Document of The World Bank FOR OFFICIAL USE ONLY Report No.: 125443-HR THE REPUBLIC OF CROATIA SYSTEMATIC COUNTRY DIAGNOSTIC (P161992) May 4, 2018 International Bank for Reconstruction and Development (IBRD) Europe and Central Asia THE REPUBLIC OF CROATIA FISCAL YEAR January 1 – December 31 CURRENCY EQUIVALENTS (Exchange Rate Effective as of May 1, 2018) Currency Unit Croatian Kuna USUS$1.00 HRK 6.12 ABBREVIATIONS AND ACRONYMS AES Adult Education Survey NCDs Non-communicable diseases BEEPs Business Environment and NEETs Neither employed nor in education or training Enterprise Performance Survey BERD Business expenditure on R&D NPLs Non-Performing Loans CAR Capital Adequacy Ratio OECD The Organisation for Economic Co-operation and Development CCA Croatian Competition Agency PIACC Program for the International Assessment for Adult Competencies CEE Central Eastern Europe PIM Public Investment Management CEPEJ European Commission for the PISA Programme for International Student Efficiency of Justice Assessment CERP Restructuring and Sale Center PMR Product market regulations EBRD European Bank for Reconstruction PPP Purchasing Power Parity and Development EC European Commission PTSD Prevalence of post-traumatic stress disorder ECA Europe and Central Asia R&D Research and Development ECEC Early childhood education and care RCA Revealed comparative advantage ECI Economic complexity index RIA Regulatory impact assessment EIF European Structural and Investment SAFE Access to Finance of Enterprises Funds EU European Union SAOs State Attorneys Offices FDI Foreign direct investment SCD Systematic Country Diagnostic GDP Gross Domestic Product SCIs Sites of community interest GEM Global Entrepreneurship Monitor SOE State-owned enterprises GERD Gross domestic expenditure on R&D SPAs Special protection areas GFKF Gross fixed capital formation STEM Science, technology, engineering and mathematics GVCs Global Value Chains TIMSS Trends in International Mathematics and Science Study LGBTI Lesbian, gay, bisexual, transgender, TFP Total Factor Productivity intersex MoSE Ministry of Science and Education TFPR Revenue Total Factor Productivity MPK Marginal Product of Capital TVET Post-secondary vocational training MPL Marginal Product of Labor USD United States Dollars MST Minimal Spanning Tree ZSE Zagreb Stock Exchange ii IBRD Vice President: Cyril Muller Regional Director: Arup Banerji Practice Managers Luis Felipe Lopez Calva, Marialisa Motta Country Manager Elisabetta Capannelli Task Team Leader: Moritz Meyer, Javier Suarez iii TABLE OF CONTENT ........................................................................................................................... Acknowledgments......................................................................................................................................... 1 Executive Summary....................................................................................................................................... 2 1. Introduction ....................................................................................................................................... 10 2. Boosting output growth and productivity .......................................................................................... 11 A. Recent economic performance ....................................................................................................... 11 Sectoral and sub-national patterns..................................................................................................... 15 Trade performance ............................................................................................................................. 17 Financial sector developments ........................................................................................................... 20 B. Productivity patterns ...................................................................................................................... 21 Aggregate productivity ....................................................................................................................... 21 Firm-level productivity analysis .......................................................................................................... 22 C. Key drivers for boosting output potential and productivity growth .............................................. 24 Improving the business environment ................................................................................................. 25 Strengthening the competition environment ..................................................................................... 30 Reducing State’s footprint in the economy ........................................................................................ 31 Enhancing the innovation ecosystem ................................................................................................. 33 Meeting private sector skills needs .................................................................................................... 36 3. Enhancing inclusion ............................................................................................................................ 40 A. Performance on the Twin Goals ..................................................................................................... 40 Low levels of labor income for the poor and vulnerable .................................................................... 43 Disparities across regions and by household demographics .............................................................. 46 B. Boosting participation in labor markets ......................................................................................... 48 Removing disincentives to labor market participation ....................................................................... 49 Reducing frictions to spatial mobility ................................................................................................. 51 Enabling a healthy and productive aging ............................................................................................ 52 C. Building Resilience .......................................................................................................................... 54 4. Ensuring sustainable paths of growth and inclusion .......................................................................... 56 A. Fiscal sustainability ......................................................................................................................... 58 B. Social sustainability ......................................................................................................................... 61 C. Environmental Sustainability .......................................................................................................... 64 5. Priority areas for policy change .......................................................................................................... 66 A. Enabling the emergence of a dynamic enterprise sector ............................................................... 67 B. Boosting participation and contribution of individuals to economic and social development ...... 68 C. Enhancing the performance of the public sector ........................................................................... 69 References .................................................................................................................................................. 74 List of Figures Figure 1: Stagnant Convergence with the EU ............................................................................................. 10 Figure 2. Croatia’s growth after the crisis was worse than most middle-income countries ...................... 11 Figure 3. Investment and private consumption drove growth before the crisis ........................................ 12 Figure 4. External balances deteriorated before the crisis ......................................................................... 12 Figure 5. External debt continued to rise well after the crisis hit ............................................................... 12 Figure 6. Investment in Croatia went to construction and services more than in other EU countries ...... 13 Figure 7: Unemployment shot up with the crisis ........................................................................................ 14 Figure 8: Gross Value-Added decomposition, ............................................................................................ 15 Figure 9. GDP per capita varied across counties......................................................................................... 16 Figure 10. Industrial production is highly concentrated ............................................................................. 16 Figure 11. Richer counties grew slightly faster than poorer counties ........................................................ 16 Figure 12. Wages grew faster in counties with already higher wages ....................................................... 16 Figure 13. Croatia’s exports increased more slowly than peers ................................................................. 17 Figure 14. Export performance has improved ............................................................................................ 17 Figure 15: The rise in exports has mainly been to the EU12 ...................................................................... 18 Figure 16: Croatia’s tourism sector purchases few inputs from other sectors .......................................... 18 Figure 17: Croatia’s FDI performance is low ............................................................................................... 19 Figure 18: Croatia’s banking sector non-performing loans are high .......................................................... 21 Figure 19. Productivity made a negative contribution to growth .............................................................. 22 Figure 20. Potential output in Croatia is below CEE peers ......................................................................... 22 Figure 21: Changes in TFP levels are driven mainly by the “within” component ....................................... 23 Figure 22. TFP decelerated in all sectors after the GFC .............................................................................. 24 Figure 23: The misallocation of capital rose ............................................................................................... 24 Figure 24: Croatia‘s institutions are far weaker than top performers........................................................ 25 Figure 25: Business environment has a direct impact on firm productivity ............................................... 25 Figure 26: Croatia scores poorly on the quality of institutions................................................................... 26 Figure 27. Croatia’s governance is among the worst in the EU .................................................................. 27 Figure 28: Croatia has low competition scores compared to other ECA countries .................................... 30 Figure 29: Croatia has more restrictive product market regulations than peers ....................................... 30 Figure 30: Sectors with more competition have a higher TFP median and lower TFP dispersion ............. 31 Figure 31: The productivity gap between state and private firms is large but narrowing ......................... 31 Figure 32: Sectors with low State presence have higher productivity and allocative efficiency................ 32 Figure 33. Innovation indicators are low in Croatia .................................................................................... 34 Figure 34. GERD is low and stagnant in Croatia .......................................................................................... 34 Figure 35. Business expenditures on R&D are concentrated in large companies ...................................... 34 Figure 36. Public-private co-publications are limited in Croatia................................................................. 35 Figure 37. The economic complexity of Croatia’s export basket is lower than peers ................................ 36 v Figure 38: Firms moving up the TFP distribution are reducing jobs ........................................................... 37 Figure 39. PISA scores in Croatia are below peers ...................................................................................... 38 Figure 40. Tertiary education is limited in Croatia...................................................................................... 39 Figure 41. Participation in lifelong learning is low in Croatia ..................................................................... 39 Figure 42. Income growth in the poorest decile was slow before the crisis .............................................. 41 Figure 43. Rich and poor suffered declines in income following the crisis................................................. 41 Figure 44. The rise in poverty was largest in Croatia among new member states ..................................... 42 Figure 45. Incomes fell in Croatia following the crisis by more than in peers ............................................ 42 Figure 46. The share of the vulnerable population is larger in Croatia than in peers ................................ 42 Figure 47. The share of labor income in the total is lower among the poor .............................................. 44 Figure 48. Dependency ratios are highest among the poor ....................................................................... 46 Figure 49. Lack of recent work experience is a major barrier to employment........................................... 48 Figure 50. Employment rates are lower among women than men at most ages ...................................... 49 Figure 51. The share of temporary contracts in employment has increased, particularly for the poor .... 50 Figure 52. Part time employment is low in Croatia .................................................................................... 50 Figure 53. Employment rates vary substantially by county ........................................................................ 51 Figure 54. Unemployment rates also vary by county ................................................................................. 51 Figure 55. In Croatia, emigration is as common as moving between counties .......................................... 52 Figure 56. Older working-age adults are less likely to be employed in Croatia than the EU average. ....... 53 Figure 57. The duration of working life is shorter in Croatia than in peers ................................................ 53 Figure 58. Social protection spending in Croatia is comparable to peers .................................................. 55 Figure 59. Social assistance coverage of the poor is low in Croatia ........................................................... 55 Figure 60. Croatia’s transfers make a smaller contribution to poverty reduction than in peers ............... 55 Figure 61. The contribution of social assistance to poverty reduction is small .......................................... 56 Figure 62. The fiscal deficit remained high after the crisis ......................................................................... 60 Figure 63. Public debt increased after the crisis ......................................................................................... 60 Figure 64. Taxes exceeded cash benefits for all but the poorest 10 percent ............................................. 62 Figure 65. Taxes and benefits, on net, increased poverty .......................................................................... 63 Figure 66. Taxes and benefits, on net, increased poverty in all household types ...................................... 63 Figure 67. The share of protected areas in natural capital has risen since 1995 ....................................... 64 Figure 68. Concentration of protected areas is associated with lower poverty......................................... 66 Tables Table 1: EU accession improved Croatia’s ability to export........................................................................ 19 Table 2: General Government Expenditures by Economic Classification, Percent of GDP ........................ 58 Table 3: Selected Priorities ......................................................................................................................... 70 Boxes Box 1: New international poverty thresholds and ICP estimates ............................................................... 43 Box 2: Europe 2020 social inclusion indicators ........................................................................................... 45 Box 3: Subnational welfare disparities across Croatia ................................................................................ 47 Box 4: Knowledge Gaps............................................................................................................................... 73 vi ACKNOWLEDGMENTS This report was prepared by a team co-led by Moritz Meyer and Javier Suarez, and comprising Zoran Anusic, Nina Arnhold, David Bernstein, Paul Andres Corral Rodas, Ana Paula Cusolito, Joao Pedro Wagner De Azevedo, Francesca De Nicola, Ivan Drabek, Jakob Engel, Josip Funda, Stjepan Gabric, Georgia Harley, Mariana Iootty De Paiva Dias, Levent Karadayi, Jonathan Karver, Austin Kilroy, Sanja Madzarevic-Sujster, Craig Meisner, Todor Milchevski, Magdalena Mishkovska, Ana Maria Munoz Boudet, Natalie Nicolaou, Georgiana Pop, Ismail Radwan, Daria Taglioni, Shawn Tan, and Ljiljana Tarade. Editorial support was provided by William Shaw. Vanja Frajtic provided communications support. Djamilya Salieva supported the team throughout the process. The team benefited from invaluable guidance from World Bank Group management and peers, including: Arup Banerji (Country Director); Elisabetta Capannelli and Carlos Pinerua (Country Managers); Paulo Correa, Luis-Felipe Lopez-Calva, and Marialisa Motta (Practice Managers); Thomas Lubeck (IFC Manager); Christian Bodewig, Andrea Liverani, Jean-François Marteau, Rogier van den Brink, and Isfandyar Zaman Khan (Program Leaders); Andrea Kucey (Country Program Coordinator); and Maria Davalos and Ivailo V. Izvorski (peer reviewers). The report also benefitted from insights from European Commission staff in DG ECFIN and DG Regio. The team is grateful to counterparts in the Government of Croatia, the Central Bank of Croatia, the private sector, academia and civil society for their insights during the various rounds on consultations. EXECUTIVE SUMMARY I. INTRODUCTION Croatia has made remarkable progress in economic reforms and living standards since independence. In the couple of decades since peace was restored, Croatia created a liberal democracy, established a market economy, achieved the status of upper-middle income country, and on July 1, 2013 joined the European Union (EU). Gross Domestic Product (GDP) increased by more than 4 percent a year from 1992 to 2008, and GDP per capita (in PPP nominal terms) reached US$22,000 or 63 percent of the EU28 GDP per capita level. During the same period, all income groups experienced welfare improvements, with the growth rate for the bottom 40 of the income distribution being higher than for the total population. Among the bottom 40, however, the increase benefitted mostly the second quintile, while the first quintile (the bottom 20 percent of the income distribution) did not experience significant improvements and poverty incidence remained largely unchanged at some 4.2 percent of the population. The global economic crisis ushered in a severe, lengthy recession and derailed convergence towards EU living standards. Growth prior to the 2008 crisis was driven by an increase in aggregate demand, fueled by an expansionary fiscal policy, capital accumulation and household consumption, financed by abundant liquidity in the global financial markets. The current account deficit, along with the debt of households, firms, and the public sector, rose sharply. With the onset of the crisis, borrowing costs shot up, capital flows plummeted, and external demand for Croatia’s exports fell. Households and firms reduced their purchases to limit the deterioration in balance sheets, so that investment and private consumption collapsed. Unemployment increased sharply, further depressing demand and reducing business and consumer confidence. The recession lasted for six years, reducing output by 12 percent. Income levels of the bottom 40 percent decreased by more than 2.3 percent per year in 2009-14, whereas incomes for the total population decreased by 2.4 percent. The share of the population living on less than US$10 PPP 2005 (in poverty or vulnerable) increased from 26.6 percent in 2009 to 33.4 percent in 2014. The severity of the Croatian recession, and the slowness of the recovery since 2014, has meant that Croatia is falling further behind the income levels of its Eastern European peers, driving public discontent and contributing to high levels of emigration of young and skilled workers, which has reduced productivity and the size of the labor force. Slow growth in Croatia is driven by limited productivity gains. Rapid growth prior to the crisis reflected rising labor force participation and massive investment, rather than increases in Total Factor Productivity (TFP). Total factor productivity made a significant, negative contribution to growth from 2002 to 2014, and only a small positive contribution in 2015-16. By contrast, in the same period growth in Croatia’s Eastern European peers, was mostly based on rising productivity. The decline in Total Factor Productivity during the recession largely reflected a slowdown in productivity performance of the most productive firms. This was driven by the tightening of credit conditions and policy uncertainty which depressed investment, and a poor business environment which impaired product market competition and limited the growth of more productive firms. For individuals, the sharp increase of poverty (measured at the household level) during the crisis and the modest reduction of poverty even during the period of high economic growth points towards limited labor force participation and wage growth of the bottom 20 and lack of resilience to economic shocks. Limited productivity improvements are reflected in Croatia’s lackluster export performance. The reforms undertaken for EU accession led to a significant rise in Croatia’s export market shares once the global recovery was underway. Most of Croatia’s leading export sectors have achieved increasing shares in global exports in 2011-15. Nevertheless, export growth from 2006 to 2016 remained significantly 2 below that of regional peers, and much of the rise in exports has been to lower-income EU members rather than to higher-income markets. Croatia remains at the margin of global value chains, thus missing important opportunities for raising productivity through learning. Services exports continue to be dominated by the tourism industry, which has picked up strongly since 2015 helped by difficulties facing competitors in North Africa and Turkey. The relatively high weight of the tourism sector, an industry which compared to knowledge-intensive services is less innovative and has fewer backward and forward linkages to other domestic industries, hampers the prospects of competitiveness enhancements in the rest of the economy. Transforming the role of the state is paramount to reignite the process of economic and social convergence. The success in reshaping how enterprises, individuals and the state interact will very much depend on the determination of the policy makers to balance the role of the state. This particularly refers to its role of the regulator, to steer competition, and the fine line to determine where in the public sector it should stay in charge and where it would do better by withdrawing. There will be specific areas where the state’s involvement is fundamental, such as in protecting the vulnerable and work on their integration. Ultimately, Croatia has an opportunity and an obligation to not only compare with its EU peers, but also to outperform where possible, in organizing a state that excels in service provision to its citizens, in the major public service areas, from judiciary and business environment to education and human development. II. ENTERPRISES Firm productivity patterns in Croatia show evidence of a lack of dynamism which hinders the process of industrial renewal and aggregate productivity growth. Firm exit is contributing to reduce aggregate productivity in most sectors, suggesting that more productive firms are exiting the market or that incumbents become less productive. Simultaneously, the contribution of firm entry, generally a key contributor to aggregate productivity growth, remains limited. In addition to global market conditions and policy uncertainties, these patterns are generally associated with a cumbersome business environment, constrained access to finance, low levels of product market competition, lack of competitive neutrality, and weak firm innovation capabilities. Despite improvements in recent years, a cumbersome business environment continues to burden firms, inhibiting private sector investment and distorting resource allocation. Croatia lags best performers in various Doing Business indicators. The key challenges faced by firms include regulatory instability, a high administrative burden, low transparency and predictability of administrative bodies, and long judicial procedures. In its action plan for 2017 the Government has identified a set of 104 measures to improve the business environment, including steps to reduce redundant administrative costs and to improve competition in the professional services market, which are expected to save up to 1.5 billion kuna for enterprises. Inefficiency, unpredictability, and delays in court processing cases are among the greatest impediments to business. Courts of general jurisdiction and commercial courts are generally perceived as too slow in processing cases. The World Economic Forum ranks Croatia 135th out of 137 economies in terms of the efficiency of the legal framework in settling disputes, a ranking that continues to slide as other economies reform to become more competitive. The number and disposition time in litigious civil and commercial cases remain among the highest in the EU. The extensive demands on judicial services are compounded by the provision of non-litigious administrative services (e.g. company registration, land ownership registration, etc.) and the lack of an effective fast-track procedure for resolving minor disputes. 3 Firms face difficulties in accessing finance. Croatian firms view access to finance as a more binding constraint than the average in the European Union (EU). The numerous government programs to support access to finance appear to be poorly coordinated, and targeted segments and financial products overlap substantially. Most programs also lack adequate monitoring and evaluation mechanisms. Moreover, equity markets are underdeveloped, constrained by small market size and little local institutional investor appetite. The limited availability of risk capital particularly affects new, innovative firms due to information asymmetries and appropriation risks, and is exacerbated by a cumbersome insolvency framework and an incomplete regulatory framework for the venture capital industry. The competitive environment is weaker than in many Europe and Central Asia (ECA) countries. The core elements of the competition law are broadly in line with EU practices, and Croatia has a solid competition policy enforcement mechanism. However, the regulatory framework appears to be less conducive to competition than in many ECA countries. The Organisation for Economic Co-operation and Development (OECD) finds that product market regulations (PMR) in Croatia are more restrictive than in peer countries. Regulatory restrictions appear particularly burdensome in the services sector, notably in network economies and professional services, although legal changes since the PMR scores were collected in 2013 may have eased some restrictions. Perceptions of the effectiveness of anti-monopoly policy in Croatia are far below the ECA average, calling in question the effectiveness of the Croatian Competition Agency (CCA). A still-inefficient insolvency framework obstructs exit and re-entry of business into markets, impairing the efficiency of resource allocation among firms. The large role of the state in commercial activities limits competition. Enterprises either partially- or wholly-owned by the central government operate in numerous sectors, including rail, road and air transport, hotels and restaurants, food processing, pharmaceuticals, water supply, financial services and services of motor vehicles. Nearly 700 companies report to sub-national, regional and municipal authorities. The State-Owned Enterprises (SOEs) account for over 10 percent of employment, a fifth of total turnover and a third of total assets. SOEs contribute directly to government deficits, with a net average borrowing of 0.6 percent of GDP between 2011-2014. The resources diverted to maintain the survival of SOEs may be better applied towards more efficient uses. Non-financial SOEs have a 40 percent higher leverage ratio than private firms but are less profitable: in 2014, the average return on equity was 4.5 percent for private companies but almost zero for SOEs. Rates of return on assets are below that of SOEs in other CEE countries, and the rate of state subsidies is higher than the EU average. SOEs affect factor returns, influence output prices through product market competition and impair market incentives to become competitive. There is some indication that high SOE presence impairs growth, as productivity and allocative efficiency are higher in sectors with low state presence. Finally, rigid wage setting practices in SOEs, where wages are higher (controlling for employee characteristics) than in private sector firms, may distort wage setting in the private sector. Innovation in Croatian firms is limited. Croatia is a moderate innovator and was ranked 32nd out of 36 countries in the 2016 European Innovation Scoreboard, and 106th out of 137 countries for the innovation pillar of the 2017-2018 Global Competitiveness Index. Both rankings have been on a declining trend in recent years. Croatia is also falling behind its peers in the level of gross domestic expenditure on research and development (GERD), which has yet to recover the level of 2008. The share of Croatian enterprises engaged in innovative activities is below the EU average, and they tend to favor non-R&D versus R&D innovation activities. In line with peers, Business Expenditure on R&D (BERD) is also concentrated in large companies (mainly pharmaceutical, telecommunications, agricultural, and food and beverage industries). Croatia also stands out for the small share of R&D expenditures by small firms. Key factors explaining the low level of R&D-driven innovation, especially among medium and small firms, include limited access to internal and external resources (both funds and qualified personnel), limited 4 information on technology and markets, the concentration of tax incentives with large firms, modest research excellence, barriers to science-industry collaboration, and pervasive weaknesses in the innovation ecosystem governance. The weak performance of the innovation ecosystem is reflected in the limited complexity of Croatia’s goods export basket. III. INDIVIDUALS Higher participation of the labor force and better skills among working-age individuals are fundamental to raise the growth potential of Croatian economy. Disincentives to labor market participation, low educational attainments and skill gaps, deficits to lifelong learning, and the relatively high prevalence of chronic and non-communicable diseases among the working population, undermine the capacity of individuals to contribute to economic and social development and to prosper. Disincentives to work reduce labor market participation for various groups. The need to pay additional income taxes and social security contributions, coupled with losing various social benefits, impose a marginal tax rate for poor individuals that is roughly comparable to the EU average. However, the marginal effective tax rate is particularly high for potential low-wage single earners with children. Generous eligibility criteria and the co-existence of multiple pension schemes, early retirement pensions, survivor’s pensions, long-career, or disability pensions can trigger an early exit from the labor market, and high payments through disability insurance reduce participation. The limited formal care for children and the elderly, coupled with low flexibility in working arrangements (only 6 percent of Croatia’s employed population worked part-time in 2016, compared to nearly 20 percent in the EU as a whole) create a particular burden for women in reconciling work and family responsibilities. Greater efforts are necessary to boost labor market participation by older workers. Continuous delays in the drafting and implementation of new legislation have impeded retraining of an aging work force , including (long-term) unemployed and inactive workers. Life expectancy is higher than EU peers with similar income levels. Nevertheless, greater emphasis on preventative health services could reduce the incidence of chronic and non-communicable diseases, which is 45 percent higher than the EU average, enabling individuals to work longer and more productively. The working age population (aged 15 to 64) is projected to decline by 30 percent by 2050, hence promoting healthy aging is critical to moderating the fall in the workforce. Croatia’s hospital-centric health system and services delivery network is not well suited to the greater emphasis required on preventative services. Labor market reforms have facilitated a rise in temporary employment. The share of individuals with temporary contracts increased to 19.3 percent of those employed between 15 and 64 years in 2016 (the EU average is 12.0 percent). The share of temporarily employed was highest among poorer working-age individuals. More than 60 percent of all unemployed who transitioned into a job worked under temporary contracts, which illustrates how labor market deregulation can provide an entry path into permanent employment. However, the rise in temporary employment also risks segmenting the labor market. Deficiencies in education and training systems hamper the development of human capital and contribute to slow productivity growth and limit social mobility. Low PISA and TIMSS scores compared to some peer countries, particularly in mathematics and science, reflect quality gaps in schooling. The number of tertiary education graduates in science, technology, engineering and mathematics (STEM) is low due to limited attention to STEM classes in the curriculum throughout the education pipeline and deficits in training for teachers. Only 29 percent of 30-34-year-olds had tertiary education in 2016, far below the EU average of 39 percent. While participation in post-secondary vocational training (TVET) is high, the training curricula do not always reflect employers’ demand for skills, so that nearly half of those with vocational training work outside of their field of specialization. The divide between the skills 5 demanded by firms and those provided by the workforce is magnified by the limited private sector input into programming and funding for TVET programs. Only 3 percent of adults between 25 and 64 years participate in some form of workforce education or training, far below many peers in the EU, which limits employability in a quickly changing labor market. Planned curricula reforms for early childhood education, primary to tertiary education, and TVET have been delayed due to lack of coordination, cooperation and commitment among stakeholders and limited policy effectiveness. A pilot phase of implementation is now expected to begin in September 2018. Social assistance was insufficient to avoid a sharp rise in poverty following the crisis. The share of the population living on less than US$10 PPP 2005 (in poverty or vulnerable) increased from 26.6 percent in 2009 to 33.4 percent in 2014. General government spending on social protection accounts for 14.2 percent of GDP, which is in line with other Member States in the European Union. However, Croatia spends a higher fraction of social protection on contributory social insurance benefits, including disability pensions, sickness benefits and old age pensions. This reduces the available budget for transfers to families and children, and for other non-contributory social assistance programs. Thus, the coverage of the bottom 20 percent through social assistance (around 56 percent) is below the average for the European Union. Fragmentation of public transfer programs, different sets of eligibility criteria for different services, the lack of monitoring and information systems, and weak coordination between agencies at the central and regional level also limit the effectiveness of social protection spending. Regional differences in incomes and labor market outcomes are substantial and persist, with no sign of convergence. The share of the population living below the US$10 PPP threshold in 2013 was 45.2 percent in rural areas but only 28.6 percent in urban areas. Employment rates in the 15 to 64 age group are lower, unemployment rates are higher and poverty rates are higher in the eastern part of the country. Convergence in labor market outcomes between regions might be expected as workers leave high- unemployment/low-wage areas to improve their job prospects. However, workers in lagging regions often face high transactions costs in moving, given Croatia’s limited rental market and declining property prices in lagging regions following the crisis. The scarcity of rentals in prosperous regions also encourages external rather than internal migration, particularly given the large wage gap between Croatia and the traditional countries of destination for migration (including Germany and Austria) and social networks in these countries that encourage external mobility. Partially as a result, internal migration is below the EU average, while external migration is among the highest in the EU. Declining population in some rural areas also impairs the sustainability of social services, given the high fixed costs. IV. THE STATE Successful public policy making and implementation is hampered by ineffective public sector institutions. Croatia’s public sector performance lags EU peers across most governance indicators. The frequent changes in government are accompanied by changes in technical staff, making it difficult to carry through with consistent policies and reducing the ability of the authorities to exercise their oversight functions for SOEs. Insufficient coordination and cooperation between agencies and levels of government impairs policy coherence. Poor legislative quality has led to frequent amendments to address shortcomings, increasing uncertainty and raising the costs of compliance. Public sector is particularly weak at the local and regional self-government units level, where high fragmentation raises costs and reduces the quality, effectiveness and sustainability of services delivery. Powerful groups, and perhaps public attitudes, have obstructed reforms. A substantial number of autonomous, self-organized groups have considerable ability to prevent generally beneficial changes that threaten their privileges. The deep politicization of the civil service, the prevalence of the SOE sector and weak governance structures provide a terrain favorable to clientelism and capture. Croatia receives low scores in the institution indicator in the Global Competitiveness Report of the World Economic Forum. 6 In addition, the 2016 Life in Transition III survey suggests that public support for a market economy is among the lowest in ECA, and the Global Entrepreneurship Monitor for Croatia 2016 (GEM) finds that successful entrepreneurs do not hold high social status and their activities are mostly not covered by the media. These attitudes may play a role in undermining support for reform. Fiscal weaknesses threaten sustainability and impair growth. The level of public debt, which has practically doubled since 2008, peaked at some 86 percent of GDP in 2014. While the reduction of the fiscal deficit and economic recovery are reducing the debt-to-GDP ratio, it is projected to remain close to 80 percent in 2018. The bulk of the debt accumulated during the crisis was either issued abroad or issued domestically in or indexed to the euro, although currency risk exposure is mitigated by the tightly managed float of the kuna’s exchange rate against the euro. Credit default swaps for Croatia is higher than for most of its peers, indicating the precariousness of Croatia’s sovereign borrowing conditions. The recently adopted strategy for debt management ignores the financing needs of extra-budgetary entities, which have contributed to a large extent to the build-up of debt. Croatia’s debt position could be eased, and growth-supporting expenditures increased, by a reduction in the levels of subsidies and public wage bill. Current expenditures are significantly higher than most EU peers, perhaps reflecting inefficient consumption of inputs (e.g., energy consumption, space renting) or higher unit prices due to insufficiently competitive public procurement. The quality and sustainability of public infrastructure raises concerns. Croatia made large infrastructure investments during the 2000s in response to the need for a visible integration of Croatian territory after independence, efforts to spur industrial growth and tourism, and the desire to integrate into the broader European network. Despite heavy investments in the road network, as well as some improvements in ports and airports, Croatia underperforms all EU peers in the World Bank Logistics Performance Index. This is largely due to underdeveloped infrastructure such as port-rail interfaces, slow and unreliable rail operations, and cooperation and communication failures among stakeholders. In addition, the impetus to develop public infrastructure often led to over-investment, with design based on high standards which result in high maintenance costs as the infrastructure ages. The need to complete important EU transport network, and comply with EU environmental standards especially in the solid-waste and wastewater sector, poses a clear fiscal burden unless utilities reforms are accelerated and adequate design, taking into account efficiency and cost-recovery concerns, are introduced to avoid excessive costs. The situation is compounded by weak corporate governance, low profitability, and high indebtedness of infrastructure SOEs. The road sector, in particular, faces high debt stock relative to earnings, which result in ratios of debt to cash flow available for debt service exceed industry averages. The companies have limited access to long term financing, leading to a mismatch between short debt tenor and the long life of road assets. They also face large currency risks, with the bulk of the debt denominated in euros. Measures are needed to adjust the level of service to correspond to demand, and reestablish operational efficiency and sustainability. Croatia needs to enhance the efficiency and growth impact of public infrastructure funds, including the projected large inflow of EU funds. Measures to enhance the efficiency and growth impact of public infrastructure funds, including the projected large inflow of EU funds, involve strengthening of public investments management (PIM) planning, contracting, and implementation capacities, along with better strategic planning and establishment of a medium-term budgeting framework. Dispersion of public investment across various levels of government and SOEs pose additional challenges in terms of coordination, rationalization and effectiveness of investment decisions. The government also faces challenges in ensuring the preservation of natural capital, which is critical to growth. Including indirect income effects, the contribution of tourism to GDP rose to over 25 percent in 2016, and is projected to reach 32 percent of GDP by 2027. The increase in visits to the major national 7 parks will require an increase in their carrying capacity. Internal population growth and/or migration to areas that have greater economic benefits (i.e. protected areas) also is placing additional pressure on local resources. Only 18 percent of municipal waste is being recycled, so substantial efforts are required to meet the EU recycling target by 2020 of 50 percent. The expected completion of new waste management centers has been delayed from 2018 to 2023, complicating the planned closure of unsanitary, illegal landfills. Resilience to natural hazards poses further challenges, especially for the agriculture and tourism sectors which are the most exposed to the impact of climate change and occurrences of extreme weather events. V. CONCLUSION A comprehensive reform program is essential to raise productivity growth. Without a step-up in productivity, Croatia could experience a deterioration of social conditions and prolonged economic stagnation or deterioration. Emigration could accelerate as the most qualified workers take advantage of their mostly free access to richer and faster-growing economies elsewhere in the EU. Population aging (by 2050, the working age population, aged 15 to 64, is expected to decline by 30 percent) could further depress output. Dissatisfaction with limited economic prospects and continued barriers to social mobility could increase political and social tensions. Institutional weaknesses and the inability to achieve the necessary reforms to continue rapid growth limits development in many formerly successful economies, and is often referred to as the ‘middle-income trap’. Transforming the role of the State to set the right incentives for individuals and enterprises to be productive and prosper, while guaranteeing the long-term sustainability of its distributive goals, is the cornerstone of the reform agenda. Building ‘efficiency-enhancing’ institutions will require changing how the government provide for public services and how it regulates the capital, product and labor markets. The need for public sector transformation is compounded by today’s world of rapid change, which poses renewed and complex challenges. The key priority reform areas, retained based on the analysis presented in the report and their critical role to allow Croatia to resume and sustain convergence towards living standards in the European Union and summarized in the Table below, can be articulated around the following three objectives: (A) Enabling the emergence of a dynamic enterprise sector; (B) Boosting participation and contribution of individuals to economic and social development; and (C) Enhancing the sustainability and performance of the public sector. Time is pressing. Rapid technological change is increasing the returns to skills and boosting productivity in more successful and sophisticated economies. The longer the delay in improving skills and enabling Croatian firms to participate effectively in innovation-led productivity growth, the more difficult it will be to catch up. Government’s objective to adopt the Euro in the next 7 to 8 years offers an opportunity to accelerate the structural reform agenda and, beyond the macro-fiscal convergence criteria, strengthen Croatia’s institutional capacities and address domestic competitiveness constraints and achieve resilient convergence. Greater efforts should be made to generate public support for the reform program. High public debt, an aging population, low labor force participation, high levels of emigration, strict rules controlling in- migration, and an institutional and policy framework which is fragmented and that impairs productivity and limits social mobility threaten the considerable economic and social progress achieved over the past few decades. Powerful groups and frequent political changes that have undermined the continuity and effectiveness of policy have limited reforms that are essential to growth and development. Transparent communication, as well as greater attention and resources devoted to explaining the benefits of reform could increase public understanding of the policy changes that would maintain the sustainability of the economic and social system. 8 Summary of key objectives and related priorities Objective A: Enabling the Objective B: Boosting Objective C: Enhancing the emergence of a dynamic participation and contribution of performance of the public enterprise sector individuals to economic and sector social development Priority 1. Foster a more competitive Priority 4. Improve learning results Priority 7. Pursue efforts to reduce environment and skills of the workforce fiscal and debt vulnerabilities Priority 2. Boost justice system Priority 5. Foster labor market Priority 8. Improve quality and performance participation efficiency of public administration Priority 3. Unleash firm innovate Priority 6. Ensure productive aging Priority 9. Ensure the preservation capabilities through lifelong learning, healthy of natural capital aging and promotion of longer working lives Note: Table 3 in the main text details specific actions that would help to achieve objectives and priorities summarized here. 9 1. INTRODUCTION 1 1. Croatia has made remarkable progress since independence, but daunting challenges impede the reforms required for sustainable development. In the space of a couple of decades since independence and its subsequent destructive war, Croatia has established a liberal democracy and a market economy, and achieved upper-middle income country. EU accession played a key role in these accomplishments by supporting legislative improvements, strengthening institutions and increasing the availability of funds. Nevertheless, implementation of the reform program continues to face significant obstacles. Capture of public institutions by powerful interest groups has limited the effectiveness of reforms. And frequent changes in government, each accompanied by changes in technical staff in public administration, have reduced momentum for reform and the consistency of government policies over time. 2. The global economic crisis started for Croatia one of the longest economic recessions in modern European history. The boom prior to 2008 was supported by large capital inflows that financed a sharp rise in investment and private consumption, while productivity deteriorated, external imbalances widened, and the liabilities of households, firms and the public sector increased. The boom ended abruptly with the crisis, as global demand fell and capital inflows dried up. Households and firms sharply reduced expenditures in the face of declining revenues, large levels of debt and rising borrowing costs. Over the next six years, GDP fell by 12 percent, household incomes dropped across the income distribution, and poverty and vulnerability increased sharply. The recovery since 2014 has been slow, and, unlike Central and Eastern European peers2, Croatia has yet to resume convergence with EU income levels (Figure 1). 3. The recession had a profound impact on Figure 1: Stagnant Convergence with the EU public sentiment and shapes low expectations of GDP per capita at EU28 at Purchasing Power Standards improved living standards Perceptions with regards to present and future welfare and socioeconomic mobility are notably pessimistic, consistent with other countries in the European Union.3 When asked to place themselves on a ten- step welfare ladder today, four years prior, and four years into the future, more than half of respondents feel that their welfare has remained unchanged since 2012 and will remain the same in 2020. And less than 5 percent believe that their welfare has improved since 2012 and will continue to improve through 2020. Croatia is among the five Source: WB staff calculations using Eurostat, WEF data. EU countries in the sample with the lowest level of perceived welfare improvement over time. 4. Lack of convergence with the EU threatens to increase public dissatisfaction. Many Croatians tend to see Western Europe as their reference point, reflecting close links through large migrant communities in Austria and Germany. Unless the Government can deliver on people’s aspirations for income levels that converge with those of Western Europe, emigration may rise further, given Croatians’ 1 The Report relies on the data available until November 2017. 2 CEE peer group countries: Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovak Republic, Slovenia. 3 Upcoming EU flagship on inclusive growth. 10 unlimited access to working in EU countries. Fostering the creation of new and sustainable jobs and enhancing the public sector’s performance will thus be critical to earn public trust and avoid the risk that net emigration and slower convergence generate a vicious circle, undermining prospects for sustained growth. 5. This Systematic Country Diagnostic (SCD) provides an assessment of where Croatia stands in terms of poverty reduction and shared prosperity (namely the twin goals), and how it could maximize progress towards these goals. The SCD’s purpose is not to assess the Government’s development plans but rather to articulate an independent identification of, and rationale for, priorities for achieving the twin goals. The analysis is meant to inform subsequent engagement between Croatia and the World Bank Group on the Country Partnership Framework. The document is structured as follows: chapter 2 presents key determinants and constraints of growth; chapter 3 presents determinants and constraints for enhancing inclusion; chapter 4 discusses financial, social and environmental sustainability, and identifies key sustainability-related constraints on achieving the twin goals over the longer term; and Chapter 5 presents the priorities for reform and concludes. 2. BOOSTING OUTPUT GROWTH AND PRODUCTIVITY A. Recent economic performance 6. Building on the market-oriented structural reforms introduced during the 1990s and regained stability, Croatia enjoyed a period of sustained economic growth before the global economic crisis. GDP per capita rose by 4.2 percent per year at constant prices from 2000 to 2008, and GDP per capita (in Purchasing Power Parity (PPP) nominal terms) more than doubled to above US$22,000 reaching 63 percent of the EU28 GDP per capita level (at PPS).4 This performance was modest, though, compared to many middle-income countries as well as Eastern European peers, including some with similar or higher initial income levels. This relative underperformance has worsened since the global financial crisis, as Croatia was not only hit severely but faced a longer recession period than most peers (Figure 2). Figure 2. Croatia’s growth after the crisis was worse than most middle -income countries Source: WB staff calculations. 4 Refers to purchasing power standard, according to the Eurostat methodology. 11 7. Economic growth between 2000 and 2008 was driven mainly by domestic demand, fueled by large capital inflows. The main drivers of expansion were capital accumulation and private consumption (Figure 3). Capital accumulation, proxied by the ratio of gross fixed capital formation (GFKF) to GDP, averaged 25 percent, which compares favorably with upper middle-income countries and other fast- growing economies in the region. Rapid growth of domestic demand triggered an import surge, which explains the negative contribution of net foreign demand throughout the period, and resulted in a fourfold increase of the current account deficit (Figure 4). External funding relied mainly on debt financing, which led to an increase in gross external debt (Figure 5). Figure 3. Investment and private consumption drove growth before the crisis (Contribution to GDP growth in percentage terms, 2002-2016) Source: CBS, and WB staff calculations. Figure 4. External balances deteriorated before the Figure 5. External debt continued to rise well after the crisis crisis hit (Percent of GDP 2006-16) (Percent of GDP, 2008-16) Source: WB staff calculations. Source: Croatian National Bank. 8. The composition of GFKF, however, was biased towards non-tradable sectors rather than the tradeable and productivity-enhancing manufacturing sector. The share of total GFKF devoted to manufacturing from 2002 to 2015 was well below the EU average, while the share of GFKF in non- tradeable sectors was among the highest in the EU (Figure 6). Croatia had the second-highest share of GFKF in construction and the highest share of GFKF in finance, insurance, accommodation and restaurant services. The low share of GFKF going to manufacturing may indicate that opportunities for productivity growth are being missed. 12 Figure 6. Investment in Croatia went to construction and services more than in other EU countries (Gross fixed capital formation in selected sectors, 2002-2008 and 2009-2015) Note: Average of the GFKF share of GDP is taken over the two periods, 2002-08 and 2009-15. Some countries do not have the data for 2014 and 2015 and the average is taken from 2009 to the latest years. Source: Eurostat and Croatia Bureau of Statistics. 13 9. The global crisis exposed fundamental weaknesses of this growth pattern and pushed Croatia into a prolonged recession. Borrowing costs increased, capital inflows dried up, and external demand for Croatia’s exports dropped, leading to a sharp fall in investment and credit-driven private consumption. The economy was pushed into a vicious circle of a sharp contraction in employment, private sector illiquidity and eroding business and consumer confidence. A six-year long recession reduced output by 12 percent and investment by 33 percent, while unemployment doubled to 17.3 percent by 2013 and youth unemployment reached 50 percent. 10. A more favorable external environment, with low energy prices and more global liquidity, along with EU accession, helped turn the tide. A strong contribution of exports–underpinned by a good tourist season and recovery of merchandise exports to neighboring countries–and a pick-up of capital investment and private consumption–boosted by low energy prices, income tax cuts, and the conversion of Swiss franc denominated loans–led to the return of positive GDP growth (2.3 percent) in 2015. GDP growth, which is spread across sectors, is estimated at 3.0 percent in 2016, 2.8 percent in 2017 and projected at 2.6 percent in 2018. Private consumption and exports of goods and services are expected to remain the key drivers of aggregate demand in coming years. Gross fixed capital accumulation is also expected to gradually increase and support GDP growth, although its contribution will likely remain lower than during the pre-crisis period. The contribution from government spending to growth is expected to remain subdued in the context of efforts at fiscal consolidation. 11. Unemployment rose sharply Figure 7: Unemployment shot up with the crisis during the crisis, but has begun to (Percent of working age, 15-64) decline. The Croatian labor market adjusted more slowly to the fall of economic activity during the crisis than in other EU States from Central and Eastern Europe.5 The lack of flexibility in nominal and real wages led to major labor shedding and a surge in unemployment, which had doubled by 2013 to reach 17.3 percent of the labor force (Figure 7). Employment contraction affected most sectors and was especially pronounced in sectors, such as construction and retail, Note: June 2007 is the average of four quarters to June. Source: which are dependent on buoyant Croatian Bureau of Statistics. domestic demand. Employment levels in economic activities dominated by the public sector remained stable (for public administration and education) or even increased slightly (for health and social work). Labor market outcomes have begun to improve, as total employment rose, and the unemployment rate fell sharply in 2015. However, the decline in the unemployment rate since 2013 is more due to early retirement, population aging and net migration outflows than the modest increase in the number of individuals employed. At about 65 percent, the labor force participation rate remains some 8 percentage points lower than the EU average. 12. Labor market reforms supported a recovery in employment levels. In an effort to facilitate labor market adjustment, authorities introduced reforms to the Labor Code in 2013 to increase flexibility in the use of temporary contracts and in 2014 to reduce rigidities in employment protection for regular contracts. The rebound in employment observed from 2014 has been largely driven by an increase in 5 Orsini and Ostojic (2015). 14 temporary contracts. The extent to which this trend will continue and lead to a segmentation of the labor market along permanent/temporary contract lines remains unclear. The increased flexibility in the use of regular contracts introduced in 2014 might lead firms to increase hiring on a permanent basis once economic activity and confidence strengthen further.6 Sectoral and sub-national patterns 13. Overall, structural transformation since 2000 has remained modest, with a slight increase of the contribution of the services sector to total value added matched by the progressive decline in relative contribution of agriculture. Services sectors accounted for about 70 percent of VA in 2016, with agriculture and industry representing respectively 4 and 21 percent of the total. Industry, construction and trade have been the key sectoral drivers of gross value-added variations, with large contributions to growth prior to the crisis and a sharp contraction during 2009-2013. The recovery after 2015, appears broad-based with a surge in export-oriented industry, and a recovery of trade and hotel and restaurant services, boosted by a robust tourism performance (Figure 8). Figure 8: Gross Value-Added decomposition, (Percent, 2002-2016) Source: CBS, and WB staff calculations. 14. Employment by sector reflect these trends, with the highest shares of employment corresponding to the industry, the trade services, and the public administration and social services. The employment in Public administration and social services account for a disproportionate share of the total and is the only sector wherein the absolute number employed grew in the immediate aftermath of the crisis. The number of employees in agriculture, while low in absolute terms, was cut in half between 2000 and 2016, reflecting a transition of unskilled workers to more industrial activity throughout the period. Meanwhile, wages have been highest in the services sector since 2000, though these observed a disproportionate drop at the onset of the crisis after observing an increase of nearly 50 percent between 2000 and 2008. Wages in industry have remained relatively low, but have grown steadily since 2000 in spite of the crisis. Wages in public administration, while declining after the crisis, appear to have been less sensitive to the crisis. 15. Income levels and labor market outcomes differ greatly among regions. GDP per capita averaged 10,228 euros in 2014, with regional estimates ranging from less than 6,000 Euros in Virovitica- Podravina and Slavonski Brod to nearly 18,000 Euros in the City of Zagreb (Figure 9). GDP per capita in the eastern part of the country is at least half the level in the capital city, and poverty rates are higher. 6 Brkic (2015). 15 Disparities in the level of economic activity (Figure 10) reflect differences in the structure of the economy but also point towards lower levels of education and lower employment rates in some parts of the country. For example, the share of the adult population (24-64 years of age) with tertiary education is twice as large in Zagreb and surrounding areas than in Slavonski Brod and surrounding areas. Employment rates (Figure 53) and unemployment rates (Figure 54) of working-age individuals (between 15 and 64 years of age) also differ substantially across the country. The three counties with the highest sales value of industrial production combined are home to 33 percent of the population, yet generate 44 percent of total production in the country. Figure 9. GDP per capita varied across counties Figure 10. Industrial production is highly (Euros, 2014) concentrated (Million of Kunas, 2014) Note: GDP per capita and the total value of industrial production are reported at the county level (21 counties in Croatia, NUTS3). Horizontal line on left panel refers to national (population weighted) average. Yellow bars represent Zagreb and Zagreb City, red indicates continental Croatia (without Zagreb), and blue indicates Adriatic Croatia. Source: WB staff calculations using Croatian Bureau of Statistics data. Figure 11. Richer counties grew slightly faster than Figure 12. Wages grew faster in counties with poorer counties already higher wages (GDP per capita, levels in 2000 and growth between (Net wages, levels in 2000 and growth between 2000 2000 and 2014) and 2014) 4.5 6 3 13 20 21 2 5 9 % Annualized growth % Annualized growth 10 2 4 3 4 19 21 16 12 5 4 6 5 1 8 15 12 4 17 9 10 7 7 20 3.5 8 16 15 3 11 14 18 1 18 13 11 14 6 17 3 2 19 2000 4000 6000 8000 10000 2800 3000 3200 3400 3600 3800 GDP per capita in 2000 (Euros) Net wage level in 2000 (Euros) Note: GDP per capita and net wages are reported at the county level (21 counties in Croatia, NUTS3). Bubble size is proportional to population size. Fitted (weighted) line in black. Numbers to identify counties: Bjelovar-Bilogora 1, City of Zagreb 2, Dubrovnik- Neretva 3, Istria 4, Karlovac 5, Koprivnica-Križevci 6, Krapina-Zagorje 7, Lika-Senj 8, Međimurje 9, Osijek-Baranja 10, Požega- Slavonia 11, Primorje-Gorski kotar 12, Sibenik-Knin 13, Sisak-Moslavina 14, Slavonski Brod-Posavina 15, Split-Dalmatia 16, Varaždin 17, Virovitica-Podravina 18, Vukovar-Sirmium 19, Zadar 20, Zagreb 21. Source: WB staff calculations using Croatian Bureau of Statistics data. 16 Trade performance 16. Croatia’s export performance remains Figure 13. Croatia’s exports increased more slowly lackluster. Exports grew at 3.9 percent per year than peers from 2006Q1 to 2016Q2, well below the rate of (Annual average growth in exports, 2006Q1-2016Q2) regional peers (Figure 13). Croatia’s share of the world export market–a measure of export competitiveness–increased by only 0.3 percent over that period. Croatia did achieve some improvement in trade performance compared to the global average: by 2014-15 Croatia’s exports as a share of GDP had risen to slightly above the expected level, given its per capita income (Figure 14). However, Croatia’s Eastern European peers achieved much higher levels of exports. Also, this measure does not control for Source: WB staff calculations using UN Comtrade. size of the economy, and a small economy such as Croatia is expected to have higher trade openness, to finance the variety of goods that cannot all be produced domestically at efficient scale. 17. Accession to the European Union has helped boost export growth. Croatia’s reform program before EU accession increased the country’s supply-side export capacity (last column of Table 1). However, the country’s export market share fell from 2006 to 2012 due to both price-related factors, and most significantly, a fall in demand from Croatia’s core export markets in the Eurozone with the global crisis. The pay-off to improved export capacity came with global recovery, as export market share increased by almost 5 percent in the three years following EU accession. Much of the rise in exports since 2005 has been to lower-income EU members (countries who joined the EU after 2004), whose share of Croatia’s exports rose from 15 percent in 2005 to 23 percent in 2015 (Figure 15). Croatia has been successful in these countries thanks to a small set of strong, multi-product exporters, both in traditional sectors and new activities. These exporters have also invested in production facilities in South Eastern Europe establishing themselves not only as successful exporters but also as regional multinationals. Figure 14. Export performance has improved (Export/GDP versus per capita GDP, 2007-08 and 2014-15) Source: World Development Indicators. 17 18. Growth in world imports in sectors of Figure 15: The rise in exports has mainly been to the Croatia’s specialization is picking up pace. Most of EU12 Croatia’s leading export sectors have achieved (Share of exports, percent) increasing shares in global exports in 2011-15. The majority have seen increased shares in global exports in sectors where global imports are increasing. These sectors include pharmaceuticals products, electrical machinery, vehicles, apparel, footwear and wood products. Only a few of Croatia’s top 20 export sectors have both lost market share in world exports and faced lower import demand. In addition, Croatia had a few products – aluminum, electrical machinery and equipment and plastics – with growing world demand and declining world market export shares. Source: WB staff calculations using UN Comtrade data. These results suggest a remarkable change from 2008-12, when the majority of Croatia’s top exports were in declining sectors and Croatia’s main export markets in the EU experienced both a deeper and more protracted recession than the rest of the world (including the US, and particularly the emerging economies of Asia, Africa and Latin America). 19. Croatia has larger services exports and a greater share of travel and tourism services than its peers. Croatia’s services exports are larger in value to its merchandise exports. Two-thirds of Croatia’s services exports are travel and tourism, while communication and transport services exports only make up a small share and financial services exports are marginal. Among peers, Bulgaria is the only other where travel makes up more than half of total services exports. By contrast, transport services exports play a larger role in most other peers, particularly in Lithuania, Estonia, and Latvia. The strong performance of the tourism sector in recent years is reflected in the marked increase in the revealed comparative advantage (RCA) index for personal and recreational services (which includes tourism), while the RCA indices for other services sectors generally associated with higher value added, such as financial, telecom, or other business services, have remained stagnant. The importance of the tourism Figure 16: Croatia’s tourism sector purchases few sector is one reason why the sophistication of inputs from other sectors Croatia’s services exports remains largely behind Accomodation and Food (2014) .015 that of peers’ services exports. 20. The tourism sector’s limited of integration Backward linkages .01 with the local economy suggests that the sector may not enhance Croatia’s competitiveness. The tourism sector earns foreign exchange and provides .005 jobs, especially to the low- and medium-skilled, who tend to be more vulnerable to unemployment. LTU However, Croatia’s hotels and restaurants sector HUN SVK BGR LVA POL HRV SVN 0 sells few intermediate inputs to other sectors 0 .002 .004 .006 Forward linkages .008 .01 (forward linkages), and only produces final products. Moreover, Croatia’s tourism sector purchases fewer inputs from other sectors Note: Indirect forward and backward linkages in export (backward linkages), compared to the level of value added as a share of total exports. Source: WB staff purchases by the tourism sector in other countries calculations using World Input-Output Database. 18 (Figure 16). Finally, note that the heavy reliance on one single sector can expose the entire economy to considerable risk. Table 1: EU accession improved Croatia’s ability to export Export market Pull factor Push factors share change Geographical Sectoral Values Price Volumes 2006Q1 –2008Q4 0.26 1.65 -0.51 -0.88 0.69 -1.57 2009Q1 –2012Q4 -3.69 -4.59 -0.08 0.98 -1.58 2.55 2013Q1 –2016Q2 4.78 -0.65 1.01 4.42 -0.40 4.82 Note: Indicators are expressed in log-difference form, which allows for additivity across indicators. Source: World Bank Measuring Export Competitiveness Database. 21. Croatia remains on the periphery of networks of global value chains (GVCs). 7 Croatia’s trade with the main centers of global trade in intermediates, China, Germany and the United States, is marginal. The EU – in particular, Italy – is the most relevant trading partner for Croatia and Croatia is the most important trade partner in intermediates goods for Bosnia Herzegovina, its close geographical neighbor. One reason for Croatia’s marginal position in GVCs is the absence of an automobile manufacturing sector. The transport vehicles, parts and equipment sector is one of the most dynamic product groups in world trade, partly because of the ease of production fragmentation across borders. The sector’s heavy reliance on just-in-time production and high weight-to-value ratio of auto parts and components motivates suppliers to locate closer to auto manufacturers, resulting in the creation of clusters. Croatia missed much of the foreign direct investment (FDI) from Western Europe Figure 17: Croatia’s FDI performance is low and Asia that entered neighboring countries in (FDI inflows, percent of GDP) Central Europe in the 1990s to build new plants or acquire old factories from the Socialist era.8 Nonetheless, Croatia’s emerging success in several industrial machinery and electronic products, where sourcing products across borders (or longer distances) is common, can provide avenues to enter new GVCs. The extent of Croatia’s participation in GVCs in the non- auto industries is unclear. The country’s exports of parts and components increased by a respectable 13 per cent per year in the past decade, although this is low compared to peer countries in the region such as Latvia Source: WB staff calculations using World Bank World (22.4 percent), Lithuania (17 percent), Poland Development Indicators. (18 percent), Serbia (24 percent) and the Slovak Republic (22 percent). 7 A country’s export of intermediate products can be represented in a Minimal Spanning Tree (MST). Each country’s export of intermediate products can be represented in a network, where each country is linked to each other. The link weights are transformed to reflect distances between nodes, larger bilateral trade flows are portrayed by closer distances between nodes, and the most connected countries represent the roots of the tree. 8 Example include Fiat’s purchase of Poland’s FSM in 1992; Volkswagen’s acquisition of Skoda in the Czech Republic in 1991; and Renault’s purchase of Dacia in Romania in 1998. 19 22. The contribution of FDI to Croatia’s trade performance is limited. The annual average ratio of FDI inflows to GDP since 2006 has been somewhat below the EU as a whole and well below that of regional peers (Figure 17). The investment promotion authority in Croatia (AIK) focuses mainly on attracting greenfield projects rather than promoting brownfield investments or mergers and acquisitions (M&As). However, M&As account for the large majority of global FDI activity; international experiences of M&As in the telecommunications and pharmaceuticals sectors show that such firms both continue to absorb new technology, engage in active R&D as part of the global networks of their parent firms, and have strong export performance. Financial sector developments 23. Underpinned by conservative prudential policies, the financial sector, dominated by banking, has remained stable throughout the long recession. Financial assets amounted to 172 percent of GDP in 2016, with the banking sector accounting for about 70 percent of the total financial sector assets. The remainder corresponds to pension funds (about 14 percent), insurance companies (6 percent), other financial intermediaries (4 percent), money market funds (2 percent) and other (4 percent). The banking sector is deep, well-capitalized, profitable and highly liquid. The banking sector was growing relative to the economy until 2011, but has declined since then due to the deleveraging that followed the financial crisis. Private credit equaled 65 percent of GDP in 2015, exceeding its Central European peer group average of 50 percent, albeit still below the EU average of almost 87 percent. 9 The average capital adequacy ratio (CAR) for the banking sector was 22 percent in 2016, one of the highest among CE countries. Stress testing for the banking sector from 2014 to 2015, conducted by the Croatian National Bank (CNB), indicates that the banks’ capital buffers would have been sufficient even in a severe stress scenario. The banking sector was generally profitable in the period 2008-2016, with the exception of 2015 when performance was adversely affected by the legislated mandatory conversion of CHF- denominated loans. The sector recovered quickly, and in 2016 return on assets had risen to 1.6 percent. The ratio of liquid assets to deposits and short-term funding amounted to 21 percent, in line with the peer group average. 24. Private sector debt, at corporate and household levels, and exposure to foreign currency risks remain high. The housing boom and expectations of rapid convergence to EU average income levels fueled the surge of households’ debt prior to the financial crisis. Deleveraging has been relatively slow, hindered by unfavorable labor market conditions and shrinking disposable incomes. The conversion of Swiss franc loans in late 2015 helped accelerate households’ debt deleveraging. With improved labor market conditions, stabilizing real estate prices, and interest rates at historic lows, pressures on further reducing household debt have eased, even though exposure to interest and currency risks remain high. For the corporate sector, pre-crisis high investment rates led to a rapid accumulation of debt. The low level of investment since has facilitated a decline in the debt to GDP ratio by some 10 percentage points. With about 77 percent of corporate debt denominated or indexed in foreign currency, exposure to currency risks remain high, especially for companies in the non-tradable sector. As of 2016, transaction data point to a recovery of bank placements to both households and corporates. The composition of the credit portfolio has also significantly shifted since before the crisis, with credit to government and the SOE sector increasing from 9 percent of the total to 20 percent in 2016. Total credit to Government and State-owned enterprises as a percentage of GDP amounted to about 30 percent in 2015, significantly higher than the average of peer countries (18 percent) and the EU average (16 percent). 25. The bulk of banking sector loans remains denominated in foreign currency. Domestic currency loans picked up in 2016, supported by the negative experience with Swiss franc-denominated loans, 9 CE peer group countries: Bulgaria, Czech Republic, Hungary, Poland, Romania, Slovak Republic, Slovenia. 20 decreasing interest rates on kuna loans, and National Bank measures. Nevertheless, foreign denominated currency loans, mainly in euros, still represented 61 percent of banking sector loans as of December 2017. These loans represent a significant source of vulnerability, as most borrowers are unhedged (e.g. about 73 percent of housing loans are either denominated in or linked to a foreign currency). As of December 31, 2017, foreign currency deposits amounted to 60 percent of total deposits in the system. The exchange rate risk is mitigated by the National Bank’s exchange rate policy of stable EUR/HRK exchange rate. 26. Non-performing loans remain above Figure 18: Croatia’s banking sector non-performing pre-crisis levels and high relative to the EU loans are high average, and continue to pose a risk to financial stability. Non-performing loans (NPLs) have been on a declining trend, with the ratio of NPLs to total loans at 13.8 percent at the end of 2016. However, the NPLs ratio is still significantly above the pre-crisis level and high compared to other EU member states (Figure 18). The level of NPLs is significantly higher for the corporate sector, where NPLs have been hovering at some 30 percent of total loans for the past four years. In stress tests conducted by the Croatian National Bank, the NPLs ratio is expected to reach between 15.6 and 20 percent Source: World Development indicators. of total loans, depending on the macroeconomic scenario and assumptions on Agrokor group’s restructuring process. resolution still requires a coordinated approach by different authorities, including an improvement of the legal, regulatory, and tax regime for NPLs sales, transfers, bankruptcy and restructuring. B. Productivity patterns Aggregate productivity 27. Productivity has made little, or a negative, contribution to growth in Croatia over the past 13 years. According to a growth accounting exercise, total factor productivity made a negative contribution to output growth during the period before the global crisis (2002 to 2008). By contrast, capital accumulation accounted for 5 percentage points of output growth during this period and labor about one percentage point. Falling TFP accounted for 3.7 percentage points of the fall in output during the 2009-14 recession, while the contribution of capital to growth fell as foreign capital inflows dried up, but remained positive. Finally, during the 2015-16 recovery, capital accumulation was again the principal source of growth, while the contribution of both TFP and labor turned marginally positive. (Figure 19). 28. The negative contribution of TFP to growth contrasts sharply with Croatia’s Eastern European peers, where growth in the years preceding and after EU accession was mostly based on productivity gains. TFP in Croatia may have fallen because capital accumulation occurred mainly in consumption- related and inward-oriented sectors rather than in the tradable sectors, where investments are generally associated with higher productivity gains. The low contribution of labor is explained by unfavorable demographics which drove slow growth of the working age population, as well as low levels of labor market participation. The dismal TFP performance has been reflected in the slow growth in potential 21 output, which has consistently been at the lower range of potential output among Central Eastern Europe (CEE) peers since 2004 (Figure 20).10 Figure 19. Productivity made a negative contribution Figure 20. Potential output in Croatia is below CEE to growth peers (% of GDP, 2002-16) (% of GDP, 2002-16) Source: CBS, WB staff calculations. Source: European Commission (2017). Firm-level productivity analysis 29. The median TFP level of Croatian firms fell over the period 2008-2013. Three factors contributed to this decline, including (on the demand side) deflation due to protracted weak global demand, and (on the supply side) efficiency losses at the firm level due to lack of technological upgrading and balance sheet vulnerabilities, which encouraged firms to engage in low risk-low return investments. Several other EU countries, including the new EU member states (Bulgaria, Czech Republic, Poland, Romania and Slovenia, but not Slovakia) also experienced a decline in firms’ median TFP levels.11 Data limitation does not allow us to identify which side – prices versus marginal costs – has made a higher contribution to explaining the observed U-shaped pattern for most of the sectors.12 30. Firms’ median TFP levels declined in all sectors after the GFC and only began to rise in 2012.13 TFP levels remain lower in Croatia than in Central and Eastern European (CEE) peers in most industries. Averaging across all industries, the median TFP level of Croatian firms was estimated at about 10 percent lower than firms in CEE peers in 2014. 14 The gap was highest in manufacturing (almost 20 percent), mining and quarrying, and construction. The only industry in which Croatian firms’ productivity was above the peer average was energy. Median TFP levels in Croatia, disaggregated by industry, only began to rise in 2012. 31. Most of the reduction in TFP at the sectoral level reflects lower productivity at incumbent firms rather than misallocation of factors between firms. The TFP growth between 2009-2015 of each sector can be decomposed into two margins: (i) the “within component”, which relates to changes in productivity by incumbent firms; and (ii) the “between” component, reflecting the reallocation of factors of production and economic activity toward more efficient firms (capturing both entry/exit dynamics and 10 Central Eastern European (CEE) peers used by European Commission in its analysis include Czech Republic, Hungary, Poland and Slovakia. 11 Adler (2017). The global decline in TFP following the crisis is well documented (Syverson, 2017). 12 Lack of information on firm-level product prices means that our measure of TFP does not exclusively reflect efficiency. Thus, the TFP measure used in the analysis reflects firms’ profitability, and is usually known as “revenue total factor productivity”. 13 The TFP measure includes confounded factors both from the demand- and the supply-side. 14 European Commission (2017). 22 the reallocation of activity across incumbent firms).15 The TFP reductions observed in many sectors are dominated by the “within” component, which was negative in most sectors over the perio d 2009-15 (Figure 21). This is consistent with evidence from the Enterprise Survey showing that the observed reduction in TFP dispersion in Croatia is mainly driven by a slowdown in the TFP performance of the most productive firms.16 The negative contribution of the “within” component is often associated with several factors, including: (i) limited investment due to increased policy uncertainty, (ii) a sharp tightening of credit conditions; (iii) low levels of investments in intangible assets such as R&D,17 (iv) low levels of product market competition, which impair firms’ incentives to become more competitive, and (iv) a cumbersome business environment that prevents more productive firms from growing faster than competitors. The large negative “within” contribution to TFP in the primary sector (crop and animal production) reflects outdated farming methods and low capital intensity in agriculture in Croatia. The lack of business dynamism hinders the process of industrial renewal and Croatia’s productive structure is aging - the average firm age increased from 12.5 years in 2009 to about 14.5 in 2015. Figure 21: Changes in TFP levels are driven mainly by the “within” component (2009-2015) Note: The decomposition of the TFP growth between 2009 and 2015 was performed using the methodology discussed in Melitz and Polanec (2011). The TFP growth is decomposed into the “within” and “between” components, as well as components related to the entry and exit of firms. Source: WB staff calculations using Croatian firm census. 32. There are some signs of improved resource allocation across Croatian firms. The dispersion of TFP levels across firms fell over the period 2008-2013, which may reflect improved resource allocation.18 However, given the available data, this decline may also be driven by lower volatility of sales coupled with reduced adjustments costs in capital.19 Improvements in allocative efficiency mainly reflect a better allocation of labor towards the most productive firms, which is captured by a reduction in the dispersion 15 Melitz, M. & S. Polanec (2015). 16 Correa, P., Cusolito, A., and P., Jorge (2017). 17 Adler (2017). 18 This interpretation requires restrictive assumptions about product demand and marginal costs, including monopolistic competition and constant marginal costs (Hsieh, C., and P., Klenow (2009)). 19 Collard-Wexler, A., and J., De Loecker (2014). 23 of the marginal product of labor (MPL) (Figure 23).20 However, there are also signs of increased misallocation of capital, as the dispersion of the marginal product of capital (MPK) has increased since 2009. Dispersion of MPL and MPK may reflect the presence of economic distortions or heterogeneity in policy treatment at the firm level, which causes misallocation of factors of production. For example, financial frictions, such as constraints on the availability of information that leads banks to lend based on collateral rather than expected profitability, may have resulted in the misallocation of capital inflows towards firms with higher net worth that are not necessarily more productive.21 These findings are in line with recent studies of other countries in the EU region.22 Figure 22. TFP decelerated in all sectors after the GFC Figure 23: The misallocation of capital rose Croatia TFP Median by 4 sectors Dispersion of Marginal product of labor/capital, Manufacturing 10.2 1.6 10 1.4 9.8 Median TFP 1.2 9.6 1 9.4 .8 9.2 2009 2010 2011 2012 2013 2014 2015 .6 YEAR 2008 2009 2010 2011 2012 2013 1.Agriculture 2.Mining year 3.Manufacturing 4.Services dispersion(MRPL) dispersion(MRPK) Source: WB staff calculations using Croatian firm census. Source: WB staff calculations using Orbis data. C. Key drivers for boosting output potential and productivity growth 33. Croatia needs to boost its output growth potential to resume and sustain economic and social convergence towards EU living standards. Boosting Croatia’s growth will require improving resource allocation and supporting firm-level productivity improvements, while increasing the supply of labor and capital. Borrowing from Schumpeterian endogenous growth theory, this can be achieved by a transition to an innovation-led growth model, involving the presence of competitive markets for both products and factors, availability of higher education and research, and capital market-based finance. In addition, government must maintain a sound macroeconomic environment while supporting strategic interventions through its public investment and research and development policy.23 The remainder of this chapter discusses key elements to support this transition, organized around the business environment, the competition environment, the footprint of the state in the real sector, the innovation ecosystem, and higher education and skills. Complementary elements relating to sustainability of public finances and efficiency of public sector are discussed in Chapter 4. 20 Cusolito, A. and S., Tan (2017). 21 Gopinath, G., Kalemli-Ozcan, S., Karabarbounis, L. and C., Villegas-Sanchez (2017). 22 Gamberoli, E., Giordano, C., and P. Lopez-Garcia (2016). 23 Aghion, P. and U. Akcigit (2015). 24 Improving the business environment 34. Despite progress in recent years, companies continue to face a cumbersome business environment that inhibits private sector investment and distorts resource allocation. Croatia continues to lag best performers in key Doing Business indicators, including access to credit, resolving Figure 24: Croatia‘s institutions are far weaker than insolvency, and obtaining construction permits, top performers which show the largest distance to the frontier. (Global Competitiveness Index, 2017-2018) While Croatia’s performance in obtaining construction permits and resolving insolvency has improved in the last five years, its performance on access to credit has deteriorated. The poor performance in these indicators are reflected in Croatia’s low scores in the institution indicator in the Global Competitiveness Report of the World Economic Forum, which captures the state of property rights, ethics and corruption, undue influence, public sector performance, security, and business ethics (Figure 24). Croatia also lags top Source: World Economic Forum, Global Competitiveness performers in the Global Competitiveness Report Report, 2017-2018. in the market size and innovation indicators. 35. Business environment dimensions have a diverse impact on firms’ productivity. Estimates of the marginal effect of policy variables on revenue total factor productivity (TFPR) show large dispersion, with larger negative impact for red tape-related variables (frequency of tax inspection, delays in customs, number of days to obtain permits) and for the intensity of competition (Figure 25).24 Variables with a large positive impact on firms’ productivity include provision of training to employees, access to credit, being an exporter, and presence of private ownership. The estimated negative impact of competition on productivity can be explained by the fact that competitive pressures affect firm’s productivity through two different mechanisms, a negative price-effect, and an ambiguous efficiency-effect that depends Figure 25: Business environment has a direct on the composition of the universe of firms, as impact on firm productivity competitive pressures encourage leading firms to (Marginal impact on TFPR for median firm, 2013) upgrade their internal capabilities to become more efficient and escape competition (positive effect), but they discourage laggard firms to engage in efficiency-enhancing investments because lower profits do not allow them to cover the fixed costs of innovating, adopting new technologies and better managerial practices (negative effect). From a policy perspective, this finding means that in the absence of government support like state-aid geared to help laggard firms, competition will reduce prices, force the exit of laggard firms, Source: Correa and al. (2017). 24WB staff calculations using methodology developed by De Loecker and applied to World Bank Enterprise Survey data (cf. Correa P., Cusolito, A., and P., Jorge (2017)). 25 encourage the entrance of productive firms, and induce the reallocation of factors of production towards the most productive firms. 36. Weak performance of the public sector increases inefficiencies and administrative burdens facing firms. Firms are directly affected by weak legislative planning, which reduces the quality of new regulations. In some cases, legislative quality may have been impaired by the urgent legislative procedures adopted (about 80 percent of all “harmonized” laws had only one parliamentary reading) to accomplish the huge changes required by EU accession, involving the harmonization of about a quarter of Croatia’s laws, the creation of implementing institutions, and increasing financial resources and staff skills.25 Although Croatia implemented a regulatory impact assessment (RIA) which can improve the legislative process, most legislation does not use it due to these fast track legislative processes. Even when performed, the RIA is often not properly implemented, due to weak quality control and Figure 26: Croatia scores poorly on the quality of lack of high level commitment. Poor legislative institutions (2016-2017) quality, in turn, leads to frequent amendments to address shortcomings and complications for SMEs. The red tape and weaknesses in public administration affect businesses mainly in the form of inconsistencies in local-level decision making and lack of strict timeframes for issuing opinions on tax issues. In its action plan for 2017 the Government has identified a set of 104 measures to improve the business environment, including steps to reduce redundant administrative costs and to Note: Scores for each category is from 1 to 7 (best) and the top performers are the top score for each category. Source: World improve competition in the professional Economic Forum, Global Competitiveness Survey, 2016. services market, which are expected to save enterprises up to 1.5 billion kuna. 37. Croatia’s public sector governance indicators are worse than the vast majority of countries in the world. Limited transparency, corruption and undue influence remain a concern, as Croatia persistently scores below the average in the Global Competitiveness Index for public institutions (Figure 26). Of the 137 countries ranked by the 2017 Global Competitiveness Indicators, Croatia is ranked 135th in challenging regulations, settling disputes, as well as burden of government regulation, 128th in the transparency of policy making , 122nd in the favoritism in decisions of government officials, and 114th in judicial independence.26 Similarly, Worldwide Governance Indicators rates Croatia the worst in the EU for regulatory quality and among the worst in other key public sector governance indicators. 25 This accelerated process stands in contrast with a more gradual approach in neighboring Slovenia, which resulted in better implementation of legislation. 26 The Global Competitiveness Report 2016-2017, World Economic Forum, Geneva, 2016. 26 Figure 27. Croatia’s governance is among the worst in the EU (Percentile rank, 2016) Source: Worldwide Governance Indicators. 38. Inefficiency, unpredictability, and delays in court processing cases continue to be among the greatest impediments to business in Croatia. Despite progress, the system struggles with a legacy of bureaucracy and red tape. Courts of general jurisdiction and commercial courts are generally perceived as too slow in processing cases. The number of pending cases and disposition time in litigious civil and commercial cases remain among the highest in the EU. For example, proceedings of commercial cases in first instance took on average 724 days (in 2015)27, and more than 300,000 cases are officially backlogged (meaning that they have been pending for more than three years). Almost 64,000 cases in Municipal Courts are officially backlogged, representing approximately 25 percent of all Municipal Court cases. In commercial courts, approximately 15 percent of cases are backlogged. Unsurprisingly therefore, the Global Competitiveness Index 2017 of the World Economic Forum ranks Croatia 135th out of 137 economies in terms of the efficiency of the legal framework in settling disputes, a ranking that continues to slide as other economies reform to become more competitive. According to the Doing Business Report 2018, contract enforcement takes 650 days, which is slower than the OECD and Europe and Central Asia averages. Resolving insolvency in Croatia takes up to 37 months, much longer than the ECA average of 27 months. Similarly, the 2016 the European Commission for the Efficiency of Justice (CEPEJ) report finds that the disposition time of civil and commercial litigious cases in 2014 was 380 days, which is significantly higher than the European average of 237 days. In the 2014 World Bank Business Environment and Enterprise Performance Survey (BEEPS), a mere 15 percent of firms report that the court system is quick, less than half the ECA average of 31 percent. The extensive demands on judicial services are compounded by the sector’s provision of non-litigious administrative services (e.g. company registration, land ownership registration, etc.) and the lack of effective fast-track procedures for resolving minor disputes. 27 European Commission (2017). 27 Simplifying and streamlining procedures for minor disputes could have positive effects on the economy by improving access to justice for smaller firms and helping courts to diminish backlogs by resolving cases more easily.28 39. Transparency, corruption, and uneven quality of justice across Croatia remain a concern and undermine users trust in the system. In the 2014 BEEPS Survey, only 36 percent of firms in Croatia report that their court system is fair, impartial and uncorrupt. This lags the EU averages, and this view has remained unchanged for at least the last 5 years. Similarly, the WEF ranks Croatia 122nd out of 140 economies in terms of undue influence and 114th in terms of judicial independence. Excessive variation in practices across the judicial administration is also a problem in both courts and State Attorneys Offices (SAOs). Workloads are distributed unevenly between courts and between judges of the same court. The variations impair efficiency and predictability and undermine trust in the system. Clearer and stricter rules on ethical standards are needed, and should be accompanied by training and monitoring to ensure that they are mainstreamed into daily practice, and with clear sanctioning and control in courts and prosecution offices. Court presidents and heads of State Attorney Offices (SAOs), under the leadership of the Councils, have a clear role to improve the operational efficiency, but also improve the system’s transparency. 40. This low efficiency and poor quality in the justice system persist despite an over-abundance of judges and staff permanently employed in courts around the country. According to the 2016 CEPEJ report, Croatia has 41 professional judges per 100,000 inhabitants, almost double the European average of 21 judges. In addition to the high number of judges, Croatia’s judiciary employs 166.5 non-judge staff per 100,000 inhabitants, more than double the European average of 70. The average ratio of staff to judges in Croatia is 4:1, more than 20 percent higher than the CEPEJ average. Their cumulative effect is to create a bloated wage bill that crowds out other expenditures, leaving little room for much-needed investments in innovation, ICT and training. The bloated size of the sector relative to other European states and the sub-optimal performance suggest that there are serious inefficiencies in the system. Efficiency gains can be achieved though improvements in processes and performance measures at each level of the system and a reduction of the scope of judicial competences in areas that are administrative in nature. Such reforms could both improve performance and ease budgetary pressures. 41. The private sector also appears to face greater difficulty in obtaining finance than in European peers. About 15 percent of SMEs identified access to finance as the most pressing constraint, compared to 9 percent on average for the EU. Similarly, 26 percent of Croatian SMEs did not manage to get the full bank loan financing they had planned, higher than the EU average (18 percent), and only 15 percent of SMEs use trade finance compared to 35 percent in the EU.29 According to the Global Entrepreneurship Monitor survey of 2014, lack of access to finance was the second largest cause for cessation of business activity, responsible for 23 percent of the cessation cases (twice the average level for the EU). Croatian firms also rely more on retained earnings for financing their operations (37 percent of firms compared to 27 percent on average for the EU), as well as grants and subsidized loans (40 percent compared to 32 percent in the EU). 42. There are numerous government programs in support of access to finance, albeit their impact and efficiency remain largely unclear. These programs, which amount to some 2 percent of the total budget in 2017, appear to be poorly coordinated and to have substantial overlaps, including with respect to target segments and financial products offered.30 Most also lack regular and systemic monitoring and 28 See Fast-tracking the Resolution of Minor Disputes: Experience from EU Member States, World Bank, 2017. 29 European Central Bank, Access to Finance of Enterprises (SAFE) survey (https://www.ecb.europa.eu/stats/ecb_surveys/safe). 30 There are over 40 government programs which support access to finance, mostly for SMEs, and which are administered through 7 ministries (Ministry of Finance, Ministry of Economy Entrepreneurship and Crafts, Ministry of Tourism, Ministry of 28 evaluation mechanisms. Given the scale of existing programs and the availability of vast EU funds, improving coordination and monitoring mechanisms is critical to enable the private sector to fully benefit from such programs. In the latest SAFE survey, SMEs mentioned that improving access to public funds would be the most critical factor in their companies’ financing in the future. 43. Equity markets are constrained by small market size, limited local institutional investor appetite, limited new capital market issuances and low free float of listed companies. The corporate bond market remains underutilized and is limited to a few large issuers. As of the first half of 2017 there were 13 corporate bonds listed on the Zagreb Stock Exchange (ZSE), with total amount outstanding (market capitalization) amounting to about 26 percent of GDP and turnover equal to only 0.1 percent of GDP. The European Commission’s Capital Markets Union framework provides a good anchor to implement institutional reforms and upgrade infrastructure for capital market development. 44. Access to start-up financing is key to encouraging innovative firms and achieving levels of business R&D to GDP closer to that of Croatia’s peers. Innovative firms tend to be underfinanced due to information asymmetries and appropriation risks. This affects in particular new entrants, which lack a track record of performance and collateral. The limited availability of risk capital financing in Croatia is exacerbated by a cumbersome insolvency framework which discourages risk taking, as well as an incomplete regulatory framework for the venture capital industry. 45. Despite large infrastructure investments, Croatia’s logistic performance remains low. Croatia has one on the highest density motorway networks in Central and Eastern Europe. This network was developed during the 2000s in response to the need for a visible integration of Croatian territory after independence, efforts to spur industrial growth and tourism, and the desire to integrate into the broader European network. While the bulk of public infrastructure was devoted to roads, significant investments were also launched for the main passenger and commercial ports, and the Zagreb international airport was concessioned out to a private consortium. Croatia’s geographical location at the crossing of major European corridors makes it well positioned to become a logistics hub for Central and South-Eastern Europe. Rijeka port has evolved from a feeder port for small vessels to a port of call for containers, resulting in a tenfold increase in container throughput between 2002 and 2015. Yet Croatia’s logistics performance remains lackluster – ranked 51 out of 160 countries and underperforming all EU peers in the World Bank Logistics Performance Index – largely due to underdeveloped infrastructure such as port- rail interfaces, slow and unreliable rail operations, and cooperation and communication failures among stakeholders. 46. While Croatia has relatively good broadband internet connectivity, the price and affordability of broadband internet lags other EU countries. The fixed broadband coverage and take-up in Croatia approaches the EU-28 average and is above some peer countries, such as Romania, Poland, and Bulgaria. However, the speed of broadband internet in Croatia is one of the lowest in the EU, above only Italy and Greece, and the price of broadband internet in Croatia is one of the highest in the EU. Slow and expensive broadband internet can limit exports of ICT, financial and professional services, and impair firms’ ability to integrate the internet into business operations to increase operational efficiency. As a result, the integration of digital technology into businesses in Croatia is below the EU-28 average and the use of digital technology in government services is among the lowest in the EU. Regional Development and EU Funds, Ministry of Labor and pension System, Ministry of Demography, Family, Youth and Social Policy, Ministry of Science and Education) and three agencies (HBOR, HAMAG-BICRO, Croatian Employment Bureau). 29 Strengthening the competition environment 47. Croatia has a solid competition legal framework, but a poor competition environment. Croatia adopted a competition law in 2009 with many components in line with EU practices, and an independent Competition Agency is in place. Nevertheless, Croatia scores below the ECA average on the extent of market dominance and the effectiveness of anti-monopoly policy, and only marginally above the average for intensity of local competition, in the perception-based Global Competitiveness Report rankings (Figure 28). An OECD product market regulation analysis shows that Croatia has the most restrictive Figure 28: Croatia has low competition scores compared to other ECA countries (2016-17) Note: Scores are normalized between 1 and 7 (best). Source: Global Competitiveness Report 2016-2017, World Economic Forum. economy compared to its peer countries (Figure 29). The largest contribution to Croatia’s product market restrictions is state control, then barriers to entrepreneurship, and, to a lesser extent, barriers to trade and investment. Exposure to competition is critical to create incentives for incumbent firms to adopt new technologies and become more productive, and increase the exit of less productive firms. 31 32 Using the Hirschman-Herfindahl index of concentration as an indication of competition, sectors with high levels Figure 29: Croatia has more restrictive product of competition (lower HHI index) have higher TFP market regulations than peers median and lower TFP dispersion (Figure 30). (Economy-wide PMR score in 2013) 48. Regulatory restrictions appear particularly burdensome in the services sector, notably in network economies and professional services. Prices are regulated for professional services, international wholesale roaming rates, and local loop unbundling. Advertising and marketing are prohibited for the legal, engineering and architecture professions. However, the regulations on some service sectors have eased since the PMR scores were collected in 2013, due to the changes to laws regarding the conduct of engineers and architects between 2013 and 2016. Planned changes to laws regarding conduct of architects, Note: OECD top 5 countries are the Netherlands, the United Kingdom, Austria, Denmark and New Zealand. lawyers and auditors in 2017-2019 may further Source: WBG/OECD PMR data 2013-2014. 31 Aghion, P., R, Blundell, R. Griffith, P. Howitt, and S. Prantl (2009). 32 Holmes, T. and J. Schmitz (2010). 30 reduce the burden of regulations in these sectors. A reduction in the stringency of services regulations would have a sizable impact on productivity. For instance, a reduction of overall restrictions in the services sector in Croatia to the average level of the three most deregulated European economies would increase the level of firm productivity by over 5 percent.33 Figure 30: Sectors with more competition have a higher TFP median and lower TFP dispersion All firms TFP Median by sector groups All firms TFP Dispersion by sector groups 10.2 3.5 TFP 90-10 pct dispersion 10.1 Median TFP 10 3 9.9 2.5 9.8 2009 2010 2011 2012 2013 2014 2015 2009 2010 2011 2012 2013 2014 2015 YEAR YEAR high HHI sectors low HHI sectors high HHI sectors low HHI sectors Note: HHI refers to the Hirschman-Herfindalh index, where a higher value represents more concentration in the sector. A sector is defined according to the NACE classification and a high HHI sector has a HHI above the median for each year. Source: WB staff calculations using Orbis data. Reducing State’s footprint in the economy 49. State-owned enterprises (SOEs) play an important role in the Croatian economy. The central government is a majority owner of 74 companies – 39 are classified as “special state interest” or “strategic” – and it is a minority owner in 381 additional companies (with 1 being of a special state interest). Nearly 700 companies report to sub-national, regional and municipal authorities. These firms operate in 20 sectors, including sea, rail, road and air transport, hotels and restaurant, food processing, pharmaceuticals, financial services and services of motor vehicles. SOEs hire over 10 percent of the labor force, contribute a fifth of total turnover and possess a Figure 31: The productivity gap between state third of total assets, with the highest share in the utilities and private firms is large but narrowing and transportation sectors. (2009-2015) 50. SOEs impose a large burden on the economy. Croatia TFP Median by ownership 10.2 SOEs contribute directly to government deficits, with a net average borrowing of 0.6 percent of GDP between 10.1 2011-2014. Resources that are diverted to maintain the Median TFP 10 survival of SOEs may be better applied towards more 9.9 efficient uses. SOEs affect factor returns, influence output prices by limiting product market competition 9.8 and reduce private firms’ incentives to become 9.7 competitive. Private firms in sectors with low SOE 2009 2010 2011 2012 YEAR 2013 2014 2015 presence have higher TFP levels (better performance) State Private and lower TFP dispersion (lower misallocation of resources) than those in sectors with high SOE presence Source: WB staff calculations using Croatian firm (Figure 31). SOEs are less productive than private firms, census. albeit the productivity gap has been narrowing since 33 Van Der Marel, Kren and Iootty (2016). 31 2013 (Figure 31). The non-financial SOEs are more indebted than private companies, with a 40 percent higher leverage ratio than private firms. Yet, SOEs are less profitable: in 2014, the average return on equity was 4.5 percent for private companies, but almost zero for SOEs. And SOEs’ rates of return on assets are below that of SOEs in other CEE countries, resulting in a higher than EU average rate of state subsidies. 51. SOEs and the public sector at large substantially influence wage determination in the private sector. The average SOE is highly unionized and able to influence labor conditions, given the uncoordinated and decentralized wage bargaining system. The wage-setting mechanisms in SOEs are rigid and do not ensure alignment of workers’ wages with enterprises’ economic fundamentals. Wages in SOEs exceed those of private sector firms, controlling for employee characteristics, by an average of about 7 percent.34 Rigid collective agreements and political vested interests make it difficult to downsize and rationalize operations. Close to 40 percent of total employees (excluding self-employed) work in public administration, public services or SOEs.35 Empirical evidence indicates that public sector wages in Croatia have a larger impact on private sector wage setting than in most EU10 countries, and appear to lead long-run wage dynamics. Figure 32: Sectors with low State presence have 52. The state-owned Croatian Bank for higher productivity and allocative efficiency Reconstruction and Development (HBOR) is equivalent in assets to the sixth-largest commercial Non-state firm TFP Median by sector groups 10.2 bank in the country. HBOR focuses on infrastructure, exports, and SME financing. It operates as a second- 10.1 Median TFP tier financial institution and also as a direct lender (47 10 percent of its total gross loan portfolio as of September 2016). While HBOR appears to follow commercial 9.9 banks practices, it is not subject to the same regulatory 9.8 and governance structure as other banks, and does not 2009 2010 2011 2012 2013 2014 2015 YEAR appear to meet the highest standards of transparency, high state-firm emp. share low state-firm emp. share accountability, and independently exercised supervision. The board is primarily composed of Ministers and members of Parliament. The authorities Non-state firm TFP Dispersion by sector groups 3.2 conducted an asset quality review, and may consider changes in HBOR’s regulatory and governance TFP 90-10 pct dispersion 3 structure. 2.8 53. Croatia has recently undertaken measures to 2.6 reform the management and corporate governance of its SOEs. The prolonged crisis and weak recovery has 2.4 exposed major corporate governance weaknesses, 2009 2010 2011 2012 YEAR 2013 2014 2015 including with respect to their planning and high state-firm emp. share low state-firm emp. share operational capabilities, and their accountability and governance frameworks. Corporate governance Note: the presence of state firms is weighted by the mechanisms appear particularly weak for transport employment shares of the State firms. Sector Codes correspond to NACE classification. A sector has a high sector SOEs and enterprises owned by local state representation if the share of state firms is above authorities. In 2013, Croatia adopted the Act on the median for each year. Source: WB staff calculations Management and Disposal of State Assets, which using Croatian firm census. 34 Nestic et al. (2014). 35 Ibid. 32 created two agencies dealing with the management of SOEs: the Ministry for State Property Management (formerly the State Office for State Assets Management, DUUDI) manages the companies of special state and strategic interest; and the Restructuring and Sale Center (CERP) manages the minority-owned companies. Also, there are plans to establish registers of centrally-owned SOEs and of managerial appointments. 54. Some progress in the privatization of public assets has been made. Between 2009 and 2015, 257 state firms were converted to private firms. Sectors with the largest number of firms included wholesale trade (36 firms), accommodation and hotels sector (16 firms), and architectural and engineering activities sector (14 firms). However, estimates of the performance of the firms privatized during that period show no significant improvements, notably with respect to productivity, revenues, and average wages. These results may reflect the time required for productivity gains to materialize, and call for caution in projecting the short-term benefits of privatization. 55. There is still a lack of regulatory neutrality with regards to the procurement rules for SOEs. The Public Procurement Act, which took effect on January 1, 2017, sets a high standard for public administration. However, SOEs and legal persons other than the public authorities, which accounted for around half of the total value of public procurement for 2008-13, are subject to weaker control mechanisms under this framework. 56. The total amount of state aid allocated by Croatian authorities has declined. The 2005 State Aid Act and related bylaws were adopted as preparation for EU accession. The unwinding of support for and restructuring of the shipbuilding industry, carried out as a condition for Croatia’s accession to the EU, is the main reason for the decline in state aid since the early 2000s. However, the railways sector has received increasing support in recent years. The use of state aid in specific sectors should be reviewed, with the goal of minimizing distortions to competition. Enhancing the innovation ecosystem 57. Croatia’s overall innovation performance is worse than most of its European peers and has been deteriorating in recent years. The European Innovation Scoreboard ranked Croatia 32nd out of 36 countries in 2016, surpassing only Bulgaria, FYR Macedonia, Romania and Ukraine. 36 And key sub- indicators, for example those related to measures of innovation in SMEs and linkages between stakeholders in the innovation ecosystem, have deteriorated since 2010. Indirect innovation outcome indicators, such as scientific production, new trademarks, industrial designs and patents also remain lackluster. Similarly, Croatia was ranked 103rd out of 138 countries for the innovation pillar of the 2016-2017 Global Competitiveness Index. Both rankings have declined in recent years. Whereas the overall number of Croatian publications rose by more than half over the last decade, the percentage of highly-cited publications in 2015 is below the EU average and that of most peer countries (Figure 33). 37 Croatia has among the lowest levels of patenting intensity, with approximately 3.43 patent applications to the European Patent Office per million inhabitants, compared to an EU average of 111.97.38 Similarly, Croatia underperforms peers with respect to triadic patent families.39 Croatia is also falling behind its peers with respect to the level of gross domestic expenditure on R&D (GERD) and has made no progress towards 36 European Innovation Scoreboard 2017, http://ec.europa.eu/growth/industry/innovation/facts-figures/scoreboards_hr. The 36 countries included are EU28, Norway, Iceland, Turkey, Serbia, FYR Macedonia, Switzerland, Israel, Ukraine. 37 According to Scopus, the number of citable documents rose from 3722 in 2006 to 5772 in 2016. http://www.scimagojr.com/countrysearch.php?country=hr 38 Eurostat data, 2014. 39 A triadic patent family is defined as a set of patents registered in various countries (i.e. patent offices) to protect the same invention. Triadic patent families are a set of patents filed at three of these major patent offices: the European Patent Office (EPO), the Japan Patent Office (JPO) and the United States Patent and Trademark Office (USPTO). 33 reaching its GERD target for 2020, while most of its peers have not only progressed but have set up more ambitious targets. Bulgaria, Slovakia, Hungary or Poland, which were trailing or at par with Croatia’s level of GERD in 2008, have now surpassed Croatia (Figure 33). Figure 33. Innovation indicators are low in Croatia (2015) Source: European Innovation Scoreboard 2017. 58. Croatian enterprises are not sufficiently involved in R&D. The share of Croatian enterprises engaged in innovative activities is below the EU average and they tend to favor non-R&D versus R&D innovation activities. Non-R&D innovation expenditures equaled about 1.2 percent of turnover, compared to 0.76 percent for the EU on average. By contrast, business sector expenditures on R&D Figure 34. GERD is low and stagnant in Croatia Figure 35. Business expenditures on R&D are (Gross expenditure on R&D, 2008-15) concentrated in large companies (2014) Source: Eurostat. Source: Eurostat, Community Innovation Survey. (BERD) was estimated at 0.44 percent of GDP, less than half the EU average and lower than most peers. In line with peers, BERD (Business Expenditure on R&D) is also concentrated in large companies (mainly pharmaceutical, telecommunications, agricultural, and food and beverage industries). However, Croatia stands out for the small share of R&D expenditures by small companies (Figure 35). Key factors explaining the low level of R&D-driven innovation, especially among medium and small firms, include limited access to internal and external resources (both funds and qualified personnel), limited information on technology and markets, the concentration of tax incentives with large firms, modest research 34 excellence, barriers to science-industry collaboration, as well as pervasive weaknesses in the innovation ecosystem governance.40 59. The lack of public-private cooperation is a core weakness of the innovation system. Croatia has only 5.7 public-private co-publications per million population, compared to the EU average of 28.7, and is well behind EU average and most regional peers (Figure 36). The lack of adequate linkages between research institutions and the business sector has been identified as a key area requiring improvement in the Croatian R&D sector.41 The public research sector is characterized by a high fragmentation, frequent institutional changes, weak coordination, and rigid funding mechanisms, which hinder excellence and the creation of linkages with the business sector. The crisis and the resulting fall in expenditures had little impact on the structure of the sector, although the number of researchers dropped by some 15 Figure 36. Public-private co-publications are limited percent between 2010 and 2014. A review of the in Croatia 2014 data for the state budget expenditures by (2015) the Ministry of Science and Education (MoSE) reveals that over 90 percent of the public funds for higher education were destined to salaries and operating costs, leaving little room for project financing for research or innovation activities or performance-based supplemental funding options. The resources are allocated mainly based on number of staff, independently of performance, and institutions have limited flexibility in using the funds. The multitude of institutions coupled with funding rigidities hamper coordination efforts and limit the Source: European Innovation Scoreboard 2017. opportunities for linkages with industry.42 60. The weak innovation ecosystem is likely one reason for the low level and limited improvement of the complexity of Croatia’s merchandise export basket. Croatia’s Economic Complexity Index (ECI) stagnated over the past decade and still stands below peers like Slovenia, Hungary, Slovak Republic and Poland (Figure 37). For Croatia’s four largest sectors at the SITC1 level, the quality of Croatia’s exports generally exceeds that of most comparators and has improved significantly for both machinery and transport equipment and miscellaneous manufactures. Similarly, of the nearly 200 products where Croatia has a revealed comparative advantage (RCA), about half are clustered around the middle deciles of sophistication. This suggests that the low aggregate level of complexity reflects a few sophisticated products co-existing with many other exports which rank at a low to medium level. In addition, while services play a large role in Croatia’s exports, the sophistication of services exports remains well below that of peers’ services exports. 40World Bank (2015d). 41Rio Country Report 2015, JRC Science for Policy Report, EC. 42 See for instance “Strategy for Education, Science and Technology” Narodne novine (2014): 124/2014. www.novebojeznanja.hr/UserDocsImages/datoteke/KB_web.pdf, and Duke, Chris et al. OECD Reviews of Tertiary Education: Croatia. OECD, 2008.