HARNESSING H U M A N C A P I TA L FOR GROWTH I N C R O AT I A Unleashing Potential for Economic Takeoff Amid Demographic and Technological Change Harnessing Human Capital for Growth in Croatia Acknowledgments © 2023 International Bank for This report has been prepared by a team led by Nithin Umapathi and Reconstruction and Development / consisting of Malathi Velamuri (Consultant), Clement Joubert (World The World Bank Bank), Lea-Karla Matić (EIZ), Tibor Kovac (EIZ), Vedran Recher (EIZ), Maria Ustinova (World Bank), Nicholas Gailey (Consultant), Boryana Gotcheva 1818 H Street NW, (Consultant), Daniel Garrote Sanchez (World Bank), Lucia Brajkovic (World Washington DC 20433 Bank), Zoran Anusic (World Bank), and Predrag Bejakovic (Consultant). Telephone: 202-473-1000 It has been prepared under the oversight and management of Gallina Internet: www.worldbank.org Andronova Vincelette (Country Director for the European Union), Fadia Saadah (Regional Director, Human Development), Jehan Arulpragasam (Country Manager for Croatia), and Cem Mete (Practice Manager, Social Rights and Permissions Protection and Jobs). Fiona Mackintosh edited the report, and Elizaveta The material in this work is subject Tarasova provided the graphic design. Ruzica Jugovic provided admin- to copyright. Because The World istrative support. The team would like to acknowledge people who have Bank encourages dissemination of provided important input and suggestions during the preparation of its knowledge, this work may be the report. The team thanks the peer reviewers Teo Matkovic, Pablo A. reproduced, in whole or in part, for Acosta, Ciro Avitabile and Josip Funda for their invaluable feedback. noncommercial purposes as long as Rafael De Hoyos Navarro offered instrumental guidance as program full attribution to this work is given. leader, Juul Pinxten made important contributions to the social protec- tion chapter, Mauro Testaverde gave important suggestions on migra- Any queries on rights and licenses, tion, and Iva Tomic shared valuable insights on the Croatian labor market, including subsidiary rights, should be which was greatly appreciated. Anna Koziel shared recent work on Cro- addressed to World Bank Publications, atian long-term care, and Mirey Ovadia shared her experience on ALMP. The World Bank Group, 1818 H Street Ana Simundza gave valuable all-around assistance and insights. Nga Thi NW, Washington, DC 20433, USA; Viet Nguyen and Diego Ambasz provided helpful comments. fax: 202-522-2625; e-mail: pubrights@worldbank.org. This work is a product of the staff of the World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments that they represent. The World Bank does not guarantee the accuracy of the data included in this work. ii Harnessing Human Capital for Growth in Croatia List of acronyms ALMMs Active labor market measures ALMP Active labor market policies The Quality of Life Scale assessing quality of life in individuals in early old CASP-19 age. CEE Central and Eastern Europe CES Croatian Employment Service CROQF Croatian Qualifications Framework CSW Case worker CVT Continuing vocational training DESA Department of Economic and Social Affairs (UN) DESI Digital Economy and Society Index DIOC Database on Immigrants in OECD and Non-OECD Countries ECA European and Central Asia ECEC Early childhood education and care ER Employment rate ESI European Skills Index EU European Union EU-27 Group of 27 EU member states GDP Gross domestic product GFC Global financial crisis GMB Guaranteed minimum benefit GTS Graduate Tracking System ICT Information and communication technologies LFP Labor force participation LFPR Labor force participation rate LFS Labor force survey LTC Long-term care NAV Norwegian Labour and Welfare Administration NEET Not in education, employment, or training NMS New member states NRRP National Recovery and Resilience Plan OECD Organisation for Economic Co-operation and Development PIAAC International Assessment of Adult Competencies PISA Programme for International Student Assessment PTR Participation Tax Rate SA Social assistance SHARE Survey of Health, Aging, and Retirement in Europe SILC Survey on Income and Living Conditions STEAM Science, technology, engineering, arts, and mathematics STEM Science, technology, engineering, and mathematics TALIS Teaching and Learning International Survey TFP Total factor productivity TIMSS Trends in Mathematics and Science Study TVET Technical and Vocational Education and Training UB Unemployment benefit VET Vocational Education and Training WBL Work-based learning YEI Youth Employment Initiative iii Harnessing Human Capital for Growth in Croatia Table of Contents Acknowledgments ii List of acronyms iii Introduction 1 1. Croatian Labor Market: Trends, Features, and Challenges 4 Introduction 4 Skills and Spatial Gaps 9 Balancing work with care responsibilities – A lack of flexible work options 14 Barriers to Work: Institutional Features Dampening Labor Demand and Supply 18 The Impact of Non-Labor Income: The “Airbnb” Effect and Remittances 20 2. Aging and Croatia's Large Employment Gap for Older Workers 22 Introduction 22 Examining Croatia’s Employment Gap for Older Workers 23 Are Retirement Patterns Different in Croatia? 27 The Role of the Pension System 28 Understanding Other Causes of the Old Age Employment Gap 34 Policy Implications 40 3. Education and Skills Needed to Prepare for the Future Job Market 42 Introduction 42 Foundational Skills in Early Childhood: Access and Importance 43 Elementary and Secondary Schools: Uneven Skill Development 46 Navigating Skill Acquisition in Secondary Schools: Current Challenges and Policies 48 Adolescence and Youth: Challenges and Opportunities for Vocational Education and Training 50 Higher Education: The Problem of Skill Mismatches Continues 52 Adult Training: Skills Development Throughout a Working Life 54 Policy Implications 56 4. Migration and Human Capital 58 Introduction 58 Migration Effects in Croatia 59 Policy Implications 61 5. Labor Activation Policies Can Strengthen Participation and Human Capital 63 Introduction 63 The Croatia’s Active Labor Market Measures 63 Characteristics of and Challenges for the Current System 65 iv Harnessing Human Capital for Growth in Croatia Lessons From Current and Past ALMM Programs 69 Policy Implications 71 6. Social Protection and Human Capital 74 Introduction 74 Enhancing the Effectiveness of Social Assistance in Reducing Poverty 76 Ensuring Proper Design of Activation and Graduation Components in Safety Nets 77 Navigating the Low-Income Support (GMB) Program: Activation and Graduation 80 Policy Implications 81 References 83 List of Figures: Figure 1: Croatia’s “business-as-usual” vs labor market and 1 Figure 17: Reasons offered for current unemployment 15 human capital reform scenarios. Potential GDP status, men and women per capita growth and growth decomposition, 2020-2050 Figure 18: Main reason reported by the part-time employed 16 Figure 2: Report’s Main Policy Focus Areas for Enhancing 2 Figure 19: Average Total Early-stage Entrepreneurial 17 Labor Productivity and Participation in Croatia Activity (TEA) rates by gender, 2021 Figure 3: Employment rates and Labor force participation 4 Figure 20: Prevalence of flexible work schedules (left) and 17 rates of Croatia and other countries in the region work-from-home option (right) (2010-2022) Figure 21: Employed persons working from home as a 17 Figure 4: Employment rates by age group, for men and 5 percentage of the total employment, 2022 (%) women, Croatia and EU-27, 2008-2022 Figure 22: Compensation of employees paid by general 18 Figure 5: LFPR by educational attainment, men and 6 government: Croatia and EU women Labor market status of Croatian men and women, 2022 Figure 23: Real GDP and nominal average gross wage 19 growth by sector Figure 6: Unemployment rates by age group, men and 7 women Figure 24: Share of temporary to total employment (left, 19 20-64) and share of involuntary temporary Figure 7: Youth unemployment rate (left) and NEET rate 7 employment to total employment (right, 15-64), (right), % and by age % Figure 8: Youth unemployment rate (left) and NEET rate 8 Figure 25: Average marginal effects of various control 20 (right), % and by age variables on labor force participation Figure 9: NEET rates among 15–24-year-olds 8 Figure 26: Personal remittances received (1993-2021) wage 21 growth by sector Figure 10: Share of young adults aged 25-34 living with 9 their parents, by sex (%) (EU-SILC) Figure 27: Projected elderly dependency ratio 2020-2050, 22 selected European countries Figure 11: Trends in the estimated average age of young 9 Croatians leaving the parental household, by sex Figure 28: Croatia’s employment gap 2010-2022 22 Figure 12: Trends in LFPRs in Croatia and EU-27 by 10 Figure 29: The employment gap vs. the EU-27 average and 23 educational attainment selected European countries, 2010-2021, by age groups Figure 13: Online vacancy index, (top), Online vacancy index 11 by skill-level (bottom), seasonally adjusted Figure 30: Employment rates of older workers by age groups 24 (2018-2021) Figure 14: DESI index 2022 – Relative performance by 13 dimension Figure 31: Employment rates by gender and age groups 24 (1995-2021) Figure 15: Principal reasons offered by respondents for 14 inactivity (percentage)* Figure 32: Employment rates by age (50-80) in Croatia 25 Figure 16: Employment rates by age of youngest child in 15 Figure 33: Weekly hours worked (top) and part-time 25 household employment rates (bottom) by age group v Harnessing Human Capital for Growth in Croatia Figure 34: Shares of male and female older workers 26 Figure 57: Participation rate in education and training 54 employed in different occupations, by age group (last four weeks), people from 25 to 64 years, percentage Figure 35: Expenditures on pensions, by country 2020 31 Figure 58: Participation rate in education and training 55 Figure 36: Distribution of retirement ages in Croatia, by 31 (previous four weeks) in 2021 by age, percentage gender Figure 59: Enterprises providing CVT training - % of all 55 Figure 37: Age profiles of employment rate (%) of men and 33 enterprises women, by birth cohort Figure 60: Types of skills studies through CVT course, per 55 Figure 38: Share of individuals with self-reported poor 35 type of enterprises health, by age group and country Figure 61: Destination countries for Croatian migrants (A: 58 Figure 39: Share of individuals with at least 1 chronic disease 35 left) and Emigration by age group and by time (B: (left) and share of individuals having difficulties right) with at least 1 activity of daily living (right), by age group and country Figure 62: Education and migration status among working 59 age population (25-64) Figure 40: Quality of life in older age, by age group and 36 country Figure 63: Trend in the stock of doctors in Croatia 60 Figure 41: Fraction of 50-65 y.o. not working and fraction 36 Figure 64: Share of emigrants in 2010 who returned by 2014 60 of healthy 50-65 y.ol not working in Croatia vs. by education and age groups comparators Figure 65: Categories of ALMMs being implemented in the 63 Figure 42: Fraction of individuals over age 50 providing 37 Republic of Croatia, 2023 informal care, by country Figure 66: Registered unemployed, ALMM users and new 64 Figure 43: Providing daily help by individuals 50-65, by 37 entrants, 2017-2022 country and employment status Figure 67: New entrants in ALMMs as share of registered 64 Figure 44: Providing daily help in Croatia by employed and 37 unemployed, 2017-2022 (%) not employed individuals aged 50-65, by type of help (1) and recipient of help (2) Figure 68: Spending on ALMMs, total and by category, 64 2017-2021 Figure 45: Reasons for retiring (top) and reasons for leaving 38 the last job (bottom), by country Figure 69: Spending on ALMMs by type of program, 65 selected European countries, 2020 (% of GDP) Figure 46: Percentage wage gap between young and 39 prime-age workers (top) and between senior and Figure 70: Distribution of spending on ALMMs, 2021 (%) 65 prime-age workers (bottom), by country Figure 71: New entrants by category of ALMMs, 2017-2022 65 Figure 47: Percentage wage gap between Youth and Prime- 40 age workers and between Senior and Prime-age Figure 72: Distribution of entrants in ALMMs by age, 66 workers, by occupation educational background and duration of unemployment spells (2022, %) Figure 48: Key business environments constraints in Croatia, 43 overall and by firm size Figure 73: Targeting performance: share of total 74 beneficiaries across income deciles Figure 49: Participation in ECEC over time – Croatia and EU 44 Figure 74: Spending on SA by type of program as a share of 75 Figure 50: Duration in early childhood education and care for 45 GDP in 2021 Croatia PISA-2022 takers and their achievements Figure 75: Total social assistance spending and share of 75 Figure 51: Croatia’s PISA scores by subject (2006-2022) 47 which are means tested Figure 52: Collaborative problem solving, PISA 2015 47 Figure 76: GMB has the most poverty reduction per budget 76 spent. (Cost-effectiveness of major direct Figure 53: Croatian students spend the fewest hours in 49 transfers in reducing poverty, figures show school in the EU. poverty reduction per HRK million spent) Figure 54: Croatia age profile employment rate, 2021 51 Figure 77: The strictness of activation requirements, overall 78 score, first-tier unemployment benefits, 2020 Figure 55: Educational attainment in tertiary education 53 among people aged 30-34, % of the population Figure 78: Key incentives and disincentives for activation 79 and graduation in the UB design Figure 56: Number of tertiary programs in Croatia per field 53 of study, 2022 Figure 79: Incentives and disincentives for activation and 80 graduation in the GMB design vi Harnessing Human Capital for Growth in Croatia List of Tables: Table 1: Selected demographic characteristics for Croatia 4 – actual (2019) and projected Table 2: Educational attainment levels in Croatia by 6 gender, 2022 Table 3: Educational composition of the population by 10 age group, EU-27 (2020) and Croatia (2022) Table 4: Labor force status in 2018 by cohort and highest 11 degree, Croatia (%) Table 5: Croatia’s key human capital indicators compared 12 to EU-27, 2020 Table 6: County-level registered unemployment rates 12 Table 7: Part-time and self-employment rates (%), Croatia 16 and EU-27, 2021 Table 8: Probit model of employment among 50-65 in 29 Croatia vs. select neighbouring countries (Poland, the Czech Republic, Slovenia and Austria) Table 9: Example of present value of retirement benefits 32 for early and full age pension Table 10: Average effective retirement age of new old-age 32 retirees, Croatia Table 11: TIMSS 4th 2019: Students Attended Pre-primary 44 Education, Croatia Table 12: Regional differences in learning outcomes, PISA 48 2015-2018 Table 13: Averages for age 15 years PISA results: General 50 and Vocational in Croatia Table 14: Cost per beneficiary of ALMMs, 2021 68 List of Boxes: Box 1: Employment at older ages, a framework 27 Box 2: 21st-century skills 42 Box 3: Migration policies in Malaysia designed to 62 enhance domestic human capital Box 4: Key indicators for assessing social assistance 75 performance vii TOC Harnessing Human Capital for Growth in Croatia Introduction Croatia faces two major trends that will significantly If Croatia implements measures to increase its pop- affect its society and economy. The first trend is an ulation’s human capital and economic participation, aging and shrinking population, and the second is it could greatly boost its long-term growth. If Cro- the impact of rapidly changing technology on the atia can achieve substantial increases in labor par- nature and organization of work. These trends pres- ticipation and human capital, it will greatly enhance ent Croatia with both challenges and opportunities its ability to seize the opportunities and mitigate that will shape employment, wages, and economic the risks. As shown in Figure 1 (top), implementing development in the future. Figure 1: Croatia’s “business-as-usual” vs labor Two main challenges are emerging as a result of market and human capital reform scenarios. these trends. The aging of the population asso- Potential GDP per capita growth and growth ciated with rising dependency rates could have decomposition, 2020-2050 a negative impact on Croatia’s economic perfor- Average growth rates by decade, percentage change in mance as there will be fewer working-age individ- GDP per capita uals to support a growing elderly population. The Baseline Moderate reforms package share of people aged 65 and older in the total pop- Ambitious reforms package ulation rose from 8 percent in 1960 to 22 percent in 3.50% 2021, and by 2050, one in every three Croatians will 3.0% be aged over 65. Also, the accelerating growth of 3.00% technology is presenting a challenge to low-skilled 2.50% workers as employment is now increasingly consist- 2.2% ing of knowledge-intensive occupations, with the 2.00% associated risk of job losses and rising inequality as 1.6% workers with low skills find it difficult to retrain and 1.50% adapt. 1.00% On the other hand, the opportunities presented 0.50% by these trends are also significant. Robots and 0% algorithms will increasingly perform routine tasks 2025-2050 within firms' production processes, and as tech- nology evolves, new occupations will emerge. The GDP per capita growth decomposition, Contribution of shift towards digitalization, automation, and green each reform to incremental GDP per capita growth 1.8% energy initiatives could change the processes of value creation and increase productivity. Aging 1.6% 0.3% also could create a thriving market for long-term 1.4% and health care services and facilitate longer and 0.4% Percentage points 1.2% 0.4% healthier working lives, cohesive communities, and 1.0% family-based support. To reap the benefits of tech- 0.8% nological advancements and to tackle economic 0.8% 0.5% 0.7% hurdles brought about by an aging population, Cro- 0.6% atia must focus on both increasing the quantity and 0.4% 0.1% 0.1% 0.1% 0.1% improving the quality of its workforce. According 0.1% 0.3% 0.2% 0.1% to the World Bank's economic projections (World 0.3% 0.1% 0.2% Bank, 2022a), under the business-as-usual baseline 0% 0.1% scenario, Croatia's productivity and per capita GDP 2020 2025 2030 2035 2040 2045 2050 are expected to grow at rates much lower than in Private investment Public investment TFP previous decades, achieving only modest growth of Tertiary education Pre-tertiary education LFP 1.6 percent over the period from 2025 to 2050. Source: World Bank (2022a). 1 TOC Harnessing Human Capital for Growth in Croatia reforms that improve the quality and increase the tively impact some large demographic groups and quantity of the workforce could double Croatia’s disadvantaged groups. Reducing these disparities projected average growth from 1.6 to 3 percent. could have significant impact. For example, despite Figure 1 (bottom) shows that ambitious efforts to Croatia’s shrinking working age population, women, enhance human capital and workforce participa- young people, and older individuals work less than tion could have as significant an impact as reforms their European peers. When it comes to education aimed at increasing total factor productivity (TFP). and skills, primary and secondary school enrollment This would accelerate Croatia’s convergence with rates are high, but half of all underprivileged stu- the average per capita income level for the EU by dents are functionally illiterate. Also, while Croa- several decades. In simpler terms, Croatia must tia has one of the highest rates of participation in find a way to consistently increase productivity and VET programs in the EU, VET graduates face the economic participation if it is to achieve economic prospect of lower wages and high unemployment. parity with wealthier European Union nations in the Furthermore, the country’s scores for knowledge near future. application and advanced skills on international edu- cation assessments at the secondary and VET levels This Human Capital Review focuses on the bar- are below average. Both productivity levels and the riers that are getting in the way of increasing the size of the workforce are shrinking due to Croatia's Croatian labor productivity rather than the prob- high emigration rate, as many working-aged people lems restricting the number and quality of jobs. are leaving the country to seek employment else- Croatia's productivity is far lower than the regional where in Europe. A sizable number of these emi- frontier represented by Germany and is catching grants are highly skilled, which means that not only up only very slowly. The World Bank’s recent Cro- does emigration shrink the available workforce, but atia Country Economic Memorandum (World Bank, it also contributes to skill shortages in the domestic 2022a) highlighted the importance of the country’s labor market. On the other hand, migration can be a low within-sector productivity. Even if Croatia had way to attract skilled labor to Croatia from overseas the same economic sector composition as the more with the right policies in place. In the area of public productive Germany, it would still be 57 percent employment services, there is scope to extend less productive. Croatian firms are less productive active labor market policies (ALMPs) such as job because of factors such as market inefficiencies matching and retraining measures to upskill more and entry and exit barriers, but above all it will be of the labor force. Figure 2 summarizes the reports essential to increase the size of the workforce and to policy focus areas for strengthening Croatian pro- equip it with the skills needed to support the devel- ductivity and economic participation. opment of higher productivity sectors. At the same time, in the current labor market, there The specific policies for enhancing the workforce is a high risk that this increased human capital that are explored in the report were carefully will be underused. Therefore, it is crucial to enable chosen for their potential to generate and utilize people to remain economically active by enacting human capital along the life-cycle and to make a various accompanying measures such as increas- significant contribution to enhancing individu- ing the availability of high-quality childcare, ensur- als' labor market participation. Given that Croatia ing that pension and social assistance systems are faces major labor supply obstacles, the report offers free of disincentives to work, facilitating the tran- insights into the barriers to labor supply and old age sition from school to work, providing flexible work employment, to skills acquisition, and to migration arrangements, and providing appealing opportu- as a form of human capital reallocation. In addi- nities for adult training and effective employment tion, it discusses two key instruments – active labor support services. market and social protection policies – that have the potential to shape economic participation. The The Government of Croatia has established an report also discusses the country’s unequal edu- ambitious plan of reforms to tap into the underused cation and skills acquisition outcomes that nega- potential of the labor force. There are many rea- Figure 2: Report’s Main Policy Focus Areas for Enhancing Labor Productivity and Participation in Croatia Activation through Skills and Human Migration Higher Economic Reducing Labor public services Capital Participation Market Barriers Increase Expand access to Encourage "brain Incentives for later Labor de-regulation experimentation in early childcare and circulation" retirement Promote more ALMP education 2 Strengthen programs Enable workers to flexible forms of work Strengthen the Increase the scope of for the diaspora to stay productive Strengthen non- integration with apprenticeships return and invest longer traditional work social services Strengthen career Strengthen Expand public arrangements Strengthen guidance immigration childcare TOC Harnessing Human Capital for Growth in Croatia sons to feel positive about Croatia’s economic accomplishments over the last few years. Croatia has overcome the economic turbulence stemming from the COVID-19 pandemic and the energy crisis precipitated by the war in Ukraine. The country also joined the Eurozone and the Schengen areas and has made considerable progress on income conver- gence with the EU. Looking forward, the govern- ment has produced a National Recovery and Resil- ience Plan (NRRP) that sets out an ambitious reform agenda for achieving higher economic growth and that contains many of the policies highlighted in this report. However, it is important to note that these plans have been created at a general level, which means that carrying out these reforms will need to be guided by detailed analysis and policy experi- mentation. This report aims to provide some of that detailed analysis by identifying the barriers that exist to expanding employment and skills acquisi- tion. The report is structured as follows. Chapter 1 dis- cusses the labor market constraints facing dif- ferent segments of the working-age population, which can inform how to make work attractive. Chapter 2 explores the factors that are discourag- ing older people from re-entering the labor market and the need to optimize pension design to pro- mote healthier and longer working lives. Chapter 3 explores ways to enhance human capital formation through the lifecycle, starting with early childcare and education, followed by upper secondary school, secondary VET, tertiary education, TVET and adult training. Chapter 4 identifies the enabling condi- tions needed to ensure that net migration can have a positive impact on the growth of economic partic- ipation and productivity. Chapter 5 focuses on ways to leverage public-private partnerships to provide training and activation services and on how ALMPs could best be used to increase labor market par- ticipation. Finally, Chapter 6 explores the role that social protection can play in stimulating participa- tion in the labor market. 3 TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Introduction Table 1: Selected demographic characteristics for Croatia – actual (2019) and projected This chapter discusses the current state of the Cro- 2019 2030 2040 2050 2060 2070 atian labor force and identifies the obstacles that Fertility rate (%) 1.43 1.48 1.51 1.54 1.56 1.59 working-age individuals often face when seeking Population employment. Table 1 shows some demographic growth, annual -1.0 -0.9 -0.9 -1.0 -0.8 -0.6 (%) characteristics in 2019 (actual) and estimated values Life expectancy for the next few decades. Like other EU member Males 75.3 77.3 79.3 81.1 82.7 84.3 countries, Croatia is characterized by low fertility Females 81.6 83.2 84.7 86.2 87.5 88.8 rates and increasing life expectancy. These factors, Net emigration along with emigration flows that increased con- as % of the -0.1 0.0 0.0 0.1 0.1 0.2 siderably following the country’s accession to the population EU but have since moderated, resulted in a rapidly Working-age population 2,433 2,165 1,986 1,788 1,643 1,533 shrinking working-age population. The share of (20-64) (in the working-age population is declining dramat- thousands) ically faster in Croatia than in its regional neigh- Old-age dependency 34.8 44.5 50.4 57.2 61.5 64.6 bors. The old-age dependency ratio is projected to ratio 20-64 rise sharply and steadily each decade, growing to Source: The 2021 Ageing Report: Economic and Budgetary 57 percent by 2050 and 65 percent by 2070. This Projections for the EU Member States (2019-2070). situation calls for the optimal use of the country’s Note: Old-age dependency ratio = population aged 65+ as a % of the population aged 20-64. available human capital, to blunt the adverse effects of these demographic trends on its long-term pro- the global financial crisis, but they remain among ductivity and growth. the lowest in the EU (Figure 3). While there appears to have been an increasing trend in both the ER and The average duration of a working life in Croatia is the LFPR in all countries in the region since 2010, less than 33 years compared with the EU average this growth has been slowest in Croatia, result- of 35.4 years and is significantly lower than that of ing in the country falling behind over time (World many new member states. Croatia’s employment Bank, 2019e). Given its rapidly aging population and rate (ER) and labor force participation rate (LFPR) shrinking workforce, the sustainability of the coun- have both been increasing from the lows caused by try’s social security system, which depends on a Figure 3: Employment rates and Labor force participation rates of Croatia and other countries in the region (2010-2022) 80 80 Labour force participation 75 Employment rate, % 75 70 65 70 rate, % 60 65 55 50 60 45 40 55 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 CZ HR HU SI CZ HR HU SI SK ME MK RS SK ME MK RS Note: CZ - Czech Republic, HR - Croatia, HU - Hungary, SI - Slovenia, SK- Slovakia, ME -Montenegro, MK – North Macedonia, RS - Republic of Serbia. Source: Eurostat. This chapter was prepared by Malathi Velamuri, Lea-Karla Matić, Vedran Recher, and Nithin Umapathi. 4 TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia pay-as-you-go arrangement, will become increas- (Figure 4). The rates for men in the 25 to 29 age ingly unviable. group and for women in the 30 to 49 age group have overtaken the corresponding EU-27 rates in Low employment rates are more prevalent among recent years. For those in the older age group (65 to women, young people, and older cohorts. Follow- 74), employment rates in Croatia were above those ing the economic recession that lasted from 2009 of the EU-27 for both genders until 2012 before to 2013, employment rates have been increasing falling sharply below the EU rates. The low employ- gradually for the prime-aged, especially for men ment rates among young people are also notable. Figure 4: Employment rates by age group, for men and women, Croatia and EU-27, 2008-2022 Men Women 30 Age 15-24 20 85 80 Age 25-29 75 70 65 60 Employment rate, % 85 Age 25-49 80 75 70 70 Age 50-64 60 50 40 12.5 10.0 Age 65-74 7.5 5.0 2.5 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 EU-27 MEN WOMEN Source: Eurostat, 2022 5 TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia Table 2: Educational attainment levels in Croatia by gender, 2022 Age group Males 25-39 40-59 60-64 Less than primary, primary, lower secondary education (levels 0-2) 0.06 0.13 0.19 Upper secondary, post-secondary non-tertiary education (levels 3 and 4) 0.69 0.69 0.62 Tertiary education (levels 5-8) 0.25 0.19 0.19 Females Less than primary, primary, lower secondary education (levels 0-2) 0.04 0.15 0.26 Upper secondary, post-secondary non-tertiary education (levels 3 and 4) 0.53 0.60 0.59 Tertiary education (levels 5-8) 0.43 0.25 0.14 Source: Eurostat. Women’s employment rates are below those of men Survey (LFS), Figure 6 depicts the labor market at all ages, with no convergence over the lifecycle. status of Croatian men and women by age groups. A notable feature is the high rate of early retire- Croatian women have an average working life of ment among middle-aged men, largely among war just 30.7 years compared to 34.5 years for men, veterans (World Bank, 2022a). Low activity rates despite having made strong gains in educational among the 55 to 64 age group is a consequence attainment in recent years. Croatia’s gender parity of discouraged workers leaving the labor force, with index for educational enrollment is high, with the share of discouraged workers in this age group women having higher enrollment rates at the pri- having risen from 2.9 percent in 2007 to 5.9 per- mary, secondary, and tertiary levels.1 The distribu- cent in 2013 (Ostrovidov Jakšić & Jakšić, 2019). This tion of educational attainment by age in Table 2 is because incentives for early retirement that are reveals a significant gender gap among the younger embedded in the pension system and have enabled cohort (25 to 39) in favor of women. A considera- this trend. The World Bank has prepared a report on bly higher share of women in this age group has a the adequacy of pensions (World Bank, 2023b) in tertiary education, though men make up a notably support of the government’s implementation of the higher share of those with a secondary education. NRRP. The World Bank’s report proposes lengthen- However, women with less than a tertiary educa- ing the duration of working lives as the main way tion have low LFPRs that are also significantly lower to increase the labor supply in Croatia and thus than the corresponding rates for men (Figure 5). increase the adequacy of pensions in the long run. The success of such efforts would also depend on Early retirement is a salient feature of Croatia’s labor demand increasing sufficiently to absorb older workforce. Using the 2021 national Labor Force workers. Figure 5: LFPR by educational attainment, men and women Men Women 80 80 60 60 % % 40 40 20 20 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 Less than primary, primary and lower secondary education Tertiary education Upper secondary and post-secondary non-tertiary education Source: Eurostat. https://www.ceicdata.com/en/croatia/social-education-statistics/hr-gender-parity-index-gpi-primary-and-secondary-school-enrollment-gross- 1 6 TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia Figure 6: Labor market status of Croatian men and women, 2022 Men Women 1.00 1.00 0.75 0.75 Share of age group Share of age group 0.50 0.50 0.25 0.25 0.00 0.00 15-24 25-34 35-44 45-54 55-64 65+ 15-24 25-34 35-44 45-54 55-64 65+ Employed Other inactive Retired Unemployed Source: LFS 2021. Unemployment has come down sharply, spurred by unemployment not only negatively affect the cur- the economy’s recovery since 2014, and by emigra- rent livelihoods of young people but also lead to tion flows but youth unemployment remains high. “scarring,” whereby these individuals fare systemat- The economic recovery following the long recession ically worse in the labor markets as adults in terms of 2009 to 2013 increased aggregate demand and of lower pay and higher probabilities of unemploy- brought down unemployment markedly (Figure 7). ment. A favorable external environment, along with the positive effects of EU accession in 2013, helped Croatia’s NEET rates are high, indicating that jobs to reignite growth in the 2015 to 2019 period and and opportunities are scarce for young people. to reduce unemployment. Another reason for the Figure 8 compares youth unemployment rates and drop in unemployment was the increase in emigra- the rates of young people who are not in education, tion following Croatia’s accession to the EU in 2013. employment, or training (NEET) in Croatia with the Youth unemployment is a common challenge in average for all EU member countries. NEET rates several EU member states and in other regions with generally tend to increase with youth unemploy- young people being disproportionately represented ment, and both have declined significantly in Croatia among the unemployed. They are also more vulner- since 2014. The youth unemployment rate fell to the able to economic shocks and downturns. At the EU-27 average level in 2019 before diverging from peak of the global financial crisis, the youth unem- the EU average again as a result of the economic ployment rate in the OECD countries was 16.8 per- impact of the COVID-19 pandemic. Both rates now cent on average, while it was only 7.1 percent among remain above the respective EU averages. Figure 9 adults (Dayioglu et al, 2022). There is robust empir- shows that, while Croatia’s NEET status in 2021 had ical evidence that frequent and repeated spells of declined relative to 2010, it was still the fourth high- Figure 7: Unemployment rates by age group, men and women Age 15-24 Age 25-54 Age 55-64 50 Unemployment rate, % 40 30 20 10 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 Men Women Source: Eurostat. 7 TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia Figure 8: Youth unemployment rate (left) and NEET rate (right), % and by age 60 25 50 20 40 15 30 10 20 5 10 0 0 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 Croatia (15-24 years of age) EU-27 (15-24 years of age) Croatia (15-24 years of age) EU-27 (15-24 years of age) Source: Eurostat. Figure 9: NEET rates among 15–24-year-olds 25 20 NEET rate, % 15 10 5 0 Sweden Netherlands Iceland Germany Slovenia Czechia Belgium Denmark Luxembourg Austria Portugal Finland Ireland Malta Latvia Spain France Hungary Estonia Cyprus Greece Slovakia Lithuania Poland Croatia Bulgaria Romania Italy 2021 EU-27 2021 2010 Source: Eurostat. est in the EU as of 2021. High NEET rates for out- demand from foreigners for apartments in Croatia of-school youths means that they lose out on labor (particularly in Zagreb and the coast) has put upward market experience and wages and have poorer job pressure on prices, making them less affordable for prospects, all of which are associated with antisocial Croatians (Bejaković and Mrnjavac, 2019). Figure 11, behavior (Henderson et al, 2017). which plots the yearly trend in the estimated age at which young Croatian men and women leave their Among all EU countries, Croatia has the highest parents’ households, reveals a gradually increasing share of young adults living with their parents, likely trend, indicating that a skills mismatch and scarcity reflecting the skills mismatch and affordable hous- of affordable housing may be the primary reasons ing for Croatian youth. Figure 10 ranks EU countries for this pattern. The gender gap, while notable, is by their shares of young adults aged 25 to 34 living not unique to Croatia – overall, EU men leave home with their parents and reveals considerable hetero- later than women.2 geneity across the continent. Given the expectation that most people in this age group would be work- In the following sections, this chapter will describe ing and getting established financially, this share some of the frictions that hinder labor reallocation is very high in several European countries and is in Croatia and prevent the economy from making highest in Croatia. Moreover, a large share of young full use of the human capital potential of the people work on temporary contracts, and the lack of workforce. The analysis will focus on the economic stable employment makes them ineligible for mort- and social barriers for the groups with the lowest gage loans. This situation has worsened since the levels of labor market activity – young people and 2008 economic crisis, disproportionately affecting women.3 The analysis focuses mainly on the supply young people. Despite low mortgage interest rates, side of the market. 2 https://www.euronews.com/2017/05/04/long-goodbyes-how-men-in-croatia-are-the-eu-s-oldest-to-fly-the-nest 8 3 Chapter 2 analyzes the labor market for older Croatians in detail. Hence the analysis in this chapter focuses on the other two major groups that have high inactivity rates – young people and women. TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia Figure 10: Share of young adults aged 25-34 living with their parents, by sex (%) (EU-SILC) 80 70 Men Women 60 50 40 30 20 10 0 Croatia Greece Bulgaria Portugal Italy Poland Malta Romania Slovenia Spain Ireland Latvia Hungary Cyprus EU 27 Czechia Lithuania Belgium Luxembourg Austria Estonia France Germany Netherlands Finland Denmark Sweden Source: Eurostat, based on the EU-LFS "Educational attainment level and transition from education to work.” https://ec.europa. eu/eurostat/databrowser/view/EDAT_LFSE_28/default/table?lang=en Figure 11: Trends in the estimated average age of young Croatians leaving the parental household, by sex 37 35 33 31 29 27 25 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Men Women Source: EU – SILC survey. Skills and Spatial Gaps a growing skills divide among the country’s work- force. Figure 12 depicts trends in LFPRs in Croatia Educational attainment has been increasing in Cro- by educational attainment alongside average rates atia over time,4 which is a promising trend for the for the EU-27. The LFPR of those with a tertiary future, but the changing educational composition education is consistently high in Croatia, at over 85 is creating a mismatch in the short term due to percent, which is very similar to the EU-27 average, high demand for lower skilled labor. Table 3 com- but it is consistently below the EU average for those pares the educational attainment of the Croatian with upper secondary and primary levels of educa- population by age group with the EU-27 average. tion. Together, the trends in educational attainment Croatia has lower shares of the less educated (levels and activity rates imply that the low levels of activ- 0 to 2) and of those with a tertiary education than ity among those with less than a secondary level the EU-27 and a much higher share of those with of education is largely driven by older workers. As a secondary education. Notably, the share of the these workers age and retire, the composition of low-educated is much higher among older cohorts the workforce is likely to shift towards the more in Croatia, highlighting the increase in educational educated. This presents an opportunity to increase attainment in the population over time.5 However, the productivity of the economy with an educated the Croatian economy is still heavily skewed toward workforce, as long as there is a concomitant increase low-skilled services, including tourism (World Bank, in demand for their skills. This will require the econ- 2022a). This has created sizable gaps in activ- omy to shift away from low value-added services ity rates between groups with different levels of towards more skill-intensive, high value-added sec- educational attainment, which has contributed to tors such as manufacturing and ICT. 4 https://ec.europa.eu/eurostat/databrowser/view/LFSA_PGAED__custom_6135595/bookmark/table?lang=en&bookmarkId=f8b1d139-db50- 9 42a7-914c-2633f402578a 5 The annual population with tertiary education peaked in 2018 at 567,000 and dropped to 545,000 in 2022. https://ec.europa.eu/eurostat/ databrowser/view/LFSA_PGAED__custom_6135595/bookmark/table?lang=en&bookmarkId=f8b1d139-db50-42a7-914c-2633f402578a TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia Table 3: Educational composition of the population by age group, EU-27 (2020) and Croatia (2022) Ages25-39 Ages 40-59 Ages 60-64 Educational attainment (ISCED levels) EU-27 Croatia EU-27 Croatia EU-27 Croatia Less than primary, primary and lower secondary 0.15 0.05 0.22 0.14 0.29 0.23 education (levels 0-2) Upper secondary and post-secondary non-tertiary 0.43 0.61 0.46 0.64 0.48 0.61 education (levels 3 and 4) Tertiary education (levels 5-8) 0.42 0.34 0.32 0.22 0.23 0.16 Source: https://ec.europa.eu/eurostat/databrowser/view/LFSA_PGAED__custom_6299650/default/ table?lang=en&page=time:2022 Trends in job vacancies reflect a high demand for Figure 12: Trends in LFPRs in Croatia and EU-27 medium-level skills and a continuing and stable by educational attainment demand for low-skilled occupations, some of which 100 is being met with migrant labor. The top panel of Labour force participation rate, % Figure 13 presents a seasonally adjusted online 90 vacancy index (OVI), which is a monthly index of 80 online job advertisements, posted on a major web 70 portal, Mojposao.6 The annual trend line indicates a 60 steady increase in vacancies from 2010 until mid- 2018, followed by a gradual fall until early 2020 50 and a much sharper fall until early 2021, presumably 40 induced by the COVID-19 pandemic. After that, the 30 index displays a reasonably steady increase, indicat- 20 ing continuous growth in labor demand on an annual basis. The bottom panel of Figure 13 presents a 10 decomposition of vacancies by the schooling levels 0 mentioned in the advertisements. The demand 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 for medium-level education/skills is considerably higher, though vacancies for low-skill jobs have EU27 - Less than primary, primary and lower secondary education been increasing since 2015. According to the web- EU27 - Upper secondary and post-secondary non-tertiary education site, demand for the five most sought-after occu- EU27 - Tertiary education pations – salespersons, waiters, cooks, warehouse HR - Less than primary, primary and lower secondary education workers, and drivers – has remained stable, which HR - Upper secondary and post-secondary non-tertiary education reflects a continuing demand for low-skilled work- HR - Tertiary education ers (bottom panel). The low and declining LFPR of low-skilled Croatians (Figure 12) has created chronic Source: Eurostat. labor shortages in sectors such as tourism, catering, izens with undergraduate or graduate degrees in and construction. Croatian businesses in these sec- eight member countries with the purpose of under- tors are addressing this issue by recruiting foreign standing barriers and constraints faced by those workers. Immigration flows had increased consider- citizens in their respective labor markets (Rimac, ably before the COVID-19 pandemic, in response to 2020).10 The survey surveyed individuals from two labor shortages.7 More recently, in 2021, in response cohorts – those who completed their studies in the to calls from employers’ associations, the Croatian academic year 2012/13 and those who completed government did away with quotas for foreign work- their studies in the academic year 2016/17. The par- ers, allowing companies much greater flexibility to ticipants included those who were in the labor force hire foreigners to fill vacancies.8 as well as those who were inactive (Table 4). Among the participants in Croatia, about 35 percent stated Among workers with a tertiary education, the per- that they were engaged in jobs that did not use all ception of being over-qualified for their job (ver- of the competences and abilities that they devel- tical mismatch) or working in a job outside their oped during their higher education. The majority field of expertise (horizontal mismatch) is high.9 of this group identified as vertically mismatched In 2018, the EUROGRADUATE pilot survey, spon- professionals, among whom 40 percent were also sored by the European Commission, surveyed cit- horizontally mismatched. These mismatched pro- 6 https://www.eizg.hr/indices-351/ovi-index/356. The index is calculated by an enumeration of single new job advertisements whose application deadlines end within 10 the same month for which the index is being calculated. 7 https://www.euractiv.com/section/politics/news/croatian-companies-import-more-workers-to-tackle-shortages/ 8 While foreign workers are required to take a labor market test to work in certain sectors, in others there is no test requirement and employers have considerable flexibility in importing foreign workers. https://www.hzz.hr/app/uploads/2023/03/Lista-zanimanja-izuzetak-od-provedbe-testa-trzista-rada_2023.pdf 9 Vertical mismatches are mismatches between a person’s job and their level of education. In the Eurograduate survey, however, only overqualified cases were considered as a vertical mismatch, in other words, graduates who assume that a lower educational qualification would be sufficient to perform their current job. Horizontal mismatches are mismatches in content. Thus, the field of education does not match the job according to the self-assessment of the graduates. 10 The eight countries were Austria, Croatia, the Czech Republic, Germany, Greece, Lithuania, Malta, and Norway. TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia Figure 13: Online vacancy index, (top), Online vacancy index by skill-level (bottom), seasonally adjusted 400 350 300 250 200 150 100 50 0 Jan 2005 Jan 2006 Jan 2007 Jan 2008 Jan 2009 Jan 2010 Jan 2011 Jan 2012 Jan 2013 Jan 2014 Jan 2015 Jan 2016 Jan 2017 Jan 2018 Jan 2019 Jan 2020 Jan 2021 Jan 2022 Jan 2023 OVI (2015=100) OVI (trend) 80 70 60 Share in % 50 40 30 20 10 0 Oct-00 May-01 Dec-01 Jul-02 Feb-03 Sep-03 Apr-04 Nov-04 Jun-05 Jan-06 Aug-06 Mar-07 Oct-07 May-08 Dec-08 Jul-09 Feb-10 Sep-10 Apr-11 Nov-11 Jun-12 Jan-13 Aug-13 Mar-14 Oct-14 May-15 Dec-15 Jul-16 Feb-17 Sep-17 Apr-18 Nov-18 Jun-19 Jan-20 Aug-20 Mar-21 Oct-21 May-22 Dec-22 Low-skilled Medium-skilled High-skilled Source: https://www.eizg.hr/indices-351/ovi-index/356 Note: A number of job vacancies do not have education requirements. The shares by skill levels in the bottom panel are based on vacancies that report specific education requirements. fessionals were likely to work part-time and in fixed need to improve the quality of training and educa- contractual engagements. The fact that those with tion.11 Croatia’s share of early leavers from educa- graduate degrees had higher unemployment rates tion and training is the lowest in the EU (2.2 per- than those with undergraduate degrees was a con- cent versus 9.9 percent in EU-27) and well below sequence of their unwillingness to accept vertically the EU-level target of 9 percent by 2030, but even mismatched jobs, which is clear evidence of the skills those who graduate have low levels of basic skills. mismatch that exists in the Croatian labor market Table 5 reveals that the percentage of low-achieving (Rimac, 2020). 15-year-olds in reading, math, and science is higher than the EU-level target of 15 percent. Moreover, Most young Croatians enter the labor market with the rates of low-achieving 15-year-olds in math weak foundational skills, suggesting that there is a (31.2 percent) and science (25.4 percent) are among the highest in the EU. Meanwhile, the percentage Table 4: Labor force status in 2018 by cohort and of students enrolled in general secondary schools highest degree, Croatia (%) in Croatia (31 percent) is among the lowest in the EU. In Croatia’s vocational education and training Cohort 2012/13 Cohort 2016/17 (VET) schools, two-thirds of pupils attend four-year B.A. level M.A. level B.A. level M.A. level programs, which constitute an alternative route towards entering tertiary education. However, the Employed 86.3 85.4 36.6 74.8 educational experience and curricula are different Unemployed 7.2 10.4 6.8 17.8 in VET schools than in general secondary schools. Out of labor force 6.5 4.1 56.6 7.3 Among all of the students who took the secondary school leaving examination in 2019-2020, 44 per- Total 100.0 100.0 100.0 100.0 cent were from general secondary schools and 56 Source: Rimac (2020). percent from VET schools, but the failure rates were 11 See Chapter 3 for a detailed analysis of skills attainment over the lifecycle. 11 TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia Table 5: Croatia’s key human capital indicators compared to EU-27, 2020 EU-level targets 2030 target Croatia EU-27 Participation in early childhood education >96% 79.4% 92.8% (from age 3 to starting age of compulsory primary education) Early leavers from education and training (age 18-24) <9% 2.2% 9.9% Reading < 15% 21.6% 22.5% Low achieving 15-year-olds in: Maths < 15% 31.2% 22.9% Science < 15% 25.4% 22.3% Pupils in general secondary schools 31% 52% Tertiary educational attainment (age 25-34) >45% (2025) 36.6% 40.5% Source: https://op.europa.eu/webpub/eac/education-and-training-monitor-2021/en/croatia.html 3.9 percent and 37.8 percent respectively.12 In 2020, Significant heterogeneity exists in these rates, with the percentage of 25 to 34-year-olds with a ter- the lowest rates registered in the city of Zagreb tiary education was 36.6 percent, which was below (6.7 percent in 2021), while some rural counties the EU average of 40.5 percent and also below the have rates close to 19 percent. While some degree EU-level target for 2030 of 45 percent. The low of regional variation can be expected, the magni- percentage of students in general secondary edu- tude seen in Croatia is unusual for a small country. cation combined with the high failure rates among students completing the upper secondary VET, is Table 6: County-level registered unemployment likely to constitute a bottleneck for increasing the rates tertiary attainment rate in the country.13 Country of 2017 2018 2019 2020 2021 The limited involvement of employers in VET cur- Republic of Croatia 13.9 11.1 9.1 8.9 9.4 ricula and planning has exacerbated the mismatch Zagreb 11.3 8.0 6.3 5.8 6.7 between the demand for and the supply of skills in Krapina-Zagorje 10.5 8.1 6.4 6.1 6.6 the Croatian labor market. Croatia has a strong VET Sisak-Moslavina 29.0 24.3 19.9 18.7 17.5 tradition. It has one of the highest numbers of stu- Karlovac 16.6 13.2 9.9 9.1 9.1 dents participating in these programs at the upper secondary level in the EU. These VET programs are Varaždin 6.9 5.1 4.0 4.2 4.0 school based but also include some practical train- Koprivnica-Križevci 13.0 8.7 6.8 6.4 6.8 ing in the workplace, though this accounts for less Bjelovar-Bilogora 22.7 18.3 13.7 12.4 11.8 than 10 percent of the practical training component Primorje-Gorski on average (CEDEFOP, 2020a). Employers play only 9.9 8.0 6.7 7.2 8.2 kotar a limited role in the planning and funding of the VET Lika-Senj 19.4 15.8 12.5 12.7 12.1 sector, which is consistent with the labor shortages Virovitica- found in sectors such as tourism, catering, con- 28.8 23.2 19.6 18.8 18.5 Podravina struction, and manufacturing as the VET schools Požega-Slavonia 19.0 16.5 12.7 12.9 12.9 are not producing graduates with the skills that the Slavonski Brod- labor market needs.14 Another consequence of the 22.4 19.9 15.7 14.6 16.2 Posavina VET curricula not being driven by demand is that Zadar 13.3 10.7 9.2 9.2 8.5 almost half of VET graduates work in a field outside their specialization. For older workers, a lack of adult Osijek-Baranja 24.9 21.2 17.6 16.8 16.8 retraining is a factor in their low LFPRs. Only 3 per- Šibenik-Knin 19.9 15.4 14.2 14.3 14.5 cent of those in the 25 to 64 age group participate Vukovar-Sirmium 25.1 20.5 16.0 15.1 14.5 in some form of workforce education or training Split-Dalmatia 21.4 18.2 15.4 14.6 16.6 (CEDEFOP, 2020a). Istria 6.2 4.9 4.6 6.3 6.8 There are discernible variations in labor market Dubrovnik-Neretva 16.5 13.0 12.7 13.1 14.0 dynamics between regions due to spatial immo- Međimurje 9.7 6.5 5.4 5.5 5.6 bility, the seasonality of tourism, and migration City of Zagreb 6.4 4.7 3.7 3.7 4.3 patterns. Table 6 presents the registered rate of Source: Croatian Bureau of Statistics. unemployment by county for the 2017-2021 period. 12 Passing the secondary school leaving examination, Matura, is obligatory for all those who wish to continue their education in one of the 12 higher education institutions. 13 https://op.europa.eu/webpub/eac/education-and-training-monitor-2021/en/croatia.html 14 In response to requests from employers’ associations, the government has done away with quotas for foreign workers since 2021. While foreign workers are required to take a labor market test to work in certain sectors, in others there is no test requirement, and employers have considerable flexibility in importing foreign workers. https://www.hzz.hr/app/uploads/2023/03/Lista-zanimanja-izuzetak-od-provedbe- testa-trzista-rada_2023.pdf TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia This is compounded by the lack of internal migra- sectors, currently accounts for approximately 5 tion. The predominance of the tourism industry in percent of GDP. Projections based on a detailed the country’s economic structure skews demand analysis – which took into account the composi- in local labor markets during the holiday season. tion of the workforce, the expected rate of adop- Cities and towns along the coastline benefit from tion of automation based on available technologies tourism, while old industrial and farming regions and the economic feasibility of their implementa- do not. Most seasonal workers do not consider it tion – indicated that the digital economy could be worthwhile to permanently relocate to the tourist the country’s new growth driver and contribute up areas given that there is little employment there in to 16 percent of GDP as early as 2025.16 While the the off-season. Another factor that hampers inter- impact of the COVID-19 pandemic may require this nal mobility is that, in the absence of an adequate timeline to be adjusted, Croatia has the potential welfare state that takes care of older adults, chil- to reach this goal. This will require citizens to reskill dren are expected to look after their elderly relatives and upskill to be able to take up the new jobs cre- (Håkansson and Bejakovic, 2020). While high levels ated by these modern technologies. Croatians are of emigration have exacerbated some of these broadly well-placed to take on this challenge albeit regional inequities, immigrant flows have allevi- from a low base. In 2022, Croatia was ranked 21st ated labor shortages in Zagreb, and coastal regions out of 27 EU member states in the annual Digital that receive significant numbers of tourists. Thus, Economy and Society Index (DESI) rankings17 and regional variation in the demand for specific skills, was among the lowest-performing EU member combined with property market distortions create states in terms of digitalization (CEDEFOP, 2020c). labor market rigidities – spatial and skill mismatches However, among the four dimensions that underlie – and prevent Croatia’s labor market from respond- the DESI score,18 Croatia ranks 9th among the 27 EU ing to economic conditions. countries in terms of human capital, scoring higher than the EU-27 average (Figure 14). This means that A well-functioning housing market would facilitate the level of digital skills among the Croatian popu- labor mobility within Croatia and enable the econ- lation is above the EU average in several key indica- omy to adjust to adverse shocks. As noted above, tors, such as the share of the population with basic the housing sector in Croatia has been a major digital skills and basic digital content creation skills. source of labor market rigidity in Croatia (World If appropriate policies were put in place, Croatia Bank, 2021a). While the homeownership rate is very could harness these advantages to create a digi- high, the distribution of housing is highly skewed, tally skilled workforce. A more digitalized economy with an oversupply of housing units in economi- will also increase labor market flexibility by enabling cally depressed areas and a shortage in the boom- more employees to work remotely, thus boosting ing urban areas. High transaction costs are a factor labor supply and increasing workers’ attachment to limiting internal mobility, as is the insufficient supply the labor market.19 of services such as preschools, schools, and elderly care in many regions (Håkansson and Bejakovic, The National Coalition for Digital Skills and Jobs in 2020). Computerizing the land registry and cadas- Croatia, established in December 2018, is focused ter and moving to fully online registration of immov- able property will increase efficiency, while making Figure 14: DESI index 2022 – Relative for more reliable records and a more efficient and performance by dimension transparent processes. Harmonization of land regis- DESI 2022 - relative performance by dimension try and cadaster data will further support more effi- Country with lowest score cient land transactions. This in turn would encour- age internal migration driven by labor demand, which would enable the economy to adjust to eco- nomic shocks more effectively. Croatia’s e-Citizens initiative, launched in 2014 in accordance with the highest score Country with European interoperability framework, is expected to foster this transition via the digitization of infor- mation and by increasing the transparency of the public sector in providing public services.15 DESI 1 Human 2 Connectivity 3 Integration 4 Digital capital of digital public Automation and digitalization offer an opportu- technology technology nity to create a more productive and competi- Croatia EU tive Croatian workforce. Croatia's digital economy, Source : https://ec.europa.eu/newsroom/dae/redirection/ which includes all digital activities in all economic document/80477 15 https://gov.hr/en 13 16 https://www.total-croatia-news.com/business/57305-croatian-workers 17 https://digital-strategy.ec.europa.eu/en/policies/desi-croatia. 18 DESI includes four dimensions: (i) connectivity (fixed and mobile broadband, prices), (ii) human capital (Internet use, basic and advanced digital skills) and use of Internet services (citizens' use of content, communication, online transactions), (iii) integration of digital technology (business digitalization, e-commerce), and (iv) digital public services. 19 Under the National Recovery and Resilience Plan (2021-2026), digital work platforms have been introduced as a new form of work, and the provisions related to digital work platforms will enter into force on January 1, 2024. See https://assets.kpmg.com/content/dam/kpmg/hr/pdf/2023/02/Promjene-u-hrvatskom-radnom- zakonodavstvu-I-Changes-to-Croatian-Employment-Legislation.pdf TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia on preparing the workforce for the future of work. Survey (LFS), among the inactive population aged While Croatia does not have a strategic framework between 15 and 64, nearly 10 percent of women focused on digitalization, the National Coalition for attributed their labor market inactivity to care Digital Skills and Jobs in Croatia, which is coordi- duties. Among inactive men, this reason was given nated by the Croatian Employers' Association, has by fewer than 0.5 percent (Figure 15). a mission to increase the employment and retention of qualified professionals in Croatia. The Coalition’s Women tend to drop out of the labor market once charter proposes to steer the unemployed toward they have children and are more likely than men ICT careers, increase the number of ICT specialists in to be inactive or unemployed. Some of the gender the labor market, increase the digital competencies gap in employment rates is driven by child-rear- of professionals in non-ICT occupations, support ing responsibilities. Figure 16 presents the average digital leaders, and foster creative digital society (in employment rates of men and women according science, technology, engineering, arts, and mathe- to the age of the youngest child residing in their matics or STEAM). The Charter also advocates for household. Interestingly, for both men and women, the continuous professional development of digital employment rates are higher among households specialists through lifelong and workplace learning with children. However, the gender gap is highest and mentorship. It aims to promote digital skills when the youngest child is less than 6 years old and careers by focusing on primary and secondary and gets narrower as children grow older, which is schools and by targeting women to consider dig- an indication that women take on the predominant ital careers. These proposals, if successfully imple- share of care responsibility by dropping out of the mented, would propel Croatia towards a full-scale labor force. Croatia’s parental leave policies, while digital transformation that would drive innovation generous in terms of duration, are not financially and growth. They would also prepare citizens for the adequate. The policies offer up to six months of future of work and increase the dynamism of the maternity leave with job protection, and additional labor market (CEDEFOP, 2020c). six months of parental leave if taken up by both parents or six months of parental leave if used by Balancing work with care one parent. However, the share of fathers taking parental leave is very low in Croatia compared to responsibilities – A lack of the shares in other EU countries (Robayo-Abril et flexible work options al, 2020). These policies, while seeking to offer ben- efits for parents, may have further institutional- Social norms place a disproportionate amount of ized the gendered division of care (Dobrotić, 2022). informal caring duties on women, which contributes Women are likely to face wage penalties as a result to their high economic inactivity rates. A lack of of their extended absences from the labor market, adequate formal and informal child and elder care with evidence suggesting a sizable negative associ- in Croatia means that women are required to take ation between men’s participation in parental leave on a disproportionate share of household chores and the gender gap in labor force participation and and child and elder care in keeping with prevailing employment (Thévenon and Solaz, 2013 and Grim- gender norms. According to the 2021 Labor Force shaw and Rubery, 2015). The expansion of kinder- Figure 15: Principal reasons offered by respondents for inactivity (percentage)* Retirement Own illness or disability Other reasons Care for children or incapacitated relatives Other personal reasons Other family reasons Education or training 0 10 20 30 40 50 60 70 80 Share of inactive people Men Women Source: Labor Force Survey 2021; * - excludes students. 14 TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia Figure 16: Employment rates by age of youngest child in household 100 90 80 Employment rate, % 70 60 50 40 30 20 10 0 Total No children Less than 6 years From 6 to 11 years 12 years or over Men Women Gap Source: LFS 2021. gartens and the recent increase in the number of irregular and uncertain earnings, contributing to instructional hours required in schools aimed at women’s socioeconomic disadvantage. Women’s improving learning outcomes should also help to lower LFPRs and tenuous attachment to the labor increase women’s labor participation rates.20 force translate into a pension gap. women in Croatia received, on average, pension or old-age safety net The Croatian government’s recent changes to the payments that were 28 percent lower than those pension system under the National Recovery and received by men, marginally below the EU aver- Resilience Plan (NRRP) are intended to redress age gender pension gap of 29 percent. Given their some of the negative consequences of women’s longer life expectancy than men, these disparities tenuous attachment to the labor market. On aver- render women more financially vulnerable as they age, working women are more likely to be marginally age. Until recently, a widowed housewife inherited employed than their male counterparts, doing less only 70 percent of her husband’s pension. stable jobs and experiencing more frequent unem- ployment spells. A higher share of working women Recent reforms to the pension system under the is engaged in short-term, temporary contracts than National Recovery and Resilience Plan (NRRP) are men. As Figure 17 shows, among the unemployed in aimed at increasing the minimum pension. They 2021, about 43 percent of women reported being also make it possible to factor part of the survivor’s unemployed due to the end of a fixed-term con- pension in addition to their personal pension into tract compared to the 33 percent share among the calculation of their benefit level for lower-in- men. Women are also more likely to work in tem- come pensioners. These changes are expected to porary seasonal jobs, which are characterized by be implemented from 2023.21 Figure 17: Reasons offered for current unemployment status, men and women End of the temporary contract Seasonal job Got fired Other reasons The firm closed down Other personal reasons Own illness or disability Further education Care for children or incapacitated relatives The firm went bankrupt Not stated Other family reasons Early retirement Retirement 0 5 10 15 20 25 30 35 40 45 Share of unemployed people Men Women Source: LFS 2021. 20 See Chapter 3 for a more detailed description of this program. 15 21 https://commission.europa.eu/system/files/2022-05/national_reform_programme_croatia_2022_en.pdf TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia Part-time employment rates are low relative to Table 7: Part-time and self-employment rates (%), the EU-27 average. Only a meager share of the Croatia and EU-27, 2021 employed in Croatia work in part-time jobs, which is a common feature of post-socialist countries. Table Croatia EU-28 7 shows that, of the total female workforce, 6.1 per- Female Male Female Male cent are in part-time employment in Croatia, com- Part-time employment pared to 28 percent in the rest of the EU. For men, 6.1 3.6 28.3 8.2 rate (%) the rates are 3.6 percent and 8 percent respectively. Based on the 2021 LFS, among those in part-time Self-employment 7.8 14.3 9.4 16.1 rate (%) employment, 34 percent of women and 33 percent of men attribute their part-time status to their ina- Source: Eurostat; * - Q3 data. Note: The rates for Croatia are bility to find full-time jobs, which again indicates the from the LFS 2021 while the EU rates are from Eurostat for 2022. shortage of full-time jobs in the economy (Figure 18). For men –- especially young men – the limited higher than that of several other countries. To the opportunities for part-time employment hinder extent that this is symptomatic of the challenges their ability to gain valuable work skills that would involved with setting up one’s own business, this is enable their transition to full-time jobs. For women, problematic, especially for more vulnerable groups the scarcity of part-time employment compounded such as those who are less mobile or less skilled. by the limited availability of formal care centers for children and the elderly makes it particularly chal- Flexible working arrangements are rare in paid lenging to combine working and family obligations, employment. Inflexible work arrangements limit which limits their labor market participation and people’s ability to combine their work with their employment potential. other obligations, a constraint that is likely to be more pressing for women. Using data from the There are few self-employment or entrepreneur- 2021 LFS, Figure 20 indicates that over 80 percent ship activities in Croatia, especially for women. of employed women have no control over their work Part-time work, self-employment, and entrepre- hours, which are set by their employer. Among men, neurial activities are characteristics of a flexible labor this share is 82 percent. Moreover, about 87 per- market. Croatia compares poorly with the EU in this cent of women and 91 percent of men stated in the respect, with self-employment rates below the EU survey that they had no option to work from home averages for both men and women. The gender (teleworking). These are very high shares, given that gap within Croatia is also notable, with the female teleworking has become commonplace in many self-employment rate being about half of the rate high-income countries following the disruption among men (Table 7). Figure 19 gives a cross-country induced by the COVID-19 pandemic. In 2022, fewer comparison of the total early-stage entrepreneurial than 5 percent of Croatian workers were working activity (TEA), a key measure of entrepreneurship.22 from home, compared to the EU-27 average of 10 While Croatia performs well overall, ranking in the percent (Figure 21). This is, to some extent, due to middle of the high-income group of EU countries the tourism-centric nature of the economy, which on this measure, its gender gap in TEA is noticeably limits the ability to telework. Figure 18: Main reason reported by the part-time employed Cannot find a full time job Other reasons Other personal reasons Care for children or incapacitated relatives Other family reasons Education or training Illness or disability 0 5 10 15 20 25 30 35 40 Share of people working part-time Men Women Source: LFS 2021. The TEA is measured as the percentage of the adult working-age population aged 18 to 64 who are either in the process of starting a new 22 16 business but have not paid wages for more than three months (nascent entrepreneurship) or have businesses that are older than three months but younger than 42 months (early-stage business activity) (GEM, 2022). TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia Figure 19: Average Total Early-stage Entrepreneurial Activity (TEA) rates by gender, 2021 45 Average Total early-stage Entrepreneurial Women 40 Men 35 Activity (TEA) (%) 30 25 20 15 10 5 0 Morocco Iran Egypt India Sudan Russian Federation Romania Belarus Turkey Colombia South Africa Kazakhstan Brazil Panama Guatemala Dominican Republic Poland Norway Italy Spain Japan Greece Slovak Republic Slovenia Germany Luxembourg France Finland Cyprus Sweden Israel Switzerland Hungary Croatia Ireland United Kingdom Oman South Korea Qatar United Arab Emirates Netherlands Latvia United States Saudi Arabia Canada Uruguay Chile Low & lower-middle Upper-middle High income income income Source: GEM (2022); https://www.gemconsortium.org/reports/womens-entrepreneurship. Figure 20: Prevalence of flexible work schedules (left) and work-from-home option (right) Never Not flexible Sometimes (Partially) flexible Frequently 0 20 40 60 80 100 0 20 40 60 80 100 Share of employed people Share of employed people Men Women Men Women Source: LFS 2021. Figure 21: Employed persons working from home as a percentage of the total employment, 2022 (%) 30 25 20 15 10 5 0 Cy tia un ce oa y Po rus ov d Li It a u y e a Ic chia Sp nd rt in Sl rwa l ov y Sw L nia D rla a m d M rk to a m 27 lg ia Gr aria Fr nia Ge ust e rm ria x lg y N em ium rl rg e s nl n Ire and nd o a Sw and Cr ar th al Lu Be an i Cz ani e i Es lt A nc Fi de Sl lan en n ak N ug itz tv Bu an Po a a he u H ee Ro U- a a la e p g et bo a a el E Source: https://ec.europa.eu/eurostat/databrowser/view/LFSA_EHOMP/default/table?lang=en 17 TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia Barriers to Work: Institutional ued to increase despite the reduction of non-taxa- ble allowances, and in the private sector, there were Features Dampening Labor significant reductions in non-taxable allowances, Demand and Supply especially in activities related to tourism (Figure 23). Despite spending more than its peers in Central Public sector wage policies have raised unit labor and Eastern Europe (CEE), Croatia ranks very low costs in both the public and private sectors, low- in several dimensions of public sector performance, ering Croatia’s productivity. While the govern- implying that public spending is inefficient (World ment made efforts to reduce the public wage bill Bank, 2021). in the period following the global financial crisis, these cost containment measures were effective High and sustained increases in public sector wages only during the 2009 to 2014 period (World Bank, are affecting public sector efficiency and limiting 2021a). Since 2015, the public sector wage bill has the dynamism of the private sector (World Bank, been increasing due to wage raises and an increase 2021a). The share of public sector employment23 in the number of public sector employees. In 2018, was 23 percent in Croatia in 2020, well above the the share of compensation for general government EU average of 16 percent.24 In contrast with the employees in GDP surpassed the level before the public sector, private sector employment fell during global financial crisis (Figure 22). While the pro-cy- the recession, with the number of private sector clical nature of public sector wages was also evi- employees declining by almost 17 percent between dent in other countries, by 2019, the share of GDP 2008 and 2014. While employment increased mod- accounted for by the public sector wage bill in erately thereafter as the economy recovered, in EU-27 countries was lower than in pre-recessionary 2019, it was still 6.8 percent lower than the 2008 2008. Furthermore, Croatia’s wage bill continued to level (World Bank, 2021). Private sector firms have increase significantly in 2020 and 2021 despite the to compete with the public sector for workers, deterioration in the country’s fiscal situation due including educated workers, because these workers to the economic downturn related to the COVID- are drawn to the high wages and benefits offered 19 pandemic. The wage increase in Croatia’s private by the public sector. This is evident from the fact sector has been even steeper over the last decade. that core public sector employees have, on average, The average wage in the private sector did not fall two more years of completed education than those in any single year after the 2009 financial crisis. In employed in the private sector or those in the public 2010, it remained the same as in 2009, and in the companies.25 Furthermore, the majority of full-time next couple of years, it grew by up to 2 percent per employees with a tertiary education are employed year and then accelerated with the subsequent in either the core public sector or in public compa- economic recovery. Between 2018 and 2019, total nies. This puts upward pressure on wages in the compensation (including non-taxable allowances) in private sector. High wages that are not driven by the private sector increased by 6.6 percent in 2018 productivity increases reduce the competitiveness and 5.8 percent in 2019, which was much higher of firms and inhibit the reallocation of labor to more than the wage increases in the public sector. How- productive activities. They also constrain the ability ever, in 2020 and 2021, public sector wages contin- of firms to make optimal capital investments and Figure 22: Compensation of employees paid by general government: Croatia and EU 14.0 13.5 13.0 12.5 In % of GDP 12.0 11.5 11.0 10.5 10.0 9.5 9.0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 EU - % of GDP CEE - % of GDP Croatia - % of GDP Source: World Bank (2021a). 23 The scope of the government sector varies among EU member states. For example, in some countries, jobs in education or health are part of 18 government employment but not in others. 24 https://ec.europa.eu/eurostat/cache/digpub/european_economy/bloc-4d.html?lang=en 25 ‘Core’ public services refer to public administration, education, health and social care sectors, with ‘public companies’ constituting the rest of public sector. TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia Figure 23: Real GDP and nominal average gross uncertain income flows (World Bank, 2019e). The wage growth by sector share of temporary employment declined stead- 9.0% ily between 2016 and 2021, going up again to 12.4 7.0% percent in 2022 (compared with the EU-27 share of 5.0% 11.1 percent). The share remains among the highest 3.0% in the EU. While temporary employment may offer 1.0% flexibility for some workers, Croatia has a high share -1.0% of the involuntary temporarily employed – employ- -3.0% ees who could not find a permanent job or whose -5.0% job is only available with a temporary contract – as -7.0% a percentage of total employees. This indicates that -9.0% most workers with this work status do not desire 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021q1 this type of employment. The Netherlands, in con- trast, has a much smaller share of involuntary tem- Public sector wages Private sector wages porary employment but a larger share of temporary Real GDP employment, indicating that much of this type of Source: World Bank (2021a). employment is driven by workers’ preferences.26 On January 1, 2023, the government amended the ultimately have a negative impact on their capacity Labor Law as part of the NRRP in an effort to reduce to innovate and grow (World Bank, 2021a). labor market inequalities and increase worker pro- tections. These reforms address several aspects Croatia’s 2013-2014 labor market reforms resulted of employment relations including the mandatory in the segmentation of the labor market, which new provisions of employment contracts, limits to the reforms under the NRRP are seeking to address. length and frequency of fixed-term contracts, reg- The 2013-2014 labor market reforms liberalized ulations concerning remote work and work through employment protection provisions for both tem- digital platforms, changes to the structure of paren- porary (2013) and permanent (2014) jobs. Empiri- tal leave benefits, and measures to prevent and cal analysis suggests that these reforms resulted in sanction undeclared work.27 an increase in temporary employment (Figure 24). Moreover, specific groups of the population – such While the tax wedge in Croatia is slightly lower as females, young people, foreigners, and the low- than the EU average, especially for lower-income skilled – had a higher probability of getting jobs earners, single households with children face with temporary contracts, suggesting that the labor disincentives in transitioning from inactivity to market was becoming increasingly segmented employment. Since the global financial crisis, the (Tomić, 2020). Commentators have characterized government’s tax strategy has been to support this market as a dual labor market of “insiders” – employment by shifting the tax burden away from those with stable, permanent jobs with benefits – labor and towards consumption. Several changes to and “outsiders” – those on temporary contracts with the personal income tax (PIT) system over the last Figure 24: Share of temporary to total employment (left, 20-64) and share of involuntary temporary employment to total employment (right, 15-64), % 20 20 18 18 16 16 14 14 12 12 10 10 8 8 6 6 4 4 2 2 0 0 14 15 16 17 18 19 20 21 22 14 15 16 17 18 19 20 21 22 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 EU-27 Croatia EU-27 Croatia Source: https://ec.europa.eu/eurostat/databrowser/view/TESEM110__custom_6428992/default/table?lang=en; https:// ec.europa.eu/eurostat/databrowser/view/TESEM190__custom_6429184/default/table?lang=en 26 The online vacancy index shows that the share of advertisements offering permanent employment increased significantly from 40.6 19 percent of all job advertisements in April 2022 to 45.2 percent in April 2023, while there was a notable decrease in the shares of fixed-term employment contracts and seasonal employment advertisements. Whether this represents a change in trend remains to be seen. https:// www.eizg.hr/indices-351/ovi-index/356 27 https://assets.kpmg.com/content/dam/kpmg/hr/pdf/2023/02/Promjene-u-hrvatskom-radnom-zakonodavstvu-I-Changes-to-Croatian- Employment-Legislation.pdf TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia decade have resulted in lowering the tax wedge in for women to participate in the labor force (Robayo- Croatia, relative to other EU member states, espe- Abril et al, 2020). Phasing out of the child allowance cially among for those with low earnings (World benefit more gradually by having an explicit income Bank, 2021a).28 The extent of the “inactivity trap,” in disregard could encourage women to re-enter the which individuals would lose their social assistance labor force after taking time away to have and raise benefits and potentially become liable for more tax- children. ation if they were to transition into employment, is largely the same as the EU average (see Chapter 6). The Impact of Non-Labor Furthermore, since 2016, regular increases in the statutory minimum wage have increased incentives Income: The “Airbnb” Effect and to work (Christiaensen et al, 2019). There is, however, Remittances a clear inactivity trap for single parent households with two children whose earnings are at 67 percent Croatia has a very high rate of homeownership. In of the average wage. For this group, the inactivity 2021, 90.5 percent of people were owners of the trap was as high as 63 percent in 2022, about 10 dwelling where they lived, giving Croatia the fourth percentage points above the EU average, although highest homeownership rate in the EU.30 A principal it has been on a declining trend.29 This is largely a reason for this is that Croatia does not levy prop- consequence of the reduction in the guaranteed erty taxes on owner-occupied dwellings. Given that minimum benefit (GMB), the housing benefit, the 20 tourism is the most important economic sector in percent deduction for pension social security con- the country, income from renting property is a sig- tributions, and, to a lesser extent, the child allow- nificant source of income for many households. ance. Payments of the GMB and the housing bene- While holiday homeowners pay property tax, the fit are phased out over time, but the child allowance amounts are very low, incentivizing rental activities benefit is means-tested, with no special provisions in the country. In the 2018 Survey on Income and as employment income is included in the means Living Conditions (SILC), 6.4 percent of households test. The abrupt withdrawal of the child allowance in Croatia reported receiving income from renting benefit combined with the mandatory social secu- out their property. This is likely a gross underesti- rity contributions creates a significant disincentive mate. In the 2021 LFS, for example, among house- Figure 25: Average marginal effects of various control variables on labor force participation # of household members HH member aged 80 and over HH member aged under 18 Urbanisation Months spent caring Household income Experience Sex Never married Widowed Independent variable Married Have rental income Tenant with reduced rent Tenant Dwelling owner Chronic condition Health Post-secondary education Upper secondary education Lower secondary education Primary education Age 25-34 Age 35-44 Age 45-54 Age 55-64 Age 65+ -0.8 -0.6 -0.4 -0.2 0.0 Average marginal e ect Reference categories are: Sex: Female, Marital status: Divorced, Tenant status: Tenant not paying rent, Education: Lower than primary education, Age: 19-24 Source: EU-SILC 2018. This table reports the average marginal effects of each of the listed independent variable on the probability of being in the labor force, estimated using a Probit regression model. Note: Months spent caring=number of months spent fulfilling domestic tasks and care responsibilities (in the income reference period). 28 The tax wedge measures the tax burden on labor income. It is the difference between the total labor cost of employing a worker and their 20 net earnings and is defined as the sum of personal income taxes and employee and employer social security contributions net of family allowances, expressed as a percentage of total labor costs for the employer. 29 https://europa.eu/economy_finance/db_indicators/tab/ 30 https://www.statista.com/statistics/246355/home-ownership-rate-in-europe/. See Bejaković & Mrnjavac (2019) for a comprehensive analysis of changes in housing policy in Croatia since the end of World War II, and the impact of recent changes on internal migration and labor market dynamics. TOC Chapter 1: Croatian Labor Market: Trends, Features, and Challenges Harnessing Human Capital for Growth in Croatia holds that reported having positive capital income, Figure 26: Personal remittances received (1993- 15.3 percent reported receiving rental income. 2021) wage growth by sector Based on estimates from a probit regression using $6 000 000 000 8 SILC data (Figure 25), the average marginal effect 7 (AME) of the household having positive rental $5 000 000 000 income is associated with a 4.8 percent decrease in 6 the probability of a working-age adult in the house- $4 000 000 000 5 hold participating in the labor force. This is consist- 4 ent with the analysis presented in the public finance $3 000 000 000 review, which states, “Tax rates on income earned 3 $2 000 000 000 from labor and capital are much higher than income 2 from immovable property, fostering rental activities $1 000 000 000 and providing disincentives to work” (PFR, 2021). 1 $0 0 Emigration may be impacting the labor market 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 activity of non-migrants in Croatia via its income effect coming from remittances. High emigra- Current US$ % GDP tion rates invariably have significant effects on the Source: World Bank staff estimates based on IMF balance of labor market of the sender country. In addition to payments data. the direct loss of workers, emigrants may repress well-designed ALMPs and lifelong learning pro- the labor market activity of non-migrants by send- grams, designed in collaboration with employers to ing them remittances that reduce their incentive to ensure that these programs teach the skills that are seek work. There is robust empirical evidence for relevant to the market and to ensure that workers this in various countries around the world (Chami et are given practical work experience. The demand al, 2018). While there is no direct evidence of such for skills also needs to increase to absorb Croatia’s a causal relationship in Croatia, the large outflow increasingly well-educated workforce. Investing in of labor in recent years combined with an inflow of the digital economy will be a crucial component of remittances suggests that such a relationship might the policy package to move Croatia up along the exist. As Figure 26 reveals, there has been a marked value chain, from a low value-added services econ- increase in remittances since Croatia’s accession to omy dominated by the tourism sector to a more the EU, which is likely to have had a negative effect diversified and productive economy. Studies have on the labor market activity of non-migrants from found that the main factors supporting the efficient sending households. However, this outcome needs adoption of digital technologies are: (i) increasing to be balanced against the potential benefits of emi- access to high-speed internet; (ii) upgrading tech- gration, such as a reduction in the unemployment nical and managerial skills; and (iii) implementing rate, a rebalancing of the demand for and supply of product and labor market reforms to facilitate the skills in the labor market, and the positive impact of reallocation of resources in the economy (Sorbe et remittances on households’ human capital acquisi- al, 2019). Increasing the supply of affordable child tion and entrepreneurship. While a sharp decline in and elderly care arrangements is a complementary the unemployment rate has been associated with policy that would allow workers, especially women, increased emigration flows in Croatia, there is less to take advantage of flexible work options and direct evidence of the receipt of remittances having increase their LFP. The current NRRP reforms of any positive impact on the incidence of self-em- pension and tax policies to eliminate disincentives ployment. Chapter 4 undertakes a more detailed to work and of labor laws and regulations to facili- analysis of the likely impact of Croatia’s emigration tate remote work will help to build a more resilient flows on human capital acquisition at home. and adaptable economy better prepared to nav- igate future crises. And crucially, Croatia’s current In summary, the challenges facing the Croatian social protection system, which has been predicated labor market require integrated solutions. Enhanc- on the standard formal employment relationship, ing human capital skills will be a critical component will need to be modified to address the challenges of this policy package. The nature of the future of associated with automation and digitalization and work will require two types of changes from the to foster an inclusive labor market for the future. workforce: (i) upskilling, in which workers gain addi- tional skills to help them to carry out their cur- rent roles and (ii) reskilling, in which they develop new capabilities to take on different or entirely new roles.31 Achieving these two goals will require https://www.mckinsey.com/capabilities/operations/our-insights/building-the-vital-skills-for-the-future-of-work-in-operations 31 21 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Introduction The aging of Croatia’s population has the poten- ency ratio, which is the ratio of working individuals tial to hamper the country’s economic growth as who can finance the consumption of the non-work- there will be fewer working-age individuals to ing population. Croatia’s employment rate was support the increasing number of elderly citizens. around 5 percentage points lower than the EU-27 Two critical trends are underway. First, over the average in 2020 and about 10 percentage points next 30 years, the Croatian population is expected lower than the rate in neighboring Austria. There- to decrease from 4 to 3.3 million. Second, by 2050, fore, it will be crucial to increase employment rates every third Croatian will be aged over 65, up from 8 to mitigate Croatia’s soaring dependency ratio. percent in 1960 and from 22 percent in 2021. The Increases in the number of adults in jobs need to elderly dependency ratio, which compares the size be supplemented by productivity increases and of the population aged 65 and over with the size of a willingness on the part of older adults to extend the working age population (those aged between their working lives. Failing to achieve these out- 16 and 64), is nearly 10 percentage points higher in comes would significantly diminish Croatia’s eco- Croatia than the EU-27 average in 2023. That gap is nomic potential. Not only would this limit economic projected to remain roughly constant over the next growth, but it is also a significant financial risk as 30 years (Figure 27). Therefore, fewer working-age the social security obligations on employers and Croatians will have to finance the costs of a growing employees, and savings in general would have to elderly population. increase to ensure adequate retirement income. The impact of aging is made worse by the fact that This chapter is based on the proposition that much Croatians work less than their European counter- of Croatia’s relative employment gap and potential parts (Figure 28) and are less productive. While “unused productive capacity” (Gruber and Wise, not all working-age individuals work for pay, unpaid 1998) happens after the age of 50. While participa- employment rates increase the effective depend- tion rates among working-age women and young Figure 27: Projected elderly dependency ratio Figure 28: Croatia’s employment gap 2010-2022 2020-2050, selected European countries 0.7 80 75 Dependency ratio (population aged 65+/working-age population) 0.6 70 Employment rate, % 65 60 0.5 55 50 0.4 45 40 0.3 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2020 2025 2030 2035 2040 2045 2050 Croatia Serbia Slovenia CZ HR HU SI EU-27 Austria Hungary SK ME MK RS Source: UN population projections. Source: Eurostat LFS. This chapter was prepared by Clement Joubert, Tibor Kovać and Lea-Karla Matić. 22 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia people in Croatia may also need to be boosted (see that are embedded in Croatia’s pension system and Chapter 1), the labor participation rates of older the extent to which they can account for the over- Croatians lag well behind those found elsewhere 50 employment gap. Lastly, drawing from descrip- in Europe and therefore they may be easier to acti- tive evidence from the SHARE survey and the exist- vate. The chapter examines empirical evidence on ing literature, it examines how Croatia fares on other retirement behavior in Croatia and analyzes the role determinants of early retirement, including deterio- of pension incentives, care obligations, and labor rating health, care obligations, and the lack of jobs demand for older workers with the aim of propos- for older workers. ing policies that promote longer work lives in the context of the factors that are pushing older Cro- Examining Croatia’s atians out of the labor force. The chapter is organ- ized as follows. The first section shows that Croatia’s Employment Gap for Older employment rate gap is largest and most persis- Workers tent for older Croatians at rates that are greater than in its peer countries in Europe. It then explores Workers over the age of 50 account for the bulk which subgroups of older workers have the lowest of Croatia’s low overall employment rates, making employment rates and provides a broad description this an important group for counteracting rising of the employment characteristics of Croatians who dependency ratios (Figure 29). In 2021, the employ- are over 50. The second section analyzes possible ment rate of older workers (aged 50 to 64) in Croa- causes of the employment rate gap, and it intro- tia was 54 percent, 12 percent lower than the EU-27 duces a framework to analyze the determinants average of 66 percent. Since older workers repre- of older workers’ decisions to retire. Data from the sent a third of Europe’s working-age population, Survey of Health, Aging, and Retirement in Europe that age group accounts for roughly three-quarters (SHARE) are used to estimate which individual char- of Croatia’s overall employment gap of 5 percent- acteristics are associated with a greater likelihood of age points. The rest of the gap comes from young working after the age of 50. These results are com- people, who have an employment rate that is 6 per- pared with equivalent results obtained for Croatia’s centage points less than the EU-27 average. In con- European peers to gain insights into the specific trast, the employment rates of Croatians between determinants of retirement decisions in Croatia. It the ages of 25 and 49 are comparable to the EU-27 then discusses the institutional incentives to retire average. Figure 29: The employment gap vs. the EU-27 average and selected European countries, 2010-2021, by age groups Y15-24 Y25-49 Y50-64 PP di erence from EU27 average 15 10 5 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Y15-24 Y25-49 Y50-64 75 Employment rate, % 50 25 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 HR EU27 Comparable countries HR EU27 RS AT SI HU Source: Eurostat LFS. 23 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia If the employment rate of older workers could be Figure 31: Employment rates by gender and age raised, this would effectively decrease Croatia’s groups (1995-2021) dependency ratios by increasing the income-gen- WOMEN MEN erating fraction of the population and decreasing 60 the fraction of the population that is dependent. While old-age dependency ratios are computed 40 Y15-24 purely based on the relative size of different age groups, an effective dependency ratio would count a retired 55-year-old as a dependent rather than 20 being part of the working-age population. As a rule of thumb, increasing old-age employment rates by 1 percentage point reduces the effective depend- 90 Employment rate, % ency ratio by around 0.5 percentage point. Bridging 80 Croatia’s 12 percent point old age employment gap Y25-49 with respect to the EU-27 average could reduce its 70 effective dependency ratio by 6 percentage points to 17 percent. Bringing Croatia’s old-age employ- 60 ment rate up to Hungary’s level would reduce it by 9 percentage points to 25 percent. 80 Croatia's employment rate for individuals over 50 Y50-64 60 has increased in the past decade, but the deficit relative to the average for other European coun- 40 tries has also grown from 7 percentage points to 11.5 percentage points. Employment rates for those 20 aged 50 to 64 increased by 6 percent between 2010 1995 2000 2005 2010 2015 2020 1995 2000 2005 2010 2015 2020 and 2021 (Figure 29). In recent years, the largest increase in employment rates in Croatia has been for those in the 55 to 59 age group, particularly HR EU27 RS AT SI HU women, a 10 percent increase between 2018 and Less than primary, Upper secondary Tertiary education primary and lower and post-secondary 2021 (Figure 30). This corresponds to the ages at secondary non-tertiary education education which workers become eligible for early retirement. 80 In contrast, employment rates for those aged 50 to Y15-24 54 and 60 to 64 remained relatively stable over the 60 same period. Rising employment rates among older 40 workers are not specific to Croatia but rather have been part of a universal trend among developed 20 0 Figure 30: Employment rates of older workers by Employment rate, % age groups (2018-2021) 90 Y25-49 MEN WOMEN 80 70 60 60 50 Employment rate, % 80 40 60 Y50-64 40 20 20 1995 2000 2005 2010 2015 2020 1995 2000 2005 2010 2015 2020 1995 2000 2005 2010 2015 2020 2018 2019 2020 2021 2018 2019 2020 2021 50-54 55-59 60-64 HR EU27 RS AT SI HU Source: LFS 2018-2021. Source: Eurostat LFS. 24 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia countries that began in the 1980s for women and The employment rate in Croatia starts declining in the mid-1990s for men (Blundell et al, 2016). As slowly at age 50 and then sharply around age 55 a result, Croatia’s deficit relative to other European for women and 60 for men (Figure 32). At age 50, countries for that age group has also grown by 4.5 employment rates are just under 70 percent for percent since 2010. While the employment gaps women and 75 percent for men, which are both for young and prime-age workers were compara- relatively low by international standards. Blundell ble to or higher than the gap for older workers in et al (2016) reported employment rates at age 50 2010, they have since been cut significantly or have as being between 80 and 90 percent for men and disappeared. Therefore, country-wide macroeco- women in France, the UK, and the US in 2012. The nomic factors are unlikely to explain Croatia’s overall inflexion points for women and men are ages 55 employment deficit. Instead, either labor demand and 60 respectively, which were the ages at which for older workers has contracted or there are now early retirement was allowed for women and men fewer incentives to remain in the labor force in Cro- respectively until 2010. Since that date, women’s atia compared to other European countries. early retirement age has been gradually increased with the aim of converging it with men’s by 2030. The increased employment gap for older work- ers in Croatia affects men and women equally and Figure 33: Weekly hours worked (top) and part- decreases as educational attainment rises. For time employment rates (bottom) by age group both genders, the gap widened during the first half 40 of the decade (2010 to 2015) in the aftermath of Average working hours per week the financial crisis (Figure 31, top). The gap stabilized 35 after 2015, with employment rates for men and 30 women growing in parallel with the EU-27 average. 25 Patterns differ by education (Figure 31, bottom). The gap is around 12 to 13 percentage points for work- 20 ers with less than a tertiary education. In contrast, 15 tertiary-educated Croatians over the age of 50 10 have employment rates close to 80 percent, which 5 is comparable to its European peers. These patterns by education also reflect the influence of the 2008 0 financial crisis. Employment among low-educated 50-54 55-59 60-64 older workers in Croatia declined substantially more MEN 20 than in its European peers in the years following the recession and, contrary to what happened with 15 younger or more educated workers, recovered only slowly and partially after 2015. 10 Figure 32: Employment rates by age (50-80) in Part-time employment rate, % Croatia 5 80 70 0 2018 2019 2020 2021 60 WOMEN Employment rate, % 50 20 40 15 30 10 20 10 5 0 0 50 55 60 60 70 75 80 2018 2019 2020 2021 Age Women Men 50-54 55-59 60-64 Source: LFS 2018-2021. Source: LFS 2018-2021. 25 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia Fewer than 15 percent of men and 10 percent of Among older female workers, the most common women in the workforce are employed past the age occupations are white-collar and pink-collar jobs, of 65. Blundell et al (2016) found similar rates for whereas blue-collar occupations predominate France in contrast with rates of around 40 percent among older male workers (Figure 34). Higher-skill in the US and the UK. white-collar occupations such as managers, pro- fessionals, and technicians together account for Croatians tend to exit the labor force abruptly around 37 percent of older female workers and rather than gradually. The availability of attractive 30 percent of older male workers. Lower-skilled early retirement conditions, as well the possibility pink-collar occupations in services and sales or in of working informally, may explain the prevalence clerical support together account for 37 percent of of abrupt retirement patterns in Croatia. Alterna- older female workers and 15 percent of older male tively, perhaps Croatian employers are unwilling workers. Older men are also heavily represented in or unable to accommodate part-time work, thus blue-collar occupations such as crafts and related pushing workers to go “cold turkey” in retiring early. trades and plant and machine operators (37 percent). This explanation is consistent with the fact that self-employment rates increase as retirement age Women in high-skilled occupations are much less nears. In 2019, older workers were twice as likely likely to exit the labor force early than those in to be self-employed than prime-age workers (20.4 pink-collar occupations. The share of older women percent versus 9.8 percent). Note though that a in professional occupations increases sharply with similar increase was observed for the EU-27 group age (17 percent among those aged 50 to 55 versus in general (21.3 percent versus 13.3 percent).32 The 26 percent among those aged 60 to 65), while number of hours worked declines only slightly after the share of lower-skilled, pink-collar occupations age 60 (Figure 33, top). Consistent with this, part- declines from 25 percent among those aged 50 time working is rare before the age of 60 and is still to 55 to 14 percent among those aged 60 to 65. unusual at age 60 to 65, with the share of part-time In contrast, the occupational composition of older workers increasing to 12 percent for that age group male workers remains stable between the ages (Figure 33, bottom). By way of comparison, part- of 50 and 65. This indicates that higher levels of time working at age 65 in France, the UK, and the human capital, and possibly higher associated earn- US was around 40 percent for males and 60 per- ings, reduce early retirement among women but not cent for females in 2012 (Blundell et al, 2016). among men. Figure 34: Shares of male and female older workers employed in different occupations, by age group Armed forces occupations Clerical support Craft and related trades Elementary occupations 25 20 15 10 5 0 50-54 55-59 60-64 50-54 55-59 60-64 50-54 55-59 60-64 50-54 55-59 60-64 Managers Plant and machine operators Professionals Service and sales Employment rate, % 25 20 15 10 5 0 50-54 55-59 60-64 50-54 55-59 60-64 50-54 55-59 60-64 50-54 55-59 60-64 Skilled agricultural forestry Tech. and assoc. professionals and fishery 25 20 15 10 5 0 50-54 55-59 60-64 50-54 55-59 60-64 WOMEN MEN Source: LFS 2021. 32 Eurostat 26 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia Are Retirement Patterns Different in Croatia? What drives older Croatian workers to exit the labor force earlier than their European peers? We draw on data from two waves of the Survey of Health, Aging, and Retirement in Europe (SHARE) to identify the fac- tors associated with being employed at older ages in Croatia compared to other European countries. We estimated the descriptive regression model described in Box 1 using data from wave 7 of SHARE for Cro- atia, Poland, the Czech Republic, Slovenia, and Austria for men and women together and for each gender separately. We interacted the variables with a Croatia dummy to detect effects that could explain the lower employment rates in Croatia as opposed to its neighbors. Our results are presented in Table 8. Box 1: Employment at older ages, a framework As a simple conceptual framework for the analysis presented in this section, consider older worker i deciding whether to work for one more year t. This worker compares the best wage offer available to them wit, and a “reservation wage”, or wage threshold wit R , which captures the relative utility obtained by not working/retiring. Workit = 1 ↔ wit > wit R Workers with higher earnings opportunities on the labor market can be expected to retire later. Earning opportunities may decline with age due to skill obsolescence, prejudices against older workers, or institutional incentives to let them go. We assume wage offers wit reflect a worker’s human capital and current labor market conditions in a classic linear Mincerian specification, denoting as xit and εit observable and unobservable characteristics capturing a worker´s skills and local labor market conditions. wit = Xit ' βt + εit The variables in our data that capture skills and labor demand include age, gender, urban/rural location, college, medium and high computer literacy, work experience, fraction of years employed since age 25, and previous employment sector (self-employed, civil sector, or private sector). On the other side of the equation, a worker’s reservation wage or pickiness will be affected by the availability of other sources of income (primarily pension benefits but also rental income, for example) or the ability to spend time caring for family members. Workers may also care about the option value of not retiring if institutional or economic factors make it difficult to return to the labor force after an interruption. For example, Croatia only recently allowed workers to combine labor income and pension benefits and only if they work less than four hours daily. Denoting as Zit and εit R observed and unobserved determinants of the reservation wage, we can express it as: wit R = Zit βt + εit ' R R Variables that could affect the reservation wage include eligibility for early retirement, being a homeowner, owning financial assets, having a retired or working spouse, and living near one’s children. Factors related to health may affect both earning opportunities (through reduced productivity) and disutility of work and would therefore be included in both equations. We measure chronic diseases, limitations in activities of daily living, and life satisfaction (CASP-19 index). Combining equations (1), (2) and (3) and assuming normal errors in the earnings and reservation wage equations yields the following probit model of whether individual i at time t with characteristics X_it and Z_it is observed to work: p(Workit = 1) = p(uit > [Xit│Zit ]' γt ) = Φ([Xit│Zit ]' γt / σu ) where γt = [-βt ,βtR ] , uit = εit – εit R with variance σu 2 , and Φ is the cdf of a standardized normal distribution. 27 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia Croatia is detectably different from its neighbors the opposite effect. The effect of care obligations in three determinants of whether older workers (as measured by the presence or geographical prox- are employed. First, public sector workers in Croa- imity of children in the household on employment at tia are less likely to work at older ages than in other old age, was not detected in the survey data. Inter- countries. Second, life satisfaction as measured by estingly, homeownership and ownership of financial the CASP-19 index has a stronger positive associ- assets are positively associated with employment. ation with work in Croatia than elsewhere (in other This could be because more productive individuals words, employed older workers in Croatia report will tend to accumulate more assets, which would having greater life satisfaction than their counter- offset the classic negative effect of wealth on labor parts in the other countries). One interpretation is supply. that the availability of early retirement in Croatia means that only those who have fulfilling jobs stay The standalone dummy variable for Croatia also in the labor market longer. A second interpretation has a significantly negative coefficient, indicating is that replacement rates in Croatia are low, which that other unmeasured factors explain why Cro- could result in relatively low levels of life satisfaction atia’s employment rates are lower at older ages. among Croatian retirees as opposed to those who For instance, participation conditional on health continue to work. Lastly, homeowners and individ- or education may be the same in Croatia and its uals with financial assets are more likely to retire in neighbors, but the stocks of health and education Croatia than elsewhere. This could again reflect the in their respective populations could be different. fact that complementary resources are needed to Our model also does not capture differences in retire comfortably given the low overall level of Cro- incentives for early retirement or in labor demand atian pensions. Owning property may also generate for older workers. We discuss these factors in the significant income in regions with tourism activity, following sections. allowing some older individuals to supplement their modest pensions. The Role of the Pension System In most other dimensions, the results were consist- ent with theoretical predictions and were compa- A large number of studies has documented ways rable in Croatia and in the other countries. Higher in which pension rules affect work and saving human capital tends to be associated with a higher decisions (Blundell et al, 2016). This is not surpris- probability of employment in old age, specifically ing considering that the rules of a pension system having a college degree, good computer skills, and regulate a person’s access to much of their lifetime good health. Retirement is associated with having resources. Most studies confirm the hypothesis that chronic diseases and with low quality of life. Limita- the availability of pensions are the primary determi- tions of daily activities are not significant predictors nants of the likelihood of retirement and the quan- of employment, but they are still rare among those tity of work of those who continue to work (Coile aged 50 to 65. Note that the variable “experience,” and Gruber, 2001). This section discusses whether which measures an individual’s lifetime number of the rules of Croatia’s pension system could explain years of employment, has a negative correlation the low labor force participation of older workers. with employment. Experience is traditionally inter- preted as human capital, which would normally Basic Features of the Croatian Pension command a positive coefficient, but here it also System captures how early an individual started working, which could increase their pension rights or their Until the early 2000s, Croatia had a pay-as-you-go disutility of work. Workers with stronger previous pension system offering defined benefits. How- attachment to the labor force also have higher ever, as contributions collected through a 20 per- employment rates after age 50. This is captured in cent payroll tax failed to cover pension outlays, the the variable “percentage employed since 25,” which system became a major source of deficits for the computes the fraction of years since the age of 25 central government. The system has been reformed in which the individual was employed. It measures several times over the years to improve its financial a person’s attachment to the labor force through- outlook and introduce an element of capitalization out their lifecycle, which can indicate both more (both mandatory and voluntary). Some features of skills and greater employability as well as a relatively past reforms are still being gradually phased in so low disutility of work. Furthermore, the data sug- their effects on employment at older ages are only gest that couples like to retire together with having partially reflected in the data. a retired spouse reducing a person’s probability of being employed, while having a working spouse has 28 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia Table 8: Probit model of employment among 50-65 in Croatia vs. select neighbouring countries (Poland, the Czech Republic, Slovenia and Austria) VARIABLES (1) Overall (2) Male (3) Female Age -0.137*** (0.0143) -0.113*** (0.0241) -0.146*** (0.0179) Female -0.306*** (0.0520) Marital status 1. Employed spouse 0.055 (0.0725) 0.069 (0.1180) 0.004 (0.0960) 2. Not employed spouse -0.124** (0.0612) -0.110 (0.1080) -0.174** (0.0764) Child(ren) living nearby (1km) -0.056 (0.0681) -0.185* (0.1080) 0.045 (0.0897) Urban 0.183*** (0.0625) 0.221** (0.1020) 0.157* (0.0808) College degree 0.144* (0.0872) 0.162 (0.1390) 0.113 (0.1140) Highschool education 0.050 (0.0636) 0.204* (0.1050) -0.028 (0.0824) Computer skill 0.116*** (0.0205) 0.088*** (0.0315) 0.148*** (0.0278) Experience -0.052*** (0.0085) -0.059*** (0.0137) -0.050*** (0.0110) % of time employed since 25 3.887*** (0.3190) 5.193*** (0.5470) 3.421*** (0.4050) Working sector 1. Self-employed 0.140* (0.0786) 0.195* (0.1170) -0.050 (0.1080) 2. Civil sector -0.030 (0.0565) -0.100 (0.0901) 0.141* (0.0746) Age greater than 60 -0.238*** (0.0881) -0.282* (0.1480) -0.271** (0.1120) Number of limitations with ADL -0.023 (0.0561) -0.066 (0.0912) 0.013 (0.0741) Number of chronic diseases -0.082*** (0.0205) -0.086*** (0.0330) -0.069*** (0.0266) CASP-19 index 0.009* (0.0052) 0.005 (0.0083) 0.011 (0.0069) Self-reported health status 0.096*** (0.0303) 0.101** (0.0481) 0.104*** (0.0398) Homeowner 0.128** (0.0603) 0.160 (0.0973) 0.120 (0.0786) Own financial assets 0.240*** (0.0508) 0.230*** (0.0806) 0.269*** (0.0669) Croatia -3.548* (1.8750) -7.306** (3.2840) -1.073 (2.4560) Slovenia -0.361*** (0.0797) -0.168 (0.1230) -0.513*** (0.1080) Poland 0.077 (0.0959) 0.326** (0.1530) -0.122 (0.1270) Czech Republic 0.431*** (0.0803) 0.770*** (0.1310) 0.216** (0.1050) Croatia*Age 0.036 (0.0341) 0.067 (0.0581) 0.010 (0.0450) Croatia*Female 0.200 (0.1280) Croatia*Marital status 1. Croatia*Employed spouse 0.183 (0.2150) 0.251 (0.3670) 0.142 (0.2900) 2. Croatia*Not employed spouse 0.171 (0.1970) 0.266 (0.3500) -0.002 (0.2580) Croatia*Child(ren) living nearby (1km) 0.109 (0.1530) 0.277 (0.2300) -0.087 (0.2300) Croatia*Urban -0.135 (0.1510) 0.158 (0.2320) -0.397* (0.2160) Croatia*College degree 0.183 (0.2400) -0.103 (0.4080) 0.472 (0.3190) Croatia*Highschool education -0.040 (0.1730) -0.245 (0.2890) 0.054 (0.2430) Croatia*Computer skill -0.050 (0.0459) -0.035 (0.0701) -0.011 (0.0674) Croatia*Experience -0.004 (0.0235) 0.039 (0.0383) -0.039 (0.0333) Croatia*% of time employed since 25 0.887 (0.8650) 0.276 (1.4190) 1.715 (1.2370) Croatia*Sector 1. Croatia*Self-employed -0.222 (0.2330) -0.273 (0.3710) -0.513 (0.3390) 2. Croatia*Civil sector -0.277** (0.1350) -0.143 (0.2040) -0.375* (0.1980) Croatia*Age greater than 60 0.214 (0.2240) -0.400 (0.3650) 0.592* (0.3130) Croatia*Number of limitations with ADL 0.071 (0.1200) 0.165 (0.1720) -0.023 (0.1970) Croatia*Number of chronic diseases -0.062 (0.0598) -0.0383 (0.0926) -0.091 (0.0868) Croatia*CASP-19 index 0.027** (0.0126) 0.059*** (0.0198) -0.000 (0.0183) Croatia*Self-reported health status -0.085 (0.0686) -0.039 (0.1040) -0.083 (0.0989) Croatia*Homeowner -0.333* (0.1920) -0.447 (0.2860) -0.178 (0.2770) Croatia*Own financial assets -0.277* (0.1670) -0.395 (0.2700) -0.250 (0.2300) Constant 5.950*** (0.8420) 3.629** (1.4130) 6.789*** (1.0710) χ2 2554.86 1055.47 1584.57 LL -2094.20 -852.29 -1185.84 Observations 4,896 1,991 2,905 Probit model estimated on all age 50-65 SHARE respondents in Croatia, Poland, Czech Republic, Slovenia, and Austria. Dependent variable is whether the individual is currently employed. Austria is used as a referent country. The referent categories for “Marital status” is “Not married”, and for “Working sector” is “Private sector”. “CASP-19 index” measures the quality of life in older age. Higher values of “Computer skill” and “Self-reported health status” represent better skill/health. (Standard errors in parentheses); *** p<0.01, ** p<0.05, * p<0.1 29 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia In 1998, the government introduced paramet- to the system for 41 years; (ii) to gradually raise the ric reforms to the pay-as-you-go system in the legal retirement age to 67 years by 2038; and (iii) so-called “small pension reform” to reduce out- to allow old-age pension beneficiaries to work part- lays. This reform raised the legal retirement age to time without losing their pension rights. In Septem- 60 and 65 years for women and men respectively ber 2019, facing opposition from trade unions, the over the following 10 years. The early claiming age Croatian government annulled the increase in the was also raised to 55 for women and 60 for men statutory retirement age to 67 and reduced the respectively, with an actuarial reduction factor of early retirement penalty. 0.34 percent per month. The formula for calculating pension benefits was revised to take an individual’s According to the current Pension Insurance Act, a entire work history into account, instead of the ten statutory early retirement pension is available up years with the highest earnings. Pensions were no to five years prior to reaching the standard retire- longer indexed solely on wage growth but also on ment age. However, early retirees receive a pension a weighted average of wage growth and inflation that is permanently lower by 0.2 percent per month (“Swiss formula”). Lastly, the eligibility criteria for (or 2.4 percent per year) than the pensions of those disability pensions became more stringent. who work until their full retirement age. To be eli- gible for early retirement, men must have worked Simultaneously, a mandatory and a voluntary pillar, for a minimum of 35 years, while women must both based on privately managed individual cap- have worked for 33 years, with this being gradually italization, were added to the system through the increased to 35 years by 2030. These requirements 1998 Pension Insurance Act. For pension contribu- are relatively generous by international standards. tors under the age of 40 at the time of the reform, a If their employer goes bankrupt, an affiliate can ful- quarter of the 20 percent payroll tax (5 percent) was fill the statutory early retirement conditions when diverted to the mandatory capitalization scheme. they have been unemployed for at least two years The switch was optional for those aged 40 to continuously prior to retirement with no reduction 50, while those older than 50 remained in the old of pension benefits. The old-age pension is also system. available to those who reach 60 years of age (61 years from 2027) and 41 years of service without The Law on Homeland War Veterans and their any reduction in benefits. Other than Croatia, this Family Members (Official Gazette no. 121/2017 and option is offered by another six EU countries, but its amendments) provides specific disability and their retirement age and years-of-service require- survivor pensions for some groups, mostly veter- ments are much more stringent (World Bank, 2019). ans of the 1990-1996 war. Croatian Homeland War veterans are entitled to a pension of a minimum 45 Despite a low average replacement rate, the pen- percent of the average net wage, provided that they sion system remains relatively costly with outlays served in military units for at least 100 days, with just under 12 percent of GDP. While Croatia’s public this percentage increasing with the length of their expenditure on pensions is in line with the EU-27 service (World Bank, 2019). These groups accounted average (World Bank, 2021a), the net replacement for around 14 percent of all Croatian pensioners in rate for men (the average pension as a percentage 2019. These provisions could explain some of Croa- of pre-retirement earnings) is 53 percent, notice- tia’s old age employment gap and are likely to con- ably lower than the EU-27 average of 68 percent tinue to result in early exits from the labor markets as of 2018.33 High early retirement rates and the by direct beneficiaries and their dependents for long duration retirement explain Croatia’s relatively some time. high outlays on the pension system compared to it’s the EU-27 average. Public pension spending The government has made smaller parametric is expected to fall by at least 2 percentage points amendments to the pension system since 1998 to of GDP over the next 50 years, and the European increase the adequacy of pension benefits. In 2007, Commission has classified Croatia as low fiscal risk a pension supplement of up to 27 percent was for pension spending in both the medium and long introduced to offset some of the reductions in pen- terms (European Commission, 2019). However, this sion benefits associated with the 1998 parametric is because the real value of pension benefits are reform. In 2010, the Constitutional Court introduced on a path to decrease over time, raising questions a transitional period until 2030 for equalizing the about the adequacy of pension payments in the retirement age for women and men. In 2014, several future (World Bank, 2019) further changes were introduced, the most signifi- cant being: (i) to allow individuals to retire at the age Already, elderly Croatians are much more vulner- of 60 with a full pension if they have contributed able to poverty than the rest of the population, https://data.oecd.org/pension/net-pension-replacement-rates.html 33 30 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia Figure 35: Expenditures on pensions, by country 2020 18 16 14 (%, relative to GDP) 12 10 8 6 4 2 0 EU EL IT FR AT PT ES FI BE DK NL DE PL SE HR SI LU CY CZ RO SK EE BG LV LT HU MT IE CH IS NO RS BA ME MK TR Source: Eurostat (online data code: spr_exp_pens). unlike in most other European countries. While the retirement ages. The distribution of retirement ages fraction of people at risk of poverty or social exclu- among current retirees shows a mass of points at sion in 2022 was similar to the EU-27 average for all the earliest ages at which early and regular retire- individuals over 18 years of age (21.6 percent), the ment are now or were previously allowed (Figure indicator for Croatians aged 65 and over was much 36). For men, the mass points occur at 55, 60, and higher at 32.1 percent. This contrasts with the EU-27 65, corresponding to the early and standard retire- where the elderly are no more at risk than the rest ment ages that current pensioners faced when they of the population. The economic vulnerability of the approached retirement. For the same reason, the elderly and the adequacy of their pensions remain distribution of women’s retirement ages shows a major challenges for policymakers. As the aging of mass of points at ages 50, 55, 60, and 65, reflecting the population continues to tighten Croatia’s fiscal previous and current legal requirements. Individu- space, extending the duration of Croatians’ working als can retire prior to those ages by qualifying for a lives appears to be one of the few remaining strate- disability pension. The fact that many workers exit gies for addressing these issues. the labor force as soon as they are allowed to do so is not specific to Croatia. Statutory retirement How strong are the incentives to retire ages have been shown have this “binding” effect in early? many contexts (Rust and Phelan, 1997), reflecting less-than-fair actuarial adjustment factors, liquidity The rules of the pension system regarding the constraints, or present bias on the part of retirees. legal retirement ages as well as disability pension eligibility strongly shape the timing of Croatians’ Disability pensions used to be a pathway into early decisions about when to retire. Croatians tend to retirement for many people in Croatia, but reforms retire when they reach either the early or statutory over the last decade have reduced this number Figure 36: Distribution of retirement ages in Croatia, by gender Females Males 16% 16% 14% 14% % of retired males % of retired males 12% 12% 10% 10% 8% 8% 6% 6% 4% 4% 2% 2% 0% 0% 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 Retirement age Retirement age Source: SHARE Wave 6, Börsch-Supan (2022). Survey of Health, Ageing and Retirement in Europe (SHARE) Wave 6. Release version: 8.0.0. SHARE-ERIC. Data set. DOI: 10.6103/SHARE.w6.800. 31 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia Table 9: Example of present value of retirement benefits for early and full age pension Years receiving a Sum of benefits PVof benefits Type of pension Monthly pension Annual pension pension received (discounted at 3%)34 HRK 2,712 HRK 32,548 HRK 846,251 HRK 581,855 Early retirement 26 (EUR 360) (EUR 4,322) (EUR 112,370) (EUR 77,263) HRK 3,082 HRK 36,987 HRK 776,717 HRK 491,821 Normal retirement 21 (EUR 409) (EUR 4,912) (EUR 103,143) (EUR 65,308) Source: the author's calculation according to the data by the Croatian Pension Insurance Institute. closer to EU-27 averages. In 2011, disability pen- until the statutory age to retire. Furthermore, the sions accounted for 25 percent of total pension present value of the early retirement pension ben- outlays in Croatia, compared to only 15 percent in efits, at the time when the early decision is made, is the rest of Europe. Homeland war veterans on disa- 18 percent higher than that of the statutory retire- bility pensions accounted for some but not most of ment pension benefits. the gap (Badjun, 2017). The disability claim system has been reformed over the last decade, with the Econometric estimates available in the literature medical screening procedure being improved, the suggest that making adjustment factors "actu- introduction of a stricter assessment of disability, arially fair" can have a large behavioral and finan- and more frequent reassessments. The govern- cial impact (Gruber and Wise, 2003). For example, ment established the Single Body of Expertise in in Germany, introducing (almost) actuarially fair 2015 to unify disability assessments and to reduce adjustments (6 percent for each year of delaying fraud. The Independent Review Sector, a special retirement) would increase the average retirement unit within the Ministry of Labor and the Pension age of German men by about three years and two System, then reviews positive assessments passed months, with effects of about half that size for on by the Single Body of Expertise (Badjun, 2017). women (Gruber and Wise, 2003). Estimates show As a result of these reforms, Croatia’s disability pen- that savings of approximately 1.2 percent of GDP sions as a percentage of all pensions had nearly would be achieved in Germany (Gruber and Wise, converged with the EU-27 average of 18 percent in 2010). Cross-country evidence also exists that 2020 according to Eurostat data. shows implicit pension taxes increasing employ- ment rates in old age. In France, the Netherlands, At 0.2 percent per month (or 2.4 percent per year), Belgium, and Italy, the high level of unused produc- the permanent reduction of Croatia’s pension for tion capacity of the elderly population (between early retirees is much less than is actuarially fair 60 and 70 percent for people aged 55 to 65) can and therefore implicitly taxes workers who con- be linked to high implicit taxes from the pension tinue to work beyond the early retirement age. The system in these countries. Conversely, countries maximum early retirement penalty is 12 percent with relatively low implicit taxes such as Japan, the over five years, which is very low by international USA, Sweden, Canada, and Spain also have signifi- standards. Many other countries deduct around 0.5 cantly lower levels of unused productive capacity of percent per month (or 6 percent per year) from the early retirement pension, which experts consider Table 10: Average effective retirement age of new to be closer to being actuarially fair (Börsh-Supan old-age retirees, Croatia et al, 2016). Thus, the pension benefit increase for Year Male Female Total a Croatian who delays retirement and contributes for one more year is relatively small, given that they 2006 63 02 59 00 61 04 will contribute for longer, draw their pension later, 2007 63 01 58 01 61 02 and for a shorter period on average than the early 2008 64 01 60 05 62 03 retirees. For instance, in Table 9, we consider an indi- 2009 64 01 60 03 62 01 vidual deciding whether to retire early. We assume 2010 63 02 59 06 61 02 that, if he retired at the statutory retirement age, 2018 63 05 61 04 62 06 he would benefit from the average net pension for 2020 63 07 62 00 62 11 new beneficiaries in 2019, which amounted to HRK 2022 August 63 08 62 06 63 02 3,082 (EUR 409). If claimed he were to retire five Source: Hrvatski zavod za mirovinsko osiguranje: Statističke years earlier, his pension benefit level would be 12 informacije Hrvatskog zavoda za mirovinsko osiguranje, percent lower, or HRK 2,712 (EUR 360). However, various years, for 2018, Table 19d, page 41, for 2020, No. 8, Table 19d, page 98, Available on https://www.mirovinsko.hr/ he would receive that pension for five more years UserDocsImages/statistika/statisticke-informacije/2020/12/ so that his total payments would amount to HRK Statisticke-informacije-HZMO-a-12-2020-sijecanj-2021.pdf for 2022. No. 8, Table 19d, page 104. https://www.mirovinsko.hr/ 69,535 or EUR 9,234 more than if he had waited UserDocsImages/listalice/mediji/statistika_kolovoz/index.html PV(receiving 32,548 from t=1 to t=26)=32,548*((a-a^27)/1-a), where a=1/1.03 34 32 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia the elderly population (just over 20 percent in Japan the requirements in other European countries that to below 50 percent in Spain). offer such an option. This option is designed to ben- efit individuals who started working early, often in With adjustment factors favoring retiring as early physically demanding occupations that are associ- as allowed, the average effective retirement of ated with lower longevity at retirement. Therefore, it women over the past decades reflects the impact should be recalibrated carefully so that the option of successive increases in their early and statutory remains available to more disadvantaged groups retirement ages (Table 10). Until 2010, the average but not to others. effective retirement ages of new male and female old-age retirees were similarly stable, except for a What are the incentives to delay temporary increase around 2008. Since 2010, when retirement past the age of 65? retirement eligibility conditions for women began to be gradually aligned with those of men, women’s Due to differences in longevity and the low and retirement ages have increased steadily whereas declining level of pension replacement rates, men’s have remained largely constant. As another focusing only on penalizing early retirements could piece of evidence of later retirement of women, be inequitable and face political opposition. Early whereas the employment rates of women born retirement is more common among those with before 1963 declined steeply at the age of 55, they lower education levels. Low education is typically remained stable past that age for younger cohorts associated with shorter life expectancy, more stren- who are subject to an older early retirement age uous lifetime employment, and worse labor market (Figure 37). In contrast, the age profile of employ- opportunities in old age. After studying data on 23 ment rates for men did not change between suc- OECD countries, Fabrice Murtin et al (2017) found cessive cohorts. a gap in life expectancy between highly educated and poorly educated people of 3.5 years for men Reducing the availability of and incentives for early and 2.5 years for women at the age of 65. Longevity retirement is a powerful way to extend working inequality can even offset the redistribution to lower lives, mitigating the negative impact of population earners embedded in pension benefit formulas. aging on growth and pension finances without fur- When considering the present value of future pen- ther endangering pension adequacy. One option sion benefits, and after accounting for differences in for doing this might be to shorten the gap between life expectancy, low earners may be at a disadvan- the early and statutory retirement ages to three tage despite higher replacement rates. Such con- years instead of five. A second possibility might be siderations may provoke strong political resistance to increase the early retirement adjustment factor to uniformly extending working lives. to an actuarially fair level. To avoid further eroding low pension levels, this should be accompanied by Allowing and encouraging people to delay retire- an increase in pension benefits for lower earners. ment past the age of 65 for those willing and able A third possibility might be to increase the number to do so might be a more equitable option. Employ- of contribution years needed to retire early without ment rates after age 65 for both men and women incurring a pension benefit adjustment. The current are currently very low in Croatia. People who are requirement is 41 years, which is low compared to over the eligibility age for an old-age pension are Figure 37: Age profiles of employment rate (%) of men and women, by birth cohort WOMEN MEN 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 1954-1958 1959-1963 1954-1958 1959-1963 1964-1968 1969-1971 1964-1968 1969-1971 Source: LFS 2018-2021. 33 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia allowed to defer retirement as long as they have at Croatians past the age of 65. This provision could least 35 years of contributions. Croatia has a pen- help employers to retain workers with crucial skills sion benefit bonus for deferred retirement, and as on a part-time basis while the workers themselves of January 2023, this increased from 0.34 percent can supplement their pension income and exit the to 0.45 percent for every month of deferment, up to labor force gradually. According to data from the a maximum of 27 percent over five years. This puts Croatia Pension Institute,38 as of the end of June Croatia in the middle of the range of OECD coun- 2022, there were 19,767 retired people (of whom tries but remains less than is actuarially fair, mean- 11,903 or 60.2 percent were male and 7,864 or 39.8 ing that it penalizes delaying retirement. percent were female) who worked up to four hours per day without losing their pension rights. Most of The age of 65 has long constituted the expected them (3,562 or 18.0 percent) worked in wholesale age for retiring and has enshrined in legislation and retail trade or the repair of motor vehicles and and has been reinforced by disincentives and legal motorcycles, followed by those in professional, sci- restrictions on working longer that have been par- entific, and technical activities (2,930 or 14.8 per- tially rescinded in recent years. For example, stay- cent) and the manufacturing sector (2,586 or 13.1 ing employed after 65 used to be disincentivized percent). These three sectors tend to experience because the responsibility for paying sick leave to recurring labor force shortages, as identified by the workers over 65 was entirely borne by the employer Employers’ Survey conducted in 2020 by the Cro- instead of being taken over by the Health Insur- atian Employment Services39 and employ almost ance Institute after 42 days as for workers under half (45.6 percent) of all retired people.40 However, 65. However, the latest amendments to the Law the same reasons that make part-time work rare in on Compulsory Health Insurance raised that age Croatia could limit the overall effect of this reform. threshold to 70. Croatian law also explicitly restricts Therefore, an alternative might be to allow workers the employment of older workers in the public past retirement age to combine full-time work with sector. Article 137 of the Act on Civil Servants and the receipt of partial pension benefits. Employees35 explicitly stipulates that a civil servant must terminate his or her civil service when he or she reaches 65 years old and has contributed for at least Understanding Other Causes of 15 years, unless otherwise agreed by the service. A the Old Age Employment Gap similar regulation is in the Act on Civil Servants and Employees in local and regional self-government.36 Declining Health Due to a serious shortage of medical staff, Article 67 of the Health Care Act states that health workers Bad health is one of the main causes of early exits in the public health service network have the legal from the labor force around the world, either right to perform private practice until they reach 65 through early retirement or by claiming disability years old. This is slightly mitigated by paragraph 3, insurance (Blundell et al, 2016). Declining health according to which “if the provision of health care may directly reduce a worker’s inclination to con- is endangered, the Minister may approve an exten- tinue to work, for example, because of fatigue or sion to the right of health workers to work in private painful movements. It can also make workers less practice until they reach 70 years old." productive, resulting in lower productivity and earnings. Also, if health conditions reduce a work- Allowing individuals to continue in wage work while er’s expected longevity then this may also reduce drawing a pension might enable skilled and pro- the probability of them outliving their savings, thus ductive workers to stay in employment for longer making it optimal to retire earlier. Lastly, disability without causing harm to those with less favorable insurance programs often preclude beneficiaries employment and longevity prospects. A promising from working while receiving benefits. development is that, since the latest changes to the Pension Insurance Act,37 full-time pensioners can However, individuals with health issues are not sig- now work up to four hours per day without losing nificantly more likely to exit the labor force in Cro- their pension rights. Whether this provision will atia than in other countries. The regression results increase the number of older people who continue presented earlier in this chapter confirmed that to work is not yet clear. It could incentivize work- health is a significant determinant of employment ers who have retired to continue working part-time, in old age. For example, suffering from a chronic but conversely it might cause workers who would illness has the same effect on employment rates have stayed in the labor force full time to partially as being eight months older. However, the effects retire. However, this second negative effect is prob- of health-related variables are not significantly ably small given the low employment rates among stronger in Croatia than in the other countries that 35 Zakon o državnim službenicima i namještenicima, OG 27/01…61/17. statisticke-informacije/2022/6/Statisticke-informacije-HZMO-a-6- 34 36 Zakon o službenicima i namještenicima u lokalnoj i područnoj 2022-srpanj-2022.pdf?vel=15099745 (regionalnoj) samoupravi - OG 86/08…04/18. 39 Hrvatski zavod za zapošljavanje, 2021. 37 Zakona o mirovinskom osiguranju, OG 157/13… 102/19, 84/21, 119/22. 40 Hrvatski zavod za zapošljavanje (2021). Anketa poslodavaca 38 Hrvatski zavod za mirovinsko osiguranje (2022). Statističke informacije 2020., Hrvatski zavod za zapošljavanje, Zagreb available (in 6, Table 7b, page 21, Hrvatski zavod za mirovinsko osiguranje, Zagreb, Croatian), on https://publikacije.hzz.hr/media/7902/hzz-anketa- available on https://www.mirovinsko.hr/UserDocsImages/statistika/ poslodavaca-2020-020321.pdf. TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia Figure 38: Share of individuals with self-reported ex-post justifications. Therefore, cross-countries poor health, by age group and country comparisons may not reflect objective differences in 45% health conditions. Nevertheless, Croatians’ subjec- 40% tive perceptions of their bad health should be taken % of population with poor 35% into account because they may be as important as self-reported health actual health status in driving individuals’ decisions 30% about working. 25% 20% In addition to perceiving themselves to be in bad 15% health, older Croatians self-report having a lower 10% quality of life compared to the responses given by 5% their peers in other countries. The CASP-19 index is a quality-of-life measure comprising four domains: 0% 50-54 55-59 60-64 65-69 70+ control, autonomy, pleasure, and self-realization Age group (Sims et al, 2011). Croatia’s ranking on the CASP- 19 index is relatively low and drops noticeably with Austria Czech Republic Poland Slovenia Croatia the age of the respondents (Figure 40). In addition, Source: SHARE Wave 6 Börsch-Supan (2022). positive perceptions about quality of life, as meas- we considered. Therefore, this does not explain why ured by the CASP-19 index, are strongly correlated Croatia’s employment rate for older ages is lower with being employed in all countries considered than elsewhere. but significantly more so in Croatia. These observa- tions are consistent with the fact that pensions in Croatians aged between 50 and 65 tend to self-re- Croatia are low and older workers face high risks of port being in worse health than their European poverty and social exclusion. These correlations also neighbors, but objective health measures show appear to rule out the idea that the employment them at about the average. For example, 14 per- rate of older people in Croatia is low because work cent of Croatians aged between 55 and 59 self-re- conditions are poor or because work is not fulfilling ported being in poor health, which is 7 percentage and reduces life satisfaction. It is also possible that points higher than in Slovenia and 10 percentage those who chose to retire early were faced with only points higher than in Austria (Figure 38). However, substandard employment opportunities that would the incidence of limitations to daily activities for all have reduced their life satisfaction even more. age groups in Croatia is about average (7 percent at ages 55 to 59, Figure 39, right). The fraction of Health status and life satisfaction are not sufficient Croatian individuals with at least one chronic ail- to explain the low employment rate for Croatia’s ment is also at the mid-range among the group of older workers. It is notable that the large employ- countries for all ages (66 percent at ages 55 to 59, ment gap persists for healthy older workers in Cro- Figure 39, left). Self-evaluations of health or of qual- atia when compared to its European peers (Figure ity of life can often reflect cultural norms or involve 41).41 If health were a key dimension in which Croatia Figure 39: Share of individuals with at least 1 chronic disease (left) and share of individuals having difficulties with at least 1 activity of daily living (right), by age group and country 100% 30% with at least 1 activity of daily living % of population having di culties 95% % of population with at least 1 25% 90% 85% 20% chronic disease 80% 75% 15% 70% 10% 65% 60% 5% 55% 50% 0% 50-54 55-59 60-64 65-69 70+ 50-54 55-59 60-64 65-69 70+ Age group Age group Austria Czech Republic Poland Slovenia Croatia Austria Czech Republic Poland Slovenia Croatia Source: SHARE Wave 6 (Börsch-Supan, 2022). 41 Defined as healthy if their self-reported health status is "Good” or higher. 35 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia Figure 40: Quality of life in older age, by age Childcare obligations can compound the burden of group and country elderly care, particularly on women near retirement 42 age, who may be expected to care for both young grandchildren and elderly parents. Conversely, 40 when family members do exit the labor force or Value of CASP-19 index reduce the quantity or quality of the labor that they 38 supply, this leads to underusing and underinvest- ing in human capital, which can hamper economic 36 growth. For example, a caregiver may refuse a pro- 34 motion or switch to an occupation that offers more time flexibility but for which they are over-qualified, 32 thus receiving lower wages (Carmichael and Charles, 2003; Heitmueller and Inglis, 2007; and Parker, 30 1990). 50-54 55-59 60-64 65-69 70+ Age group Formal care options have emerged to compensate Austria Czech Republic Poland Slovenia Croatia for the shrinking supply of informal care, but there Source: SHARE Wave 6 (Börsch-Supan, 2022). is a lack of capacity in both public and private child and elderly care in Croatia. In 2019, just 18 percent Figure 41: Fraction of 50-65 y.o. not working of people older than 65 in need of LTCused formal and fraction of healthy 50-65 y.ol not working in home care, and less than 3 percent were cared for Croatia vs. comparators outside the family in organized forms of residential 80% care. The number of requests for nursing homes 70% placements was four times greater than the number of places available in all of Croatia’s nursing homes 60% Employment rate (Badun, 2017). Therefore, a natural question to ask 50% is whether the lack of public or market-provided 40% formal care options is pushing older Croatians out of the labor force. Studies have found that provid- 30% ing informal elderly care is associated with a 4 to 20% 6 percent decrease in labor supply but attempts 10% to make causal estimates have suggested that this effect could be much greater (Norton, 2016; Crespo, 0% 2006; and Heitmueller, 2007). Few papers have Czech Austria Poland Slovenia Croatia Republic examined the effects of introducing or expanding subsidized childcare programs on family members’ Employment rate Healthy employment rate labor supply. Attanasio et al (2022) found a large Source: SHARE Wave 6 (Börsch-Supan 2022). impact (in the region of 20 percentage points) on lags behind its peer countries, the employment gap the labor force participation rates of grandparents, would be much narrower if we considered only 50 though this evidence from a middle-income coun- to 64-year-olds in good health. Instead, employ- try may not be directly comparable to the situation ment rates exhibited almost identical patterns in all in Croatia. the countries in the analysis, whether conditioned on good health or not. Only a small minority of Croatians aged between 50 and 65 provide informal care so frequently that Care Obligations it would be incompatible with participating in the labor force. The extent to which Croatians over the Countries with aging populations such as Croatia age of 50 provide informal care is comparable to are simultaneously facing an increase in the demand other European countries (Figure 42). Around 20 for elderly care and a decline in the supply of tradi- percent of individuals aged over 50 provide unpaid tional informal care. Increases in life expectancy are care at least weekly in Croatia (Tur-Sinai et al, 2020). increasing the number of elderly individuals that live Figure 43 shows the potential effects of providing with chronic illnesses and require varying levels of informal care specifically on the employment of assistance in their daily lives. Lower fertility rates working age individuals aged 50 to 65 who provide lead to increases in female labor participation which care every day. According to this definition, 11 per- implies that fewer family members are available to cent of employed Croatians of this age provide daily provide help to their elderly parents on a daily basis. care while 14 percent of the unemployed provide 36 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia Figure 42: Fraction of individuals over age 50 Figure 44: Providing daily help in Croatia by providing informal care, by country employed and not employed individuals aged Portugal 50-65, by type of help (1) and recipient of help (2) Denmark 35.00 Spain 7% % of (not) employed people 30.00 Belgium Slovenia Employed Not employed 25.00 6% providing daily help 20:00 15.00 5% Luxembourg Poland 10.00 4% 5.00 0.00 3% Croatia Estonia 2% Czech Republic Italy 1% 0% Sweden Greece Personal care Practical Help with household help paperwork Germany Austria Type of help SHARE EHIS EQLS AV Source: Reproduced from Tur-Sinai et al (2020). % of (not) employed people 6% such care, and these percentages are comparable Employed Not employed providing daily help 5% to the shares in Croatia’s neighboring countries. 4% While the frequency with which individuals aged 3% 50 to 65 provide care is comparable between those who are employed and unemployed, the type and 2% recipients of care differ. Employed individuals are 1% less likely to provide personal care and are more likely to help with paperwork. Employed individ- 0% uals are also less likely to provide daily help to a Partner Parents Brother Child Grandchild Others / sister partner. These results indicate that many forms of care, even when provided daily, are compatible with Recipient of daily help employment. Therefore, employment is most likely Source: SHARE Wave 6, weighted (Börsch-Supan 2022). to be hindered by more intensive forms of personal care that cannot easily be shared with other rela- Even in the absence of strong effects on employ- tives. Only 7 percent of unemployed individuals in ment rates, care obligations can have a negative our sample provided daily personal care, which sug- impact on the productivity and human capital use or gests that the impact of care obligations on Croa- investment of older workers. A caregiver may have tia’s employment gap is modest. an incentive to remain employed either to mitigate the psychological toll of providing care, a phenom- enon known as the “respite effect” (Carmichael and Figure 43: Providing daily help by individuals Charles, 2003) or to reduce the share of the burden 50-65, by country and employment status that will fall on them as opposed to other rela- tives. However, their care obligations could affect 20% their ability or desire to invest in advancing their % of (not) employed people career or could lead them to switch to a more flexi- providing daily help 15% ble occupation. Therefore, increasing access to and the availability of formal childcare and elderly care should be an important part of any drive to increase 10% labor force participation in Croatia in response to the population aging. To the extent that infor- 5% mal care provision does not appear to be a major reason for the low labor force participation rates of 0% older workers, it would be useful to explore inno- Czech Slovenia Croatia Austria Poland vative solutions to increasing the compatibility of Republic working while also providing informal care at home. Employed Not employed One example of such a solution is a “cash and coun- Source: SHARE Wave 6, weighted (Börsch-Supan 2022). seling” program, in which informal caregivers receive 37 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia cash transfers and counseling on how to make use the effects of macroeconomic shocks such as the of available care resources in the community. 2008 financial crisis and its ramifications were also apparent in the time-series data presented earlier Lack of Demand for Older Workers in this chapter, suggesting that fluctuations in labor demand can often prevent workers from continuing Even if disincentives to work related to health status in employment in old age, at least in the short run. or pension benefits could be removed, the question remains whether there is enough labor demand in The existence of barriers to the employment of Croatia to absorb any additional older workers who workers past the age of 50 have been well doc- would wish to delay retirement. In the short run, umented around the world. Reports from older older workers will benefit from the demand cre- workers of age discrimination on the job have been ated by the economic conditions and the exit from supported by objective evidence of older workers the labor market of large cohorts of baby boomers. having higher separation rates and longer spells More broadly, the literature considers that labor of unemployment. In addition, audit studies have supply and, particularly, pension regulation are the shown that older jobseekers with otherwise com- two most important factors that explain why old parable resumes receive much lower call-back rates age employment rates differ between countries than younger applicants (Neumark et al, 2019). The (Blundell et al, 2016). This is because, even with skills of some older workers can become obsolete in different labor market institutions, many countries a changing technological environment, but the lit- are still have experiencing the same long-term erature is divided with regard to the extent of this trends, such as the near-universal increase in old issue (Crépon et al, 2003 and Friedberg, 2003). An age employment rates in recent decades. However, alternative perspective is that older workers, being Figure 45: Reasons for retiring (top) and reasons for leaving the last job (bottom), by country 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Became eligible for Own ill health Was o ered an early Other Made redundant public/private pension retirement with bonus Austria Czech Republic Poland Slovenia Croatia Source: SHARE Wave 6 (Börsch-Supan 2022). 80% 70% 60% 50% 40% 30% 20% 10% 0% Voluntary Involuntary Other Austria Czech Republic Poland Slovenia Croatia Source: SHARE Wave 7 (Job Episodes Panel), weighted; SHARE Job Episodes Panel. Release version: 8.0.0. SHARE-ERIC. Data set. DOI:10.6103/SHARE.jep.800. 38 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia more experienced and skilled, can benefit most off or because their plant or office closed). from innovation, but empirically, the evidence on this is also mixed (Hairault et al, 2010, p. 2). A related question is whether older workers are too expensive, prompting firms to substitute them The reasons given by workers for retiring are unu- with prime-age or younger workers. Specifically, the sual in Croatia compared to other countries because wages of older workers may not be aligned with of the widespread practice of agreed layoffs in Cro- their productivity due to efficiency wages or incen- atia (Figure 45). Twenty-five percent of Croatian tive structures based on seniority. Estimating the retirees report having been offered a bonus by their productivity of older workers is challenging, but the employer to agree to retire early, a much higher literature suggests that productivity tends to follow proportion than found in other Eastern European wages closely until the age of 55 but is inconclusive countries. The fact that severance payments to about older ages (Aubert and Crepon, 2003). older workers are exempt from taxation has encour- aged employers to push older workers toward early The age wage gaps, which compare the wages of retirement in exchange for a one-time bonus. This older workers with those of either prime-age or practice was particularly common in large state- younger workers, are not particularly large in Cro- owned companies and in a significant number of atia. Older-aged workers earn roughly the same as privatized companies in the telecommunication and prime-age workers, which puts Croatia close to the financial sectors. Older people have accepted this EU-27 median (Figure 46, bottom). Western and model of dismissal because they often continued to southern European countries tend to have age gaps work in another company or moved in the informal of 10 to 20 percent in favor of older workers com- economy instead. Older workers in Croatia are also pared with prime-age workers, whereas eastern more likely than workers in other countries to report European countries tend to have negative age wage having left their last job involuntarily (by being laid gaps for older workers. Younger workers in Croatia Figure 46: Percentage wage gap between young and prime-age workers (top) and between senior and prime-age workers (bottom), by country Youth over prime-age workers 70 60 Percentage gap, % 50 40 30 20 10 0 R G LV LT EE O SK CZ U PL T SE FI IT SI ES E PT FR K T LU EL CY L IE E N M A B D H D B H R EU-27 Senior over prime-age workers 40 30 Percentage gap, % 20 10 0 -10 -20 -30 R ES E SI IE SE K FI O T U CZ PL SK G LT EE LV EL LU T IT E PT CY FR L N M A B D H D B H R EU-27 Source: Calculations based on Structure of earnings survey 2018, Eurostat. 39 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia Figure 47: Percentage wage gap between Youth and Prime-age workers and between Senior and Prime-age workers, by occupation 70 60 Senior over prime-age 50 Percentage gap, % 40 Youth over prime-age 30 20 10 0 -10 ns rs rs rs rs rs na te ls rs s bl rs ke y er na or tr ke ke ke ke ke ke sio cia tio m to s rs ls ag w es er sio or or or or or or se ra es o pa an ry for of ss w lw w lw tw lw as pe es cu M pr d a s es of ua ua ua nd o he l, or le oc fis ra ad ne Pr sa an an an an pp d ltu ry tr hi M m m d su s an cu ta ac an an d d on en te a al i m r ci ille e ag ic la N em ni ic nd er Sk re rv ch d Cl El ta ille Se d Te an an Sk ft Pl a Cr Source: Calculations based on the Structure of earnings survey 2018, Eurostat. Policy Implications earn around 30 percent less than the two other age An effective strategy for extending work lives and groups on average, but this is again in line with EU-27 increasing pension finances is to reduce the incen- averages. In other words, older workers are not par- tives for early retirement while maintaining the ade- ticularly expensive in Croatia relative to elsewhere. quacy of pension benefits. To ensure fairness, it is advisable that the early retirement adjustment factor Occupations with high education requirements have be increased from 0.2 percent per month (which is larger wage gaps between older and younger work- lower than international standards) at least to a more ers. However, as discussed earlier in this chapter, the reasonable and actuarially fair level around 0.3 per- employment rates of tertiary-educated older work- cent end up to 0.5 percent. To avoid further eroding ers in Croatia are high and in line with EU-27 aver- the already low pension levels, this should be com- ages. The old-age employment rate gap concerns pensated for by increasing pension benefits for lower mostly low-educated workers in Croatia, but the earners. An actuarial adjustment of 0.5 percent per age wage gap is small in low-skill occupations such month implies that a worker working until age 65 as services, sales, and manual jobs. Therefore, the would be 30 percent higher than if he had retired at substitution of younger for older workers solely to 60, instead of 12 percent higher as is currently the minimize the wage bill is unlikely to explain Croatia’s case. This adjustment compensates for the fact that employment gap. early retirees contribute for fewer years but receive their benefits earlier and for a longer time than those Although older Croatians may face non-wage bar- retiring at the statutory age. riers to continuing to work, increases in the early retirement age around the world have typically A more direct way of achieving longer work lives is increased the employment rates of older workers. to shorten the gap between the early and statu- For example, Staubli and Zweimuller (2013) found tory retirement ages to three years instead of five. that a two-year increase in Austria’s early retire- Assuming the statutory retirement age remained 65, ment age raised the employment rate of those who men would then be able to retire early after turning had lost eligibility by around 10 percentage points, 62 instead of 60. Besides sizeable fiscal benefits, this mostly among healthy individuals and those with would lead to higher pensions benefits, though as higher levels of education. However, the employ- discussed in this chapter, Croatia’s system does not ment rates of low-educated or unhealthy individ- currently rewards delaying retirement enough. uals declined instead. This suggests that increasing the early retirement age creates at least some extra It is important to acknowledge that pushing back the demand for older workers. early retirement age impacts low earners much more 40 TOC Chapter 2: Aging and Croatia's Large Employment Gap for Older Workers Harnessing Human Capital for Growth in Croatia than high earners and that it is likely to be perceived work for up to four hours per day without losing their as a welfare cost. This is because more educated pension rights. Whether this provision will increase workers already tend to work until the age of 65 and the number of older people who continue to work is have a higher longevity so that they enjoy their pen- not yet clear. It could incentivize workers who have sion benefits for longer. Academic research also sug- retired to continue working part-time, but conversely gests that some of those forced to delay retirement it might cause workers who would have stayed in the through this proposed measure would switch from labor force full time to partially retire. However, this early retirement to unemployment or disability insur- second negative effect is probably small given the ance because they are unable or unwilling to con- low employment rates among Croatians past the tinue working. Therefore, part of the fiscal benefits age of 65. This provision could help employers to of delaying retirement should be invested into dis- retain workers with crucial skills on a part-time basis positions improving the employment prospects and while the workers themselves are able to supplement conditions for those who remain longer in the labor their pension income and exit the labor force grad- force. This includes allowing more flexibility to adjust ually. However, the unusually low rates of part-time hours worked, improving the safety, ergonomy and work in Croatia could limit the overall effect of this comfort in the workplace for older individuals with measure. The barriers that explain this phenomenon milder disabilities, combatting ageist discrimination, should be investigated to enable workers to exit the and providing specific training focused on allowing labor market more gradually. In addition, the legal older workers to keep contributing their accumulated provisions that prevent people from working past skills in salaried work or in flexible self-employment. the age of 65 or making retirement at 65 a default should be removed. The adjustment in the pension It is also possible to increase the number of contri- level received by those retirees who delay claiming bution years required to retire early without incur- their retirement benefits was recently increased but ring a pension benefit adjustment. The current is still less than is actuarially fair and should therefore requirement is 41 years, which is low compared to be revised upwards significantly. Choosing a level the requirements in other European countries. This that is more than actuarially fair could be justified option is designed to benefit individuals who started because of the country's demographic challenges working early, often in physically demanding occu- and the low levels of benefits offered by the Croatian pations associated with lower longevity at retire- pension system. ment. For example, someone who started working at the age of 19 could claim full benefits when they Besides strengthening institutional incentives to reach 60 or less if they qualify for extended service persuade workers to extend their working lives, a period (hazardous and arduous occupations). Main- broader set of complementary measures is needed taining the possibility of qualifying for a full pension to adapt the work environment to the needs of based on many years contributed appears econom- an older workforce and to shift social and cultural ically fair and can help to soften some of the equity norms. By putting in place an older early retire- concerns related to the two previous reform options. ment age and less rigid retirement requirements, Therefore, this measure should be adopted only after firms should find it economically more worthwhile a careful examination of the types of occupations to invest in older workers’ human capital. However, where workers who started working in their teens are these reforms need to be accompanied by more found. The number of such workers should be eval- age-appropriate training opportunities, employ- uated to determine whether the employment gains ment services adapted to the specific needs of older achieved by this option are sizeable enough. people, and improved working conditions driven by higher health and safety standards. While employers’ Enabling workers to work for wages while also attitudes and practices may adapt endogenously receiving pension benefits can enable skilled and to rising labor market participation by older work- productive individuals who wish to continue work- ers, legislation and public action can help by using ing for longer to do so without negatively affecting public information campaigns and guidelines and the those with less favorable employment and longevity implementation and enforcement of legislation pre- prospects. Workers past the retirement age could be venting age discrimination in employment (Švaljek, allowed to continue working full-time while collect- 2008). Beyond the direct impact of each measure, ing a percentage of their pension benefits. This hap- a coherent, salient, and well-coordinated strategy pens in Slovenia, for example, where such workers to promote the participation of older workers would receive 40 percent of their pension benefits while still help to shift social norms and perceptions of older in employment. A promising development in Croatia people in the workplace. is that according to the latest changes in the Pen- sion Insurance Act,42 full-time pensioners may now Zakona o mirovinskom osiguranju, OG 157/13… 102/19, 84/21, 119/22, 42 41 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Introduction Croatia must prepare for the fast-paced techno- VET, and higher education. This is essential to max- logical and business organizational changes that imize Croatian competitiveness in the global race come with the changing nature of work. Govern- between technology and the education system. This ments worldwide are trying to modify their educa- is especially important for the fast-aging Croatian tion systems to align with the major shifts that will economy, which needs to increase and improve its shape the future of work. As technology evolves, workforce in the face of its shrinking working-age new occupations appear, and the skill mix required population. Therefore, it will be vital to help people of workers constantly changes. Digitalization, auto- to strengthen their human capital over their lifecy- mation, and the green transition mean a steady cles. This chapter outlines the specific skills chal- evolution in how value is added. Employment is lenges faced by Croatia by assessing how human skewing toward knowledge-intensive occupa- capital is built, protected, and used throughout the tions that present countries with new challenges lifecycle, starting with early childhood development, for enabling their workers to acquire the required then childhood and youth, and finally adulthood. skills. Demand for skills is changing rapidly as some jobs are disappearing, such as middle-skilled occu- Box 2: 21st-century skills pations that can be automated. This means that Foundational skills: Basic literacy, numeracy, increasingly workers will need not just the skills and transferable skills, which are the building needed to perform routine cognitive tasks – which blocks for a life of learning. are likely to be displaced by computers (Acemoglu and Autor, 2011) – but also the higher-order skills Cognitive skills: Logical, intuitive, and creative (such as problem-solving and teamwork) necessary thinking, such as verbal ability, numeracy, prob- to perform work in flexible forms of organizations lem-solving, memory, and mental speed. and to adequately react to skill-biased organiza- Socio-emotional skills: Beliefs, personality tional change (Caroli and Van Reenen, 2001). In the traits, and behavioral skills. future, education systems will need to be designed Technical skills: Manual dexterity and use of to nurture a virtuous feedback loop by increasing methods, materials, tools, and instruments, skills for new job opportunities, which in turn will including those related to specific occupations enhance the business environment for more knowl- or trades. edge-based job opportunities. In its World Devel- opment Report 2018, the World Bank proposed Chile is a great example of a country that pri- a set of cognitive, socio-emotional, and technical oritizes 21st-century skills in its education skills that education systems should aim to develop system. These skills are seen as crucial for the in their students to prepare them for the future of comprehensive education of students and are work, including lifelong learning (World Bank, 2018b integrated into all subjects, along with knowl- and Shmis et al, 2021). edge and crosscutting attitudes. The National Curriculum in Chile provides equal opportuni- Many countries have recognized the importance ties for students to pursue their life goals and of this 21st-century skills approach and have incor- interests. porated it into their curricula for preschool, school, Another notable model for 21st-century skills vocational, and tertiary education. Today’s chang- education comes from Portugal. Their curricu- ing nature of work demands that the educational lum emphasizes developing a sense of personal systems develop in their students the skills needed and social identity, valuing diverse forms of for the 21st century (Box 2). While Croatia has initi- communication and expression, and nurturing ated important policy initiatives, especially in early intellectual curiosity and a work ethic, among childhood development and general schooling, other dimensions. more must be done in the fields of adult training, This chapter was prepared by Maria Ustinova and Nithin Umapathi. 42 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia Improving the skills base in Croatia is important as is not high enough to meet the required achieve- the country is currently ranked as "middle achiev- ment level for each occupation in each industry. ing" in the European skills index. Croatia is ranked 12th (out of 31) on the European Skills Index (ESI), The next issue is the mismatch between the skills which has three pillars: (i) skills development; (ii) skills that are needed and those that are available, activation; and (iii) skills matching (CEDEFOP, 2022). which is a major obstacle for business develop- Croatia ranks highly in some sub-areas, such as VET ment. According to the World Bank’s 2019 Enter- participation (3rd) and “early leavers from training prise Business Survey for Croatia skills deficit is the and education” (1st), and its ranking in "computer second most crucial business development chal- skills" has improved significantly from 27th position lenge facing employers after taxation (Figure 48). in 2020 to 15th in 2022. However, the country still A common complaint made by employers in the performs poorly in “recent graduates in employ- survey was that their employees were insufficiently ment” (with a score of 26 out of 100), labor market trained or lacked practical experience. Larger busi- participation of the existing workforce aged 25 to nesses were especially affected by this with 26 per- 54 (with a score of 26 out of 100), and “labor market cent of large Croatian enterprises and 20 percent of participation of the existing youth workforce aged medium enterprises declaring that having an inad- 22 to 24” (with a score of 29 out of 100). Croatia’s equately educated workforce was a critical obsta- performance in skills activation (in which it is ranked cle in the business environment (Figure 48) (World 22nd) is average, which shows that there is scope to Bank, 2019b). Furthermore, Croatian employers improve the transition from education to work. One considered a lack of practical experience to be the area that needs urgent attention is the “qualification second most important reason for labor shortages, mismatch,” in which Croatia ranks 29th, which indi- after the lack of suitable education or specialization cates that Croatian workers’ education attainment (World Bank, 2019b). Figure 48: Key business environments constraints in Croatia, overall and by firm size 60 50 48 40 % of Firms 30 20 16 10 10 5 4 3 3 3 3 2 0 Tax rates Inadequately Tax Political Labor Transportation Courts Practices of Corruption Business educated administration instability regulations the informal licenses and workforce sector permits Croatia2019 Europe & Central Asia Source: 2019 Enterprises Survey, World Bank, 2019. URL: https://www.enterprisesurveys.org/content/dam/enterprisesurveys/ documents/country/Croatia-2019.pdf 43 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia Foundational Skills in Early Figure 49: Participation in ECEC over time – Croatia and EU Childhood: Access and Importance Children age 3 to the starting age of compulsory education at primary level, % of the population of the corresponding age Developing foundational skills during the early 100 years of life is crucial for facilitating effective life- 90 long learning and developing higher-order skills. Early childhood education and care (ECEC) services, 80 such as kindergarten and pre-primary education, 70 provide the basic knowledge and skills that help children to learn later in life. Each layer of the knowl- 60 edge supplied in primary and then in secondary 50 schooling is amplified by the quality of a student’s 2013 2014 2015 2016 2017 2018 2019 2020 foundational skills. This creates a basis for further skills development – “as skills beget skills” over a life- European Union - 27 countries (from 2020) time. Conversely, economic returns to human capi- Croatia tal investment diminish as children age (Heckman, Source: Eurostat (EDUC_UOE_ENRA21). 2006). In other words, early investments can matter more than late investments, and it is not always Although enrollment rates in ECEC in Croatia have possible to remediate early skill deficits thoroughly. increased recently, they still need to be much higher More broadly, early childhood investments raise compared to the EU. In 2020, around 79 percent of the level of a child’s human capital in such a way Croatian children aged 4 to 6 attended ECEC pro- as to increase the productivity of later childhood grams. This is lower than the current 93 percent investments. This kind of self-productivity is crucial average enrollment rate for the EU-27 (Figure 49). for cumulative learning processes such as science, It also lower than the EU’s goal to have 96 percent mathematics, and language, where concepts build of 3 to 6-year-olds enrolled in ECEC by 203043 as upon one another. Those who receive high-quality well as the revised Barcelona targets for childcare.44 ECEC are also more likely to develop higher-order In 2020, participation in ECEC by Croatian children reasoning, teamwork, and problem-solving skills aged between 6 months and 3 years old was also that are increasingly economically valuable (OECD, significantly lower than in the EU-27, only 65 per- 2001). Substantial evidence supports the value of cent of 3-year-olds as opposed to an average of skill investments in early childhood, and research on 88 percent in the EU. Croatian 2-year-olds were an the effects of programs for young children suggests exception, as their 47.2 percent enrollment rate was an extremely high return on investment. According higher than the EU average of 35.2 percent. Also, to Lynch and Vaghul (2015), every US$1 spent on encouragingly, the enrollment rate of Croatian chil- ECEC can deliver an eightfold return on the invest- dren under the age of 2 has been steadily increasing ment in the long term, which is likely to be much and, in 2021 was 16.2 percent, higher than the EU higher for disadvantaged children. Other studies average of 9.4 percent.45 such as Hendren and Sprung-Keyser (2020) have shown that, based on the “marginal value of public To its credit, the government is trying to expand goods,” ECEC “pays for itself” since it improves out- ECEC access, and this is yielding results. Croatian comes for recipients when they become adults. students who benefited from ECEC are achieving Table 11: TIMSS 4th 2019: Students Attended Pre-primary Education, Croatia Attended Attended Attended Subject Did not attend Three years or more Two years One year Percentage Average Percentage Average Percentage Average Percentage Average of students achievements of students achievements of students achievements of students achievements Math 60 520 9 500 14 498 16 494 Percentage Average Percentage Average Percentage Average Percentage Average of students achievements of students achievements of students achievements of students achievements Science 60 533 9 515 14 509 16 515 Source: Mullis et al (2020). 43 https://education.ec.europa.eu/es/education-levels/early-childhood-education-and-care/about-early-childhood-education-and-care 44 44 The European Council adopted the Barcelona objectives in 2002, which included targets for ECEC enrollment of 90% of children from age 3 up to mandatory school age, and an enrollment rate of 33% of children under 3 years old. On average, these have been reached in the EU, and the targets are currently being revised to 96% and 45% respectively. The European Council set two targets, i.e., https://eur-lex.europa. eu/legal-content/EN/ALL/?uri=COM:2022:442:FIN 45 https://ec.europa.eu/eurostat/databrowser/view/EDUC_UOE_ENRP07__custom_6215829/bookmark/table?lang=en&bookmarkId=10 18c285-fa2f-458d-ac97-2413c4159d3d TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia more robust academic results in international edu- around a 21-percentage point difference in attend- cational assessments. For example, the 2019 anal- ance at a pre-primary level between advantaged ysis of data by Matković et al collected for Croatia and disadvantaged Croatian students. This is one of over several cycles of PISA and TIMSS between the largest gaps among 15-year-old students in the 2009 and 2015 showed that longer-term ECEC OECD (OECD, 2018d). There is clearly a need and an attendance contributed to students' cognitive opportunity for policy interventions aimed at clos- achievements in TIMSS and PISA tests, which could ing this gap. not be explained by the education or occupation of their parents, their migration origin, their house- The government has begun to increase the avail- hold composition, or the characteristics of the local ability of ECEC services. By taking a pro-active environment. For example, Croatian students who approach to early childhood development, it aims attended pre-primary education recorded better to reduce disparities in coverage and improve learn- results in the mathematics and science assessments ing outcomes. Its initiatives consist of a mixture of of TIMSS (Trends in Mathematics and Science Study) regulatory changes, infrastructure improvements, 2019 than 4th-grade students who did not receive inclusion efforts, and incentives for private sector any ECEC (Table 11). involvement: Children in Croatia who attended ECEC have had Amending the law to improve the quality of ECEC better learning outcomes in high school educa- provision. In 2022, the government approved tion ranking competitions. According to the latest amendments to the Preschool Education Act to PIRLS 2021 assessment of reading comprehension, increase access to ECEC for 6-month-old chil- Croatian 4th graders who had attended ECEC pro- dren. (European Commission, 2022). grams when they were younger produced better learning outcomes than those who did not. The Improving ECEC infrastructure. Croatia’s National share of students who scored below minimum pro- Recovery and Resilience Plan (NRRP) aims to ficiency was only 1.6 percent, compared with 4.3 increase the supply of ECEC through large-scale percent for students who did not attend such pro- infrastructure investments funded by the EU. This grams. In the most recent PISA in 2018, 15-year-old is expected to create 22,500 additional ECEC students who had attended ECEC outperformed places and increase participation to 90 percent those who did not. The PISA 2018 data also show of children eligible for ECEC, thus helping close that the more years spent in early childhood educa- most of the gap between advantaged and disad- tion, the lower the chances a student had of being vantaged children by 2026. among low performers in PISA in all three subjects: reading, mathematics, and science. However, par- Increasing inclusion through financing initia- ticipation in ECEC in Croatia is closely associated tives. These initiatives include providing subsidies with higher household socioeconomic status (PISA, to socially disadvantaged families and national 2018). According to the PISA data analysis, there is minorities, providing transport to ECEC facilities Figure 50: Duration in early childhood education and care for Croatia PISA-2022 takers and their achievements 520 500 480 460 440 420 400 380 Attended Attended Attended Attended Attended Attended Attended ECEC for less ECEC for at ECEC for at ECEC for at ECEC for at ECEC for at ECEC for at than a year least 1 but less least 2 but less least 3 but less least 4 but less least 5 but less least 6 but less than 2 years than 3 years than 4 years than 5 years than 6 years than 7 years Reading Maths Science Source: Author’s calculation based on PISA 2022 data. 45 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia for children from remote areas, and prioritizing tion, which allows for primary education teachers to infrastructure financing to areas with no ECEC step in if enough ECEC-qualified teachers are not facilities. available. However, these primary education teach- ers will have to pass any requisite exams within two Notwithstanding these initiatives, the lack of years to qualify as ECEC teachers. This measure access to ECEC for disadvantaged children has the potential to increase the number of future requires a more robust policy response. As the ECEC teachers, but the teaching profession in Croa- ECEC expansion policy progresses, it will be cru- tia still needs to become more attractive, according cial to prioritize children from underdeveloped to the OECD’s Teaching and Learning International regions and rural areas (Dobrotić and Matk- Survey (TALIS) (OECD, 2018b).47 Therefore, it will be ović, 2023), low-income families, and minori- necessary to implement additional support, such as ties since these groups have the greatest diffi- wage subsidies and grants for teachers from lag- culty in accessing ECEC (Matković et al, 2019). A ging regions. large variation in ECEC access between different counties in Croatia highlights the extent of the inequality of opportunity around the country. In Elementary and Secondary 2016, the ECEC participation rates for children Schools: Uneven Skill aged 0 to 3 years old ranged from 6 to 40 per- Development cent depending on where they were located, while for children aged between 4 and 6 years Croatia’s elementary education students rank old, the range was even wider at 24 to 83 percent among Europe’s best in reading and math, and (World Bank, 2019d). This is a significant concern, secondary school students excel in skills needed as the gap between children at risk of poverty for a harmonious life, and have some of the highest and social exclusion and more advantaged chil- school completion rates. According to the results of dren is equal to 36 percentage points. This gap is the 2021 Progress in Reading Literacy Study (Mullis higher in Croatia than in any other EU countries et al., 2023), Croatian fourth graders’ reading skills (Education and Training Monitor, 2022). Among are among the best in the ECA region (557 points, children from national minorities, only children compared to a mean score for ECA of 460 points). from Roma backgrounds receive fully funded The share of these children who scored below the ECEC support (a parental fee subsidy and a free minimum proficiency level was only 2 percent, while preschool program for children two years before those scoring at or above the advanced proficiency primary education who did not participate in level was 59 percent. According to TIMSS 2019, Cro- ECEC).46 However, despite this support, only 24 atian 4th-grade students are performing at a level percent of Roma children attend ECEC, and the that is either equal to or higher than their peers in extent to which they participate varies in every other EU countries in both mathematics and sci- municipality (Education and Training Monitor, ence. In fact, Croatia's students increased their 2022). The inequality in access magnifies the risk average score from 490 in 2011 to 509 in 2019. that children who live in low-income families will However, while there was an improvement in their be trapped in life-long poverty themselves. This science scores from 516 in 2011 to 533 in 2015, there is why the returns to investing in ECEC are par- was a slight decrease to 524 in 2019. Furthermore, ticularly high for these groups. according to the PISA 2018 global competence assessment, Croatian secondary school students Further efforts to rapidly increase the supply of are highly skilled in intercultural issues that pro- ECEC teachers are essential to increase access to mote a harmonious life. They achieved impressive and improve the quality of ECEC. While Croatia has scores on par with students from Canada, Hong the capacity needed to train high-quality ECEC Kong (China), Israel, Korea, Latvia, Scotland (UK), educators, the current system is not yet producing Singapore, the Slovak Republic, and Spain. Moreo- enough of them to meet the growing need. Accord- ver, they performed exceptionally well on the cog- ing to the government, about 5,660 new ECEC nitive test that gauged their ability to comprehend educators will be required by 2030 (European Com- and respect the viewpoints of others. This feat mission, 2022). While the current education system was shared by students from Canada, Hong Kong produces around 500 ECEC teachers annually, the (China), Korea, Scotland, and Spain. Furthermore, number of new entrants will still need to be increased Croatia also has low number of early school leavers, to cover the demand (Education and Training Mon- which is a positive indication of social inclusion. As itor, 2019). The government is taking steps to alle- of 2019, it scored high in school completion rates, viate the shortage of educational professionals in its dropout and repetition rates are negligible, and the country through the Act on Preschool Educa- 97 percent of Croatia’s 20 to 24-year-olds reported https://eurydice.eacea.ec.europa.eu/national-education-systems/croatia/access 46 46 The OECD’s Teaching and Learning International Survey (TALIS) is a large international survey that asks teachers and school managers about 47 their working conditions and learning environments every five years. TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia Figure 51: Croatia’s PISA scores by subject (2006-2022) PISA Reading performance PISA Mathematics performance PISA Science performance 500 500 510 490 490 500 480 480 490 470 470 480 460 460 470 450 450 460 440 440 450 430 430 440 2006 2009 2012 2015 2018 2022 2006 2009 2012 2015 2018 2022 2006 2009 2012 2015 2018 2022 Croatia OECD EU Croatia OECD EU Croatia OECD EU Source: Author’s calculations based on PISA database for 2015, 2018, and 2022: Europe and Central Asia. © World Bank, Washington, DC. https://openknowledge.worldbank.org/entities/publication/2d6e2858-07f7-51cc-a2b1-02afec3af085 having completed at least upper-secondary educa- ing (see Figure 51). Between 2012 and 2015, Croatia tion, which is higher than the EU-27 average of 78 saw a sharp decline in science skills, losing its lead in percent.48 what was previously a high-performing subject for the country. For the other major subjects of read- Despite these achievements on standardized ing and math, Croatia had largely maintained its assessments, significant deficits persist in terms underperformance gap since 2006. However, the of students’ comprehension of mathematics and most recent results from 2022 indicate that Croa- science. Only 28 percent of Croatian 4th grade stu- tian children coped with learning disruptions better dents have a conceptual understanding of mathe- than those in other countries. EU scores declined matics, and only 34 percent can apply a conceptual sharply in the aftermath of COVID-induced lock- understanding of science to solving problems. Just downs, whereas Croatian students managed either 4 percent of 4th graders can apply their understand- to maintain or improve in absolute terms and now ing and knowledge of mathematics and science to outperform the EU in each of the three major sub- various complex situations and explain their reason- jects tested. Aside from the averages, Croatia’s PISA ing (Mullis et al, 2020). Furthermore, the test results scores also indicate challenges at the lower bounds. of older students suggest this deficit continues into Gaps in student scores associated with family soci- later grades. oeconomic status are sizeable and in line with the EU average. Most of the worst-performing pupils Before disruptions from the COVID-19 pandemic, come from the lowest socio-economic quintile, a Croatian secondary school pupils scored below the significant risk for the lifetime human capital “scar- OECD and EU averages on international PISA test- ring” effect among disadvantaged groups. Critical Figure 52: Collaborative problem solving, PISA 2015 Averages for age 15 years PISA collaborative problem solving scale (2015) 600 500 400 300 200 100 0 Itaia lo ey ex a Ch o ak Gr ile Li pu e ua lic H Isra a l oa y xe at ly bo a an g Sp ce O ain p al Be lanic lg d ov m or ia us y U Ne Sw tria d rla n U De gd s te m m GeSta k rm tes ew s ny al a n d t d na ia Ko da Ja rea n h rtu D un e n d r Cr gar A wa M bi ni m vi Ze rali Re eec te he de pa ic Fr ur Fi an Es lan Ic ubl t ec Po EC Re g N en Ca on d a Sl iu ni n o th b Co urk Ki n N Au a m ni t e t e L T Lu ov Cz Sl Author’s calculations based on PISA database, 2015. Source Eurostat, indicator: TPS00186. 48 47 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia reforms to improve general primary and second- Table 12: Regional differences in learning ary education include substantially raising the total outcomes, PISA 2015-2018 hours of instruction in line with high-income coun- PISA 2015 PISA 2018 try standards (for some subjects, almost doubling the time) as well as expanding the focus of curricula Slavonia Croatia Slavonia Croatia to cover not only knowledge but critical thinking Reading 465 487 468 479 and problem-solving skills. Math 442 464 451 464 Navigating Skill Acquisition in Science 451 457 459 472 Secondary Schools: Current Source: Ramljak D. et al. Challenges and Policies region achieved lower results in all three PISA test subjects than the Croatian national average (Table While schools in Croatia have a successful track 12). The average result for students from Slavonia record of teaching children mathematics, read- on the State Matura exam in 2019 was 46, com- ing, and writing, the teaching of higher-order skills pared to the Croatian average of 58. Students from needs to be improved. Good educational outcomes Eastern Croatia have consistently achieved worse rely on more learning than is predicted by PISA results than those from other Croatian regions. Aca- reading, mathematics, and science achievement demic research attributes this to socioeconomic gains alone. A growing body of work emphasizes factors such as lower parental education and the the importance of “non-cognitive” or “soft” skills dearth of places in general gymnasium programs in like patience, self-control, conscientiousness, team- the region (Matković and Šabić, 2022). Most of the work, and critical thinking. The PISA 2015 Collab- counties in Croatia that have achieved better out- orative Problem-Solving assessment measures a comes, such as the cities of Zagreb, Varaždin, Istria, person’s ability to solve a problem with two or more and Dubrovnik-Neretva, are also those with the other people. It reflects how skilled students are in lowest at-risk-of-poverty rates. The problem with sharing their understanding with others in search sub-regional inequalities in education is that they of a solution and at pooling their knowledge, effort, usually lead to a widening gap in socioeconomic and skills to solve a problem successfully. In this status between different areas. Areas with better respect, the 15-year-olds in Croatia score low (Figure quality schools and higher educational attain- 52) and lag behind many OECD and EU countries ment will also have better job prospects, higher such as Korea, Estonia, and the Netherlands by 45 incomes, and a better quality of life overall. This can to 65 points, which is the equivalent of almost 1.5 create a cycle of disadvantages for those living in to 2 years of study.49 According to the TALIS study, the lagging areas, as they may struggle to access Croatian teachers do not use instructional meth- high-quality education and be unable to improve ods that require students to evaluate, integrate, their socioeconomic status. Furthermore, if specific and apply knowledge within the context of prob- areas are consistently disadvantaged in education, lem-solving in order to stimulate higher levels of students from those areas may feel marginalized cognition. For example, only 27 percent of teachers and excluded from the benefits of society. in VET schools and 24 percent in non-VET schools reported frequently asking students to decide on Despite Croatia having had the EU’s shortest school their own approaches to solving complex tasks closures during the pandemic, the negative impact (OECD, 2018b). Therefore, it is vital for education on student learning was especially severe for those policymakers to develop and support pedagogical from lower-income households, resulting in sig- practices that could help students to acquire high- nificant loss of human capital. In the World Bank’s er-level foundational, cognitive, and socio-emo- global analysis of COVID-19’s impact on human tional skills. capital, it estimated that the pandemic has caused alarming declines in the cognitive and social-emo- One of the critical risks to educational achievement tional development of young children. It found that is the geographical disparity in educational out- 1 billion children in low- and middle-income coun- comes. According to spatial analysis (Ramljak et al, tries missed at least one year of in-person schooling, 2022), Croatia suffers from substantial regional ine- which could translate into a 25 percent reduction in quality in learning outcomes. For example, the per- their earnings as adults (World Bank, 2023). Croatia formance of students from Slavonia on national and took an approach that minimized school closures by international standardized tests such as the State continuously assessing the epidemiological level of Matura and the PISA test is significantly below the risk in each school locality. This helped to minimize national average. School students from this lagging the need for remote learning and resulted in 98 per- A difference of 30 points in the international PISA scale is equivalent to one year of study. 49 48 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia cent of primary school teaching being delivered in With the support of the World Bank. Croatia is cur- person, while the shares for lower and upper sec- rently introducing the “Whole Day School” (WDS) ondary school were 85 percent and 71 percent. Fur- model to increase instruction time and support thermore, 90 percent of students in the final grades working parents. This fundamental reform, which of secondary education attended in-person classes has started as a pilot supported by the World Bank in their schools. However, despite Croatia’s relatively in 2023 with the expectation of a national roll-out short school closures, the COVID-19 pandemic did in 2028, will introduce the Whole Day School model have a negative impact on the learning process. for all children in grades 1 to 8. The reform will sig- According to a study conducted by the Institute of nificantly increase instruction time by extending the Social Research, teachers confirmed that the aca- school day, thus allowing teachers to implement demic results of around 60 percent of their stu- teaching methods emphasized in the new curric- dents deteriorated and that that the students were ulum that is being introduced in a parallel reform. less efficient in preparing individual assignments Among the many benefits of WDS, it is also expected than in the pre-pandemic period (ReferNet Croa- to increase equity in the system as the benefits of tia, 2021). In addition, about 37 percent of students increasing instructional time are often greatest for indicated in the survey that they felt less prepared students from disadvantaged backgrounds. Signif- to enter the labor market than they would hope to icant infrastructure investment is part of this com- be (ReferNet Croatia, 2021). prehensive reform and is complementary for imple- menting WDS. The number of instructional hours in Croatian schools is the lowest in Europe, which is a major Once implemented, the WDS model could result barrier to skills development. Croatia has just 13.3 in more equal and better opportunities for stu- mandatory hours of instruction per week in grades 1 dents and more efficient use of resources. However, to 4 and 18.7 hours per week in grades 5 to 8 (Figure for this reform to succeed, other measures will be 53). Croatia’s eight-year basic education system needed such as: (i) optimization of the school net- prescribes 525 instruction hours per annum in read- work; (ii) producing more and better data on what ing, writing, and literature. However, this is substan- students know and can do at critical stages of the tially below the 949 hours specified by EU school basic education cycle and using these data to make standards on average. Similarly, in mathematics, the improvements to the system; (iii) strengthening of number of instruction hours in Croatia amounts to education system’s accountability to increase equity only 420, much lower than the European average and reduce inefficiencies; (iv) refurbishing schools to of 628 hours. Not only is this underusing teachers’ provide more supportive learning environments for human capital but it is also depriving current and students and teachers; and (v) providing more sup- future generations of children of better educational port for teachers. opportunities. Increased instructional time and sup- port can be particularly effective for students from The Government of Croatia has also launched an disadvantaged backgrounds, vulnerable groups, ambitious curriculum reform with the aim of making and students with developmental disabilities. the education system more relevant and skills-ori- ented. The new curriculum represents the most Figure 53: Croatian students spend the fewest hours in school in the EU. Average number of hours per week in primary and lower secondary education (2018/19) 35 30 Primary Lower Secondary Number of hours 25 20 15 10 5 0 Croatia Bulgaria Poland Latvia Finland Lithuania Czech Republic Slovakia Estonia Hungary Slovenia Sweden Germany Austria* Romania EU-28 Greece UK* Italy Spain Belgium* Cyprus Malta Ireland Netherlands* Portugal Luxembourg Denmark France Source: European Commission/EACEA/Eurydice, 2019. 49 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia ambitious reform of the Croatian school system in (CEDEFOP, 2020). Only three-year dual programs the past decades. It aims to: (i) develop students’ and craft "unified" programs are run as appren- basic competencies for lifelong learning; (ii) define ticeship schemes in which about half of schooling learning outcomes that include, other than knowl- happens in the workplace, but these programs are edge, the development of skills, attitudes, inno- a minority. Sixty percent of four-year VET students vation, creativity, critical thinking, entrepreneur- enroll in a university or in vocational studies immedi- ship, and problem-solving skills; (iii) stimulate and ately upon completing their studies, most of whom develop the autonomy of education workers; and continue in the same field that they studied in high (iv) define criteria for the development and achieve- school (CEDEFOP, 2020a). For example, according ment of educational outcomes. In addition, there is to a 2016 study by CEDEFOP (the European Center renewed focus on increasing student satisfaction for the Development of Vocational Training), Croa- and teacher motivation. tia has a shortage of staff in occupations such as ICT, mechanical engineering, medicine, nursing, and The government has also made significant invest- teaching because students rarely choose to enter ments in digital tools for increasing efficiency and these education fields. One way to make the educa- improving learning outcomes in schools. These tion system more relevant is to enhance its partner- investments have mainly been financed with EU ship with industry and employers to ensure that the funds. For instance, the e-Schools project invested skills taught in the classroom align with the needs EUR 41.4 million in 151 pilot schools in its first of the labor market. This would help to ensure that phase, which was spent on equipping classrooms graduates of VET programs are well-prepared to and teachers with digital infrastructure, developing meet the demands of employers and contribute to teaching and administration applications, providing the growth of the Croatian economy. teachers with digital training, and creating digital learning materials. Improving the quality of VET in Croatia is critical for transforming the lifetime potential of students from disadvantaged backgrounds. While wealthier Adolescence and Youth: students attend general education schools (gym- Challenges and Opportunities nasiums), students from vulnerable families are for Vocational Education and significantly more likely to continue their educa- tion in vocational schools. However, according to Training PISA 2015’s economic, social, and cultural status The share of students enrolled in Croatia’s VET (ESCS) index, more than 87 percent of Croatian stu- programs is the highest among PISA participat- dents from the lowest socioeconomic quintile were ing countries (OECD, 2018). Vocational education enrolled in VET programs (53 percent in four-year and training (VET) in Croatia is critical for prepar- VET programs and 35 percent in three-year VET ing students for employment and for closing the programs), while only 13 percent attended general country's skills gap. The Croatian VET system enrolls upper-secondary education (World Bank, 2019d). approximately 70 percent of secondary school stu- dents. It offers programs at various levels, including Croatia’s VET students perform worse than general secondary and tertiary education, vocational edu- education students on the PISA tests for 15-year- cation, adult education, and training for reskilling olds. The scores of three-year VET students being to meet private sector needs. TVET (technical and even worse than those of four-year VET students. vocational education and training) is provided in the This raises the possibility that the whole VET edu- upper secondary and post-secondary levels, with students having to complete compulsory educa- Table 13: Averages for age 15 years PISA results: tion at the age of 14 or 15 to be accepted. Roughly General and Vocational in Croatia two-thirds of VET students are in four-year school- PISA 2018 PISA 2018 based programs, with fewer than one-third in three- General Vocational year programs (Matković and Šabić, 2022). Only one Reading 479 448 program, an upper secondary program for nurses, lasts five years and leads to a general nursing qual- Mathematics 464 436 ification. Science 472 443 PISA 2015 PISA 2015 There is still a discrepancy between the education profiles of VET graduates and their skills. VET pro- Collaborative Problem 473 448 grams still focus too much on theoretical knowl- Solving* edge and not enough on practical skills training Source: PISA 2018 database, * PISA 2015. 50 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia cation system might be holding back the perfor- were employed, which is significantly lower than mance of the overall education system (see Table the employment rates of medium-skilled (67.1 per- 13). In addition, VET students generally have lower cent) and high-skilled workers (84.1 percent). This scores in reading, mathematics, science, and collab- suggests the existence of skill mismatches, which orative problem-solving than the average score for points to a need for better coordination between Croatian students. industry and VET sector in balancing the supply and demand for the workforce. Many VET graduates limit their job search to the specific occupations in which they have experi- Another reason for the low employment of VET ence or training instead of exploring alternative job graduates in Croatia might be the limited amount of opportunities that could benefit from their skill- work-based learning in the VET curricula. Interna- set.50 According to research conducted on Croatian tional studies suggest that the divergence of trends students (Rafajac, 2019), a large number of young between overall employment and VET graduates is individuals require the assistance of professional smaller in countries where learning in VET systems is skills-matching services to expand their range of mostly workplace-based (World Bank, 2023). How- potential job opportunities. The most important ever, the trend difference appears larger in countries factors that motivate young people in Croatia to where the average VET student has a low share of choose a specific career are interest in that pro- work-based learning (WBL). In Croatia, the amount fession, the opportunity to have fun, their personal of WBL varies between different types of VET pro- talent, and an opportunity to meet new people. grams. For example, the WBL share is around 10 These factors may not be easily identifiable in a typ- percent in four-year school-based VET programs. In ical vacancy advertisement. As a result, 41.7 percent comparison, in three-year VET programs, the share of graduates estimate that they will have a low to of WBL ranges from around 5 percent in those with medium chance to find jobs in their chosen occu- a school-based modality to 60 percent in appren- pations (Rafajac, 2019). International assessments ticeships (CEDEFOP, 2020). In general, VET grad- also suggest that specific occupations need a uates have an advantage over general education higher supply of new specialists in select fields. Even graduates in securing employment at the beginning though there are great opportunities for engineers of their professional lives, but this advantage disap- or science professionals in the labor market, PISA pears quickly (see Figure 54). 2018 reveals that only 20 percent of high-perform- ing male students in Croatia plan on following those Survey findings from the EU indicate that trainee- careers, while only 16 percent of high-performing ships or internships are an important way for young female students have the same expectation. Simi- people to transition into employment, which sug- larly, only about one in three high-performing girls gests that Croatian policymakers should be more intends to work in health-related professions, while proactive in helping young people to acquire skills only one in eight high-performing boys expects to in this way. According to a recent EU-wide survey on do so. Also, only 11 percent of boys and only 1 per- traineeships and internships among the 18 to 35 age cent of girls in Croatia expect to work in ICT-related group, 44 percent of Croatians who had never been a professions (OECD, 2018d). Given the shortage Figure 54: Croatia age profile employment rate, of workers in these occupations, the government 2021 might consider measures such as career counselling to promote students’ interest in these subjects. 100 90 80 The TVET system produces large number of gradu- 70 ates, but the employment rate of these graduates 60 remains low within Croatia. The Croatian Employ- 50 ment Service (CES) has reported that during the 40 period from 2010 to 2015, a high percentage of 30 unemployed Croatians completed TVET programs. 20 Nearly 60 percent of those registered as unem- 10 ployed had a TVET degree. Yet recent data show 0 4 9 4 9 4 9 4 9 4 that the employment rate of VET graduates is -2 -3 -4 -5 -2 -3 -4 -5 -6 25 55 35 20 50 30 45 60 40 below the EU average, only 69.9 percent compared to the EU average of 76.4 percent (Education and Lower secondary VET Training Monitor, 2022). The situation is even more Tertiary General secondary challenging for graduates with lower skill levels. Source: Author’s compilation based on Eurostat, LFSA_ In 2021, only 42.1 percent of low-skilled workers ERGAED, 2021. https://blogs.worldbank.org/impactevaluations/are-jobseekers-looking-too-narrowly-jobs?CID=WBW_AL_BlogNotification_EN_EXT 50 51 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia trainee attributed their lack of traineeship to a short- formance of students in the VET system in Croatia age of suitable opportunities and their lack of aware- and the relatively low number of instructional hours ness of such opportunities. This was higher than the in general education, extending the general curricu- EU average of 38 percent. Among those who had lum to cover students of older ages could yield pos- been trainees, 39 percent reported that the experi- itive results. For example, providing more hours of ence had lasted less than three months. Despite this, math instruction has been shown to be an essential about 82 percent stated that the traineeship had factor in improving PISA scores in other countries. taught them skills that were useful professionally. Moreover, Croatia had the highest share (78 percent) In 2022, Croatia adopted amendments to the of paid traineeships of all EU member states. In the VET Act (ZID ZSO, 2022) to regulate work-based case of most paid traineeships (71 percent), trainees learning. These amendments regulate coopera- had access to social protection. These results sug- tion between technical and vocational schools and gest that the lack of adequate training opportuni- employers. Additionally, in 2022, the European ties stems from employers' unwillingness to bear Social Fund (ESF) and the European Social Fund Plus the costs of training. A clear implication is that the (ESF+) provided funding for new vocational curricula government could facilitate skill acquisition among based on occupational and qualification standards Croatian young people and attenuate the skill mis- set out in the Croatian Qualifications Framework matches in the labor market by subsidizing the costs (CROQF) to make the skills acquired in TVET schools of training – either by paying the trainees directly or more aligned with the labor market needs. In addi- compensating the employers for their costs. tion, VET teacher training is being reinforced with regional centers of competencies created in Croatia The Ministry of Science and Education has initiated that focus on infrastructural investments, program reforms dealing with curricula, skills mismatches, development, and the strengthening of human the quality of work-based learning and appren- resources. ticeships, and the public image of TVET. Its current priorities for TVET development focus on increas- Higher Education: The Problem ing its labor market relevance, quality, appeal, and internationalization. As the flagship initiative of the of Skill Mismatches Continues ministry, the TVET curriculum is being comprehen- The higher education sector is struggling with sig- sively redesigned in line with labor market needs to nificant challenges, including a misalignment with focus on learning outcomes, WBL, and quality of the needs of the labor market and relatively poor teaching. The new TVET curriculum has been imple- outcomes in terms of human capital use. The first mented nationwide since 2023. problem relates to the mismatch between the labor market needs and which fields of study are chosen To strengthen TVET education, the Croatian gov- by students in higher education. The second issue ernment has adopted a dual technical and voca- is related to the low higher educational attainment tional training model, making it one of its key pri- of Croatian students. For example, in 2021, higher orities in education reform. The model aims (i) to education attainment in Croatia was 35.7 percent, improve the quality and increase the relevance of below the EU average of 41.2 percent and below vocational education and training; (ii) to promote the EU-level target of a 45 percent share of 25 to the quality, appeal, and inclusiveness of vocational 34-year-olds having a tertiary education (Education education and training; and (iii) to increase the inter- and Training Monitor 2022). Finally, the employment nationalization of vocational education and training rate of tertiary graduates was 80.9 percent in 2021, by encouraging the international mobility of stu- which is the lowest rate in the EU (the EU average is dents and teachers. As a result, four dual educa- 84.9 percent). Also, there is a difference between the tion international VET experimental programs were genders in terms of their employment rates after the adopted by 14 schools in 2020/2021 after they had graduation.51 According to the Education and Train- been evaluated and revised (Education and Training ing Monitor (2022), although Croatian women tend Monitor, 2022). to achieve higher grades than Croatian men in ter- tiary education (4.06 versus 3.83), they face greater Delaying the entry of students into vocational pro- challenges in securing employment (58.8 percent grams can lead to improved learning outcomes in find jobs compared with 62.5 percent of men). basic skills in math, reading, and science, which are Moreover, they earn less income and report lower increasingly sought after by the labor market. In levels of job satisfaction. It is important to narrow Croatia, students enter vocational programs at the the gender gap and encourage young people to age of 14 or 15, which is younger than in many other make smarter choices regarding their education to EU and developed countries. Given the poor per- improve the country's use of human capital. See Chapter 2 for a detailed analysis of gender gaps in employment. 51 52 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia Figure 55: Educational attainment in tertiary ers in 2020 in this respect (Education and Training education among people aged 30-34, % of the Monitor, 2022). However, there is a surplus of grad- population uates in certain fields like political scientists, jour- 45 nalists, and philosophers in relation to demand for 40 their skills, meaning that employment opportuni- ties for graduates in these fields are scarce (CEDE- 35 FOP, 2016). Most unemployed respondents in 2021 30 graduated from biotechnical sciences (24 percent) 25 or humanities (24 percent) (Education and Training 20 Monitor, 2022). Graduates in humanities and inter- 15 disciplinary studies are most likely to get the lowest wage offers, unlike technology graduates, who tend 10 to receive the highest incomes. According to the 5 EU’s 2018 Eurograduate survey, Croatia has the 0 lowest employment rate for university graduates 2012 2013 2014 2015 2016 2017 2018 2019 2020 among all European Union countries. One year after European Union - 27 countries (from 2020) Croatia graduating, 9 percent of bachelor's degree holders Source: Eurostat, 2023 and 18 percent of master's degree holders are still While the share of STEM graduates is growing, seeking employment, while five years after gradu- other tertiary-level skills in Croatia are in surplus. ating, 7 percent of bachelor's degree holders and 10 On the one hand, Croatia has a relatively high and percent of master's degree holders are still unem- increasing share of STEM graduates – 28.5 per- ployed. Moreover, the job search period after grad- cent versus an EU average of 24.9 percent in 2020, uation tends to be very lengthy (Rimac, 2020). Pol- which was up by 4.6 percentage points from 2015. icymakers need to devise a systemic labor market The share of ICT graduates in Croatia is also above information mechanism that would allow students the EU average – 4.7 percent versus EU 3.9 percent to make their decisions about which fields of higher in 2020 (Education and Training Monitor, 2022). education to follow in full awareness of what skills Interestingly, women are more likely than men to are in highest demand in the labor market. This is choose to study STEM subjects, with 18.1 per- particularly needed as Figure 56 shows that the cent (versus the EU average of 14.1 percent) of all composition of higher education has a strong bias female graduates having STEM degrees. Women towards the social sciences. comprised 38.6 percent of all STEM graduates in 2022 compared with the EU average of 32.5 per- Croatian higher education graduates are con- cent (Eurostat, 2020). The share of female STEM cerned about their employability, skills acquisi- graduates has increased by 2.6 percentage points tion, and successful transition from university to since 2015 to 11 percent (versus the EU average of work. For example, the results of the 2021 Croatian 8.1 percent), representing the highest increase in the national graduates’ survey suggested that, even EU, and making Croatia one of the best perform- one year after graduating, graduates in all subjects needed more critical thinking skills. Furthermore, Figure 56: Number of tertiary programs in Croatia graduates in the biomedical and health sciences per field of study, 2022 had the largest deficits in practical knowledge and Interdisciplinary its application, adaptability, entrepreneurial spirit, sciences and leadership (Pažur Aničić et al., 2022). Therefore, Natural sciences it is important to strengthen the quality assurance of higher education programs and make them more Biotechnical sciences relevant to the needs of the labor market. Art The government has undertaken several initiatives Biomedicine and to remedy this situation, including: health Significant improvements to the higher educa- Humanities tion funding model that make financing condi- tional on performance. These changes include Technical studies provisions to internationalize higher educa- Social sciences tion, harmonize the content of study programs with the needs of the labor market, and reduce 0 100 200 300 400 500 600 administrative barriers to hiring and advancing Source: MOZVAG, 2022, URL: https://mozvag.srce.hr/preglednik/ teachers. In addition, the Act on Quality Assur- 53 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia ance in Higher Education and Science has intro- Figure 57: Participation rate in education and duced procedures for the accreditation of edu- training (last four weeks), people from 25 to 64 cational programs. These procedures involve the years, percentage Croatian Employment Service reviewing the rel- 12 evance of each new study program to assess it 10 alignment with labor market needs and with the 8 standards set by the Croatian Registry of Qual- ifications.52 6 4 The digital transformation of higher education 2 (the e-University project) developed under the 0 National Recovery and Resilience Plan (NRRP). 12 13 14 15 16 17 18 19 20 21 This initiative will equip public higher educa- 20 20 20 20 20 20 20 20 20 20 tion institutions with the capacity to enhance the digital competencies of higher education staff. The goal is to strengthen the capacity of European Union - 27 countries (from 2020) Croatian higher education institutions, which Croatia will enable them to launch higher-quality study Source: Eurostat, 2021. programs with a strong digital component and support teachers and students in online teach- those who lack a primary education, and therefore ing and distance learning. would benefit from it the most. The Eurostat data show that only 0.5 percent of this group had partic- The introduction of tuition fee subsidies for stu- ipated in adult education. For most individuals, pro- dents from lower socioeconomic groups, which ficiency in certain skills decreases with age so they has already yielded positive effects. Recent may need additional training to keep up with the Eurostudent data have shown that current stu- latest technology, particularly in problem-solving, dents have higher educational attainment than as was highlighted in a 2018 OECD report (OECD, their parents. The data show an increase in the 2018c). share of students whose parents did not com- plete higher education from 50 percent in 2010 Croatia has one of the lowest participation rates in to 58 percent in 2016 (Education and Training adult learning in the EU, especially by individuals Monitor, 2019). who are low-skilled, older, living in rural areas, or have been unemployed for a long period. Accord- ing to Eurostat data, the share of adults aged 25 to Adult Training: Skills 64 who participated in an adult education program Development Throughout a in the previous four weeks had increased from 3.3 Working Life percent in 2012 to 5.1 percent in 2021. However, it remained significantly lower than the EU average Those individuals who are most likely to bene- throughout the decade (see Figure 57). The groups fit from training are the least likely to participate who rarely participate in training include people according to OECD and EU data. These groups over 55, the unemployed and inactive, people with include the lower-skilled, older adults, displaced only basic education, and those employed in low- workers, those whose jobs are most at risk of auto- skilled occupations. At the same time, participation mation, and non-standard workers such as part- is significantly higher among the urban population, time and on-call workers. On average, in the OECD, the young, the highly educated, and the employed, participation by low-skilled adults is 40 percent especially managers, and production workers (Mat- below that of highly skilled adults, while older adults ković and Jaklin, 2021). The Croatians who partici- are 25 percent less likely to train than 25to 34-year- pate the most in lifelong learning are those in man- olds and workers with jobs at high risk of automation agerial positions with successful careers. Sixty-two are 30 percent less likely to engage in adult learning percent in this group participated in some form of than their peers who are less exposed to automa- learning the previous year compared with only 20 tion (OECD, 2021). Moreover, Eurostat data reveal percent of manual workers. The low levels of partic- that it is mostly individuals with higher levels of edu- ipation by the unemployed and low-skilled are par- cation who attend adult training.53 Among highly ticularly concerning given that they have the most educated individuals, 7.5 percent had participated to gain from re-skilling and up-skilling (employment in adult education within the previous four weeks. rates for low-qualified workers are significantly Unfortunately, those who participate the least are below the EU average) and given the need for https://eurydice.eacea.ec.europa.eu/news/croatia-new-act-quality-assurance-higher-education-and-science 52 54 https://eurydice.eacea.ec.europa.eu/national-education-systems/croatia/adult-education-and-training 53 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia increased productivity at the national level because Figure 58: Participation rate in education and of demographic changes. training (previous four weeks) in 2021 by age, percentage While younger people participate in training more 20 often than older workers, their share is still well below the EU average. In 2021, a mere 11.7 percent 15 of 25 to 34-year-olds had received some training during the previous four weeks (including in formal 10 education) as opposed to an EU average of 18.2 percent. Only 1.4 percent of people aged 55 to 64 5 have received training in Croatia, which is lower than the average of 5.7 percent for the same age group 0 across the EU (see Figure 58). From 25 to From 35 to From 45 to From 55 to 34 years 44 years 54 years 64 years The high cost of training is one of the biggest obstacles to participating in adult education. A European Union - 27 countries (from 2020) survey conducted in 2017 by the Agency for Voca- Croatia tional Education and Training and Adult Education Source: Eurostat, 2021. found that 30 percent of respondents gave the price of training as the reason why they were not partici- Figure 59: Enterprises providing CVT training - % pating in adult education.54 While many adult edu- of all enterprises cational programs are funded by employers and/or 90 the state and local governments, only a small share 80 of companies offers continuing vocational training 70 60 (CVT). CVT is education and training that takes place 50 during paid working hours or is partially paid for by 40 employers if training is conducted outside of work- 30 ing hours. In 2020, 48.2 percent of Croatian firms 20 10 provided their employees with CVT, compared to 0 an EU-wide average of 67.4 percent. Croatia’s share 2010 2015 2020 was lower than the share in most of its peer coun- tries and had declined since 2010. The size of the European Union - 27 countries (from 2020) enterprise is an important factor in whether or not Croatia Slovenia it provides training. Data from the 2020 Continuous North Macedonia Serbia Vocational Education Survey show that 84.6 per- cent of large enterprises in Croatia provided some Source: Eurostat, 2023. form of education for their employees, whereas Note: Data on Serbia is available only for 2020, and data on North Macedonia only for 2015-2020. only 43.2 percent of small enterprises did the same. Figure 60: Types of skills studies through CVT course, per type of enterprises NKD 1 NKD 2 NKD 3 NKD 4 NKD 5 Small enterprises Medlum-sized enterprises Large enterprises 1 629 1 400 1 200 1 000 800 600 400 200 0 General IT IT professional Management Team working Customer Problem O ce Foreign Technical, Oral or written Numeracy Other skills skills skills skills skills handling skills solving skills administration language skills practical or communication and/or skills job-specific skills literacy skills skills Source: Croatian Bureau of Statistics 2020. URL: https://podaci.dzs.hr/2023/en/60850 Note: NKD is a classification of all economic activities of the Republic of Croatia. it was designed at the end of 1994 on the basis of statistical classification of economic activities of European Union – NACE. https://eurydice.eacea.ec.europa.eu/national-education-systems/croatia/adult-education-and-training 54 55 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia Most of the vocational training (74.7 percent) was ESF+ program. Furthermore, foreign workers could conducted for staff in information and communica- fill certain labor vacancies and skills shortages, which tion or financial and insurance positions. will require the introduction of efficient procedures for recognizing foreign qualifications. Men are more likely than women to join CVT courses, especially in small enterprises. On average, 24.2 percent of employees of Croatia’s businesses Policy Implications participated in CVT courses in 2020. The most significant difference between the sexes occurred It would be beneficial to increase investments among those employed by small enterprises. The and develop targeted interventions to support share of men was 55 percent, while that of women the development of foundational and high-order was 45 percent. Men and women working in large skills among children and youth from disadvan- enterprises participated in CVT to an almost equal taged groups. The annual results of the national extent (Croatian Bureau of Statistics, 2020). Matura exam show large inequalities within the country at the upper secondary education level. Challenges related to the quality of and access to The high share of low-performing students from adult education still need to be addressed. Since lower socioeconomic backgrounds suggests that there is still no monitoring of the outcomes of adult they enter VET systems with a lack of foundational learning programs, their effectiveness and impact and higher-order skills, which reduces their ability on the employability of their students remains to learn and apply new knowledge, take informed unknown. In 2021 the Croatian Parliament adopted decisions about their next educational and profes- a new Adult Education Act (ZOO, 2021) aimed at sional pathways, conduct relevant job searches, and improving the quality of the programs offered and interact with prospective employers. Specific meas- recognizing knowledge and skills that were acquired ures to counter this might include programs spe- outside the education system. A quality assurance cifically aimed at teaching foundational, cognitive, system for adult education is also being devel- socio-emotional, technical, and life skills, tutoring, oped in Croatia based on the OECD's Program for special support services like providing books and the International Assessment of Adult Competen- learning materials, subsidies for transportation and cies (PIAAC) along with quality assurance models health care, and career counseling. and self-evaluations and external evaluations of adult education institutions. The new quality assur- It is recommended to incorporate modern peda- ance system will ensure that Croatia’s adult edu- gogical practices that support students' learning cation institutions will meet the standards set by and skills development (such as group work, team the CROQF for matching the skills developed by teaching, or co-teaching) into teachers’ daily prac- the workforce with the needs of the labor market. tice. Teaching styles represent an important com- Additionally, in 2021, the government organized ponent of schooling that greatly affects learning multiple events to promote lifelong learning to var- outcomes and the development of specific skills. ious groups of stakeholders. Croatia has also set an For example, arranging students into small groups ambitious target of doubling the rate of adults in for discussions and joint work could stimulate com- training from 26.9 percent by 2030 (Education and municative and problem-solving skills, while putting Training Monitor, 2022). a team of teachers together to work with a large group of students could enhance critical thinking The government introduced learning vouchers in and collaboration, as well as serving as a continu- 2022, with an emphasis on green and digital skills ous learning activity for the teachers themselves. development in ALMPs. These vouchers can be However, in order to introduce these methods suc- used by both employed and unemployed people cessfully, it will be necessary to upgrade the teacher for reskilling and upskilling programs and are aimed training curriculum. The use of modern teaching at reducing the considerable skills gaps in the Croa- practices would be enhanced by expanding the tian labor market. The program’s goal is for 30,000 availability and use of technology in learning and by people to benefit from the scheme over four years. improving the quality of the furniture used in class- Additionally, the program aims to target vulnerable rooms. International studies have shown that tech- groups in the labor market. Its effectiveness will be nology and comfortable furniture both play a sig- externally assessed after a certain implementation nificant role in enabling modern teaching practices period. Based on those results, the government such as group work and team teaching and enhance plans to broaden the offer of vouchers and intro- both students’ performance and their well-being.To duce the building blocks needed to implement enhance teachers' ICT and digital skills, creating and Individual Learning Accounts using funds from the implementing specialized modules in teacher train- 56 TOC Chapter 3: Education and Skills Needed to Prepare for the Future Job Market Harnessing Human Capital for Growth in Croatia ing courses is recommended.According to the 2018 tives to encourage take-up, and raise awareness TALIS, only 45 percent of teachers in upper second- about apprenticeships, informing firms about the ary education use ICT for teaching, and a mere 38 advantages and benefits to them as well as inform- percent feel adequately prepared to teach using ing workers about the high future return to these digital aids. To eliminate this gap, existing teacher kinds of WBL. training programs should be upgraded to equip teachers with the skills needed to use technology Increase the motivation of older and low-skilled in teaching as well as to provide them with ongoing adults to remain in the labor market and reduce support in their daily work. barriers to learning and reskilling. This could be done by conducting studies to identify and min- Continue investments and specific policy inter- imize barriers to the participation of older or low- ventions to reduce teacher shortages at different skilled workers in adult learning. One way for the levels of the education system. Possible measures government to encourage employers to offer more to achieve this might include raising teacher’s sala- training opportunities to their employees is by pro- ries, offering help with transportation, accommoda- viding financial support for small and medium-sized tion, and relocation to those willing to fill vacancies enterprises (SMEs) and offering tax incentives. in remote/lagging regions, promoting employment opportunities for young graduates of teacher train- ing programs, upgrading of teacher preparation and training programs, providing continuous learning opportunities to teachers, and creating regional or local partnerships with local authorities, the private sector, and school communities to regularly monitor and understand the local needs of the teachers and provide them with the required support. Creating a Graduate Tracking System (GTS) would bring many benefits. This system would provide val- uable data on the employment outcomes of gradu- ates, guide students in choosing fields of study that align with labor market demand, help colleges and universities to redefine their course offerings, and offer yield insights into salary levels of graduates by sector and discipline. Enhance the career guidance given to students before they enroll in higher education degrees and after they graduate from VET and tertiary educa- tion. To identify ways to improve career guidance services, the government might consider mapping and assessing the impact of existing career guid- ance practices and initiatives in VET and higher education institutions and developing a set of rec- ommendations and policy measures based on the findings. Increase the quantity and improve the quality of work-based learning, such as apprenticeships and internships. This would enhance the learning oppor- tunities available to VET students and ensure their smooth transition into the labor market. In addition, WBL requires stronger cooperation between edu- cation institutions and employers and the creation of conditions that will give employers and young employees an equal incentive to provide or partic- ipate in WBL. The government could increase the scope of apprenticeships, provide financial incen- 57 TOC Chapter 4: Migration and Human Capital Harnessing Human Capital for Growth in Croatia Chapter 4: Migration and Human Capital Introduction Migration has the potential to create an upward ever, to maximize these benefits, the host country economic spiral in Croatia by amplifying positive needs to have policies in place for managing the trends in the human capital environment and in skills of the migrant workforce. In sum, this means economic opportunity. One of the channels of this that strengthening Croatia’s human capital and positive feedback loop involves both the sending labor market while also implementing sound migra- and receiving countries exchanging knowledge and tion policies could yield cumulative and sustained skills via “brain circulation” (Boeri et al, 2012 and economic benefits. Mayr and Peri, 2008). Another key channel is the labor reallocation effect from low-productivity to According to UN DESA, Croatia has one the highest high-productivity countries and sectors (Clemens et shares of emigrants in the EU, as 23 percent of the al, 2008 and Clemens, 2011). These benefits become Croatian-born population lives abroad. Emigration available when labor markets are flexible and when from Croatia has been a common phenomenon since education can expand in response to higher returns the 1990s as 820,000 Croatian-born citizens out to education, allowing the upward occupational of a total of 4.2 million were living abroad by 2000, mobility of native workers (Bossavie et al, 2022). many of whom were fleeing the conflict. Since then, However, when labor markets are rigid, migra- a dearth of economic opportunities drove more tion might not result in benefits such as reduced Croatians to emigrate to Western European coun- unemployment and greater economic efficiency tries. This process got a further boost when Croatia (D’Amuri and Peri, 2014). This can force a negative joined the EU in 2013, as 230,000 Croatian emi- growth feedback loop, as it stimulates even more grants took advantage of the free mobility in the emigration, and these emigrants are less likely to following four years (Draženović et al, 2018). Around return and invest in their country of origin. In other 86 percent of Croatian emigrants settled in Europe. words, the structural drivers that increase economic Over the last decade, attracted by the freedom of growth, such as flexible labor markets and human movement within the EU and higher wages, migra- capital investments, also produce desirable effects tion outflows (Figure 61A) have been concentrated with regard to migration flows and outcomes. How- in the EU (72 percent and 50 percent in Germany Figure 61: Destination countries for Croatian migrants (A: left) and Emigration by age group and by time (B: right) 100% 35 8% 90% 16% 2% 8% 30 80% 9% 10% 25 70% 60% 15% 29% 20 Percent 50% 7% 1% 15 40% 7% 30% 5% 10 20% 50% 32% 5 10% 0% 0 Stock of emigrants Migration outflows 0-14 15-24 25-34 35-44 45-54 55-64 +65 (2019) (2013-19) Age group Germany Austria Other EU Rest of World Emigrants (2013-18) Non-migrants (2013-18) BiH Serbia Other Europe Emigrants (2010) Source: Emigration (database), Eurostat, European Commission, reference years 2013/18 [migr_emi4ctb]; Demographic profile of emigrants in 2010: DIOC (Database on Immigrants in OECD and Non-OECD Countries), reference years 2010/11, OECD; Demographic profile of nonmigrants: Population (database), Eurostat, European Commission, reference years 2013/18; Stock of emigrants (2019): International Migration (database), Population Division, Department of Economic and Social Affairs, United Nations; Emigration flows 2013-19: Immigrant and emigrant population to/from Croatia, Croatian Bureau of Statistics. This chapter was prepared by Daniel Garrote Sanchez. 58 TOC Chapter 4: Migration and Human Capital Harnessing Human Capital for Growth in Croatia alone). While before Croatia’s EU accession, Cro- In the short term, these outflows of migrants rep- atian emigrants tended to be older, with over half resent a loss of about 18 percent of the tertiary-ed- being aged 55 or older, those migrating during the ucated population. Since these emigrants are of last decade were more likely to be of prime age (25 working age, the loss of human capital directly to 44) and more motivated by income disparities reduces the availability of skills in the Croatian labor and the ability to move for work freely (Figure 61B). market. This is partially compensated for by the The Baltic countries have demonstrated that reduc- arrival of immigrants from other countries, although ing income gaps can decrease emigration rates and there are fewer of them than Croatian emigrants even result in a net influx of people. Other than eco- (half a million versus 1 million), and they are slightly nomic reasons, emigration has been stimulated by less educated. other factors, such as the high occurrence of cor- ruption and the challenging job market in Croatia Migration is expected to have a positive effect on (Jurić, 2017 and Draženović et al, 2018). Croatian human capital in the long run as it incen- tivizes workers to invest in high-demand skills. Take Migration Effects in Croatia the case of the emigration of qualified profession- als, such as health workers, which is a major concern Migration has costs and benefits for Croatian in Croatia (Gruber et al, 2020). Since Croatia joined human capital. The costs and benefits vary in the the EU in 2013, about 0.8 percent of its doctors short and long terms due to the changing compo- have emigrated each year, which is lower than the sition of migration and return flows and the indirect average rate of 1.1 percent for all EU’s new member effects of migration on households’ human capital states (NMS) during the same period. Despite this investment decisions. Typically, in the short term, continuous outflow, the number of doctors in Cro- emigration reduces the stock of human capital as atia has steadily increased, surpassing the average educated individuals leave the country. The inten- rate for the NMS. This growth can be attributed to sity of this “brain drain” depends on what skills are the rise in medical graduates, who were encour- held by the emigrating population. According to aged to go to medical school by the prospect of the OECD-DIOC data, while Croatian emigrants have high wages paid overseas (Figure 63). These higher similar education levels to those who remained in wages paid for specific occupations overseas in the country (Figure 62), not enough recent data are raise the returns to education in those fields. This, in available to allow for a more up-to-date analysis of turn, has increased the number of Croatians invest- emigrants’ skill profiles. However, some studies have ing in their own skills, thus helping significantly to shown that, after Croatia’s EU accession, emigrants offset the impact of emigration in specific sectors. have tended to have slightly lower education levels, This phenomenon is quite widespread; there is no with a larger increase in outflows of less educated correlation across all of the NMS between migration workers (Župarić-Iljić, 2016). outflows and changes in the number of doctors as a higher number of graduates compensates for the skills lost to emigration. However, this requires that Figure 62: Education and migration status among countries scale up the supply of their education ser- working age population (25-64) vices. Romania has successfully met the growing 70% demand for medical education, which has not been 61% the case in other countries where inflexible training 60% 57% systems may have impeded growth in education 53% 50% services. However, despite the replacement effect, the emigration of highly skilled workers has a cost to 40% their country of origin, which loses the investment it made in their expensive education and ends up 30% subsidizing the richer migrant-receiving countries 24% 23% 23% 24% 19% (Glinos, 2015). Furthermore, more medical pro- 20% 16% fessionals are becoming available at the national 10% level in Croatia, there is evidence of understaffing in some primary care practices in rural areas (World 0% Bank, 2020a). While working in certain occupations Emigrants Non-migrants Immigrants abroad may provide higher returns for emigrants, this tends to reduce the availability of skilled work- Low Medium High ers in their home country. As a result, students may Source: DIOC (Database on Immigrants in OECD and Non- choose to specialize in high-return fields, which may OECD Countries), reference years 2015, OECD. not be beneficial overall for the country. For exam- 59 TOC Chapter 4: Migration and Human Capital Harnessing Human Capital for Growth in Croatia Figure 63: Trend in the stock of doctors in Croatia Stock of doctors (2013=100percent) Changes in the stock of doctors in HR 115% 600 110% 400 105% 200 Number 100% 0 95% -200 90% -400 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 20 7 20 8 09 10 11 12 13 14 15 16 17 18 19 20 0 year 0 20 20 20 20 20 20 20 20 20 20 20 20 Graduates Retirees Croatia EU NMS-13 Net migration Change in stock Source: Eurostat and Regulated Professions Database, European Commission. ple, a study conducted in the Philippines found However, several factors limit the potential gains to that the sudden emigration of nurses resulted in Croatia from migration. First, the return migration a rise in student enrollment in nursing. However, rate is exceptionally low in Croatia. When migrants this increase was due to existing students chang- return to their home country, they bring new working ing their choice of degree, which led to a rise in the knowledge and practices with them that enrich the number of college graduates overall but a decline in country’s stock of human capital (Bahar and Rapo- some fields (Abarcar and Theoharides, 2021). port 2018) and increase the value of foreign direct investment (Kugler and Rapoport, 2007) and trade Large flows of remittances can positively affect (Orefice et al, 2022 and Peri and Requena-Silvente, human capital. Croatia has the largest share of 2010). These advantages translate into the availa- remittance inflows of all EU countries. Over the bility of more and better employment opportunities last decade, the value of remittances received by in the country. However, only about 8 percent of all Croatian households has doubled and now repre- emigrants from Croatia in 2010 had returned home sents 7 percent of GDP. Remittances are one way by 2014, a ratio significantly lower than the rate of in which education is financed. Experience from 27 percent for other EU new member states. The other countries tends to suggest that the receipt of disinclination to return among Croatian migrants remittances has a positive effect on household edu- may be partly due to their profile, as they tend to cational attainment (Amuedo-Dorantes and Pozo be older on average and more likely to have spent 2010; Bansak and Chezum, 2009; Dustmann and many years abroad and thus been deeply integrated Glitz, 2011; and Yang, 2008). However, emigration into their host countries. This is likely to be especially can also negatively affect school attendance and true of older emigrants who fled the country during attainment given the family disruptions resulting the Balkan War. However, even younger Croatian from the absence of one or both parents (McKenzie emigrants are less likely to return than their peers and Rapoport, 2011). from other NMS countries (Figure 64). While two Figure 64: Share of emigrants in 2010 who returned by 2014 by education and age groups 60% 50% 40% 30% 20% 10% 0% 25-34 35-44 45-54 55-64 65+ 25-34 35-44 45-54 55-64 65+ 25-34 35-44 45-54 55-64 65+ 25-34 35-44 45-54 55-64 65+ All education Low education Mid education High education Croatia NMS13 Source: DIOC (Database on Immigrants in OECD and Non-OECD Countries), reference years 2010/11, OECD, and EU-LFS ad-hoc module of 2014. 60 TOC Chapter 4: Migration and Human Capital Harnessing Human Capital for Growth in Croatia in five young NMS migrants (ages 25 to 44) return and assistance with planning their return. Reinte- to their home countries within four years, the ratio gration programs for returning migrants could also for Croatians is only one in five. The gap in rates of help to smooth their transition into the domestic return is especially wide in the case of emigrants labor market by, for example, supporting self-em- with high school diplomas, whereas higher-skilled ployment and startups. Croatians are as likely to return home as those in other peer countries. As more Croatians who Strengthen programs to help the Croatian dias- migrated to the EU return home than do those who pora contribute to the country’s development. For migrate to other OECD or neighboring countries, example, the Pilot Registry of Croatians Abroad is a the concentration of migration flows to EU coun- digital database using social media platforms that tries since Croatia accession is likely to increase the connects local Croatian firms and individuals abroad number of returnees. Second, emigrants returning (OECD, 2022). In 2007, the Croatian government to other Eastern European countries face positive also set up the Unity through Knowledge Fund wage premia (Co et al, 2000 and Martin and Radu, (UKF), a dedicated grant facility for joint research 2010) and are more likely than non-migrants to projects between the Croatian diaspora and institu- become self-employed or entrepreneurs and create tions in Croatia that benefit the Croatian economy jobs for non-migrants (Hausmann and Nedelkoska, or contribute to the development of the country’s 2018), but that is not the case for Croatian return- infrastructure. Since 2007, 91 projects have been ees. The lack of economic returns to returning to implemented at the cost of EUR 5.1 million. This Croatia might deter more Croatian emigrants from collaboration has strengthened the chances of Cro- returning home. However, there is a need for more atian researchers in science and technology being analysis of the reasons behind the low return rates awarded EU funds (for example, their acceptance in Croatia vis-à-vis other countries. rates to the 7th Framework Program funds have doubled). However, further efforts are needed to Policy Implications engage the Croatian diaspora in efforts to benefit their home country. Ireland is a good example of To strengthen the benefits of migration for Croa- how to take a holistic approach to this. It has devel- tia, policymakers should focus on facilitating “brain oped a diaspora ministry and a Diaspora Strategy circulation” by improving the business environ- for 2020-2025, which involves: (i) supporting Irish ment. While returnees to the EU’s NMS tend to have people living abroad (including by expanding the positive “returns to returning,” Croatian returnees network of honorary consuls to engage with the do not have the same positive labor market out- diaspora and to work with organizations to assist comes. Therefore, their choice not to return can be any Irish in distress abroad); (ii) developing business explained by barriers to reintegration, such as loss of networks of Irish emigrants (particularly in priority networks or a lack of knowledge of how to navigate trade and investment sectors as well as research Croatia’s bureaucracy. In particular, the employ- and innovation); and (iii) facilitating their return ment gap between returnees and stayers tends to to Ireland by providing them with information on be narrower for lower-skilled workers and wider for employment opportunities in Ireland, negotiating higher-skilled workers, suggesting that the latter reciprocal agreements with countries with large group experience more obstacles to returning. Irish immigrant populations on the transferability of social security, the mutual recognition of academic Find a more effective way to track the demand or professional qualifications, and the avoidance of for and supply of skills. If data could be collected double taxation, and monitoring other barriers to on the migration flows of workers from differ- returning. ent occupations, this would enable the education system to become more responsive to changes in Keep strengthening Croatia’s immigration man- labor demand. Furthermore, more efforts could be agement programs to promote the inflow of the made to attract talent from abroad, either foreign- skills demanded in the domestic economy. Immi- ers or Croatian emigrants, to fill critical occupations. gration flows into Croatia have increased recently. Malaysia is a good example of a migrant-sending In 2020, there were more than half a million for- and receiving country that has linked its migration eign-born individuals in Croatia. However, since management and workforce planning systems to the 2013 accession of Croatia into the EU, there the benefit of the domestic labor market (Box 3). has been a renewed upward trend in the arrival of Interventions that could attract more Croatian emi- foreign-born individuals attracted by Croatia’s sub- grants to return home might include introducing stantial economic progress. The main countries of fiscal incentives and providing prospective return- origin of these immigrants are neighboring Bos- ees with online information on job opportunities nia-Herzegovina and Serbia (accounting for nearly 61 TOC Chapter 4: Migration and Human Capital Harnessing Human Capital for Growth in Croatia half of the total inflows) and Germany (mostly Ger- its growing attraction as a destination. On a tactical man-born individuals of Croatian descent). Also, level, inflows of migrants can fill certain gaps in the immigrants from Asia are a small but fastest-grow- labor market, but the government needs to keep ing group, with almost 10,000 having arrived strengthening its skills management programs to between 2019 and 2021. These trends highlight the ensure the inflow of the skills needed by the domes- complex nature of migration in Croatia, which has tic economy. Box 3 gives examples from Malaysia of sizable emigration outflows and is also increasingly relevant policies that could be adapted to the Cro- becoming a migrant-receiving country, reflecting atian context. Box 3: Migration policies in Malaysia designed to enhance domestic human capital Malaysia is a country that has a large outflow of emigrants (6 percent of the population born in Malay- sia) as well as a large and fast-growing number of immigrants (who constituted 12 percent of the population as of 2020). The emigration of high-skilled professionals mostly to Singapore, Australia, and the United States has raised concerns about a brain drain and a skills gap in Malaysia, especially in the most complex occupations, which can limit technological progress. The Government of Malaysia established the TalentCorp initiative in 2011 to attract, nurture, and retain the highly skilled workforce that is needed for the growing knowledge-based economy. This initiative focuses on three segments of the workforce: (i) Malaysians in Malaysia (by improving the quality of education); (ii) Malaysians abroad (by encouraging the return migration of highly skilled professionals); and (iii) foreign talent (by instigating a dual work migration system that aims to attract high-skilled migrants) (World Bank, 2022). The TalentCorp’s Critical Skills Monitoring Committee (CSC) assesses skills imbalances in Malay- sia and maintains a Critical Occupation List (COL), which has been updated annually since 2016. The list identifies those critical occupations for which there is a mismatch between the skills demanded by employers and the supply of skills available on the market with a particular focus on shortages of mid-to-high skills. The COL’s top-down approach collects quantitative data on the current labor market that is comparable over time and across occupations and relies on sev- eral indicators to determine shortages. These indicators include three-year employment growth, three-year median wage growth, and the vacancy rate, all of which are associated with rising labor demand. The CSC also uses a bottom-up approach to complement its top-down approach by consulting with key stakeholders (such as employers and industry associations) on their perspec- tives of supply and demand. Good practice suggests that the lists should be updated regularly, be based on sound labor market evidence, be transparent, be produced by a specialized agency, and involve private and public sector stakeholders (World Bank, 2019). This is the case in Malaysia. The occupations categorized as critical can then be filled by foreign immigrants, returned emigrants, or by investing in training programs to upgrade the skills of the domestic workforce. Malaysia has a dual work permit system in place to manage foreign labor flows in a way to better respond to labor market needs (World Bank, 2020b). It distinguishes foreign workers by skill level and provides an Employment Pass for the highly skilled and a Visit Pass (Temporary Employment) for the low- skilled. The Employment Pass has three categories that are distinguishable by the minimum salary of migrants and employment duration requirement, and their holders can take up jobs for up to five years and have a pathway to become permanent residents. The Government of Malaysia also introduced the Returning Expert Program (REP) to incentivize the return of highly skilled Malay- sians from abroad. Given that jobs abroad pay higher wages than those in Malaysia, one of the key incentives offered by the program is a 15 percent flat income tax rate for five years instead of the standard progressive tax schedule, which increases the net salaries of skilled returnees. Eligible returnees also enjoy a tax exemption on the import or purchase of two vehicles and on all personal effects brought back to Malaysia, while their foreign spouses and children automatically receive permanent residency status (World Bank, 2022). Eligibility for the REP is based on a system that allocates scores according to the level of education of the returning emigrant and the work expe- rience that they gained abroad. Evidence has shown that the program increased the probability of emigrants with a preexisting job offer returning to Malaysia by 40 percent (Del Carpio et al, 2016). 62 TOC Chapter 5: Labor Activation Policies Can Strengthen Participation and Human Capital Harnessing Human Capital for Growth in Croatia Chapter 5: Labor Activation Policies Can Strengthen Participation and Human Capital Introduction Active labor market programs can improve how shortages in critical professions in education, pre- human capital is deployed in the labor market in school education, healthcare, social welfare, and Croatia. Human capital consists of the skills and culture. After having to adjust to the challenges expertise that allow individuals to work productively of the coronavirus pandemic, Croatia’s ALMMs are and that help them to do their jobs effectively and now having to respond to the economic fallout efficiently. Active labor market programs (ALMPs) and migration flows provoked by Russia’s war in can improve the way in which human capital is Ukraine, including refugees and forcibly displaced used in the labor market by providing services such people. The Croatian Employment Service (CES) as job matching and measures to help workers to is the public institution responsible for the design upskill and reskill. However, the efficacy of ALMPs is and delivery of ALMMs as a part of its overarching affected by their capacity to provide effective types mandate to implement the government’s strategic of support services, their organizational setup, the plans and programs for employment promotion. availability of funding, the existence or lack of incen- Figure 65 summarizes the portfolio of ALMMs cur- tives for private sector participation, and the avail- rently being implemented. ability of jobs. The primary role of ALMPs is to bring Figure 65: Categories of ALMMs being imple- the users of public employment closer to the labor mented in the Republic of Croatia, 2023 market. ALMPs have a social objective as well – to increase inclusion and participation in economic Self-employment subsidies and business employment expansion and social life. They incorporate active labor market Subsidies Self-employment subsidies for green and digi- for self- measures (ALMMs) which include employment ser- tal industries vices and the administration of benefits which aim Business expansion schemes (till end-2022) to reduce imbalances between labor supply and Choose Croatia - labor mobility demand, assisting unemployed people to find work, Employment subsidies and by decreasing professional and geographical Employment subsidies for green and digital mismatches. ALMPs improve labor market out- industries employment Support for comes by building the skills and enhancing the Support for employment in Sisak-Moslavina county employability of the people whom they help, espe- Traineeship / Apprenticeships / Internship cially the disadvantaged. This reduces inequality and subsidies boosts productivity, while mobilizing human capital Traineeships in green and digital industries that previously been ‘’wasted’’ because workers Traineeship subsidies in the public services have lost skills while not working, staying inactive, Training of the unemployed and other jobseekers or working on a job that does not allow them to use Training On-the-job / workplace training their available skills. Upskilling subsidies Training in institutions for adult education The Croatia’s Active Labor Activation program Market Measures Public works Direct job creation Public works for natural disaster relief in Sisak-Moslavina county Croatia provides various ALMM programs and services for job seekers and employers. For exam- Permanent seasonal worker Job retention maintenance work income and support ple, since 2020, ALMMs have been promoting the subsidies / out-of Job retention subsidies – short-time work acquisition of skills for digital and green industries schemes through self-employment, employment subsi- Job retention subsidies – textile, clothing, foot- dies, and internships. They have also been helping wear, leather and wood sectors (till end-2022) jobseekers to relocate both within Croatia and in Source: Croatian Employment Sservices (2022 and 2023), returning to Croatia as well as reducing labor supply https://mjere.hr/katalog-mjera/mjere-aktivnog-zaposljavanja/. This chapter was prepared by Boryana Gotcheva and Nithin Umapathi. 63 TOC Chapter 5: Labor Activation Policies Can Strengthen Participation and Human Capital Harnessing Human Capital for Growth in Croatia The overarching strategic direction for active labor Figure 66: Registered unemployed, ALMM users market policies is provided by the National Devel- and new entrants, 2017-2022 opment Strategy of the Republic of Croatia 2018- 250 2030 and the National Recovery and Resilience Plan (NRRP) 2021-2026. The NRRP reinforces eco- 200 nomic and social resilience by spending EUR 277 million to support employment and social inclusion. 150 For employment promotion, the funding envelope amounts to EUR 179 million, with most allocations 100 focusing on employment and self-employment, as well as the creation of an adult education voucher 50 system for training programs in new skills related to the green and digital transitions. Furthermore, to 0 tackle low labor activity, the country’s active labor 2017 2018 2019 2020 2021 2022 market policies will focus on increasing the compet- New entrants in ALMMs All ALMM users itiveness and employability of vulnerable groups, Registered unemployed the long-term unemployed, the inactive, and young people who are neither employed nor in education Figure 67: New entrants in ALMMs as share of or training (NEET). The planned interventions also registered unemployed, 2017-2022 (%) include improving the business processes of the 100% CES for monitoring labor market outcomes. In addi- 90% REGISTERED UNEMPLOYED tion, the most recent guidelines focus on building 80% a competent and adaptable workforce that will be 70% able to meet the requirements of the labor market 60% and on creating comprehensive social and employ- 50% ment services. 40% 30% The fraction of unemployed workers who benefit from ALMMs has been increasing since 2017 but 20% 24.1 29.2 26.7 remains low in comparison to the EU. Although 10% 19.5 15.4 23 the introduction in 2017 of new laws on job inter- 0% mediation and employment promotion has led 2017 2018 2019 2020 2021 2022 to a gradual increase in the coverage of ALMMs, Newly enrolled in ALMMs Other registered unemployed it remains limited (Figure 66). As of 2022, partici- Source: Ministry of Labor, Pension System, Family and Social pants in ALMMs have increased to 26.7 percent of Policy and CES (2022). the registered unemployed, which is slightly below the peak level of 2019 (Figure 67). However, even Figure 68: Spending on ALMMs, total and by in 2019, participation in ALMMs “per 100 persons category, 2017-2021 wanting to work” was low compared to the average 1400 for the EU (OECD, 2022b). Being able to reach only 1200 a fraction of those who need ALMMs and employ- ment services bears risk of losing human capital 1000 because, without receiving assistance, many job- seekers could become long-term unemployed, 800 HRK making it even harder to help them to (re)integrate 600 into the labor market. Also, scarce outreach and lack of services mean that potential participants are less 400 likely to register with the CES. 200 Croatia’s expenditures on ALMMs as a percent- 0 age of GDP are below the average for the EU-27.55 2017 2018 2019 2020 2021 Spending on ALMMs comprised 20.44 percent Employment incentives Training of the CES’s expenditures on active employment Subsidies for self-employment Public works / Direct policies in 2021 and 17 percent of its total budget and business expansion job creation (CES, 2021). This was mostly due to the scaling up Job retention subsidies TOTAL of two more complex categories of ALMMs with Source: Ministry of Labor, Pension System, Family, and Social higher costs per beneficiary, namely: (i) subsidies Policy. These estimates are based on statistical data held by the Croatian Bureau of Statistics and data provided by the Ministry of Labor, Pension 55 64 System, Family, and Social Policy. Only the spending on ALMMs is included, while expenditures on CES employment services and administration are excluded. For the calculation, GDP in current prices (in HRK) is taken from: https://podaci.dzs.hr/media/ l2wkiv1a/croinfig2022.pdf. According to Eurostat data, in 2020 (the most recent available comparative data), the EU-27 countries and the 19 countries from the Eurozone spent 0.6 percent of GDP on ALMMs (categories 2-7). According to the same statistics, Croatia spent 0.332 percent - https://webgate.ec.europa.eu/ empl/redisstat/databrowser/view/LMP_EXPSUMM/default/bar?lang =en&category=lmp_expend. TOC Chapter 5: Labor Activation Policies Can Strengthen Participation and Human Capital Harnessing Human Capital for Growth in Croatia for self-employment and business expansion and Figure 69: Spending on ALMMs by type of program, (ii) employment incentives that include the costly selected European countries, 2020 (% of GDP) internship subsidies and wage subsidies. This Sweden explains why spending on ALMMs had increased Slovenia Slovakia by the end of 2021, whereas their coverage was still Portugal lagging. Norway Netherlan… The CES’s spending on “active” interventions such Ireland as training, job rotation, employment incentives, Finland Denmark direct job creation, and entrepreneurship incentives Czechia was considerably lower than the European average Belgium – 0.33 percent of GDP in Croatia compared with 0.6 Austria percent in -the EU-27 and in the Euro area (Figure Croatia 69). Spending on labor market services was also Euro area EU27 lower in comparative terms – 0.07 percent of GDP 0 0.5 1 1.5 2 in Croatia versus 0.2 percent in EU-27. Spending Training Supported employment higher amounts does not itself guarantee success- Employment incentives and rehab ful labor market(re) integration. Compared to some Direct job creation Start-up incentives other countries, Croatia spends more on business Source: Eurostat. https://webgate.ec.europa.eu/empl/ start-up incentives and direct job creation and less redisstat/databrowser/view/LMP_EXPSUMM/default/ bar?lang=en&category= on training and employment incentives. In 2020, as a share of GDP, Croatia’s spending on start-up Figure 70: Distribution of spending on ALMMs, incentives was around five times higher than in the 2021 (%) EU-27 and the Eurozone countries, while its spend- ing on direct job creation was 2.5 times higher. At 9% 16% the same time, Croatia spent two to three times less 9% 2% than the EU average in terms of shares of GDP on 4% training, employment incentives, and employment 18% support and rehabilitation. Characteristics of and Chal- Employment subsidies Traineeship subsidies 42% lenges for the Current System Subsidies for self-employment and business expansion Training Upskilling subsidies While ALMMs on the labor demand side are Public works expanding over time, they are crowding out sup- Job retention subsidies ply-side ALMMs, which aim to increase clients’ employability and competitiveness in the labor market. Wage subsidies and subsidies for self-em- Figure 71: New entrants by category of ALMMs, ployment were allocated as much as 76 percent of 2017-2022 the ALMM budget in 2021 compared with 23 per- 2022 cent in 2017. In contrast, the share of the ALMM Number of new entrants budget spent on training and upskilling subsidies 2021 dropped sharply between 2017 and 2021 to just 6 percent (Figure 70). Similarly, over half of all new 2020 ALMMs were for employers as employment incen- 2019 tives and subsidies for self-employment. Partici- pation in training had significantly decreased even 2018 before the pandemic due to the phasing out of the program for workplace training for those without 2017 an employment contract and picked up again only 0 10 20 30 40 in 2022 (Figure 71). The recipients of self-employ- Employment incentives Training and upskilling subsidies ment subsidies comprised one quarter of all ALMM Subsidies for self-employment and business expansion beneficiaries in 2021 and 2022, while the allocated Public works / Direct job creation Job retention subsidies budget for these subsidies was as high as 42 per- Source: Data from the Ministry of Labor, Pension System, cent of Croatia’s total spending on ALMMs in 2021 Family and Social Policy, 2022. And Croatian Employment and even higher at 45 percent in 2022.56 This is con- Services, 2023. Data for 2022 are for the months January through August. 56 65 TOC Chapter 5: Labor Activation Policies Can Strengthen Participation and Human Capital Harnessing Human Capital for Growth in Croatia siderably higher than in previous years (for example, Figure 72: Distribution of entrants in ALMMs by the 12 percent that was allocated in 2017) and also age, educational background and duration of quite high by international standards. As of 2019, unemployment spells (2022, %) the average spending on ALMMs in the EU coun- Age tries was 6 percent.57 While expanding demand- side ALMMs is one approach to increasing Croatia’s 7% productivity,58 it will also be necessary to expand 6% ALMMs aimed at building new skills and reducing 20% skill mismatches.59 While all registered unemployed workers have equal legal access to ALMMs, they are dispropor- 17% tionately used by those who are relatively advan- taged being more attached to the labor market. 50% This group is characterized by people of prime age, with higher levels of education and a history of shorter unemployment spells. They tend to be more self-motivated to participate, more active in inter- Up to 24 years 25 - 29 years 30 - 49 years acting with employment counselors, and more able to meet the requirements for participation in a wider 55 years and older 50 - 54 years range of ALMMs. The key characteristics of workers who used ALMMs in 2022 are displayed in Figure Eucational background 72. Young people (those up to 29 years old) and 1% 2% prime-age people (aged 30 to 49) together com- prise 87 percent of all ALMM participants, which is 9% 18% disproportionately higher than their share of reg- istered unemployed, which is around 63 percent. Conversely, people aged 50 and older comprise 37 percent of the registered unemployed but only 13 11% 28% percent of those who use ALMMs. Only 12 percent of users of ALMMs are the long-term unemployed (those who have been out of work for 12 months or longer) who are as a rule in a greater need of 31% support, while 58 percent of ALMM users started paricipating after short spell of unemployment (up No schooling or incompleted basic school Basic school to six months) and another 20 percent have never 1 to 3 year vocational secondary school Unknown been unemployed. As many as 70 percent of ALMM 4 or more year vocational secondary school Non-university degree participants have completed secondary voca- University and post-graduate degree tional education or other non-university post-ba- sic school, which is similar to the educational profile Dspelluration of unemployment spell of all registered unemployed. However, apparently their vocational skills are obsolete and/or insufficient for staying employed in the present labor market. 20% Croatia’s ALMMs fall short of ensuring equity of access to excluded or strongly disadvantaged groups in the labor market. These include the long- term unemployed, older workers, young people 12% (particularly those who are NEET), people with dis- 58% abilities, and women. This is the result of the lim- 10% ited scope ALMMs and the fact that they are used mostly by those with fewer employment barriers in terms of age, skills, or length of unemployment spells. Only 12 percent of ALMM users have been unemployed for 12 months or longer, only 7 percent Up to 6 months 6 - 12 months are 55 years old or older, and only 3 percent have More than 12 months Not unemployed no schooling or have an incomplete basic educa- Source: CES (2022). 57 OECD Strictness of Activation Requirements Database. https://stats.oecd.org/Index.aspx?DataSetCode=LMPEXP. 66 58 The OECD recommends that, as countries recover from the COVID-19 crisis, they should increase their overall expenditure on ALMPs, and particularly support job creation by scaling up existing programs and introducing new programs and measures, which should increase demand for labor (OECD, 2021), 59 In 60 percent of OECD countries, the budgets for ALMMs expanded in 2020 and then again in 2021, most notably for employment incentives but also for training (OECD, 2022a). TOC Chapter 5: Labor Activation Policies Can Strengthen Participation and Human Capital Harnessing Human Capital for Growth in Croatia tion. The percentage of new entrants to ALMMs providing pensioners with information on available who belong to these disadvantaged groups is lower job offers, how to look for a job, and motivational than their percentage in unemployment. Their pro- messages and providing employers with informa- file characteristics frequently go hand-in-hand with tion on the kind of support for which they would low or obsolete skills. Often their labor supply-side become eligible if they were to hire a retiree.60 constraints are multiple and simultaneous and are combined with household-level constraints, espe- ALMMs are not yet targeted to young people cially in the absence of adequate social services. who are not in employment, education, or training Acknowledging these constraints, Croatia’s National (NEET). In all [youth?] age sub-groups, the share Recovery and Resilience Plan (NRRP) proposes to of NEETs is higher in Croatia than the average in develop social mentoring and job search services for the EU-27 countries. Among the NEET population hard-to-reach groups of beneficiaries such as per- in Croatia are those who are either actively seek- sons with disabilities, victims of violence, homeless ing employment or would like to do so. Around people, migrants, Roma, and young people leaving three-quarters of NEETs are registered with the CES. the social welfare system. The NRRP aims for these services to be covering a total of 26,400 people by Compared to the EU-27 countries, Croatia has a the end of 2025, helping them to find employment, high percentage of unemployed young people internships, or self-employment (especially for green with an upper secondary education and univer- and digital transition of the economy), with a focus sity graduates, which indicates skill mismatches. on young people and the long-term unemployed. Although Croatia is making some progress in fos- tering employment opportunities and social inclu- The country’s ALMMs have not yet adapted to sion for NEETs, challenges remain, specifically with the need to extend Croatians’ working lives and respect to the need to fine-tune ALMMs to the spe- to make better use of the working capacity and cific needs of young people, to improve outreach skills of older workers. It will be vital for ALMMs to strategies and targeting, and to reduce regional find ways to respond to the decline of working-age disparities in the availability and design of ALMMs population due to the emigration of many of the (European Commission, 2020). The limited capac- working-age population and the early retirement of ity of local CES offices is a barrier to providing all those who remain. The average duration of a Croa- NEETs with effective personalized guidance and tian’s working life increased from 31.2 to 33.5 years activation and to monitoring and evaluating the between 2012 and 2021, but this is still shorter than performance of different measures and the efficient the average for EU-27 (by some 7 percent in 2021). use of available funds (Tomić, 2015). The existing The widest divergence between labor force partici- partnerships between the CES and education and pation rates in Croatia and those in the EU-27 is for training institutions are designed primarily to target the over-60 age group. There are also very few older active NEETs (in other words, those who are regis- people who are self-employed or who work part- tered as unemployed). The CES has taken steps to time in Croatia. At the same time, Croatia has a high identify inactive NEETs by rolling out a NEET track- share of young pensioners as people in Croatia tend ing system that provide the basis for the develop- to retire early, which is a major motivating factor ment of measures for cohorts of NEETs who have for exiting the labor market. The CES rarely makes not been supported before. However, a peer review any effort to encourage older people to remain in of Croatia NEET policies (Tomić, 2015) revealed that the labor market, and workers aged 65 and over current policies and measures aimed at identify- are automatically removed from unemployment ing and supporting potential NEETs are limited in records. At the same time, Chapter 2 of this report scope, and are weakened by a lack of reliable basic argued that there is untapped potential to enable administrative data, weak collaboration between these workers to participate in the labor market for institutions, weak administrative capacity in those longer, especially for workers with a higher educa- institutions, and the heavy caseloads of CES coun- tion. However, there are significant disadvantages selors. to continuing to work for those with lower educa- tion, poor socioeconomic status, and health prob- Persons with disabilities are a distinct marginalized lems, even when they are willing to work. Prolonging group in the labor market that is not adequately the working lives of older workers will require public targeted by ALMMs. In 2022, persons with dis- policies that encourage them to extend their labor abilities represented only 4.3 percent of the new force participation, that reduce the barriers to their entrants to ALMPs, which was an increase over the employment, and that change the CES’s modus 2.3 percent in 2019 but was still very low. The only operandi. Initial steps are being made in 2023 with activation measure that is explicitly targeted to this the launch of a CES web portal that will be aimed at group is hiring subsidies to employers. In contrast See https://umirovljenici.hzz.hr/. 60 67 TOC Chapter 5: Labor Activation Policies Can Strengthen Participation and Human Capital Harnessing Human Capital for Growth in Croatia to the OECD’s and EU’s best practices, there are was smaller (2.8 peecentage points), meaning that no specific ALMMs that address the specific skills the prospect of women being referred to an ALMM needs of people with disabilities or manage their has decreased. Some recent changes to the ALMM up-skilling or re-skilling. Within the current ALMM portfolio61 that focus on female-dominated profes- mix in Croatia, there are also no measures for adapt- sions have the potential to increase female partici- ing workplaces or sheltered workshops for people pation in ALMMs, but more needs to be done given with severe disabilities. There are also no self-em- that around one-third of Croatian women aged ployment or entrepreneurship measures tailored between 25 and 64 are inactive because of their to persons with disabilities or legal guarantees and care responsibilities. A recent study identified care mechanisms to help them to claim disability and responsibilities and the related likelihood of having other benefits if they are not successful. the Recov- little or no recent work experience as substantial ery and Resilience Plan addresses these challenges barriers to employment for women across different by proposing several measures aimed at increasing age groups, education and qualifications, and mari- the employment of persons with disabilities, such tal status (World Bank, 2018a). as ensuring the physical accessibility of buildings and the housing environment, providing adequate The cost-effectiveness of some ALMMs should mobility services, promoting hybrid access to work- be reassessed. It is important to monitor the places, and ensuring access to digital public ser- effectiveness of demand-side ALMMs using pro- vices. In most EU countries, the labor (re)integration gram-level data and evaluations of their perfor- of persons with disabilities is a key mandate of the mance and impact.62 The costs per beneficiary vary employment services, which co-operate with other widely among ALMMs and are highest by a signif- institutions who work with people with disabilities icant amount in the case of programs for self-em- to provide them with vocational rehabilitation and ployment and for internships (Table 14). In general, other measures. higher unit costs tend to reflect a more intensive investment in building sophisticated skills or, alter- It is less likely for women to be referred to an ALMM natively, in removing more and multiple barriers than men, and over time their participation has to (re)entering the labor market. While higher unit been declining. Currently, the share of women share costs for individual programs can be justified, this in ALMM users is 6.5 percentage points lower than is only the case if these programs can also demon- the share of women among the registered unem- strate proportionally higher effectiveness, for exam- ployed. Female participation in ALMMs was slightly ple, in terms of job placements (World Bank, 2022c). higher in 2019 before the pandemic, and the gap This is why it is important to carefully monitor which with women’s share of the registered unemployed ALMMs provide the best value for money. Table 14: Cost per beneficiary of ALMMs, 2021 Category or sub-category of Expenditures, HRK Beneficiaries Average cost, HRK Average cost, EUR ALMM Employment subsidies 192 970 130.69 5 698 33866.3 4495.1 Internship / traineeship subsidies 221 010 245.13 3 636 60783.9 8067.4 Subsidies for upskilling/training 25 091 775.39 2 058 12192.3 1618.3 Subsidies for self-employment 505 919 100.60 6 332 79898.8 10605.0 and business expansion Training for unemployed and 28 956 169.74 3 599 8045.6 1067.9 other jobseekers On-the-job training 14 523 880.41 564 25751.6 3418.0 Public works / Direct job creation 108 832 005.46 3 184 34180.9 4536.8 Permanent seasonal worker 57 975 406.78 4 366 13278.8 1762.5 Subsidies for preservation of jobs 57 116 422.43 1 955 29215.6 3877.8 and/or shortened work hours Source: Ministry of Labor, Pension System, Family and Social Policy, 2022. Note: the average cost includes only the direct costs of the respective ALMMs and no additional (administrative and other) costs of CES. Beneficiaries are the new ALMM entrants throughout the respective year. Conversion to euro is at the exchange rate of EUR 1 = HRK 7.5345. 61 Like launching traineeship subsidies in the public sectors in female-dominated professions, launching training vouchers, and increasing in job 68 retention subsidies for seasonal workers and for workers in female-dominated industries like textile and clothing. 62 Program evaluations should be mandatory and implemented on a regular basis (Bejakovic, 2016). The most recent publicly available evaluation commissioned by the CES covers ALMMs from 2010 to 2013 (CES and Ipsos Puls, 2016). TOC Chapter 5: Labor Activation Policies Can Strengthen Participation and Human Capital Harnessing Human Capital for Growth in Croatia Lessons From Current and Past of the subsidy, and post-business creation services (World Bank, 2022c and Ribe et al, 2010). If Croatia’s ALMM Programs self-employment subsidy were to be redesigned to Subsidies for self-employment and business expan- be ”integrated” or “multi-component” in this way, sion dominate the ALMM mix, but these measures it would be more effective in terms of using the are expensive and are usually not adequate for human capital of people with a weaker attachment those with a weak labor market attachment. The to the labor market. subsidies for self-employment are designed to pro- vide start-up capital to those individuals or groups Wage subsidies are useful to support the employ- among the registered unemployed who decide to ment of people who are in an unfavorable position start or expand their own business. The amount of in the labor market by financing the costs of their the subsidy depends on what industry is involved labor, thus providing employers with an opportu- and is higher for start-ups in green or digital indus- nity to “test” prospective employees at lower than tries.63 The subsidy has two components: (i) a fixed full wage costs. These subsidies should result in amount to cover part of the initial costs of business a net increase in the number of employees in any registration, the cost of one full-time employee, given year. Currently, hiring subsidies in Croatia are and recurrent costs for the first few months, and (ii) available for two well defined target groups: (i) reg- a variable amount to fund the purchase of equip- istered unemployed who have been identified as ment, vehicles, raw materials, and consumables, to disadvantaged due to facing multiple barriers to purchase or lease IT programs, or to buy a franchise. employment and (ii) persons with disabilities. These The subsidy for business expansion is granted to a are target groups for whom other measures have business that has already received self-employment proved ineffective and who are left with no alter- support when it plans to expand by opening a new natives to access jobs or build their skills. These unit or facility with new hires from among the reg- hiring subsidies are time-limited and differentiated istered unemployed. Qualitative and quantitative depending on the reason for the person’s labor evaluations have found that self-employment sub- market disadvantage and completed level of edu- sidies have a strong positive net effect on employ- cation and are higher for persons with disabilities. ment, the high survival rate of the subsidized busi- The employer is obliged to keep the co-financed ness initiatives, and the high level of satisfaction employee on an employment contract and to pay of subsidy users due to their improved financial their salary and benefits after the subsidy is over for status.64 These measures have proved to be most a period equal to the duration of the subsidy. From successful when those who receive the start-up 2022, employment subsidies are explicitly targeting funds are attached to the labor market, can invest employment in green and digital industries. their own funds, including after the subsidy is over, and have specific skills, mindset, and motivation.65 When time-limited, well-designed, and tailored to However, such characteristics are not typical of needs, wage subsidies can be a cost-effective way those who have a weak attachment to the labor to reduce unemployment, strengthen the employ- market. Therefore, it is important to design these ability of workers, and support the most vulnera- subsidies in such a way that they foster successful ble (OECD, 2021). In Croatia, a quarter of all ALMM outcomes for the vulnerable groups. For example, users are employed through hiring subsidies, which it is important for the CES to build its capacity and account for 16 percent of all budgetary allocations instruments for assessing proposed business ideas, for ALMMs. In the past, an evaluation assessed the to identify which prospective recipients of the sub- net effect of the subsidy on employment as being sidy are most likely to succeed,66 and to reach out positive for both the employers and the subsidized to groups that are under-represented in the labor employees. The highest positive effect was observed market. This particularly applies to unemployed for young people with no work experience (CES and women who are less active on the labor market than Ipsos Puls, 2016).67 However, these positive effects men and much less involved in entrepreneurial activ- were often limited to the period when the subsi- ities. For example, the current temporary financial dies were paid and declined thereafter. Based on support programs could be supplemented by advi- the findings of that evaluation, the CES redesigned sory services on sustainable business development the subsidy amount to cover the worker’s full gross and by training in entrepreneurship and financial lit- salary instead of only 50 percent as in the past. This eracy. International evidence suggests that the key makes it more attractive for employers and should factors for success are access to finance beyond increase coverage of people who face multiple crit- the direct start-up subsidy, institutional support for ical barriers to employment. Wage subsidies also access to markets, land, and other physical assets or enable these subsidized workers to “learn by doing,” inputs, advisory services related to the effective use which builds their human capital. It also ensures that 63 The criteria for green and digital activities and workplaces are defined by the CES (see: https://mjere.hr/katalog-mjera/kriteriji-zelene-digitalne-djelatnosti-radna-mjesta/). 64 An evaluation of the previous version of this subsidy (Start-up Incentives for Unemployed Individuals) revealed a strong positive effect on employment. The subsidy also resulted in 69 a high survival rate of subsidized entrepreneurial activities (CES and Ipsos Puls, 2016). 65 According to the CES and Ipsos Puls evaluation, the majority of self-employment subsidy recipients were people who used to be registered as unemployed for only a short term (less than three months). One quarter of them had taken the subsidy in order to "legalize" the work that they already performed. Many "successful" users of the subsidy were convinced that it was inappropriate for the "average" unemployed, in other words, that a successful outcome depended on specific personal qualities such as "responsibility," "organization," "diligence," and “market experience,” or “market or line of work familiarity.’’ 66 A step in this direction is the workshops for self-employment organized by the CES. 67 The evaluation of hiring subsidies from 2010 to 2013 revealed that the likelihood of being employed six months after receiving the subsidy was twice as high for young participants with no work experience thn for non-participants (CES and Ipsos Puls, 2016). TOC Chapter 5: Labor Activation Policies Can Strengthen Participation and Human Capital Harnessing Human Capital for Growth in Croatia their existing skills do not depreciate over time as With an increase in funding and a stronger focus tends to happen during long unemployment spells on upskilling and reskilling, training subsidies have (Almeida et al, 2014). the potential to build skills and tackle skill mis- matches and labor shortages. Experience suggests Subsidies for traineeships, internships, and appren- that training programs improve skill matching, while ticeships are effective in giving young jobseekers vocational training programs facilitate the transi- first-job experience and improving their prospects tion from education to work. An evaluation of Cro- for subsequent employment. This kind of subsidy atian programs providing for labor market-oriented program was launched in Croatia in 2018 aimed training between 2010 and 2013 found that close at giving the registered unemployed with no work to a quarter of the participants started a job requir- experience the opportunity to work in the profes- ing the qualifications that they achieved as a result sion for which they were educated and to learn how of their training. However, for around 40 percent to work independently. Private sector employers of trainees, the probability of finding a job did not are given a monthly lump-sum per intern regardless increase significantly. Moreover, more than half of of how much they actually spend on their intern’s the trainees did not work in any job that required wage, which varies according to their level of edu- the qualifications for which they were trained. Also, cation and the possibility of tax relief. For green over 40 percent were unemployed for two years and digital industries, the amount of the subsidy after participating in the training program. The is higher. In the case of public services, the subsi- evaluation also highlighted the difficulties involved dies cover the intern’s full wage and transportation in forecasting labor market demand for the purpose expenses when the internship is in healthcare, edu- of designing training content. The responsibility for cation, preschool education, social welfare, and cul- designing training programs is delegated to staff ture. Those who are eligible to be subsidized are the at the CES regional offices to ensure that the pro- registered unemployed with less than six months grams are compatible with local labor market needs, of registered insurance-paying work experience at but these staff need more information and assess- their educational level in their public service field. ment tools. To a significant extent, training was The internship subsidies were scaled up significantly ‘’supply-driven,’’ and CES counselors were promot- in 2021, with both the number of beneficiaries and ing enrollment to keep the training programs going the level spending almost doubling. Internships in despite low demand. the green and digital sectors and in the public ser- vices commenced in 2022, albeit on a small scale. Job retention subsidies or short-time work schemes In 2021, interns accounted for 11.6 percent of all are beneficial for retaining qualified seasonal ALMM entrants, and internships accounted for 18 workers. Currently, Croatia’s income maintenance percent of the total ALMM budget. These subsidies and support for unemployed workers consists of are the second most expensive (in terms of cost per the Permanent Seasonal Worker measure and two beneficiary) after the subsidy for self-employment smaller-scale job preservation subsidies, one for job and business expansion. retention when employers are in temporary difficul- ties and, two, short-time subsidies for employers Global empirical evidence shows that internships who are forced to reduce their employees’ work- are associated with better post-internship employ- load. The Permanent Seasonal Worker measure is a ment outcomes for interns, and this has also been financial incentive given to employers to retain key the case in Croatia. They are instrumental in pro- staff at times of reduced or absent workload due viding the registered unemployed with work expe- to seasonal nature of their business, which covers rience which – according to mounting evidence – is the pension insurance of seasonal workers for six one of the most important factors that influences months. It also provides seasonal workers with employers’ recruitment decisions. Since the cost monthly financial assistance for up to six months per beneficiary is one of the highest for all ALMMs, as an incentive to stay with the same employer. it is important to prioritize these internship subsi- An evaluation of the Permanent Seasonal Worker dies, for example, by focusing on public sectors that measure revealed that take-up from employers was are experiencing or will experience labor shortages variable and, in some years, low, even though they and on regions that are lagging behind in economic could benefit from lower expenses as a result of the development. Also, in line with international expe- CES co-financing. To some extent, the reason for rience, internship subsidies could also be provided this is the scheme’s administrative complexity, as a to young people with disabilities to enable them to result of which the measure is most often used by use their human capital. This would require com- larger companies with greater operational capac- bining the internship subsidy with broader training, ity. Another reason is the existence of a limit on the counseling, and even social services. number of eligible seasonal workers as a percent- 70 TOC Chapter 5: Labor Activation Policies Can Strengthen Participation and Human Capital Harnessing Human Capital for Growth in Croatia age of regular employees, which might discourage ally enhanced their employability in the long term. smaller businesses from participating. According For the majority of users, participating in public to international evidence, job retention schemes works was most beneficial as a temporary way to should be temporary and should be designed care- overcome their financial difficulties. Roma who have fully to avoid supporting unviable jobs in companies frequently participated in public works (mostly in with structural difficulties. Also, job retention subsi- municipal public works and in works organized by dies are increasingly being combined with training non-profit Roma associations) also acknowledged that is provided when the trainee is not working or that their self-confidence had increased when is working reduced hours, with the CES reimbursing looking for another job as well as appreciating employers for the costs of training. the access to health care that comes with partici- pation in public works (CES and Ipsos Puls, 2016). Public works that target groups facing significant employability and participation constraints have Policy Implications not been able to significantly increase their prob- ability of employment. In Croatia, these programs In the coming years, policymakers should focus on are defined as providing socially beneficial work that experimentation, analysis, and evaluation. It will be takes place over a limited period to the registered important to understand the reasons for low cov- unemployed from specific target groups that are far erage rates and take-up of ALMMs and for their distant from the labor market. These public works existing targeting bias towards workers with fewer are initiated by local communities or civil society barriers to being employable. In this section, we organizations in cooperation with local govern- discuss some specific lessons from experience that ments with the aim of increasing social inclusion. could serve as the basis for experimentation and for They involve projects in the fields of social welfare, improving the design and delivery of ALMMs. education, environmental protection and preserva- tion, and maintenance and communal works. They Further integration of ALMMs with social services provide participants with earnings from work but will be essential as well as strengthening cooper- also build their ‘’soft skills’’ and self-confidence and ation and service-delivery partnerships with the encourage them in further efforts to (re)join the private sector and at the local level. This should labor force. Their target groups are: (i) people in a involve creating a network of trusted institutions disadvantaged position on the labor market regard- and public services to reduce barriers to employ- less of the length of their registration as unem- ment and difficult life situations such as ill health ployed (including parents with four or more children, and disability or family issues that prevent people parents of children with developmental disabilities from participating in the labor market. Policymakers or suffering from malignant diseases, persons with should also explore competitive ways to outsource disabilities, victims of human trafficking or domes- different labor market services to private providers. tic violence, Roma people, and homeless people); (ii) beneficiaries of the guaranteed minimum benefit; The support for disadvantaged groups in the and (iii) the very long-term registered unemployed labor market needs to be reinforced. International (for over 24 months). Public works are used to test experience suggests the need to use ALMMs and the ability to work of potential recipients of the measures that are sensitive to the needs of target guaranteed minimum income benefit. groups, especially those who are most distant from the labor market. In Croatia, these groups include Participation in and spending on public works are older employees and jobseekers, the long-term decreasing in Croatia as of 2022, and longer-term unemployed, minority groups, people with disabil- trends are not clear. Several developed countries ities, young people especially NEETS, and women. have downsized or even phased out such programs,68 mostly because their effectiveness in terms of ena- In the Croatian context, this means that following bling participants to find jobs on the open market efforts need to be made to adapt the business pro- is questionable. These programs are also unlikely to cesses, governance, and management of the Croa- address existing skills shortages. A meta-analysis of tian Employment Service (CES): evaluations of ALMMs concluded that they are gen- erally ineffective in the short, medium, and longer One direction would be to use modern technol- term (Card et al, 2018). In the case of Croatia, several ogy to profile clients and to conduct an in-depth evaluations have concluded that participating in assessment of barriers preventing people in public works did not directly increase the probabil- these groups from working such a limited liter- ity that a participant will find employment, though acy, educational and vocational skills barriers, at least it did not reduce it, and that it only margin- and spatial obstacles. For example, Denmark, Estonia, Israel, Norway, and Switzerland do not use these programs anymore. Other countries have shifted their 68 71 spending from public works towards ALMMs such as training and employment incentives, which they deem to be more effective (OECD, 2021). experience,” or “market or line of work familiarity.’’ TOC Chapter 5: Labor Activation Policies Can Strengthen Participation and Human Capital Harnessing Human Capital for Growth in Croatia Integrate ALMMs with other social services to menting the existing career guidance in Croatian provide interventions to increase labor market schools with additional CES staff and additional inclusion. This might involve targeted coun- financing. seling for the resolution of debt issues, inten- sified counseling to foster the motivation to • Strengthen Youth Guarantee and the YEI (Youth work, medical counseling, and social counseling Employment Initiative) to support young people (for example, on housing or the management of with physical and especially mental health chal- household finances). lenges. This would be a way to increase the cov- erage of those with NEET status, to target NEET The existing training programs need to be subgroups more effectively, and to ensure that adapted to provide upskilling and reskilling they are offered real job opportunities rather training to older workers and the lower-skilled than poor quality traineeships.70 unemployed who have been unemployed for a long time. • Foster the development of socio-emotional skills, which can be offered as a separate service NEETs and Young People (for example, as part of job counselling) and/or as part of different ALMMs (such as vocational Supporting young people’s employment prospects training). requires a recognition that they face different and age-specific barriers in the labor market, such as Older Workers a lack of work experience and limited professional networks. These realities need to be considered in Tailoring training methods and time horizons to the design and implementation of ALMMs. Some the learning needs and abilities of older workers.71 ideas for how to design and implement policies Short-term training programs are usually more specific to NEETs and young people are as follows: effective for older workers as assigning older individ- uals to long-term training and subsidized employ- • One-stop shops for the delivery of ALMMs, edu- ment is unlikely to be efficient. The training content cation, social services, and benefits for NEETs. must be designed to meet the special needs of CES regional offices should strengthen their older workers. Research has shown that older work- cooperation with educational institutions with a ers, including those with cognitive decline, can learn special focus on hard-to-employ groups as well new skills if the training is adapted to their learn- as their cooperation with the Centers for Social ing needs and abilities. For example, older learners Work beyond data exchange and towards pro- should be given ample time to process new knowl- viding more integrated package of services to edge. International experience has shown that fos- NEETs and young people (Tomic, 2015). tering the employment of older workers requires programs that take the following actions: • Strengthen the current outreach service to NEETs.69 NAV runs such a service, which is • Increase incentives for employers and job- implemented by the municipalities. The service seekers. Financial grants or vouchers could be cooperates with youth clubs and schools, child given to employers to encourage them to pro- welfare services, school health services, health vide training geared to older workers. This is centers for young people and their families, and likely to be more effective than tax incentives in the police to seek out young people who need increasing the provision of training and improv- employment assistance but are not inadequately ing employment outcomes. Another approach supported by other institutions. is to provide jobseekers with individual learning accounts (ILAs), which are bank accounts man- • Monitor young people who are on the verge of aged by the workers themselves to fund contin- dropping out of upper secondary school. An uous learning (Vodopivec et al, 2018). early intervention service would consist of indi- vidual tracking and follow-up of potential upper • Adapt labor market job matching methods for secondary school dropouts. The objective would older workers. Job search assistance and place- be to increase the number of learners who com- ment services should be redesigned to be more plete upper secondary education and engage in targeted towards older individuals. Job search the labor market. While NAV through the munic- assistance may be more effective for older than ipalities closely monitors low-performing stu- for younger workers, especially when older work- dents in upper secondary education to prevent ers possess sufficient skills and previous labor dropout, this could be strengthened by supple- market experience but are lacking job search skills. 69 See: https://www.cedefop.europa.eu/en/country-reports/inventory-lifelong-guidance-systems-and-practices-norway#guidance-for-neet. 72 70 European Parliament resolution of February 17, 2022 on empowering European youth: post-pandemic employment and social recovery (2021/2952(RSP)). 71 The characteristics and recommendations are summarized from Vodopivec et al (2018) and European Commission (2019a). TOC Chapter 5: Labor Activation Policies Can Strengthen Participation and Human Capital Harnessing Human Capital for Growth in Croatia • Promote flexible work arrangements, including ployed and moving into training and education, part-time jobs, whereby workers can gradually temporary work placements, or employment is reduce their working hours while remaining in the often very fragile. It is important to keep provid- workforce but with a greater sense of well-being. ing individuals with post-placement support for at least three months to minimize the chances of • Challenge and change employers’ negative per- them dropping out. ceptions about older workers. A large number of countries have introduced awareness-raising Women campaigns to tackle negative social percep- tions about older workers by highlighting the Women face more constraints to participating in strengths, benefits, and opportunities of having ALMMs and accessing employment (world Bank, an age-diverse workplace. 2022c). The most common barriers that they face include time constraints related to their higher Long-Term Unemployed and Minority burden of family and care responsibilities than Groups men, their more limited mobility, their lack of pro- fessional experience, smaller social and profes- A mix of ALMMs, services, and interventions is typi- sional networks, limited agency, restricted access cally necessary to bring the long-term unemployed to productive inputs (such as finance and land), and closer to the labor market.72 According to interna- social norms that define their domestic responsibil- tional experience, this mix could include: ities and limit their occupational choices. Therefore, ALMMs and employment services should adopt the • Personalized and intensive counselling by trained following measures that have been shown to facili- counsellors supported by psychologists, to build tate female enrollment and employment: skills such as self-confidence and to overcome challenges related to motivation as well as to sup- • Affordable childcare and transportation support port an enabling environment within the family. and/or stipends to cover the expenses asso- ciated with participating in the program. Also, • Social support services to alleviate difficult life training, and other services should be provided circumstances of the long-term unemployed. in locations that are safe and easily accessible to These services might include guidance on women. administrative procedures and about how to access healthcare, education, or training ser- • Job counselling and training activities to vices, mental health counselling, debt-counsel- enhance women’s decision-making capacity ling, marriage counselling, and housing support. and self-esteem through, for example, soft skills training and/or coaching support. • High-tech job search assistance and placement. AI-based technology can enhance skills profiling • Orientation and intermediation services to pro- and career assistance services. For example, it vide women with information (for example, on can make it easier for CES counselors to profile the wages paid for different occupations) and individuals’ skill sets and match them with career proper career guidance to foster women’s aspi- and learning pathways. It can also help CES case rations and decisions about training and career workers to operate more efficiently, to improve choices. These activities should be designed to their chances of placing clients, and to focus deconstruct gender-specific stereotypes and their attention and efforts on more vulnerable address occupational segregation, for example, candidates. by highlighting successful women and former female participants as role models. • Support to employers for providing intern- ships and on-the-job training and subsequent • Expanded and strengthened training for women employment for this target group. This involves in the skills required for non-traditional and identifying suitable companies in the local area, male-dominated occupations (which provide assessing which positions are open and what greater employment opportunities for quality skills are required and informing them about the jobs and decent earnings) as well as soft-skills benefits available to them, such as subsidies for training. For instance, topics such as self-con- hiring the long-term unemployed. fidence or negotiation skills may be especially helpful for women as social norms have meant • Post-placement support (drop-out prevention). that they are less likely than men to have devel- The transition period between being unem- oped these skills earlier. The recommendations are summarized from World Bank (2022c). 72 73 TOC Chapter 6: Social Protection and Human Capital Harnessing Human Capital for Growth in Croatia Chapter 6: Social Protection and Human Capital Introduction The goal of social protection is not just to pre- 48 percent among people above 65 living alone. vent poverty but also to protect and improve the skills and abilities of these individuals. Well-de- The safety net programs primarily target spe- signed social protection systems and tools can be cific groups of people. Improving and expanding transformative. They help to reduce economic and the safety net through income testing is crucial to individual shocks that may push a household into lessen poverty and promote human development. poverty. This way, families can maintain their human Poverty reduction via social assistance programs capital potential instead of being trapped in pov- depends on coverage, targeting, and adequacy of erty. Financial assistance and efficient social and support to the intended population (Box 4). The employment services are at the heart of boosting analysis of the targeting the low-income families human capital and can empower people to pursue (Figure 73) is especially useful to assess whether the health, education, and employment. Cash transfers safety net aimed at addressing poverty alleviation help prevent poverty traps by providing financial targets poor and low-income households and, thus, stability and preventing old-age poverty. Social whether it is cost-effective. According to the World care services stimulate the use of health and edu- Bank analysis of the Croatian safety net, there is a cation services which is especially critical for chil- significant under-coverage of the poor and those dren. Employment services and unemployment at risk of poverty. There are two key anti-poverty benefits increase participation in skills development programs in the country. First, the Guaranteed Min- programs that equip job seekers with higher-qual- imum Benefit, and second, the child benefit; both ity skills for more productive and better-paid jobs. the programs are income and asset test condi- When these are integrated, they could effectively tions. However, taken together, the child and GMB help those who, for various reasons, have dimin- (Guaranteed Minimum Benefit) benefits only cover ished earning opportunities because of a disability, around half the poorest 25 percent of the popula- care duties, or insufficient skills. tion. The child benefit provides the most coverage (50 percent of the poorest decile). The GMB pro- Although Croatia has a comprehensive social pro- tection system, poverty remains a significant threat Figure 73: Targeting performance: share of total to human capital. Around 100 centrally executed beneficiaries across income deciles social assistance programs aim to address risks 100% such as poverty, unemployment, disability, health, 90% old age, and survivorship needs. In addition, around 80% 50 to 60 social assistance programs are managed at the local government level (Pezer and Urban, 70% 2017). Yet, the national poverty rate remains high at 60% about 18 percent, and the at-risk-of-poverty was at 50% about 20 percent in 2022. Poor households in Croa- 40% tia cannot devote sufficient resources to education, 30% health, and mental well-being, especially for their 20% children. According to a World Bank survey, 38 per- cent reported difficulties making ends meet in 2021 10% (World Bank, 2021b). Single-member households 0% are a group with a substantial risk of poverty and 1 2 3 4 5 6 7 8 9 10 households with three or more children. While the GMB Child benefits risk of poverty among all households without chil- Income deciles dren was 22 percent in 2017, it increased to 44 per- Source: World Bank calculations. Note: GMB coverage for cent (almost twice as much as within the EU28) and 2022 is estimated using 2018 incidence data. This chapter was prepared by Nithin Umapathi and Juul Pinxten. 74 TOC Chapter 6: Social Protection and Human Capital Harnessing Human Capital for Growth in Croatia Box 4: Key indicators for assessing social assistance performance Coverage Coverage measures the direct and indirect beneficiaries receiving social assistance in a target group identifiable in the nationally representative household survey data. A target group might be the total population, first income quintiles, or total population in urban and rural areas. Conceptually, program coverage is a necessary but insufficient condition for a program to be effective for its target group: only if the program reaches its intended beneficiaries (targeting) and provides adequate benefits will it have a chance to ameliorate the condition for which it was initially designed. Targeting Targeting assessment is based on benefit incidence, which measures how much of the program’s ben- efits reach the poorest or desired characteristics. Benefit incidence is the transfer amount received by a group as a percentage of the total transfers received by the population. For example, a group could be the poorest decile or the lowest three deciles. In other words, benefit incidence is calculated for the total population, quintiles, or deciles. Adequacy The adequacy of a program is calculated as the value of the transfers received by a group divided by the total consumption or income of the beneficiaries in that group. Adequacy, therefore, measures the share of the welfare of recipients represented by each transfer. gram only covers around 20 percent of the poorest Figure 74: Spending on SA by type of program as decile or 2 percent of the population. a share of GDP in 2021 Child benefits 0.27% Compared to other EU countries, social assistance expenditure is relatively low. Central non-contrib- Assistance and care allowances 0.21% utory social assistance programs account for only about 1 percent of GDP, and around 0.4 percent is Personal disability allowances 0.16% targeted to low-income households (World Bank, 2019a and 2022b). The composition of non-con- Guaranteed Minimum Benefit 0.08% tributory social assistance spending in 2021 shown in Figure 74 demonstrates that the safety net Other social benefits 0.08% system prioritizes children and caregivers. The share of means-tested social assistance which targets National allowance for the elderly 0.01% low-income status includes the Guaranteed Mini- 0% 0.05% 0.1% 0.15% 0.2% 0.25% 0.3% mum Benefit (GMB) and child However, the expend- share of GDP iture levels are small relative to poverty. Source: MoLPFSP, WB staff calculations. Figure 75: Total social assistance spending and share of which are means tested 1.0% 0.9% 0.8% 0.7% 0.6% 0.5% 0.4% 0.3% 0.2% 0.1% 0.0% 2016 2017 2018 2019 2020 2021 Total social assistance Share of social assistance that is means-tested Source: MoLPFSP, WB staff calculations. 75 TOC Chapter 6: Social Protection and Human Capital Harnessing Human Capital for Growth in Croatia Enhancing the Effectiveness of tive data, an analysis of the difference between the number of applications and approvals for the GMB Social Assistance in Reducing for recent years, both overall and by locality, could Poverty provide information that would help to understand whether a home visit has any effect. The proce- Raising the percentage of means-tested aid to dure may be unnecessary as only a small fraction decrease poverty cost-efficiently through social of households typically misrepresent their welfare assistance is recommended. This would result in an status. Household visits are expensive in terms of efficient reduction of poverty compared to other administrative and staff time costs which in some programs. For comparison, Figure 76 shows that localities could lead to delays in processing because GMB and child benefits decreased poverty per of staff time and transportation constraints. If this is expenditure unit more than other transfers. Using the case, the total cost of visiting all the households these estimations, the World Bank compared how exceeds the potential savings from prevented cases much poverty could be reduced if social assistance of ineligible benefit recipients. It is also a possible coverage among the poorest 20 percent of the source of low GMB coverage and administrative population is increased through the GMB, which inefficiency and not a requirement for the means- comprised 0.03 percent of GDP in 2018. The model tested child benefit. Since each case worker (CSW) shows that if the GMB expenditure were scaled carries out household visits and incorporates the to 0.5 percent of GDP, the program would cover findings into the application assessment process, approximately 90 percent of the poorest 10 percent, it is also possible that staff in the offices may use resulting in a poverty reduction impact of roughly different standards. For example, this could lead to 3.3 percentage points. As such, the GMB could be geographic inequity; in other words, higher or lower developed to become the flagship anti-poverty standards may be applied depending on whether program in consolidation or coordination with the the CSW is in a richer or poorer region. It is also pos- means-tested child allowances. The increase in sible that the staff responsible for home visits may access can be implemented by simplifying the GMB be expected to identify evidence of extreme pov- administrative procedures since some requirements erty, with severe conditions such as lack of heating, for GMB applications are more restrictive than other water, electricity, or food for children. This could lead countries and may discourage access. Specifically, to eligible but less extreme cases needing approval one is the requirement for a property lien in case as GMB recipients. incorrectly provided benefits need to be “clawed back.” To increase access, consideration could be Implementing a risk-based approach to encour- given to eliminating the condition. Another possible age participation by using algorithms that suggest factor contributing to low coverage is stigma, with home visits selectively can reduce the likelihood of visibility within the community due to the need to ineligible individuals and fraudulent activities. The apply for the benefit in a social welfare center and basic idea of a risk-oriented approach is to select the required home visit. households for inspection visits based on risk profil- ing. This can be automated using statistical methods It is necessary to review the process of mandatory for identifying a high probability of inconsistency home visits for GMB applicants. Using administra- between reported and actual income and wealth. Figure 76: GMB has the most poverty reduction per budget spent. (Cost-effectiveness of major direct transfers in reducing poverty, figures show poverty reduction per HRK million spent) -0.002 Poverty impact HRK million spent ($5.5 international poverty line) -0.0015 -0.001 -0.0005 0 Child benefit Guaranteed minimum Disability pension & Unemployment benefit benefits benefit 76 TOC Chapter 6: Social Protection and Human Capital Harnessing Human Capital for Growth in Croatia The technique relies on identifying predictors of when calculating social assistance benefits. While household characteristics correlated with poverty the benefit formulas should be strengthened by status or fraud. There are several advantages of this allowing a higher proportion of disregarded earn- more selective approach to home visits: introduc- ings, whether this inactivity trap is as pervasive as tion of more objectiveness, administrative savings, perceived is another question. Academic research simplification of the enrollment process, avoidance has shown that high PTR is not the leading cause of inconvenience for many eligible households, and of unemployment. Lack of employment-related more cost-effective administration of the program. capabilities seems to be a more significant barrier to There are good examples of this approach used in employment that hinders access to the labor market. welfare systems in Europe. This change would also According to Bejaković et al (2014), the high-level be relevant to make the program more crisis-adap- PTR in Croatia is not associated with labor market tive; for example, any potential scale-up of the participation, and Croatian GMB recipients with program in its current form may be impeded by vocational education transition into employment the speed at which staff can implement household about twice as often as others, suggesting the key visits. role of education over financial disincentives. Another factor that can discourage work is the high Ensuring Proper Design of level of benefit generosity when people receive Activation and Graduation multiple transfers. The GMB is not very gener- Components in Safety Nets ous on its own. It is calculated based on HRK 800 (€106) and has not been revised since October 8th, Croatian policymakers are worried that welfare 2014. According to the rules, the benefit amount payments may discourage work and hinder the uti- shall not exceed the gross minimum wage (in 2022, lization of human capital. There is concern among HRK 4,687.50 (€623). However, the generosity is many in Croatia that the government may offer higher because GMB recipients are automatically insufficient social support to impoverished individ- eligible for several other cash and in-kind benefits, uals to prevent reliance on welfare. High PTR (Par- which could end up as a generous “package” that ticipation Tax Rate) rates could in theory create an includes housing allowance (cost of rent, electricity, "inactivity trap" for those unemployed and receiv- gas, heating, water, water drainage, other); energy ing benefits with low earning potential. This situ- subsidy for vulnerable energy consumers; a heat- ation occurs when employment results in little or ing allowance for recipients of GMB who use wood no increase in disposable income. This would occur for heating; supplementary health insurance; a 50 when the combined effects of benefit withdrawal percent subsidy for school textbooks for primary and higher tax burdens erode in-work income. and secondary school pupils from GMB families; According to a 2013 study, the transition from inac- coverage of transportation expenses for secondary tivity to single-earner minimum wage resulted in a education; and additional cash and in-kind benefits PTR above 100 percent for individuals from job- extended by local government administrations. less households with dependent children and sin- gle-parent families (Bejaković et al, 2014). For a sin- Although a high participation tax rate can discour- gle-person or two-person household, PTR for such age work, activation tools and "in-work" require- a transition is still substantial but lower (63 percent ments can help reduce this effect. Activation refers and 72 percent), primarily due to GMB being far to policies to support access to employment, edu- lower than the minimum wage. Similarly, in the case cation, and training for all those facing barriers to of a dual-earner couple where the primary earner entering the labor market and who are receiving earns 100 percent of the average wage, the poten- social welfare payments, including unemployment tial for an inactivity trap in Croatia is generally higher benefits, income support for people experiencing than the EU average, and for single individuals with poverty, benefits for single parents, people with two children with earnings below 67 percent of the disabilities, immigrants, and older people. Such average wage where the inactivity trap can be as conditions and assistance are aimed at overcoming high as 127 percent, more than twice the EU aver- dependency and graduating from social assistance. age, indicating strong financial disincentives to There are two approaches to activation methods: work. In other words, there is a possibility that for “work-first” and “human capital.” The “work-for-wel- such households’ work does not pay off. fare” approach comprises labor force attachment schemes that compel individuals to take formal One of the instruments to fix this issue is imple- employment as soon as possible. menting earnings disregard. This means that a portion of the income from GMB is not considered However, stricter employment conditions may 77 TOC Chapter 6: Social Protection and Human Capital Harnessing Human Capital for Growth in Croatia have limited benefits and focus on short-term the unemployment register means not only loss of solutions. The “work-for-welfare" approach aims to access to benefits but also to CES services. minimize welfare dependence by enforcing manda- tory job search requirements and imposing penal- While severe sanctions may encourage a quicker ties for non-compliance. However, stricter criteria return to work, it is important to consider their for low-income support benefits may only lead to potential drawbacks. Studies have shown that strict an increase in low-paid employment. The "human sanctions can result in job matches that are less capital" approach focuses on long-term activation stable and lower paid or lead to more part-time policies that offer comprehensive services and pri- work instead of full-time employment. Additionally, oritize training and skills development. The design if job seekers are placed in lower-quality jobs, their of activation and graduation in the unemployment skills may suffer. Hence, it is important to maintain a benefit should therefore take into account three balance in imposing sanctions. factors: Providing wider financial incentives could help 1. The employability barriers of unemployed or unemployed people to move to new locations or inactive individuals. jobs more easily. Currently, the UB recipients are entitled to financial assistance and reimbursement 2. The availability of active labor market services of expenses incurred when investing in skills and and education, health, social services, and long- new occupations through participation in adult term care for older individuals significantly influ- education and training arranged and financed by ences the effectiveness of ALMP policies. CES. The same right to financial assistance, covering travel and other costs, exists in the case of partici- 3. The activation conditions for receiving benefits pation in active employment measures. When the and graduation from benefit recipient status. unemployed person finds a job outside of the place of residence, one is entitled to one-time financial According to Figure 77, Croatia's unemployment assistance and reimbursement of travel and reloca- benefit has some of the strictest activation require- tion expenses. ments among EU countries. Figure 78 summarizes the key work-related design characteristics of the However, if someone decides to work and earn UB (unemployment benefit). In case of non-com- income, they cannot continue to receive the UB. The pliance, the legislation envisages suspension of moment they take administrative steps towards further payment of the UB and benefit recovery. work, such as becoming a member of a cooperative, There could be numerous reasons for the UB to be a seasonal worker, a freelancer, or registering a busi- suspended, for example, if the beneficiary fails to ness, the UB is terminated. This policy discourages report in person to the CES at least once every four activity-related behavior, regardless of whether it weeks; when they start working without a contract; leads to actual income generation. Furthermore, the after refusing to accept offered (suitable) job; after design of UB does not include any financial incen- declining participation in ALMM, etc. Sanctioned is tives to help ease the risks and uncertainties that the first case of non-compliance. The removal from come with entering the job market. Unfortunately, Figure 77: The strictness of activation requirements, overall score, first-tier unemployment benefits, 2020 5.0 4.5 4.17 4.0 3.5 3.22 3.0 2.5 2.0 1.5 1.0 0.5 0 … M Cy tia (2 ry ei ia Fi nia la ng ) Bu den rla g N mb ia Sw ain (u om 7) Cr ria ed 7) Re ium en lic L ly ch elg ia rt P ds al and Lu h ia (2 y ov a ov ) Sp a ( s Es rk Fr nd Ire Hu 17 Gr Ge ce Sl 017 ta ru ce an Sl aki i he r nw an n Ita B tr Li atv en 1 01 nd a u D ub n a a oa gh 0 EU R 20 an to xe ua al p a et o us m lg ee rm e ug ol (2 nl p A t e Po Cz Source: OECD Strictness of Activation Requirements Database. https://stats.oecd.org/Index.aspx?DataSetCode=SBE. 78 TOC Chapter 6: Social Protection and Human Capital Harnessing Human Capital for Growth in Croatia Figure 78: Key incentives and disincentives for activation and graduation in the UB design Activity-related (behavioral) To fulfil conditions for active job search as agreed in a job-search plan or eligibility criteria and sanctions agreement for inclusion in the labor market To demonstrate availability for work responding to all CES calls and initiatives To report to CES in person, at least once every 4 weeks, on job-search efforts Not to undertake work without a contract To enrol and complete educational programs for increasing qualification and employability organized or paid for by CES To accept 'suitable' job offers made by CES or employer in the place of residence or within a distance of up to 50 km (exceptions apply) To accept occupatioal mobility (job offers in different professions) Administrative sanctions for non compliance - suspension of UB and refund Administrative sanction - deregistration from the record of unemployed if not maintaining contacts with CES for 6 months Incentives and disincentives in The UB amount is a percentage of (lower than) the past peronal earnings the UB design for graduation which decreases over time / labor market integration of The UB amount has a statutory ceiling - cannot exceed the gross minimum claimants wage irrespective of household size The duration of UB receipt depends on past work record and appers long for laid-off long-time employees and employees close to retirement age No automatic eligibility to other rights, or cash and in-kind benefits Interaction with the To register with CES if able to work or partially able to work within 30 days employment service (how the No mandatory waiting period for the UB ‘mutual obligations’ principle Intensity of iterventions on behalf of CES is different / lower for UB recipients works?) of close to retirement age Financial compensations (cash payments) to assist the unemployed into employment • Financial assistance and reimbursement of expenses incurred in the course of education (training program arranged and financed by CES) • One-time financial assistance and reimbursement of travel and relocation expenses Other conditions Cumulation of UB with earnings from work is not possible; suspension of UB and obligation to refund apply Not possible to combine UB receipt with the status of permanent seasonal worker, with starting a business, membership in a cooperative, freelance work Source: Labor Market Act (Official Gazette, 118/2018), last amended 18/2022. this means there are no options for a gradual reduc- the remaining part of the entitlement to UB. If the tion of UB payments or the ability to receive income claimant has worked for over three months, the UB from work while still receiving UB temporarily. will be recalculated/updated with the new work record and wage. However, individuals who are As labor shortages continue to increase, it's impor- unemployed but have a lengthy work history may tant to reassess the length of time that some need more motivation to seek employment under workers can receive unemployment benefits (UB). the current UB design actively. The duration of UB Although some factors promote graduation, some benefits for those who have been laid off and have a aspects of the UB design hinder the activation of long work history is quite generous and can last up older workers. There is a legal guarantee that the to 450 days (about 1 year 3 months) for those who right to UB payment will be preserved on reemploy- have worked for over 25 years. ment and that the UB will be resumed in case of lay-off. This is a way to shorten the actual duration To make the most of all available human resources, of the unemployment benefit. If the UB claimant it's crucial to include older UB recipients in activa- takes a job before the expiration of its full term and tion and labor market inclusion policies. Individuals is subsequently laid off, they are entitled to receive who have worked for 32 years or more and require 79 TOC Chapter 6: Social Protection and Human Capital Harnessing Human Capital for Growth in Croatia up to five more years before they are eligible for an Navigating the Low-Income old-age pension will continue to receive unemploy- ment benefits until they become eligible for a pen- Support (GMB) Program: sion. These regulations prioritize long-term "passive" Activation and Graduation receipt of unemployment benefits, serving as an early retirement benefit until one becomes an old- The GMB incorporates activation elements consist- age pensioner. There are no obligations for active ent with the “second-pillar”73 of last-resort income job searching or specific labor market measures to support schemes in the EU and OECD countries. target the long-term unemployed who are older. As a result of such reforms, income support is com- Allowing up to five years of UB without requiring bined with employment and social services in an efforts to reintegrate into the job market, includ- integrated service provision model to help benefit ing behavioral expectations, incentives, and con- recipients become self-sufficient. Figure 79 presents sequences, is too lengthy. Moreover, this extended a summary of work-related design characteristics period puts older workers at risk of losing their cur- of the GMB, which describes the conditions for the rent skills and missing out on opportunities to learn activation of recipients. Effective GMB activation new ones. Therefore, it is necessary to modify various and graduation element design depends on work ALMMs and employment services to promote job incentives, activation requirements, and sanctions. seeking and the ability to work among older workers. Based on meta-evaluations, implementing job- search monitoring and sanctions can lead to more Figure 79: Incentives and disincentives for activation and graduation in the GMB design The ‘mutual obligations’ Obligation for the able to work GMB recipients to register with CES principle / rights and Obligation to accept any offer for work made by CES regardless of responsibilities qualification or experience, including temporary and seasonal jobs Obligation to participate in ALMMs, specifically in municipal/ community works and training Frequent and personalized interventions by CES during the time of benefit receipts Offer (but no obligation to participate) for people with reduced ability to work (mostly training and retraining) Incentives and disincentives The GMB amount for able to work is lower than GMB for not able to in the GMB design for work graduation / labor market The GMB amount cannot exceed the gross minimum wage integration irrespective of household size The benefit formula is ‘’the difference’’ between income threshold and own income; benefit amount is reduced with earned income at 100 percent The duration of benefit receipt is unlimited, with no downsizing or phasing out over time Presence of associated rights / related benefits (cash and in-kind), including from local government units Sanctions for non- The right to GMB can be suspended for six months if: compliance with activation • a job offer is rejected requirements • If employment is terminated Financial work incentives / When employed those receiving GMB uninterruptedly for at least one in-work benefits year will continue receiving support for 3 consecutive months (full amount in the first month of work, 75% in the second month and 50% in the third month). The right to the GMB is not suspended after three months of work if the average earnings over the period does not exceed the amount of the GMB In case of participation in public works the GMB is not suspended Source: Social Welfare Act (2013), last amended Official Gazette, 138/2020. 73 The focus of the first pillar is on improving the targeting and increasing the coverage of benefits, thus reducing leakage to the non-poor. The 80 second pillar focuses on activation and graduation conditions for benefit recipients. TOC Chapter 6: Social Protection and Human Capital Harnessing Human Capital for Growth in Croatia employment. Like UB, these provisions encourage the sanctions should not withhold adequate sup- individuals to seek actively, prepare for, and actively port from intended recipients. It is recommended accept employment opportunities. One poten- to adopt balanced conditions for compliance with tial use of these tools is to decrease the number of work requirements and use targeting and fraud people making claims, thereby lessening the strain control systems to minimize GMB claims which are on the benefits systems and employment services. not legitimate or not addressed to the target group. However, this should be achieved via a targeting system by more accurately identifying and support- It is crucial to have well-organized social services ing those who truly need it. integrated with the CES through a case manage- ment system to maximize the human potential of A distinct feature of the activation and graduation GMB recipients. There is a need for more inter-in- element of the GMB is the presence of financial stitutional coordination to address the various chal- work incentives or make-work-pay provisions. A lenges related to labor market reintegration and key aspect of the GMB's activation and graduation societal inclusion. To achieve the desired activation component is its inclusion of financial incentives or outcomes, individuals require personalized support "make-work-pay" measures. When employed, those throughout their journey to work. This is particularly receiving GMB for at least one year will continue important for GMB recipients, who may struggle receiving (in-work) support for three consecutive with long-term unemployment and require housing, months (full amount in the first month of work, 75 childcare, and health and mental health support. percent in the second month, and 50 percent in the Successful (re)integration, due to complex support third month). The right to the GMB is suspended needs, requires coordinated case management with after three months of work if the average earnings social services. Currently, several tools are available over the period exceed the amount of the GMB. In to help coordinate service provision, including the case of engagement in public works, the GMB is job-search plan, the agreement for inclusion in the not suspended. Finally, the GMB is not generous labor market, and the individual plan for improv- compared to incomes from work and thus is not ing the life situation. A good plan should consider expected to raise benefit dependency. This makes all family members' needs and limitations and pro- it less likely to discourage work than similar benefits vide activation and support options for everyone in in other EU countries.74 GMB beneficiaries who can the household. Personal factors and complex issues work or are partially qualified to work must regis- can greatly impact their situation. When conduct- ter at the CES within 30 days from the application; ing a needs assessment, it is important to consider must be ready to accept any offer of employment, various factors such as childcare, home care for regardless of their qualifications or experience, dependent relatives, financial support, and in-kind including temporary and seasonal jobs; must look support to ensure regular school attendance. When actively for employment and be available for work. creating service packages, it is important to person- Individuals under five years from qualifying for an alize them to reflect the unique needs and potential old-age pension are exempt from these behavioral for activation of each benefit recipient. conditions. The receipt of GMB for those who can work is subject to a public works requirement. The local governments organize and finance ‘’work for Policy Implications the common good without compensation’’ where participation is obligatory, takes 60 to 90 hours per To reduce poverty cost-effectively through social month, and is not paid. This is atypical and is not assistance, increasing the percentage of means- aligned with EU practice. Therefore, it is crucial to tested aid is important to help alleviate poverty ensure that demanding conditions don't become and promote human development. Compared to onerous, as this can lead to intended recipients other countries in the European Union, Croatia's being excluded from receiving financial support and social assistance expenditure is relatively low. Cur- employment services. rently, the safety net programs in Croatia primarily target specific groups of people, but they should Sanctions for failure to comply accompany require- shift to means-tested benefits by increasing the ments for GMB activation. The right to the benefit share of the budget flowing to programs targeting is suspended for six months for beneficiaries who low-income status, such as the Guaranteed Mini- refuse a job offer without a valid reason, terminate mum Benefit (GMB). their employment, or decline to meet workfare requirements. Furthermore, the amount is reduced To make it easier for people to access GMB, the by the GMB amount to which the sanctioned administrative procedures for GMB applications member has been previously entitled. However, need to be simplified. Some of the requirements for The observations are based on the qualitative comparative data on activation-related requirements in guaranteed minimum income programs 74 81 collected in the MISSOC database of the European Commission (as of 01.01.2022); https://www.missoc.org/missoc-database/comparative- tables/results/. With respect to lower-tier (social assistance) benefits, the OECD Strictness of Activation Requirements Database provided ratings for a very few countries in 2020 and earlier. Croatia’s GMB was not featured among them. TOC Chapter 6: Social Protection and Human Capital Harnessing Human Capital for Growth in Croatia GMB applications are stricter than in other coun- they lose their benefits after becoming a member tries, which can discourage people from applying. of a cooperative, a seasonal worker, a freelancer, or One such requirement is the need for a property lien a registered business owner. Therefore, the policy in case benefits are provided incorrectly and need should be more encouraging towards job seeking to be returned. To increase access, we could con- by not abruptly terminating the benefits. sider eliminating this condition. Considering older unemployment benefit recipi- It is important to review the mandatory home ents in policies promoting labor market inclusion is visits for GMB applicants to address the issue of essential. Individuals who have worked for 32 years low coverage. Stigma can be one of the reasons for or more and require up to five more years before low coverage of certain benefits, especially when they are eligible for an old-age pension will con- the application process involves compulsory home tinue to receive unemployment benefits until they visits. A better approach would be to use a prob- become eligible for a pension. The current regula- abilistic approach to suggest selective home visits tions prioritize the "passive" receipt of unemploy- based on the higher risk factors identified by pro- ment benefits for older unemployed individuals as filing algorithms. This will encourage participation an early retirement benefit until they become old- while limiting the possibility of fraudulent activities age pensioners. However, there are no obligations and ensuring that only eligible individuals receive for job searching or specific labor market measures the benefit at a lower administrative cost. to target the older, long-term unemployed. Allowing up to five years of unemployment benefits without Activation and graduation policies could be requiring reintegration into the job market, includ- strengthened by enhancing complementary ser- ing behavioral expectations, incentives, and conse- vices such as health, training, and social services quences, is too lengthy. Furthermore, this extended for those facing barriers to labor. These services period puts older workers at risk of losing their skills should aim to overcome dependency and help indi- and missing out on opportunities to learn new ones. viduals graduate from social assistance through the Therefore, it is necessary to modify various ALMMs "human capital" approach. The enhanced availability (Active Labor Market Measures) and employment of active labor market services, education, health, services to promote job-seeking and the ability to social services, and long-term care for older indi- work among older workers. viduals significantly influences the effectiveness of ALMP policies. To maximize the potential of GMB recipients, well-organized social services must be integrated with the GMB and the Croatian employment ser- vices. Inter-institutional coordination is necessary to address challenges related to labor market rein- tegration and societal inclusion. Personalized sup- port is vital for individuals on their journey to work, especially for GMB recipients who require housing, childcare, and health and mental health support. Successful reintegration requires coordinated case management with social services, considering all family members' needs and limitations. An inte- grated case management must consider personal factors and complex issues to provide personalized service packages and activation options for each benefit recipient. Financial incentives can be a helpful solution to assist unemployed individuals in relocating or find- ing new job opportunities. While Financial assis- tance is available to registered unemployed individ- uals to help them invest in new skills, cover travel expenses, and provide one-time financial support for relocation expenses, this policy can be further improved to support those who decide to work, as 82 TOC Harnessing Human Capital for Growth in Croatia References Abarcar, P., and C. Theoharides. 2021. “Medical Barca, V. and Hebba, M (2020). On-Demand Worker Migration and Origin-Country Human and Up to Date? Dynamic Inclusion and Capital: Evidence from U.S. Visa Policy.” The Data Updating for Social Assistance. https:// Review of Economics and Statistics, 1-46. s o c i a l p ro t e c t i o n .o rg /d i s c ove r/p u b l i c a- tions/demand-and-date-dynamic-inclu- Acemoglu, Daron & Autor, David, (2011). Skills, Tasks, sion-and-data-updating-social-assistance and Technologies: Implications for Employment and Earnings," Handbook of Labor Economics, Bejaković, P. (2004). The informal economy in in O. Ashenfelter & D. Card (ed.), Handbook of Croatia and economic development. SEER- Labor Economics, edition 1, volume 4, chapter South-East Europe . Review for Labour and 12, pages 1043-1171, Elsevier. Social Affairs, 69-78. Almeida, Rita, Larry Orr & David Robalino (2014). Bejaković, P (2016). How efficient are measures “Wage subsidies in developing countries as a for employment and inclusion of long-term tool to build human capital: design and imple- unemployed: the case of Croatia? Peer Country mentation issues.” IZA Journal of Labor Policy Comments Paper – Croatia, European Commis- volume 3. Article number: 12 (2014). sion, DG Employment, Social Affairs and Inclusion. Amuedo-Dorantes, C. and S. Pozo. (2010). “Accounting for Remittances and Migration Bejaković, P., & Mrnjavac, Ž. (2019). Housing policy effect on Children’s Schooling,” World Devel- and labour market in Croatia. In Investigating opment, 38 (12): 1747-1759.Attanasio, Orazio, Spatial Inequalities. Emerald Publishing Limited. Ricardo Paes de Barros, Pedro Carneiro, David K. Evans, Lycia Lima, Pedro Olinto, and Norbert Bejaković, P., Urban, I., & Bezeredi, S. (2014). Kako- Schady (2022). “Public Childcare, Labor Market raduHrvatskojucinitiisplativim? (How to achieve Outcomes of Caregivers, and Child Devel- “Does work pay” in Croatia?),7. Zagrebackiekon- opment: Experimental Evidence from Brazil.” omskiforum, Zagreb: Friedrich EbertFounda- National Bureau of Economic Research. tion, pp.5–38 Aubert, Patrick, and Bruno Crépon (2003). “La Blundell, R., E. French, and G. Tetlow (2016). productivité des salariés âgés : une tentative “Retirement Incentives and Labor Supply.” In d’estimation.” Economie et Statistique 368, Handbook of the Economics of Population no. 1 (2003): 95–119. https://doi.org/10.3406/ Aging, 1:457–566. Elsevier, 2016. https://doi. estat.2003.7300. org/10.1016/bs.hespa.2016.10.001. Badjun, Marijana (2017). “Financiranje Domova Za Boeri, T., H. Brücker, F. Docquier, and H. Rapoport Starije i Nemoćne Osobe u Hrvatskoj.” Revija Za (2012). “Brain Drain and Brain Gain: The Global Socijalnu Politiku 24, no. 1 (2017): 19–42. Competition to Attract High-skilled Migrants.” Oxford: Oxford University Press. Bahar, D. and H. Rapoport (2018). “Migration, Knowledge Diffusion, and the Comparative Boockmann, Bernhard and Tobias Brändle (2015). Advantage of Nations.” Economic Journal, 128 Integrating Older Employees into the Labor (612): F273–F305. Market – Evidence from a German Labor Market Program. CESifo DICE Report 3/2015 Bansak, C., and B. Chezum (2009). “How Do (September). Remittances Affect Human Capital Formation of School-Age Boys and Girls?” American Economic Review, 99 (2): 145-48. 83 TOC Harnessing Human Capital for Growth in Croatia Börsch-Supan, A. (2022). Survey of Health, Ageing CES and Ipsos Puls (2016). External Evaluation of and Retirement in Europe (SHARE) Wave 6. Active Labor Market Policy Measures 2010- Release version: 8.0.0. SHARE-ERIC. Data set. 2013. Summary evaluation report, February DOI: 10.6103/SHARE.w6.800 2016. Börsch-Supan, A., K. Härtl, and D.N. Leite (2016). Chami, M. R., Ernst, E., Fullenkamp, C., & Oeking, A. “Social Security and Public Insurance.” In (2018). Are remittances good for labor markets Handbook of the Economics of Population in LICs, MICs and Fragile States?. . International Aging, 1:781–863. Elsevier, 2016. https://doi. Monetary Fund. org/10.1016/bs.hespa.2016.09.002. Christiaensen, L., Ferre, C., Ivica, R., Matkovic, T., Bossavie, L., D. Garrote Sanchez, M. Makovec and & Sharafudheen, T. (2019). Jobs Challenges in C. Özden (2022). “Skilled Migration: A Sign of Slavonia, Croatia–A Subnational Labor Market Europe's Divide or Integration?” International Assessment. World Bank Group. Development in Focus, © World Bank. Clemens, M. A., C. E. Montenegro, and L. Pritchett Brown, A. and J. Koettl (2015), “Active labor market (2008). “The place premium: wage differences programs - employment gain or fiscal drain?”, for identical workers across the US border.” IZA Journal of Labor Economics, Vol. 4/12. World Bank Policy Research Working Paper, (4671). Card, D., J. Kluve and A. Weber (2018), “What Works? A Meta-Analysis of Recent Active Labor Clemens, M.A. (2011). “Economics and Emigration: Market Program Evaluations,” Journal of the Trillion-Dollar Bills on the Sidewalk?” Journal of European Economic Association, Vol. 16/3. Economic Perspectives, 25 (3): 83-106. Caroli, Eve, and John Van Reenen (2001). Skill-Bi- Co, C. Y., I. N. Gang, and M.S. Yun (2000). “Returns ased Organizational Change? Evidence from A to Returning.” Journal of Population Economics Panel of British and French Establishments, The 13 (1): 57–79. Quarterly Journal of Economics, Volume 116, Issue 4, November 2001, Pages 1449–1492, Coile, C., & Gruber, J. (2001). Social security incen- https://doi.org/10.1162/003355301753265624 tives for retirement. In Themes in the Economics of Aging (pp. 311-354). University of Chicago Carmichael, Fiona, and Susan Charles (2003). Press. “The Opportunity Costs of Informal Care: Does Gender Matter?” Journal of Health Economics Crépon, Bruno, Nicolas Deniau, and Sébastien 22, no. 5 (2003): 781–803. Pérez-Duarte (2003). Wages, Productivity and Worker Characteristics: A French Perspective. CEDEFOP (2020a). Vocational education and INSEE Paris, France. training in Croatia: short description. Luxem- bourg: Publications Office of the European Crespo, Laura (2006). “Caring for Parents and Union. Employment Status of European Mid-Life Women.” In Second Workshop Economics of CEDEFOP (2020b). Spotlight on VET. Croatia. URL: the Family at the University of Zaragoza, 9–10. https://www.cedefop.europa.eu/files/8136_ Citeseer. en.pdf Croatian Bureau of Statistics 2020. URL:https:// CEDEFOP (2020c). Vocational education and podaci.dzs.hr/2023/en/60850 training for the future of work: Croatia. Agency for Vocational Education and Training and Adult D’Amuri, F., and G. Peri. (2014). “Immigration, jobs, Education. and employment protection: evidence from Europe before and during the great recession.” CEDEFOP (2022). European Skills Index. Country Journal of the European Economic Association, Pillar: Croatia. URL: https://www.cedefop. 12(2): 432-464. europa.eu/en/tools/european-skills-index/ country/croatia?y=2022 Dayioglu, M., Küçükbayrak, M., & Tumen, S. (2022). The impact of age-specific minimum wages on CES (2022). CES Annual Report, Zagreb 2021. youth employment and education: a regression 84 TOC Harnessing Human Capital for Growth in Croatia discontinuity analysis . International Journal of of Public Employment Services. PES Strategies Manpower. in Support of an Ageing Workforce: Study report, ISBN:978-92-76-08562-1 doi: 10.2767/535792 Del Carpio, X. V., C. Ozden, M. Testaverde, and KE-02-19-469-EN-N. M. Wagner (2016). “Global migration of talent and tax incentives: evidence from Malaysia’s European Commission (2019b). “Joint Paper on returning expert program.” World Bank Policy Pensions 2019.” Research Working Paper 7875. European Comission (2021). Ageing Report: Dobrotić, I. (2022). The (in)equality dynamic of child- Economic and Budgetary Projections for the EU care-related policy development in post-Yugo- Member States (2019-2070). Institutional Paper slav countries. Journal of European Social Policy, 148. Brussels. 270-286. European Commission (2020). Employment, Social Dobrotić, I., & Matković, T. (2023). Understanding Affairs & Inclusion The Youth Guarantee country territorial inequalities in decentralised welfare by country – Croatia, October 2020. systems: Early childhood education and care system expansion in Croatia. Public Sector European Commission (2022). Croatia: Comprehen- Economics, 47(1), 89110. https://doi.org/10.3326/ sive steps to improve early childhood education pse.47.1.4 and care (ECEC). URL: https://eurydice.eacea. e c .e u ro p a .e u /n ews /c ro a t i a- c o m p re h e n - Draženović, I., M. Kunovac, and D. Pripužić (2018). sive-steps-improve-early-childhood-educa- “Dynamics and Determinants of Migration: tion-and-care-ecec The Case of Croatia and Experience of New EU Member States.” Croatian National Bank (HNB), European Parliament (2021). Empowering European Zagreb. youth: post-pandemic employment and social recovery. European Parliament resolution of 17 Dustmann, C. and A. Glitz. (2011). “Migration and February 2022 on empowering European youth: Education,” Edited in the ‘Handbook of the post-pandemic employment and social recovery Economics of Education,’ Volume: 4. Elsevier (2021/2952(RSP)). Publication. Friedberg, Leora (2003). “The Impact of Techno- Education and Training Monitor (2019). Croatia. logical Change on Older Workers: Evidence from URL: https://education.ec.europa.eu/sites/ Data on Computer Use.” ILR Review 56, no. 3 default/files/document-library-docs/et-moni- (2003): 511–29. tor-report-2019-croatia_en.pdf GEM (2022). Global Entrepreneurship Monitor Education and Training Monitor (2022). Country 2021/22 Women’s Entrepreneurship Report: analysis. Croatia. URL: https://op.europa.eu/ From Crisis to Opportunity. Global Entrepre- webpub/eac/education-and-training-mon- neurship Monitor. itor-2022/en/country-reports/croatia.html Glinos, I. A. (2015). “Health professional mobility in EUROMOD (2019). Country Report Croatia. the European Union: exploring the equity and European Commission 2016-2019. https:// efficiency of free movement.” Health Policy, 119, euromod-web.jrc.ec.europa.eu/sites/default/ 1529-1536. files/2020-04/Y10_CR_HR_Final.pdf Glunčić, M. (2022), Analiza istraživanja zapošl- EUROMOD (2022). Country Report Croatia jivosti diplomiranih studenata 2020. godine, 2018-2021. European Commission. https:// AZVO, Zagreb;https://www.azvo.hr/images/ euromod-web.jrc.ec.europa.eu/sites/default/ stories/novosti/Analiza_istra%C5%BEivanja_ files/2022-01/Y12_CR_HR_Final.pdf zapo%C5%A1ljivosti_diplomiranih_ studenata_2020._godine.pdf European Commission (2015). Employment and Social Developments in Europe 2015. Brussels: Grimshaw, D., & Rubery, J. (2015). The motherhood European Commission. ISSN 2315-2540. pay gap: A review of the issues, theory and inter- national evidence. ILO Conditions of Work and European Commission (2019a). European Network Employment Series. 85 TOC Harnessing Human Capital for Growth in Croatia Gruber, E., I. Sarajlic Vukovic, M. Musovic, D. Moravek, Economics 26, no. 4 (2007): 821–41. B. Starcevic, S. Martic Biocina, and R. Knez. (2020). “Migration of Croatian Physicians in the Heitmueller, Axel (2007). “The Chicken or the Egg?: Global Context.” Medicina Fluminensis (56) 2: Endogeneity in Labour Market Participation of 88–96. Informal Carers in England.” Journal of Health Economics 26, no. 3 (2007): 536–59. Gruber, Jonathan, and David Wise (1998). “Social Security and Retirement: An International Henderson, J. L., Hawke, L. D., Chaim, G., & Network, Comparison.” The American Economic Review N. Y. (2017). Not in employment, education or 88, no. 2 (1998): 158–63. training: Mental health, substance use, and disengagement in a multi-sectoral sample of Gruber, Jonathan, and David A. Wise, eds. (1999). service-seeking Canadian youth. Children and Social Security and Retirement around the Youth Services Review, 138-145. World. A National Bureau of Economic Research Conference Report. Chicago: University of Hyslop, D., & Stillman, S. (2007). Youth minimum Chicago Press. wage reform and the labour market in New Zealand. Labour Economics, 201-230. Gruber, Jonathan, and David A. Wise, eds. (2003). Social Security Programs and Retirement IEA (2012). PIRLs 2011 International Results in around the World: Micro-Estimation. Univer- Reading. International Association for the sity of Chicago Press. https://doi.org/10.7208/ Evaluation of Educational Achievement (IEA). chicago/9780226309989.001.0001. URL: https://timssandpirls.bc.edu/pirls2011/ downloads/p11_ir_fullbook.pdf Gruber, J., and D.A. Wise (2010). Social Security Programs and Retirement around the World: IMF (2021) Guaranteed Minimum Income The Relationship to Youth Employment Fiscal Schemes in Europe: Landscape and Implications of Reform, Chicago and London the Design. https://www.elibrary.imf.org/view/ University of Chicago Press. journals/001/2021/179/001.2021.issue-179-en. xml Håkansson, P. G., & Bejakovic, P. (2020). Labour market resilience, bottlenecks and spatial IMF. (2020). Republic Of Croatia Country Report No. mobility in Croatia. Eastern Journal of European 20/50. International Monetary Fund. Studies, 5-25. Immervoll, H. and C. Knotz (2018). How Demanding Jean-Olivier Hairault & François Langot & Thepthida Are Activation Requirements for Jobseekers? Sopraseuth, 2010. "Distance to Retirement and DELSA/ELSA/WD/SEM (2018) 14, July 12, 2018. The Job Search of Older Workers: The Case For Delaying Retirement Age," Université Paris1 Isik-Dikmelik, A., Ovadiya, M., Karacsony, S., Millán, Panthéon-Sorbonne (Post-Print and Working N., & Vandeninden, F. (2017). PORTRAITS OF Papers) hal-00517107, HAL. LABOR MARKET EXCLUSION 2.0 Country Policy Paper (CPP) for Croatia. World Bank Group. Hausmann, R., and L. Nedelkoska (2018). “Welcome Home in a Crisis: Effects of Return Migration on Jaumotte, F., Li, L., Medici, A., Oikonomou, M., the Non-migrants’ Wages and Employment.” Pizzinelli, C., Shibata, I., . . . Tavares, M. M. (2023). European Economic Review 101 (January): Digitalization During the COVID-19 Crisis: Impli- 101–32. cations for Productivity and Labor Markets in Advanced Economies. IMF. Heckman J. & Jora Stixrud & Sergio Urzua (2006). "The Effects of Cognitive and Noncognitive Jurić, T. (2017). “Suvremeno iseljavanje Hrvata u Abilities on Labor Market Outcomes and Social Njemačku: karakteristike I motivi.” Migracijske i Behavior," Journal of Labor Economics, Univer- etničke teme 24 (3): 337–71. sity of Chicago Press, vol. 24(3), pages 411-482, July. Kelmanson, M. B., Kirabaeva, K., Medina, L., Mircheva, M., & Weiss, J. (2019). Explaining the shadow Heitmueller, Axel, and Kirsty Inglis (2007). “The economy in Europe: size, causes and policy Earnings of Informal Carers: Wage Differen- options. International Monetary Fund. tials and Opportunity Costs.” Journal of Health 86 TOC Harnessing Human Capital for Growth in Croatia Kugler, M., and H. Rapoport (2007). “International Neumark, David, Ian Burn, and Patrick Button (2019). Labor and Capital Flows: Complements or “Is It Harder for Older Workers to Find Jobs? New Substitutes?” Economics Letters 94 (2): 155–62. and Improved Evidence from a Field Experiment.” Journal of Political Economy 127, no. 2 (April Lynch, R., & Vaghul, K. (2015). The benefits and costs 2019): 922–70. https://doi.org/10.1086/701029. of investing in early childhood education: The fiscal, economic, and societal gains of a universal Norton, Edward C. (2016). “Health and Long-Term prekindergarten program in the United States, Care.” In Handbook of the Economics of Popula- 2016-2050. Washington Center for Equitable tion Aging, 1:951–89. Elsevier, 2016. Growth. OECD (2001). Chapter 1. Life-long learning for all Martin, R., and D. Radu (2012). “Return Migration: policy directions. URL: https://one.oecd.org/ The Experience of Eastern Europe.” Journal of document/DEELSA/ED/CERI/CD(2000)12/ International Migration 50 (6): 109–28. PART1/REV2/En/pdf Matković, T., Dobrotić, I., & Baran, J. (2019). Što vrtić OECD (2018a). Benefit generosity and work incen- ima s tim? Pristup ranom i predškolskom odgoju tives for recipients of disability benefits in 12 i obrazovanju i reprodukcija društvenih nejed- EU Member states. https://www.oecd.org/els/ nakosti u redovnom školovanju: analiza podataka soc/benefits-and-wages/Benefit-generosi- PISA i TIMSS istraživanja. Revija za sociologiju, ty-and-work-incentives-for-disability-bene- 49(1), 735. https://doi.org/10.5613/rzs.49.1.1 fit-recipients.pdf Mayr, K., and G. Peri. (2008). “Return Migration as OECD (2018b). TALIS 2018 Results (Volume I) a Channel of Brain Gain.” NBER Working Paper Teachers and School Leaders as Lifelong 14039 (May), National Bureau of Economic Learners. URL: https://www.oecd-ilibrary.org/ Research, Cambridge, MA. education/talis-2018-results-volume-i_1d0b- c92a-en McKenzie, D. and H. Rapoport. (2011). “Can migration reduce educational attainment? Evidence from OECD (2018c). Skills Studies, Additional results from Mexico,” Journal of Population Economics, 24: the survey of adult skills. 1331–1358. OECD (2018d), Program for International Student MoLPFS (2022) Socijalne Naknade 2016-2022. Assessment (PISA), 2018 Reading, Mathematics, Internal data. and Science Assessments Mullis, I.V.S., von Davier, M., Foy, P., Fishbein, B., OECD (2021). OECD Employment Outlook Reynolds, K.A., & Wry, E. (2023). PIRLS 2021 2021: Navigating the COVID-19 Crisis and International Results in Reading. Boston College, Recovery. OECD Publishing, Paris, htpps://doi. TIMSS & PIRLS International Study Center. org/10.1787/5a700c4b-en. https://doi.org/10.6017/lse.tpisc.tr2103.kb5342 OECD (2022a), OECD Employment Outlook 2022: Mullis, I.V.S, Martin, M., Foy, P., Kelly, D., and Fishbein, Building Back More Inclusive Labor Markets. B. (2020). TIMSS 2019 International Results in Supporting jobs and incomes: An update on the Mathematics and Science. TIMSS & PIRLS Inter- policy response to the COVID‑19 crisis. OECD national Study Center, https://timss2019.org/ Publishing, Paris, https://doi.org/10.1787/06c- reports/wp-content/themes/timssandpirls/ fa2e6-en. download-center/TIMSS-2019-Internation- al-Results-in-Mathematics-and-Science.pdf OECD (2022b). Connecting People with Jobs. Reaching Out and Activating Inactive and Unem- Murtin, F., Mackenbach, J., Jasilionis, D., & d’Ercole, M. ployed Persons in Bulgaria, https://www.oecd. M. (2017). Inequalities in longevity by education org/publications/reaching-out-and-activat- in OECD countries: Insights from new OECD ing-inactive-and-unemployed-persons-in-bul- estimates. garia-7b91154a-en.htm Nestić, D., & Tomić, I. (2018). Jobless Population OECD (2022c). “Labour Migration in the Western and Employment Flows in Recession. Journal of Balkans: Mapping Patterns, Addressing Chal- Balkan and Near Eastern Studies, 273-292. lenges and Reaping Benefits” © OECD. 87 TOC Harnessing Human Capital for Growth in Croatia Orefice, G., H. Rapoport, and G. Santoni (2022). Croatia’s Lagging Region (English). Washington, “How Do Immigrants Promote Exports?” IZA D.C.: World Bank Group. Discussion Paper No. 15722. ReferNet Croatia (2021). Croatia: survey confirms Ostrovidov Jakšić, A., & Jakšić, I. (2019). How to the impact of COVID-19 pandemic on education. prolong labour market participation in the National news on VET Republic of Croatia? Public sector economics, 79-108. Rimac, I. (2020). EUROGRADUATE pilot survey: Country report - Croatia. European Commission. Ostrovidov Jakšić, A., & Jakšić, I. (2019). How to prolong labour market participation in the Robayo-Abril, M., Terskaya, A., Vaughan, K. N., & Republic of Croatia?. Public sector economics, Garcia Pena, N. (2020). Encouraging Women’s 43(1), 79-108. Economic Opportunities in Croatia: Empirical Evidence of Determinants and Policy Advice. Ovadiya, M., Vandeninden, F., Isik-Dikmelik, A., World Bank Group. Karacsony, S., & Millán, N. (2017). Portratis of Labor Market Exclusion 2.0: Country Policy Paper Robert G. Lynch and Kavya Vaghul. (2015). “The for Croatia. World Bank Group. benefits and costs of investing in early childhood education: the fiscal, economic, and societal Parker, Gillian. (1990). “Whose Care? Whose Costs? gains of a universal prekindergarten program in Whose Benefit? A Critical Review of Research on the United States, 2016-2050” Washington, DC: Case Management and Informal Care.” Ageing & Washington Center for Equitable Growth Society 10, no. 4 (1990): 459–67. Rust, John, and Christopher Phelan (1997). “How Pažur Aničić, K., Gusić Munđar, J. & Šimić, D. (2022). Social Security and Medicare Affect Retirement Generic and digital competencies for employa- Behavior In a World of Incomplete Markets.” bility — results of a Croatian national graduates Econometrica 65, no. 4 (1997): 781–831. https:// survey. High Educ (2022). doi.org/10.2307/2171940. Peri, G., and F. Requena-Silvente (2010). “The Trade Shmis, Tigran; Ustinova, Maria; Chugunov, Dmitry; Creation Effect of Immigrants: Evidence from the Melianova, Ekaterina; Parandekar, Suhas D.; Remarkable Case of Spain.” Canadian Journal of Kruske, Lucy (2021). Russian Federation - New Economics, 43 (4): 1433–59. Skills for New Century: Informing Regional Policy (English). Washington, D.C.: World Bank Group. Pezer, Martina & Urban, Ivica. (2017). An Overview http://documents.worldbank.org/curated/ of Social Protection Benefits in Croatia. Press en/522441624592935175/Russian-Feder- Releases. https://repozitorij.ijf.hr/en/islandora/ ation-New-Skills-for-New-Century-Inform- object/ijf%3A549 ing-Regional-Policy Rust, J., & Phelan, C. (1997). How social security and Sorbe, S., Gal, P., Nicoletti, G., & Timiliotis, C. (2019). medicare affect retirement behavior in a world Digital Dividend: Policies to Harness the Produc- of incomplete markets. Econometrica: Journal of tivity Potential of Digital Technologies. OECD. the Econometric Society, 781-831. Hendren, N., & Sprung-Keyser, B. (2020). A unified Poutvaara, P. (2004). “Educating Europe: Should welfare analysis of government policies. The Public Education Be Financed with Graduate Quarterly Journal of Economics, 135(3), 1209- Taxes or Income-Contingent Loans?” CESifo 1318. Economic Studies 50 (4): 663–84. Šućur, Z. et al (2016), Struktura naknada, izdaci i Rafajac O. (2019). Analysis of Youth Professional korisnici programa socijalne zaštite u Republici Orientation in the Republic of Croatia Zbornik Hrvatskoj. https://bib.irb.hr/datoteka/807966. Veleučilišta u Rijeci 7 (1), 163-183 https://doi. Izvjetaj_-_Publikacija_1.pdf org/10.31784/zvr.7.1.8 Staubli, S., & Zweimüller, J. (2013). Does raising the Ramljak, D., Brajkovic, L., and Ambasz, D. (2022). early retirement age increase employment of Education and Skills Development in Eastern older workers?. Journal of public economics, 108, Croatia: Toward Strengthening Human Capital in 17-32. 88 TOC Harnessing Human Capital for Growth in Croatia Švaljek Sandra (2008). “Enhancing employers’ Vodopivec Milan, Dan Finn, Suzana Laporšek, Matija demand for older workers: Policies, practices and Vodopivec, and Nejc Cvörnjek (2018). Increasing recommendations” in Vehovec, Maja, ed. “New Employment of Older Workers: Addressing perspectives on a longer working life in Croatia Labor Market Obstacles, Journal of Population and Slovenia.” Institute of Economics. Ageing, https://www.researchgate.net/publica- tion/330097142_Increasing_Employment_of_ Thévenon, O., & Solaz, A. (2013). Labour market Older_Workers_Addressing_Labour_Market_ effects of parental leave policies in OECD Obstacles/link/60c71c7a299bf1949f57fc76/. countries. OECD. Wheeler, C. M., Quaglietti, L., & Norfleet, J. R. (2022). Tomić, I. (2020). Employment protection reforms EU RER: Living Up to Potential. Washington D.C.: and labour market outcomes in the aftermath World Bank Group. of the recession: Evidence from Croatia. Public Sector Economics. Williams, C. C., & Franic, J. (2016). Explaining partic- ipation in the informal economy in post-socialist Tomić, Iva (2015). NEETs in Croatia – not in employ- societies: a study of the asymmetry between ment, education or training – but where are formal and informal institutions in Croatia. they? Peer Review on “Targeting NEETs – key Journal of Contemporary Central and Eastern ingredients for successful partnerships in Europe. improving labor market participation”. European Commission. Mutual Learning Programme DG World Bank (2010). Helena Ribe, David A. Robalino, Employment, Social Affairs and Inclusion. Peer and Ian Walker Achieving effective social protec- Country Comments Paper – Croatia, Oslo, 24-25 tion for all in Latin America and the Caribbean: September, 2015. from right to reality. Tomić, Iva and Ivan Zilic (2018) Working for 200 World Bank (2014). Public Finance Review. Restruc- euro? The effects of traineeship reform on labor turing Spending for Stability and Growth. market outcomes in Croatia. EIZ-WP-1804. World Bank (2015). PISA 2015: Europe and Central Tomić, I., Botrić, V., & Žilić, I. (2018). Analysis of the Asia. World Bank, Washington, DC. https:// status and recommendations for the develop- openknowledge.worldbank.org/entities/publica- ment of further activities for people in NEET tion/2d6e2858-07f7-51cc-a2b1-02afec3af085 status. Croatian Ministry of Labor and Pension System. World Bank (2018a). Portraits of Labor Market Exclusion 2.0; Country Policy Paper (CPP) for Tur-Sinai, Aviad, Andrea Teti, Alexander Rommel, Croatia. Valentina Hlebec, and Giovanni Lamura (2020). “How Many Older Informal Caregivers Are There World Bank (2018b). The World Development Report in Europe? Comparison of Estimates of Their 2018 (WDR 2018)—LEARNING to Realize Educa- Prevalence from Three European Surveys.” Inter- tion's Promise. URL: https://www.worldbank.org/ national Journal of Environmental Research and en/publication/wdr2018 Public Health 17, no. 24 (December 19, 2020): 9531. https://doi.org/10.3390/ijerph17249531. World Bank (2019a). Performance of SP in Croatia. https://documents1.worldbank.org/curated/ Umapathi, N. Wang, D. O’Keefe P. (2014). Eligi- en/172051593098632012/pdf/A-Note-on-Per- bility thresholds for minimum living guarantee formance-of-Social-Protection-Programs-in- programs: international practices and implica- Croatia.pdf tions for China. https://ideas.repec.org/p/wbk/ hdnspu/83118.html World Bank (2019b). Enterprise Surveys: What business experience. Croatia 2019 Country Vidovic, Hermine, Vladimir Gligorov, Renate Haup- Profile. URL: https://www.enterprisesurveys.org/ fleisch, Mario Holzner, Katja Korolkova and content/dam/enterprisesurveys/documents/ Monika Natter (2011). Developing Efficient Acti- country/Croatia-2019.pdf= vation Approaches and Identifying Elements for Regional Cooperation in the Western Balkans, World Bank (2019c). Investing in Opportunities for WIIW, Research Reports | 374, October 2011. All: Croatia Country Gender Assessment. World Bank Group. 89 TOC Harnessing Human Capital for Growth in Croatia World Bank (2019d). National Development Strategy The Economic Journal, 118 (528): 591-630. Croatia 2030 Policy Note: Education and Skills. URL: https://hrvatska2030.hr/wp-content/ Župarić-Iljić, D. (2016). “Emigration from the Republic uploads/2020/10/Education-and-Skills.pdf of Croatia after the Accession to the European Union.” Friedrich Ebert Stiftung, Zagreb. World Bank (2019e). National Development Strategy Croatia 2030 Policy Note: Labor Market. The World Bank. World Bank (2019f). “Monitoring Occupational Shortages Lessons from Malaysia’s Critical Occu- pations List.” World Bank. World Bank (2020a). “Health workforce mobility from Croatia, Serbia, and North Macedonia to Germany.” World Bank. World Bank (2020b). “Who is Keeping Score? Estimating the Number of Foreign Workers in Malaysia”. World Bank. World Bank (2021a). Croatia Public Finance Review 2021. World Bank Group. World Bank (2021b). Survey: How has the COVID-19 Pandemic Affected Croatian Households? https://www.worldbank.org/en/country/croatia/ brief/survey-how-has-the-covid-19-pandemic- affected-croatian-households#2 World Bank (2022a). Croatia Country Economic Memorandum. Laying the Foundations: Boosting productivity to ensure future prosperity. World Bank Group. World Bank (2022b). Commitment to Equity – Croatia results. Internal data. World Bank (2022c). Active Labor Market Measures in Armenia. Analysis and way forward. Technical note, July 2022. World Bank (2022d). “Strengthening the Critical Occupations List and Leveraging the Malaysian Diaspora.” Inclusive Growth and Sustainable Finance Hub in Malaysia. © World Bank. World Bank (2023). The Future of Work: Technology and its Implications for Equity and Growth in Europe (presentation January 23, 2023) World Bank (2023b). World Bank Croatia Pensions RAS output. Yang, D. (2008). “International Migration, Human Capital, and Household Investment: Evidence from Philippine Migrants’ Exchange Rate Shocks,” 90 Croatia Country Office Radnička cesta 80/IX, Zagreb Tel: +385 (0)1 2357 222 www.worldbank.org/croatia WorldBankCroatia