaire provided by the WB team was adapted according to cir- cumstances in Kosovo and translated into Albanian and Serbian, while taking into account the 80 60 60 60 260 linguistic peculiarities and mentality in the country. EMPLOYEES EMPLOYEES EMPLOYEES EMPLOYEES (5-9) (10-15) (16-50) (51+) TOTAL Total 140 120 120 120 500 EMPLOYEES EMPLOYEES EMPLOYEES EMPLOYEES (5-9) (10-15) (16-50) (51+) TOTAL 90 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 91 ANNEX 5: ANNEX FIGURE 2: STEP Employer Survey: Distribution of firms by key characteristics Firm characteristics Descriptive Statistics for STEP 79 96 361 19.3 21 22 457 229 19.3 Employer and Household Surveys 235 243 266 186 155 135 ANNEX FIGURE 1: STEP Employer Survey: Distribution of firms by sector Size (employees) Age (years) Innovative? Ownership Int business contacts? SOURCE: STEP Employer Survey. 94 AGRICULTURE AND INDUSTRY ANNEX FIGURE 3: STEP Household Survey: Distribution of population by key characteristics OTHER SERVICE 131 ACTIVITIES Distribution of Population in Household Survey Sample 79 59 14 17 2.94 Other 54.91 TRADE, TRANSPORT, ACCOM- 35-54 Tertiary MODATION, ENTERTAINMENT 4.56 Serbian 92.47 Female 34 2 Post sec Inactive 68 BUSINESS SERVICES 35-54 49 voc Upper sec PUBLIC SERVICES 25 41 20 25-34 Albanian 14.32 Un- employed 33 33 30.73 CONSTRUCTION 32 Male Lower Sec 15-24 or less Employed SOURCE: STEP Employer Survey. 149 Gender Age Education Ethnicy LM status SOURCE: STEP Household Survey. Note: Bias towards young, female, and unemployed compared to overall urban population. 92 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 93 STEP Household Survey: Occupation of the currently employed ANNEX TABLE 2: (aged 15-64, urban) by gender (%) 2016-2017 ANNEX 6:   Male Female Total Higher skill occupations (Type A) Legislators, senior officials and managers 9.8 5.8 8.2 Professionals Technicians and associated professionals 18.6 6.9 37.2 9.9 25.8 8.1 Determinants of employment and earnings Medium-lower skill occupations (Type B) Probability of employment ANNEX TABLE 3: a. Service (vs. non-employment, i.e. unemployed and inactive) Clerks 4.1 5.0 4.4 Service workers and shop and market sales workers 30.0 29.7 29.9 (1) (2) (3) (4) (5) (6) (7) Education and skills (z-scores relative to total population) b. Technical/Physical Adj. years of edu 0.133*** 0.112*** 0.088*** 0.104*** 0.086*** 0.102*** 0.103*** Skilled agricultural and fishery workers 1.2 0.0 0.7 Craft and related trade workers 15.8 6.1 12.0 Socio-emotional skills Plant and machine operators and assemblers 7.1 0.2 4.4 Openness 0.092 0.087 0.081 0.092 0.086 0.088 Elementary occupations 6.7 6.0 6.4 Conscientiousness 0.202*** 0.197*** 0.231*** 0.199*** 0.203*** 0.201*** SOURCE: STEP Household Survey. Extraversion 0.023 0.021 -0.045 0.037 0.034 0.030 Agreeableness -0.152*** -0.138** -0.114** -0.142** -0.136** -0.134** Emotional 0.129*** 0.106** 0.073** 0.105** 0.121*** 0.122*** Stability Grit 0.090 0.073 0.087 0.072 0.081 0.081 Decision making 0.004 -0.024 -0.054 -0.036 -0.006 -0.004 Cognitive skills Reading outside 0.227*** 0.175*** 0.220***   Numeracy -0.010 0.042 -0.013   outside Computer 0.055 0.083** 0.058 0.108** 0.111** outside Literacy 1 0.024   Literacy 2 -0.048 -0.010   Literacy 3 -0.052 94 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 95 Individual characteristics Cognitive skills   Age 0.153*** 0.145*** 0.149*** 0.164*** 0.154*** 0.150*** 0.151*** Reading outside -0.056 -0.017 -0.065   Age squared -0.002*** -0.002*** -0.002*** -0.002*** -0.002*** -0.002*** -0.002*** Female -0.459*** -0.490*** -0.523*** -0.354*** -0.525*** -0.488*** -0.487*** Numeracy outside 0.100 0.062 0.097   Married 0.200 0.233* 0.291** 0.286** 0.279** 0.233* 0.234* Female*married -0.442*** -0.427** -0.420** -0.469*** -0.418** -0.421** -0.422** Computer outside -0.064 -0.069 -0.065 -0.076 -0.075 Serbian 0.521*** 0.502*** 0.468*** 0.438*** 0.470*** 0.474*** 0.491*** (ethnicity) Literacy 1 0.014   Other (ethnicity) 0.034 0.132 0.168 -0.130 0.197 0.152 0.146 Mother’s edu sec- Literacy 2 0.167 0.150   0.283*** 0.268*** 0.234*** 0.193*** 0.242*** 0.266*** 0.266*** ondary Mother’s edu Literacy 3 0.020 0.522*** 0.524*** 0.452** 0.353** 0.460*** 0.502*** 0.497*** higher Constant -5.024*** -4.654*** -4.476*** -5.043*** -4.506*** -4.659*** -4.664*** Individual characteristics Observations 2,599 2,588 2,520 2,299 2,416 2,482 2,482 Age -0.099*** -0.089*** -0.086*** -0.100*** -0.095*** -0.098*** -0.096*** F-test 42.35 27.36 23.34 21.44 23.68 23.76 Age squared 0.001** 0.001* 0.001* 0.001** 0.001** 0.001** 0.001** Prob > F 0 0 0   0 0 0 Female 0.412** 0.472*** 0.464** 0.203 0.449** 0.456** 0.453** SOURCE: Estimates based on STEP Household Survey. Notes: All models are estimated with the “svy” command. Z-scores relative to the total Married 0.000 -0.045 -0.110 -0.157 -0.094 -0.038 -0.043 population mean are used for skills. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Omitted categories: Albanian; Mother’s education, primary and less (ISCED 0-1). Literacy 1: ETS literacy score (advanced); Literacy 2: ETS core pass dummy (i.e. levels 0-2 vs 3-8); Literacy 3: ETS core high dummy (i.e. level 8 vs other levels). Number of children, an interaction variable for married and children, as well as socio-economic status were included Female*married -0.185 -0.275 -0.268 -0.088 -0.238 -0.249 -0.244 in regressions 3 and 4 but not significant. Serbian (ethnicity) -0.543*** -0.535*** -0.560*** -0.512*** -0.578*** -0.531*** -0.524*** Other (ethnicity) 0.016 -0.038 -0.053 0.460* -0.120 -0.103 -0.140 ANNEX TABLE 4: Probability of unemployment (vs employment). Mother’s edu -0.154 -0.102 -0.085 -0.130 -0.087 -0.095 -0.096 secondary VARIABLES (1) (2) (3) (4) (5) (6) (7) Mother’s edu higher -0.387* -0.400* -0.371* -0.300 -0.371* -0.380* -0.386* Education and skills (z-scores relative to total population) Constant 3.307*** 2.848*** 2.808*** 3.348*** 2.810*** 2.824*** 2.894*** Adj.years of edu -0.104*** -0.083*** -0.078*** -0.089*** -0.075*** -0.077*** -0.075*** Observations 1,372 1,368 1,303 1,209 1,278 1,341 1,341 Socio-emotional skills  F-test 12.86 8.573 7.493 7.119 7.924 7.777 Openness -0.103 -0.108 -0.115 -0.124 -0.112 -0.107 Prob > F 0 0 0   0 0 0 Conscientiousness -0.243*** -0.235*** -0.225*** -0.236*** -0.244*** -0.246*** Extraversion -0.045 -0.037 0.110 -0.058 -0.056 -0.056 SOURCE: Estimates based on STEP Household Survey. Notes: All models are estimated with the “svy” command. Z-scores relative to the total population mean are used for skills. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Omitted categories: Albanian; Mother’s education, primary and less (ISCED 0-1). Literacy 1: ETS literacy score (advanced); Literacy 2: ETS core pass dummy (i.e. levels 0-2 vs 3-8); Literacy 3: ETS core high Agreeableness 0.050 0.046 0.004 0.056 0.044 0.039 dummy (i.e. level 8 vs other levels). Emotional Stability -0.080 -0.076 -0.048 -0.072 -0.073 -0.070 Grit -0.140 -0.130 -0.122 -0.125 -0.127 -0.125 Decision making 0.159* 0.156* 0.128* 0.177** 0.174** 0.172** 96 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 97 ANNEX TABLE 5: Determinants of Earnings (OLS) (1) (2) (3) (4) (5) (6) (7) Female*married -0.069 -0.086 -0.062 -0.099 -0.090 -0.105 -0.097 Education and skills (z-scores relative to employed population) Serbian (ethnicity) -0.025 -0.034 -0.124* -0.120* -0.095 -0.144* -0.123 Adj. years of edu 0.061*** 0.060*** 0.031*** 0.022* 0.021* Other (ethnicity) 0.040 0.066 0.105 0.149 0.150 0.142 0.143 Non-cognitive skills Mother’s edu, sec- -0.001 -0.016 0.046 0.030 0.033 0.033 0.036 ondary Openness 0.049 0.052 0.044 0.045 0.045 0.046 Mother’s edu, higher 0.016 -0.035 -0.035 -0.066 -0.071 -0.065 -0.069 Conscientiousness -0.013 -0.011 -0.017 -0.010 -0.017 -0.010 Job characteristics Extraversion -0.000 -0.003 0.029 0.029 0.029 0.029 Private formal -0.252*** -0.250*** -0.255*** -0.251*** -0.256*** Agreeableness -0.082** -0.093** -0.133*** -0.135*** -0.134*** -0.137*** Private informal -0.192** -0.189* -0.205** -0.190* -0.206** Emotional Stability 0.028 0.025 0.019 0.016 0.018 0.015 Permanent job (dum- -0.028 -0.021 -0.021 -0.020 -0.020 my) Grit 0.023 0.032 0.046 0.046 0.046 0.046 Sector Decision making -0.051 -0.046 -0.021 -0.019 -0.022 -0.020 Construction 0.143 0.135 0.134 0.138 0.137 Use of information-processing skills at work Trade, Accomm -0.097 -0.106 -0.106 -0.105 -0.104 Reading at work 0.124*** 0.130*** 0.014 0.011 0.014 0.011 Transport & ICT 0.251** 0.238* 0.214* 0.244** 0.222* Numeracy at work 0.003 0.006 0.049** 0.051** 0.048* 0.049** Finance/Real estate 0.029 0.070 0.059 0.074 0.063 Computer at work -0.042 -0.043 0.015 0.020 0.014 0.020 Public services 0.154* 0.143* 0.137* 0.144* 0.138* Generic skills at work Other services 0.121 0.121 0.105 0.126 0.111 Interpersonal skills -0.008 -0.009 0.003 0.005 0.001 0.004 Occupation Problem solving 0.078** 0.075** 0.040 0.042 0.040 0.042 Professionals 0.029 0.093 0.093 0.092 0.092 Learning 0.024 0.028 0.003 0.011 0.004 0.012 Techn & ass. prof -0.320** -0.229* -0.207 -0.233* -0.211 Autonomy 0.110*** 0.115*** 0.070** 0.072** 0.067** 0.069** Clerical support -0.169 -0.118 -0.101 -0.121 -0.105 ETS literacy score 0.004 0.021 0.024 Service and sales -0.540*** -0.397*** -0.383*** -0.396*** -0.380*** Individual characteristics Crafts -0.366** -0.312** -0.317** -0.307** -0.312** Age 0.019 0.017 0.009 0.006 0.006 0.006 0.005 Operators -0.210 -0.084 -0.068 -0.085 -0.068 Age squared -0.000 -0.000 -0.000 -0.000 -0.000 -0.000 0.000 Elementary -0.517*** -0.357** -0.356** -0.358*** -0.357** Female 0.173 0.179 0.047 0.081 0.065 0.085 0.070 Married 0.170 0.163 0.123 0.164* 0.163* 0.169* 0.168** 98 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 99 Annex Table 5 cont. (1) (2) (3) (4) (5) (6) (7) ANNEX 7: Region Gjakova -0.187** -0.244*** -0.257*** -0.256*** -0.269*** Gjilani 0.130 0.154 0.156 0.152 0.154 Mitrovica -0.069 -0.057 -0.063 -0.055 -0.060 Determinants of Skills Prizren 0.063 0.072 0.068 0.077 0.075 Peja -0.154** -0.196*** -0.195*** -0.196*** -0.195*** ANNEX TABLE 6: Cognitive skills used (reading, numeracy, computer use): ordered probit with svy Ferizaj -0.101 -0.075 -0.077 -0.066 -0.065 VARIABLES Reading overall Numeracy overall Use of computer overall Constant -0.714** -0.633* 0.454 0.516 0.808** 0.534 0.820** Female 0.239*** -0.117** -0.281*** Age -0.108*** -0.090*** -0.022*   Age squared 0.001*** 0.001*** 0.000 Observations 577 543 564 542 542 542 542 Adjusted years of education 0.148*** 0.077*** 0.135*** ECD 0.146* 0.310*** 0.318*** R-squared 0.366 0.359 0.552 0.580 0.580 0.581 0.581 Mother’s edu secondary 0.236*** 0.094 0.091 Mother’s edu higher 0.420*** 0.221* 0.340** SOURCE: Estimates based on STEP Household Survey. Notes: The dependent variable is the log of hourly wages (trimmed at 5th and 95th percentiles). Socio economic status at age 15 All models are estimated with the “svy” command. Z-scores relative to the employed population mean are used for skills. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Omitted categories: Lower secondary education or less; Mother’s education, primary and less (ISCED 0-1); Middle 0.125 0.057 -0.059 Albanians; Managers; Industry; Public (both formal and informal), Prishtina. Occupational groups “Armed forces” and “Skilled agricultural workers” (2 obs) are excluded. Agriculture, forestry and fishing (1 obs) is excluded. High 0.119 0.267*** 0.052 Asset wealth index quintile second 0.014 0.037 0.432*** Third 0.135* 0.173** 0.436*** forth 0.022 0.139* 0.441*** highest 0.208** 0.256*** 0.768*** Serbian 0.513*** -0.097 0.739*** Other ethnicity -0.324* -0.243 0.399** Gjakova 0.652*** 0.650*** 1.195*** Gjilani -0.204** -0.180* 0.105 Mitrovica 0.343*** 0.337*** -0.009 Prizren -0.037 -0.153 0.845*** Peja 0.114 0.011 0.462*** Ferizaj 0.143 -0.260*** -0.112 Constant cut1 -0.489** -1.868*** 1.347*** Constant cut2 0.331 -0.944*** 1.606*** Constant cut3 0.860*** 0.883*** 2.003*** Observations 3,019 3,018 3,019 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 100 Kosovo: findings from the STEP survey Kosovo: findings from the STEP survey 101 ANNEX TABLE 7: Socio-emotional skills: OLS with svy ANNEX TABLE 8: Literacy levels Conscien- Agreeable- Emotional Decision Coefficients Coefficients VARIABLES Openness tiousness Extraversion ness stability Grit making Marginal effects (ordered probit) (ordered probit) (OLS) Female 0.030 0.048** 0.014 0.094*** -0.010 0.077*** 0.105*** Levels from VARIABLES Levels 0/1 Level 2 Level 3 Level 4/5 Score 0/1 to 4/5 Age -0.015*** -0.001 0.005 0.002 -0.006 -0.001 0.002 Female -0.012 0.005 -0.003 -0.002 0.000 -1.680 Age squared 0.000* 0.000 -0.000 0.000 0.000 0.000 -0.000 Age -0.031** 0.012** -0.007** -0.005** 0.000 -0.715 Adjusted years of education 0.036*** 0.027*** 0.026*** 0.022*** 0.029*** 0.034*** 0.028*** Age squared 0.000 0.000 0.000 0.000 0.000 0.004 ECD 0.041 -0.003 -0.039 -0.011 0.048 0.022 0.046 Adjusted years of 0.091*** -0.035** 0.021** 0.013** 0.000 4.060*** education Mothers education ECD 0.143 -0.054 0.033 0.021 0.000 3.842 Secondary 0.084*** 0.041 0.043 0.040 0.015 0.074** 0.076* Mothers education Higher 0.109** 0.015 0.002 0.069 0.099** 0.103* 0.154** Secondary 0.045 -0.017 0.011 0.006 0.000 5.772* Socio economic status at age 15 Tertiary -0.073 0.028 -0.018 -0.010 0.000 1.492 Middle 0.032 0.048* -0.018 0.010 -0.022 -0.001 0.021 Socio-economic status at age 15 High 0.069* 0.059* 0.030 0.011 0.048 0.021 0.019 Middle 0.032 -0.012 0.008 0.004 0.000 2.230 Asset wealth index quintile High 0.126 -0.048 0.029 0.019 0.000 5.166 second 0.067* 0.005 -0.023 -0.006 0.066* 0.031 -0.020 Asset wealth index quintile third 0.071* -0.014 0.003 -0.028 0.065* 0.020 0.010 second 0.183 -0.071 0.047 0.024 0.000 7.723** forth 0.096** 0.015 0.046 -0.002 0.089** 0.070* 0.033 third 0.172 -0.067 0.045 0.022 0.000 8.405** highest 0.153*** 0.120*** 0.211*** 0.125*** 0.073* 0.175*** 0.157*** forth 0.117 -0.046 0.031 0.014 0.000 6.375* Serbian 0.167*** 0.134*** 0.152*** 0.035 -0.015 0.142*** 0.112** highest 0.421** -0.157** 0.092** 0.065** 0.000 15.800*** Other ethnicity -0.063 -0.094 -0.101 -0.026 -0.030 -0.264*** -0.139 Serbian 1.684*** -0.377** -0.129** 0.479** 0.027 48.466*** Gjakova 0.056 0.000 -0.069 -0.256*** -0.005 0.174*** 0.095* Other ethnicity 0.009 -0.003 -0.001 0.003 0.000 -13.802 Gjilani 0.083** -0.139*** -0.037 -0.081* -0.076* 0.213*** 0.090* Gjakova 0.571*** -0.216** 0.132** 0.084** 0.001 21.726*** Mitrovica 0.020 -0.199*** -0.047 -0.070** 0.085** -0.100*** -0.007 Gjilani -0.261* 0.099 -0.060 -0.038 0.000 -1.353 Prizren 0.015 0.009 -0.136*** -0.174*** -0.013 0.048 -0.109** Mitrovica -0.277** 0.105** -0.064** -0.040** 0.000 -6.282* Peja 0.104*** 0.076** 0.061 0.013 -0.046 0.134*** 0.019 Prizren -0.259** 0.098** -0.060** -0.038** 0.000 -8.586** Ferizaj 0.179*** 0.160*** 0.195*** 0.154*** -0.112** 0.367*** 0.318*** Peja -0.221** 0.083** -0.051** -0.032** 0.000 -5.161 Constant 3.214*** 3.301*** 3.140*** 3.312*** 2.855*** 3.344*** 3.376*** Ferizaj -0.534*** 0.202** -0.123** -0.078** 0.000 -18.002*** Observations 3,011 3,011 3,011 3,011 3,011 3,009 3,007 Observations 3016 3016 3016 3016 3016 3019 R-squared 0.157 0.113 0.087 0.093 0.074 0.113 0.085 Standard errors in parentheses Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 *** p<0.01, ** p<0.05, * p<0.1