86671 v1 ACHIEVING ROMA INCLUSION IN ROMANIA – WHAT DOES IT TAKE? Prepared by the Human Development and Sustainable Development Teams Europe and Central Asia The World Bank Group February 28, 2014 . Standard Disclaimer: . This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. . Copyright Statement: . The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Roma inclusion is smart economics for Romania. With an ageing population and a young and growing Roma minority, Romania cannot afford to leave Roma children and youth and their families behind. This report discusses what it would take for Romania to achieve the socioeconomic inclusion of its Roma population. After a brief look at Romania’s demographic and labor market challenges, the report identifies the most important socioeconomic achievement gaps of Romanian Roma. Next, it identifies obstacles to Roma inclusion and examines the relevant institutional framework. Finally, it draws policy recommendations based on the observed gaps in outcomes and policies, informed by evidence on what works from international experience. These recommendations focus on providing support and enhancing opportunities for the next generation of Roma while helping to improve the living conditions of the current generation. This report is a summary of a more comprehensive study titled “Diagnostics and Policy Advice for Roma Inclusion in Romania.” The study will be made available to stakeholders to inform the Romanian government’s plans for the European Union 2014–2020 programming period, the new National Roma Inclusion Strategy, sector-specific strategies, new legislations, and ministerial and municipal budget allocations, in order to help Romania achieve its social inclusion goals under Europe 2020, the European Union’s 10-year growth strategy. The study relies primarily on data from the Regional Roma Survey (RRS) carried out in 2011 by the United Nations Development Programme, the World Bank, and the European Commission. The RRS is the most comprehensive household survey effort to date to capture the situation of Roma in Romania and compare it to that of non-Roma. The survey data support comprehensive diagnostics that draw comparisons between Roma households, their non-Roma neighbors, and the Romanian population as a whole. This study also draws on qualitative information collected through field visits and through focus groups and interviews with key stakeholders, including civil society actors, representatives of Roma communities, Romanian academics, and Romanian government officials. In addition, it benefited from comments and information provided by representatives of Romanian civil society, academia, and government agencies in the course of three consultations (September 24–25, 2013; November 21–22, 2013; and February 4, 2014). For more information on the RRS and other data sources used in developing this report, please refer to Annex A. 2 ACKNOWLEDGMENTS The study was conducted under the technical supervision of Omar Arias and Elisabeth Huybens, with overall guidance from Mamta Murthi, Ana Revenga, Alberto Rodriguez, Laszlo Lovei, Elisabetta Capannelli, Roberta Gatti, Ismail Radwan, Christian Bodewig, Katarina Mathernova, and Ines Fraile-Ordonez. The development of the report was coordinated by task team leaders Kosuke Anan (kanan@worldbank.org) and Sandor Karacsony (skaracsony@worldbank.org). The report was prepared with contributions by Simona Anton, Magda Balica, Florin Botonogu, Aida Catana, Adrian Dan, Plamen Danchev, Dana Farcasanu, Celine Ferre, Richard Florescu, Vlad Grigoras, Sorin Ionita, Ana Ivasiuc, Barbara Kits, Adam Kullman, Joost de Laat, Mihai Magheru, Ashna Mathema, Cristi Mihalache, Mariana Moarcas, Larisa Ofiteru, Claudia Rokx, Iulius Rostas, Manuela Stanculescu, Rob Swinkels, Lea Tan, Dikshya Thapa, and Michael Weber. Oana Maria Caraba, Sujani Eli, Sophia Georgieva, Corina Grigore, Camelia Gusescu and Isadora Nouel also contributed to the team’s work and provided support. The team is grateful for comments received from the report’s peer reviewers, Dena Ringold, Ana Maria Munoz Boudet, and Eva-Maria Szavuj (European Commission, DG Regional Policy). The study also benefited from constructive comments by Enrica Chiozza and Koen Delanghe (European Commission, DG Regional Policy), Soraya Goga, and Siobhan McInerney-Lankford. The team appreciates the feedback and support provided by Marie-Jeanne Ghigea from the Romanian Ministry of Labor, Family, Social Protection and Elderly Persons. The study also greatly benefited from comments and inputs provided by representatives of Romanian civil society, academia, and government agencies in the course of three consultations (September 24–25, 2013; November 21–22, 2013; and February 4, 2014). 3 CONTENTS Acknowledgments........................................................................................................................... 3 Abbreviations .................................................................................................................................. 5 1 Roma Inclusion Is Smart Economics for Romania .......................................................... 6 2 Socioeconomic Achievement Gaps of Roma ................................................................. 10 2.1 Dimension 1: Education and Skills Development.......................................................... 11 2.2 Dimension 2: Earning Opportunities for Families ......................................................... 14 2.3 Dimension 3: Basic Services and Living Conditions ..................................................... 16 2.4 The Intergenerational Cycle of Exclusion ...................................................................... 18 3 Obstacles to Roma Inclusion .......................................................................................... 20 3.1 Discrimination ................................................................................................................ 20 3.2 Capacity of Service Providers ........................................................................................ 22 3.3 Institutional Framework for Roma Inclusion ................................................................. 23 4 Priority Interventions and Policy Measures ................................................................... 26 4.1 An Integrated Approach to Roma Inclusion .................................................................. 26 4.2 Targeted Interventions to Offer Equality of Opportunities for Roma ............................ 28 4.3 Policy Measures to Strengthen Accountability and Ensure Capacity to Deliver ........... 36 Annex A. Methodology and Data Sources ................................................................................... 40 4 ABBREVIATIONS € euros CLLD Community-Led Local Development EC European Commission ERDF European Regional Development Fund EU European Union EU-SILC European Union Statistics on Income and Living Conditions HBS Household Budget Survey NCCD National Council for Combating Discrimination NRIS National Roma Inclusion Strategy RDA Regional Development Agency RRS UNDP/World Bank/EC Regional Roma Survey RSDF Romanian Social Development Fund UNDP United Nations Development Programme WWC What Works Clearinghouse 5 1 ROMA INCLUSION IS SMART ECONOMICS FOR ROMANIA Romanian Roma families today constitute a large, young, and extremely poor ethnic minority group, facing exclusion from markets and services. Investments in Roma inclusion are essential for Romania to achieve its Europe 2020 social inclusion goals, and the considerable returns on such investments would lay a more solid foundation for achieving sustained, shared prosperity across Romanian society. Therefore, Roma inclusion is not only a moral imperative, but also smart economics for Romania. 1. Romania’s Roma population is a large minority group. Given the lack of availability of ethnically disaggregated data in new European Union (EU) Member States, along with the fact that Roma families are often reluctant to reveal their ethnic identity to officials, it remains difficult to precisely assess the size of the Roma population in Romania. About 613,000 Romanians, approximately 3.3% of the population, declared Roma ethnicity in the 2011 Census. However, there are concerns that this estimate is inaccurate due to underreporting. Even this lower-bound estimate would make Roma the second-largest ethnic minority in Romania, after Hungarians. Expert estimates place the number of Roma much higher: according to Council of Europe data, the Romanian Roma population in 2010 was estimated at between 1.2 million and 2.5 million, or 6.5% to 13.5% of the total population.1 2. The Roma are a young population, and an increasing share of new labor market entrants come from Roma families. Children and youth aged 0–14 years—the new generation of labor market entrants—make up almost 40% of the total Roma population, compared to 15% among the general population. At the same time, about 17% of the general population in Romania is 65 or older, a proportion that is projected to grow rapidly in the near future. The youth of the Roma thus stands in stark contrast to the fast-ageing profile of Romania’s general population (Figure 1.1). Depending on the estimates of the Roma population, between 6% and 20% of labor market entrants in Romania today are Roma. As the overall working-age population in Romania is projected to fall by 30% by 2050, the Roma share is expected to grow. 6 FIGURE 1.1: THE ROMA POPULATION IS YOUNGER THAN THE OVERALL POPULATION OF ROMANIA Roma Overall Population Percent Percent Male Female Female Male Sources: UNDP/World Bank/EC Regional Roma Survey, 2011; United Nations, Department of Economic and Social Affairs, Population Division, World Population Prospects: The 2012 Revision , DVD, 2013. 3. There is wide scope to increase employment rates and earnings among Roma. One important factor underlying the potential economic benefits of Roma inclusion is that employment rates and average earnings among Roma are currently far below those of the general Romanian population, or even non-Roma neighbors. In 2011, 66% of men and 53% of women of working age (15–64) in the general population were employed, while only 42% and 19% of Roma men and women respectively had jobs, including informal jobs. Moreover, employed Roma earn only a fraction of the average earnings of the general population. As a result of low employment rates and low wages, the labor income of working-age Roma men in Romania is estimated to be only 20% of that of the general population; the corresponding figure for Roma women is 12%. These figures are much lower than for neighboring non-Roma (Figure 1.2). 7 FIGURE 1.2: ROMA HAVE FAR LOWER LABOR INCOMES THAN OTHER ROMANIANS 100 100 100 80 Percent 60 52 40 35 20 20 12 0 Men Women General Population Non-Roma Neighbors Roma Sources: EU-SILC (2008), for earnings information of the general population; Eurostat (2012), for 2011 Quarter 2; and UNDP/World Bank/EC Regional Roma Survey (2011). Note: Labor income rates for the general population are normalized at 100%. The corresponding rates for Roma and non-Roma neighbors are relative to the general population. Income rates are estimated by multiplying the average employment rate times the average earnings for those who are employed. This gives the average earnings for an individual in the working-age population. For comparative purposes, values have been adjusted to 2011 prices using the Harmonized Index of Consumer Prices (Eurostat, 2012). 4. Roma inclusion is not only a moral imperative, but also smart economics for Romania. With an ageing population, pension and health care costs are bound to increase in the near future. Equal labor market opportunities could enable faster productivity growth and could contribute fiscal benefits through increased revenue from taxes and lower social assistance spending. According to a World Bank estimate based on 2008 data, assuming an equal number of working-age Roma men and women, and assuming that average wages in the economy remain unchanged, equalizing labor market earnings in Romania for Roma could result in potential economic benefits ranging between €887 million and €2.9 billion annually, and fiscal benefits ranging between €202 million and €675 million annually.2 While these numbers are based on overly simplified assumptions about adjustments in the economy and the labor market, they illustrate the economic potential of Roma inclusion. The challenges imposed by the very large gap in labor market outcomes gather more significance in light of the aforementioned demographic trends: Romania will experience a substantial increase in the proportion of elderly people and shrinking of the overall size of the working-age population. The labor market inclusion of the younger Roma population can be a boon that addresses the fiscal and economic challenges posed by demographic trends. 5. Moreover, the social and economic inclusion of Roma is essential if Romania is to meet its Europe 2020 targets. Roma are significantly poorer and more vulnerable than non- Roma, including their own non-Roma neighbors (Table 1.1). In fact, the vast majority of Roma in Romania, as in other Eastern European countries, continue to live in poverty, much more so than their non-Roma neighbors living in the same or nearby communities. In addition, the at- risk-of-poverty rate of Romanian Roma, at 84%, is almost three times higher than among 8 neighboring non-Roma. 3 The rate of Romanian Roma households facing severe material deprivation is also alarmingly high (90%).4 Finally, almost half of Roma households in Romania have very low work intensity.5 As a result, reducing poverty among the Roma is a crucial step toward achieving Romania’s Europe 2020 goal of reducing the number of Romanian citizens at risk of poverty by 580,000. The same is true for two other Europe 2020 goals: employment and education (including secondary school dropout and access to tertiary schooling). Indeed, the importance of Roma inclusion for social cohesion in all of Europe has been pointed out by the European Union, starting with the EU Framework for National Roma Integration Strategies up to 2020 6 and leading to the adoption, in December 2013, of the first-ever EU-wide legal instrument for Roma inclusion, the Council Recommendation on Effective Roma Integration Measures in the Member States.7 TABLE 1.1. EUROPE 2020 SOCIAL INCLUSION INDICATORS IN ROMANIA National, target National, Non-Roma for 2020 actual (2011) Roma, actual neighbors, actual At risk of poverty 21% 22% 84% 31% Living in severe material deprivation Unspecified 32% 90% 54% Living in household with very low work intensity Unspecified 6% 44% 22% Employment rate 70% 64% 30% 44% Share of 30- to 34-year-olds having achieved tertiary 27.0% 22.0% 0.4% 7.0% education Sources: EU-SILC (2011); UNDP/World Bank/EC Regional Roma Survey (2011). World Bank staff calculations. 9 2 SOCIOECONOMIC ACHIEVEMENT GAPS OF ROMA Skills development (chiefly through early childhood development and basic education), earning opportunities for Roma families, and basic services and living conditions are the three dimensions that underlie the challenge of achieving equality of opportunities for the Roma. Exclusions in the three dimensions are mutually reinforcing and perpetuated across generations. The intergenerational cycle of Roma exclusion can only be broken if the inequality of opportunities that starts early in life is adequately addressed for the next generation of Roma. 6. The socioeconomic exclusion of the Romanian Roma is the result of multiple interacting factors that result in a stark inequality of opportunities, starting early in life. In an effort to streamline policy recommendations and to highlight the most promising areas of intervention, this report focuses on three key dimensions of exclusion: skills development (chiefly through early childhood development and basic education), earning opportunities for Roma families, and basic services and living conditions. These three dimensions underlie the challenge of achieving equality of opportunities for the Roma, both the current generation and— crucially—the next generation (Figure 2.1). FIGURE 2.1: KEY DIMENSIONS OF ROMA EXCLUSION THROUGHOUT THE LIFE CYCLE 10 7. Despite their dim prospects, Roma have by and large the same aspirations as non- Roma for themselves and their children. Although Roma are often perceived to lack interest in social inclusion, the survey evidence indicates otherwise. Evidence from other countries indicates that the aspirations of the Roma are often influenced by conflicting priorities and family obligations, mainly stemming from inadequate access to resources and information.8 For example, Roma boys are often required to leave school to help their families generate additional income for daily survival. Vulnerable families also prioritize expenses on basic daily needs over transportation, clothing, or other expenses related to schools. In spite of lower access to quality health, education, and earning opportunities, Roma parents still want to see their children achieve higher levels of education: about 73% of Roma parents want their children to complete at least secondary education, compared to 92% of non-Roma parents living nearby. Regarding employment, Roma with jobs report greater levels of happiness and life satisfaction than those without work, and 74% of Roma men and 76% of women report preferring “Secure employment but low paid” over “Having a higher income but insecure and irregular.” These figures are 82% and 83% for male and female non-Roma neighbors, respectively. Similar to their non-Roma neighbors, a majority of Roma prefer “Having secure employm ent but having to be at work 8 hours a day, 5 days a week, and not having the freedom to manage your time” rather than “Having irregular employment but being free to manage your time.” 8. There is a lack of rigorous systematic evidence to inform Roma inclusion policies about what works, what does not work, and why. While European civil society has long advocated for evidence-based Roma inclusion policy making, stakeholders (government agencies, municipalities, civil society organizations, donors) in Eastern Europe typically have little or no information about whether policy efforts over the years have actually made a difference in the lives of Roma communities. As a result, considerable knowledge gaps remain on why inclusion outcomes are so dire, and there is a need to learn through the rigorous evaluation of initiatives. 2.1 DIMENSION 1: EDUCATION AND SKILLS DEVELOPMENT 9. Inequality of opportunities for Roma start early in life, with inadequate access to early childhood development services. Approximately 40% of children in Roma households are undernourished, which severely impairs their lifelong cognitive and socio-emotional development potential as it hinders their ability to attend school, learn, and socialize.9 Inadequate feeding practices are a driving factor. Exclusive breastfeeding rates are generally low in Romania, reaching 20% among Roma. Moreover, diet diversification is inadequate, with Roma mothers seldom feeding their children meat, eggs, and vegetables during their crucial early years. Only one in five Roma households can afford to eat meat every second day (compared to three in five neighboring non-Roma households). As discussed later, poor housing and inadequate access to health, water, and sanitation services are also key contributors to child malnutrition and related diarrhea, as well as to respiratory and other infectious diseases among children. BOX 2.1: THE CHANCES OF A ROMA CHILD ARE GRIM IN COMPARISON TO NON-ROMA PEERS 11 When a Roma child is born in Romania, he or she is typically born into poverty: nine out of 10 Roma live in severe material deprivation. One-half of Roma children grow up in overcrowded housing, and one-third in slum dwellings. The Roma child faces a higher chance than non-Roma peers of suffering from early malnutrition or diseases that jeopardize healthy development in the crucial first years of life. The Roma child can typically expect to start school only at age 6 or 7, missing out on crucial preschool education, and to be out of school again as early as age 16, by which time only 29% of Roma boys and 18% of Roma girls are still found in the classroom. If the child is from a neighborhood where the majority of residents are Roma, there is a high likelihood (31% to 60%) that he or she will study in a segregated school or classroom for Roma. Even Roma children who attend integrated schools are likely to be treated differently by their peers and teachers, often because of prejudices. The Roma youth’s chances of completing secondary school are small, as are his or her chances of finding a job: Roma girls have only a one in five chance to succeed in finding employment. About a third of Roma who look for a job will experience discrimination. By the time the Roma individual reaches the age of 55, he or she has a 54% chance of suffering from a chronic illness or health problem. On average, he or she can expect to live six years less than non-Roma peers. 10. Few Roma children have access to quality preschool education. Only 37% of Roma children between 3 and 6 years of age are enrolled in preschool, against 63% of their non-Roma neighbors (see Figure 2.2). This low preschool enrollment undermines the school readiness of Roma children and their chances of achieving higher educational levels and employment skills later in life. The international evidence demonstrates that early intellectual stimulation in the home and in preschools develops the foundations of cognitive and socio-emotional skills, improving chances of socioeconomic success later in life, especially for vulnerable groups.10 Available data illustrate this: for example, while only 5% of Roma children ages 4–6 who do not attend preschool can identify at least 10 letters of the alphabet, the percentage in the same age group goes up to 40% if they attend preschool.11 Unaffordability of preschool contributes to low preschool enrollment rates of Roma children. Roma parents with children in preschool report spending €7.5 per month on preschool, approximately 4% of household income, with lunch brought from home in more than three-fifths of all cases. Though this may seem a comparatively low amount, costs nonetheless are commonly cited by Roma parents as a reason for not sending their children to preschool. A small number of available places in local preschools and the generally low quality of the early education available to Roma communities further limit access for Roma children. 12 FIGURE 2.2: FEW ROMA CHILDREN HAVE ACCESS TO EARLY EDUCATION Preschool Enrollment Rate (3–6 years old) of Romanian Roma and Non-Roma Neighbors 70 60 50 Percent 40 30 20 10 0 Roma Non-Roma Roma girls Non-Roma Roma Non-Roma boys boys girls children - children - Total Total Source: UNDP/World Bank/EC Regional Roma Survey (2011). 11. Roma children’s opportunities for skills development are further hindered by the high level of segregation of Roma pupils in classrooms. Nearly a quarter of Roma children currently attending basic education are in classes where most of the children are Roma. This is worrisome, given the evidence that school segregation has strong negative implications for the quality of education and skills development among Roma. Findings from international studies suggest that students in integrated schools achieve gains on standardized reading comprehension tests and in the development of non-cognitive skills and attitudes of tolerance, both among Roma and non-Roma.12 12. Roma have low rates of secondary school completion. Secondary education in Romania lasts four years: grades 9 and 10 (lower secondary) are compulsory, while grades 11 and 12 (upper secondary) are not. Not surprisingly, only 10% of Roma adults (25–64 years of age) have completed all four years of secondary education, in stark contrast with 58% of non- Roma living nearby and the even higher completion rate in the general population. Completion rates are particularly low among Roma women. As shown in Figure 2.3, the gap between Roma and non-Roma enrollment widens among youth: only 18% of Roma women and 29% of Roma men in the age group 16–19 are still enrolled in school, and only 23% of Roma youth are in upper-secondary education (compared to 83% of non-Roma). Other indicators of professional qualifications point in the same direction: computer literacy and participation in vocational training are about half as frequent among Roma as among non-Roma neighbors.13 As discussed next, the marked early child development and education gaps in turn result in few labor market opportunities for the Roma. 13 FIGURE 2.3: THE MAJORITY OF ROMA CHILDREN DROP OUT OF SCHOOL TOO SOON Enrollment Rates among Roma and Non-Roma Neighbors, by Age Group Male Female 100 100 80 80 Percent Percent 60 60 40 40 20 20 0 0 7-15 16-19 20-24 7-15 16-19 20-24 Roma Non-Roma Roma Non-Roma Source: UNDP/World Bank/EC Regional Roma Survey (2011). 2.2 DIMENSION 2: EARNING OPPORTUNITIES FOR FAMILIES 13. As noted earlier, Roma men and women are largely excluded from labor market opportunities in Romania. Roma are employed far less often than non-Roma, even when one compares Roma to their non-Roma neighbors, who share the same regional labor markets. This reflects low labor market attachment, which in turn stems from discouragement about the low prospects of finding work. Labor force participation of working-age Roma (57% among men and 34% among women) is lower than that of non-Roma neighbors (67% among men and 42% among women). 14 In addition to lower participation rates, Romanian Roma also face higher unemployment rates, meaning that many of them are looking for work without being able to find it. Moreover, among those who are employed, jobs for the most part are unstable and informal. Household survey data show that only about 35% of Roma employment is formal (based on a written contract or legal business documents) and covered by health and pension insurance, in contrast with 81% of employment among non-Roma living nearby. Many of the jobs held by Roma require only low skill levels—likely as a result of their skills gaps—and are poorly paid. For example, many working Roma are employed in low-paying agriculture jobs. 14. Indeed, in a typical Roma household, only €101 per month is brought in through income from employment, compared to €218 among non-Roma neighbors. Roma households typically have a lower share of income coming from employment than do non-Roma neighbors. Alternative income sources are also highly limited. For example, Roma people have fewer assets for developing an agricultural livelihood. Most Roma own plots of arable land smaller than half a hectare; furthermore, they report that due to the discriminatory restitution procedure, they receive low-quality land in unfavorable locations that is of little use for cultivation. Traditional trades were strongly discouraged by the former socialist regime and, consequently, have almost disappeared. Many Roma resort to other activities such as collecting 14 empty bottles, paper, or scrap metal and returning them for refunds, or selling used clothes, livelihoods that are time-consuming, poorly paid, and extremely vulnerable. 15. Roma marry young, begin childbearing early, and have high dependency ratios. About 28% of Roma between the ages of 15 and 19 years are married, as opposed to only 2% in the general Romanian population. According to the Regional Roma Survey (RRS) of 2011, the mean desired age to start having children is 21 years for Roma women, compared to 26 among non-Roma women. In a household survey across Europe, the fraction of women aged 14–16 years that had given birth for the first time was three times higher among the Roma than among the non-Roma.15 In a recent study of Romanian Roma by the Impreuna Agency for Community Development, early age pregnancy is linked to low income, likely due to lower levels of education among low-income Roma households, as well as to a lack of information on birth control and a lack of means to afford it.16 In part because of early marriage and childbearing, the Roma population has a higher dependency ratio than the population at large, meaning that more people who are not in the labor force depend on community or family members who are gainfully employed. This, coupled with low employment rates, exacerbates poverty levels. 16. If the dramatic inequality of opportunity starting early in life and continuing throughout the schooling years is not addressed, labor market exclusion is likely to continue among future generations of Roma. Statistical analysis shows that education is a significant predictor of employment among both Roma and non-Roma: individuals who complete secondary school are more than twice as likely to find stable employment as those who do not, after controlling for background characteristics. As mentioned earlier, only 10% of Roma adults have completed secondary school. Moreover, the level of illiteracy is extremely high among Roma, about 25% of whom cannot read or write. 17 While employment levels among Roma and non-Roma are similar among those with low levels of education (up to grade 8 at most), and among those with higher (tertiary) education, Roma with a secondary education have a considerably lower probability of employment than non-Roma with equivalent schooling. This may suggest that unemployment and underemployment among Roma are due not only to their lower levels of educational attainment, but also to lower skills resulting from their early life disadvantages and lower quality of education. 17. Low participation and restricted access to stable, gainful employment reinforce precarious incomes, a high risk of poverty, and social exclusion among Romanian Roma, and curtail opportunities for their children. In 1992, it was estimated that 63% of Roma families lived in absolute poverty. Another 18% had total income greater than minimum subsistence, but not enough for a decent life.18 Since then, both the incidence and the depth of poverty have remained disproportionally high in the Roma population. High rates of joblessness and job instability not only restrict income and perpetuate poverty among adults, but also take a toll on children, leading to absenteeism and dropping out of school, malnutrition, and chronic diseases. These in turn perpetuate the low levels of labor market opportunity that Roma currently face. Because of the low and irregular incomes earned from work, many Roma households end up relying on child allowances and on irregular and low-level streams of informal income. In some cases, school-aged children are expected to contribute to the family income. As highlighted earlier, higher employment rates and earnings opportunities among Roma would not only result 15 in a more inclusive Romanian society but would also boost economic growth and result in fiscal gains. 2.3 DIMENSION 3: BASIC SERVICES AND LIVING CONDITIONS 18. Roma have worse health than the non-Roma population. Data from the RRS on the health status of Roma and the results of the Impreuna survey both point to a higher burden of infectious and chronic disease. When broken down by age group, the self-reported health status of Roma adults and elderly is much worse than that of their non-Roma counterparts (Figure 2.4). FIGURE 2.4: ROMA REPORT WORSE HEALTH OUTCOMES THAN NON-ROMA NEIGHBORS AND THE GENERAL POPULATION Percentage Reporting that Their Health was Fair, Good, or Very Good, by Age Group 100 90 80 70 60 Percent 50 40 30 20 10 0 15 to 20 to 25 to 30 to 35 to 40 to 45 to 50 to 55 to 60 to 65 to 70 to 75 to 80 to 85 to 19 24 29 34 39 44 49 54 59 64 69 74 79 84 89 Roma Non-Roma Neighbors National Sources: Roma and non-Roma neighbors: UNDP/World Bank/EC Regional Roma Survey (2011); national population: Eurobarometer 2006. 19. The majority of Romanian Roma live in segregated communities. The RRS found that 56% of Roma households live in settlements where the dominant ethnicity is Roma, indicating a high level of spatial segregation. Spatial segregation is highly correlated with lower health status, early school-leaving, low labor market attachment, and costly access to other services (public transport, health facilities, etc.). About half (51%) of Roma households living in segregated communities live in dilapidated houses or slum conditions. 20. Roma experience exclusion in the housing market. A significant proportion of Roma live in poor-quality houses with inadequate infrastructure, overcrowding, and lack of tenure security. Housing conditions for Roma households are consistently worse than for non-Roma households: the RRS finds that 30% of Roma households live in a dilapidated house or slum, compared to 4% of non-Roma households living nearby. 19 Only about half of the Roma 16 households in urban areas have access to housing of relatively good quality—either newly constructed housing, dwellings made of traditional materials in older settlements, or social housing provided by the local authorities. The remaining 40–45% live in low-quality multi- storey blocks or slums, or in temporary camps with poor-quality structures and inadequate infrastructure; only 20% of the neighboring non-Roma live in such poor conditions. In rural areas, about a third of Roma households live in poor-quality housing. 21. A large number of Roma households lack access to water and sanitation. Only about 17% of Roma households have indoor sanitation (toilet, bathroom, sewage connection), while about 44% of nearby non-Roma families have these amenities. Similarly, only about 17% of Roma households have access to drinking water indoors, compared with 34% for non-Roma neighbors. Some 36% of Roma households report irregular or no collection of waste, compared with 18% for non-Roma neighbors. 22. Roma families’ access to health services is also inadequate. Roma suffer worse health conditions than non-Roma throughout the life cycle, with a higher burden of infectious and chronic diseases. As a result, the life expectancy of Roma in Romania is on average six years lower than that of the non-Roma population. Underlying causes include poverty, poor housing and living conditions, and lack of access to nutrition and health care in childhood. Other contributing factors are high-risk behaviors such as smoking (among both men and women), poor diet, low levels of physical activity, and inadequate utilization of health services. Romania’s health system does not perform well in delivering health services to the rural poor, a population that is disproportionately Roma. About two-fifths (42%) of Romanian Roma do not seek health care when they need it (versus 25% of their non-Roma neighbors). They cite as reasons financial constraints, including lack of insurance and the high cost of medication. 23. Access to safety net programs is widespread among Roma households, but these households remain largely below the poverty line. Nine out of 10 Roma households have access to at least one social protection program in Romania, but three out of five Roma households remain in the poorest quintile of the income distribution. While safety net programs are reaching the poorest, the fragmentation and flawed design of the current social assistance programs—including their formula, which tends to reduce generosity for larger households— render them less than effective in mitigating poverty of Roma households. Crucially, the system discourages recipients from moving out of labor inactivity by significant forgone tax and social benefits, especially for low-wage and part-time earners. Even the not overly generous unemployment benefits are often abruptly withdrawn when people start to work formally. These ill-designed components of the social assistance system generate an inactivity trap for low- income families, many of whom are Roma. 24. Exclusion from the credit market constrains the Roma’s ability to improve their livelihoods and living conditions. Many lower-income households in urban areas, including Roma households, could be eligible borrowers and would like to take a loan to buy or build a house, but they have no access to financing because they are informally employed or self- employed and have no property title to provide as collateral. 20 The lack of access to formal property mechanisms is undermining the Roma’s ability to leverage the value of their assets for productive use and to increase them. 21 According to the RRS, virtually none of the Roma 17 surveyed had access to mortgage loans or to microfinance. This severely limits the ability of low- income households to start or expand a business or improve their economic situation. 25. Many Roma lack identity documents, which has further restricted their access to services and credit. One major hurdle has been the required proof of residence, which many Roma are unable to show because of their informal housing situations. Other barriers include lack of literacy, as well as discriminatory attitudes among personnel in charge of issuing the documents. The lack of an identity card prevents access to many basic services, including health care. This increases the incidence of medically unassisted births, which in turn is a further barrier to identity registration. Lack of identity documents also prevents many Roma citizens from participating in elections. 2.4 THE INTERGENERATIONAL CYCLE OF EXCLUSION 26. Exclusions in the three dimensions are mutually reinforcing and perpetuated across generations. For example, inadequate education and skills and poor health hamper one’s access to earning opportunities. At the same time, the lack of earning opportunities results in insufficient resources to support children’s continued education and to maintain living conditions conducive to good health. Data collected by the Roma Observatory show that 79% of parents with at least one child who has dropped out of school mention poverty and material deprivation as the main reason for school dropout.22 Another 7% of respondents stated that the reason for dropout was the need to involve the child in income-generating activities. Poverty is also the most commonly reported reason for not consulting a doctor when needed; among Roma households who had forgone needed medical care, “too expensive” was the reason cited by 84% of them.23 The lack of access to basic services and living conditions, such as adequate space or light, also impedes Roma pupils from studying at home, which negatively affects their performance at school. The intergenerational cycle of exclusion is perpetuated when low educational achievement and poor health limit labor market opportunities among the next generation of Roma. 27. The intergenerational cycle of Roma exclusion can only be broken if the inequality of opportunities that starts early in life is adequately addressed for the next generation of Roma. The early years of life lay the foundation for growth and determine the individual’s potential for lifelong development. For example, since the brain develops most rapidly in the early childhood years, deficiencies in nutrition, stimulus, or nurture in these years can result in irreversible impairments to the brain that can prevent complete growth and development. Children who experience an extraordinary level of stress also risk having impaired cognitive and socio-emotional abilities in the future. This means that early childhood education is crucial, but also that poor feeding practices, housing, and living conditions (lack of water and sanitation) stemming from poverty need to be tackled. Interventions during early childhood will be critical to improving the Roma’s potential to succeed in school, maintain good health, and enjoy higher earnings and opportunities in the future.24 If Romania succeeds in achieving better outcomes for its Roma children, these children will grow into productive adults who in turn can offer their children better opportunities. 28. Well-designed policy interventions can help achieve the social and economic potential of Roma inclusion in Romania. Despite this dire picture of self-reinforcing and 18 perpetual exclusion in the three dimensions, there are entry points for breaking the intergenerational cycle that have proven to deliver results in other contexts. For example, policy efforts that facilitate access to early childhood development for disadvantaged children, skills training for disadvantaged youth, and slum upgrading programs could go a long way toward achieving social inclusion of the Roma population. Section 3 identifies three overarching obstacles to Roma inclusion. Section 4 builds on this framework by offering a number of concrete policy recommendations for closing the gaps in the three main dimensions of exclusion. 19 3 OBSTACLES TO ROMA INCLUSION Closing the gaps in skills development, earning opportunities, and basic services and living conditions requires an integrated approach with targeted interventions in these three dimensions. According to extensive interviews with Roma communities, policy makers, nongovernmental organizations, and other stakeholders, three cross-cutting issues may hamper Roma inclusion: discrimination, low capacity of service providers, and inadequacy of the institutional framework. 3.1 DISCRIMINATION 29. Being of Roma ethnicity increases the risk of poverty. Analysis of the RRS data shows that a Romanian individual is 38% more likely to be at risk of poverty if he or she is of Roma origin, in comparison with a non-Roma of similar age, gender, education level, household composition, community composition, and geographic location (shown in Figure 3.1 below). Similarly, 20% of the gap in employment between Roma and non-Roma neighbors cannot be explained by measured factors such as age, gender, education level, and geographic location. This gap reflects “unmeasured” factors, which may include unobserved skills or factors such as discrimination, norms, beliefs, and values. BOX 3.1: DISCRIMINATION AS A FACTOR CONTRIBUTING TO ACHIEVEMENT GAPS This study discusses discrimination as defined by the European Commission. Direct discrimination occurs when “one person is treated less favourably than another is, has been or would be treated in a comparable situation” on grounds of gender, racial or ethnic origin, disability, age, religion or belief, or sexual orientation. Indirect discrimination occurs where an apparently neutral provision would put persons of a particular racial or ethnic origin at a particular disadvantage compared with other persons, unless that provision is objectively justified by a legitimate aim and the means of achieving that aim are appropriate and necessary.25 It is often very difficult to measure and determine the extent and influence of discrimination in a society. Theorists have tried to measure discrimination concretely by using statistical techniques, especially in relation to labor market discrimination.26 However, objective statistical evidence of discrimination is rare, and where such direct evidence exists, corroboration is even rarer. Nonetheless, self-reported experiences and feelings of unfairness and injustice, although unquantifiable, provide an important indication of the level of discrimination that exists in a society. This study examined self-reported feelings of non-Roma toward Roma, as well as self-reported experiences and feelings of Roma, and assessed how these experiences and feelings affect the well-being of Roma and their ability to access services. A more detailed, specialized research design would be required to measure the occurrence of discrimination against Roma and enhance understanding of its root causes. 20 30. Over a quarter of Roma report having experienced discrimination because of their ethnicity in 2011. In the RRS, Roma households report experiencing the highest levels of discrimination when looking to buy or rent property (31%) or when looking for paid work (30%). 31. Roma experience exclusion on the basis of their ethnicity in both service provision and markets. Results from the focus group discussions carried out with Roma communities27 indicate that the discriminatory attitudes of service providers are essential barriers that impede Roma access to services. With respect to education, for example, a Roma mother from Telechiu in Bihor County explained, “The teachers are not teaching the Roma children as they teach the non-Roma ones.” 28 In the area of health, discriminatory practices toward Roma patients are manifested in “avoiding physical contact with the patients; non-involvement of the patients and of their family in choosing the treatment; omission of the explanations concerning the risks of administering a certain type of treatment; using aggressive procedures.”29 These experiences may help explain why so few Roma complete secondary school or visit health practitioners regularly. In the labor market, Roma report being told by employers that they were not hired because they are “gypsies.”30 In the financial market, bankers are reported to hesitate making loans to slum dwellers and informal sector workers, particularly the Roma, considering them a high-risk group that “does not play by the rules.”31 32. Negative stereotypes and mistrust are a root cause of Roma exclusion. For example, non-Roma parents still scare children by saying, “If you don’t behave yourself, the Gypsies will take you away.” 32 The media, including television, movies, and advertising, perpetuate demeaning images and stereotypes about the Roma.33 Articles and images portray the Roma as provoking violence in cases of community riots or behaving aggressively. A recent study conducted by the National Council for Combating Discrimination (CNCD), indicates that 45% of non-Roma “are afraid when they meet a group of Roma on the street.”34 This points to a deep mistrust between the two populations. 33. Discrimination can be combated most effectively by adopting a two-tiered approach: combating the sources of discrimination, and sanctioning acts of discrimination. The roots of discrimination, such as negative stereotypes and mistrust, need to be tackled by reducing negative portrayal of the Roma and increasing interactions between Roma and non- Roma to foster mutual understanding. Acts of discrimination need to be strictly sanctioned on a zero-tolerance basis through effective and consistent application of the law against discrimination. Sources of discrimination at the service delivery level could be reduced by strengthening the cultural competency of public officials (teachers, health care providers, police, mayors) and by increasing the number and capacity of mediators who serve as the bridge between Roma and service providers. The next section will discuss challenges related to service providers and mediators in more detail. BOX 3.2: POTENTIAL OF THE ROLE OF ROMA MEDIATORS AND COUNSELORS Romania has been a pioneer in involving Roma mediators. They were introduced by Romani CRISS, a Roma NGO, as early as 1992 as part of a community conflict mitigation program. Romania was also the first country in the region to institutionalize the Roma health mediator program, in 2002. 21 Roma mediators serve as a bridge between Roma communities and schools or health facilities, or between Roma families and public officials. Qualitative research conducted for this report suggests that these mediators can play an important role. Local health mediators, for example, can contribute to changing social norms that have discouraged the uptake of health services. This can be done by addressing the social stigma associated with accessing counseling services, reproductive health services, and testing for sexually transmitted infections. Roma school mediators are said to have had a positive impact in a wide range of areas, contributing to a decrease in the number of school dropouts and non-enrollment cases, improvement of school attainment and academic performance of Roma students, and reduction of absenteeism among students. They have also worked to combat the segregation of Roma and non-Roma students in classes and contributed to the desegregation of schools. Roma mediators are reported to have achieved improvements in communication between schools and the Roma community, and in the attitudes of teachers toward the Roma. A regional qualitative review of Roma health mediators by the Open Society Public Health Program shows that mediators have generally had success in changing the knowledge and attitudes of health care providers. The mediators report reduction in discriminatory behaviors and use of abusive language by doctors with whom they work. They also believe that they have helped physicians gain better understanding of Roma and enhanced their ability to provide care through more effective interactions with Roma patients. Source: Roma Health Mediators: Success and Challenges (New York: Open Society Foundations, 2011), http://www.opensocietyfoundations.org/sites/default/files/roma-health-mediators-20111022.pdf. 3.2 CAPACITY OF SERVICE PROVIDERS 34. Service providers often face significant challenges in providing services to Roma. At the local level, there are often insufficient financial resources to meet the needs of Roma communities for improved service delivery. Information gaps and communication barriers add to these challenges. In some cases there is a lack of incentives to improve delivery. In other cases, service providers lack understanding of the local context of Roma communities, which undermines the quality of services. 35. Capacity issues are particularly marked in education. Schools serving predominantly Roma children suffer from lower instruction quality, stemming in turn from inadequate teacher training and learning resources. Several studies find that schools with a higher percentage of Roma students tend to have less well trained and less experienced staff, along with high staff turnover. School infrastructure and resources correlate with the quality of education and are particularly inadequate in many schools with a high proportion of Roma pupils. A concentration of Roma in a school, for example, is associated with lower quantity and quality of endowments such as computers, laboratories, sport fields, libraries, and an adequate number of textbooks per student. This results in inadequate learning opportunities for Roma children. 36. The school mediator program is promising, but it needs to be strengthened. In spite of the reported positive roles of Roma school mediators, the mediator programs have only partial coverage of Roma communities. Existing capacity could be improved by creating proper 22 incentives. The Impreuna study shows that over half (55%) of the surveyed schools lack a school mediator, including schools where more than 50% of the students are Roma.35 Between 2003 and 2013, through various programs, a total of 1,001 school mediators have been trained in Romania: out of these, the number of currently active mediators is about 400,36 suggesting that there is scope for using the existing capacity better. Some local mayors have diverted resources from the Roma school mediator program to other areas; in other cases, Roma mediators opt out due to unattractive salaries. In addition, better regulation of the role of Roma mediators is critical. For example, mediators are often pushed aside to do administrative work or cleaning. In other cases, they are assigned to teach Roma children in place of formally accredited teachers, while the formally accredited teachers only teach non-Roma. 37. Addressing the inadequate quality of basic services to Roma requires investments to strengthen service providers’ capacity to deliver. Some existing initiatives for training and sensitizing service providers have already shown promise of effectiveness. One such program educates mayors on the challenges that Roma families typically face. This program is regarded as highly successful by local NGOs. In addition, increasing local budgets for underserved schools and Roma mediator programs is an important area of opportunity. But its potential will only be realized if effective checks and balances are in place to monitor spending, so that increased budgets actually improve the quality of schools and allow Roma mediators to deliver more and better services. Furthermore, synergies among different mediation activities (education, health, employment) may be considered on the basis of successful local or international examples, so that mediators can strengthen the linkages between the communities and services or facilities through their bridging role. 3.3 INSTITUTIONAL FRAMEWORK FOR ROMA INCLUSION 38. Effective planning and implementation of Roma inclusion interventions in Romania have been hampered by a lack of clear mandates and accountability. When institutions responsible for addressing Roma inclusion do not have clear mandates, including incentive mechanisms to implement these mandates, it is difficult to hold them accountable and make concerted efforts to implement priority actions in an integrated manner. This becomes an even greater challenge when the mandate and accountability do not fully coincide. For example, the implementation of the National Roma Inclusion Strategy (NRIS) suffers from an inconsistent institutional and legal framework. Some measures proposed in this strategy do not fall under the legal mandate of local authorities, severely hindering implementation. At the national level, neither the National Agency for Roma, which is responsible for coordinating the implementation of the NRIS, nor the Inter-ministerial Working Group, which is in charge of the coordination and implementation of the public policies for Roma, has the sufficient level of statutory power to hold other institutions accountable for implementing the NRIS. The scope and complexity of interventions required in implementing the NRIS requires greater attention at the highest levels. Responsible institutions need to communicate with one another, have clear and transparent decision-making procedures, and establish budgetary links to strategies based on a clear set of priorities. Moreover, there is a crucial opportunity to implement a comprehensive monitoring framework, which will allow responsible institutions to track progress and improve interventions accordingly. 23 39. Policy measures proposed in the NRIS are often not adequately funded. The devolvement of decision-making authority to the local level has often not been accompanied by a corresponding decentralization of the budget. Moreover, centrally imposed spending cuts to local budgets have seriously affected the autonomy of local governments, since local budgets have decreased substantially. Under these budget constraints and with this high degree of unpredictability, local development strategies, including the NRIS, have become less relevant in practice. According to a civil society monitoring report on the implementation of the National Roma Integration Strategy,37 some policy measures that are not particularly aligned to the objectives of the NRIS are included in the strategy. There is a risk that these measures will burden local governments without having impacts on Roma or will result in efforts and expenditures that will not be conducive to Roma inclusion. 40. Roma are often not explicitly targeted in social inclusion programs. 38 Currently, specific social policies, programs, and investments often target disadvantaged population groups but without specific methodologies for measuring the disadvantage. While there is not always a need to target Roma exclusively, the absence of specific criteria for identifying target groups and the lack of clear methodology by which to measure these criteria makes it difficult to target social programs at the local level. The Roma Inclusion Strategy explicitly targets Roma, but data on the ethnicity of program beneficiaries are not collected systematically to monitor progress. Information on ethnicity would enable better tracking of progress in social outcomes for Roma and non-Roma communities and enable a much deeper analysis of determinants of poverty. 41. Insufficient capacity of local communities to design and implement projects, in the context of inadequate central-level support, is undermining the ability of many disadvantaged communities to benefit from EU funds.39 The EU Structural Funds, Cohesion Fund, and European Agricultural Fund for Rural Development are not sufficiently reaching marginalized Roma communities. The overall absorption rate of the EU Structural and Cohesion Funds during 2007–2013 has also been low in Romania. This is caused by a few key factors: (a) lack of alignment between the legislation regulating implementation of structural instruments and other pertinent national laws; (b) complexity of administrative procedures; (c) lack of administrative capacity both within management authorities and among beneficiaries; (d) weak accountability of local authorities to address issues faced by Roma and other disadvantaged communities; (e) lack of co-financing capability by local authorities; and (f) unclear distribution of tasks among ministries and other public entities at the national level. 42. Limited consultation around policies related to social inclusion of Roma at the local level compromises the relevance and sustainability of implementation measures. This is a missed opportunity for making sure that measures to tackle Roma inclusion better reflect local needs and opportunities. Moreover, the findings from the qualitative study for this report suggest that discriminatory attitudes of some officials toward Roma affect progress toward true Roma participation in the design and implementation of inclusion programs directed at them. 43. Local-level capacity to develop and implement integrated projects is key to improving the living conditions of Roma. Poor living conditions for Roma in Romania are manifested in diverse ways, depending on the community context, and require a similarly diverse range of interventions. For instance, simply providing social housing, while very costly, often 24 does not bring a marked improvement in Roma living conditions. In some cases, upgrading specific services (e.g., water, electricity, transport) or improving accessibility of health, education, and job opportunities is a more efficient solution, both socially and economically, than moving people into new housing. In addition, these interventions need to be integrated with measures to address sustainability risks. One prominent risk is affordability: ensuring the continued ability of residents to pay rent, loans, or utility fees might involve interventions to help them find jobs or enhance livelihoods. Such integrated interventions are best developed and implemented at the local level, as this increases the ownership and sustainability of projects by identifying stakeholders’ priorities and their ability to use, maintain, and operate the services. It also contributes to their social inclusion through empowerment. Participation of non-Roma communities is also critical to avoid stigmatization of Roma, foster interaction and cooperation between Roma and non-Roma on the basis of mutual interests, and gain non-Roma support for projects. 44. In sum, effective policies for Roma inclusion depend on clear accountability, adequate budget allocation, effective targeting of interventions, participation of Roma, and an increased capacity of local authorities to manage projects. These efforts can be accompanied by a comprehensive monitoring and evaluation system, which would allow authorities to track progress and hold responsible stakeholders accountable. A rigorous impact evaluation could also be undertaken for some of the key strategic programs to learn, share, and replicate effective interventions. 25 4 PRIORITY INTERVENTIONS AND POLICY MEASURES To make progress on Roma inclusion, an integrated package of targeted interventions and investments is needed. These should focus on improving opportunities for the next generation of Roma and their families in skills development (including health and education), earning opportunities, and basic services and living conditions. The study also suggests policy measures to strengthen accountability and ensure effective delivery of these interventions. 4.1 AN INTEGRATED APPROACH TO ROMA INCLUSION 45. Breaking the intergenerational cycle of Roma exclusion requires targeted interventions to address the multiple drivers of inequality. This means tackling barriers to opportunity for Roma children, especially those related to education and skills development; living conditions and access to health, water, and sanitation; and parents’ employment and household income. Therefore, this study recommends three targeted interventions: (a) closing the early childhood development gap through targeted investments in health, nutrition, and education; (b) investing in skills development and employability of Roma youth and young adults to facilitate their insertion in the labor market and tackle the underlying poverty of Roma households that currently limits their access to adequate diets, health, education, and housing, hindering their children’s development; and (c) supporting improvements in housing conditions and access to basic infrastructure to improve Roma families’ welfare while creating an enabling environment for early childhood development. Across these three areas, we recommend (d) improving the quality of social services to ensure the effectiveness of all interventions; (e) improving coordination, accountability, incentive structures, and monitoring and evaluation to deliver and track concrete results, including at a very high level; and (f) combating negative stereotyping of Roma and investing in projects that require interaction between Roma and non- Roma to address discrimination (Figure 4.1). 46. Complex, interconnected problems require an integrated approach and multipronged solutions. Integrated approaches create synergies and can increase the effectiveness, efficiency, and sustainability of interventions. Integration can be achieved at different levels and in different forms. Various interventions can be combined to achieve a common objective, for example, by combining early childhood development interventions with basic infrastructure interventions to increase educational outcomes. An intervention can comprise both hard and soft measures, for example, by combining physical construction of social infrastructure with training of service providers who operate the infrastructure. An intervention can combine various measures to address both supply-side impediments (e.g., lack of health providers) and demand-side challenges (e.g., lack of awareness) to achieve an intended outcome. Finally, common resources can be utilized to implement multiple interventions or activities: for example, the same Roma mediator can help the Roma access various social services while also serving as a bridge between Roma communities, public officials, and non-Roma communities, closing the communication gap and fostering mutual awareness and understanding. 47. In delivering the interventions, it is important to address both supply- and demand- side challenges. Many Roma communities are highly impoverished, and merely providing new infrastructure or services will not necessarily result in their utilization. Demand-side bottlenecks, such as those related to users’ awareness, financial means, capacity constraints, opportunity costs, 26 social norms, and risks (safety, dignity, reputational, etc.), need to be assessed and addressed. Even if a service is extended, people will not use it if they are not aware of its benefits or if it is too costly. Moreover, even when a service is provided free of charge, people may decide not to access it if the transaction or opportunity costs are considered too high: for example, using the service may entail transport costs or lost wages for time away from work. Fear of mistreatment or humiliation by service providers such as teachers and health care providers may also discourage people from accessing a service. FIGURE 4.1. BREAKING THE CYCLE OF INTERGENERATIONAL EXCLUSION THROUGH TARGETED INTERVENTIONS AND POLICY MEASURES 27 4.2 TARGETED INTERVENTIONS TO OFFER EQUALITY OF OPPORTUNITIES FOR ROMA 4.2.1 INTERVENTION 1: INVEST IN CLOSING THE EARLY CHILDHOOD DEVELOPMENT GAP 54. Skills development policies should be a key priority for Roma inclusion. Upgrading education and skills is critical for improving the labor market outlook of the next generation, as well as for narrowing the employment gap facing the current generation of Roma adults. Investments in quality early childhood development interventions in Roma communities could involve improving school infrastructure; working with parents and incentivizing participation; addressing school dropout by scaling up school mediation and making schools more friendly to all students; and improving the quality of schools with a high percentage of Roma pupils by training and incentivizing teachers and providing adequate budgets. These measures will contribute to closing the skills gap in the long run. 55. Given the high returns to early childhood development and the very large gap that exists between Roma and non-Roma in Romania, practitioners should:  Consider initiatives that incentivize families to get regular health check-ups at zero cost at facilities close to their communities. This should include actions to raise awareness of the importance of these check-ups and let families know how they can access health services free of charge. Priorities might include increasing access to and usage of family planning and counseling, prenatal, and early childhood health care.  Develop and implement outreach initiatives, such as national media campaigns, that focus on maternal and child health, a healthy diet, and prevention of risky behaviors.  Design innovative incentives that pay directly for results—such as completing the vaccination schedule for Roma infants—and provide specific bonuses for the responsible providers.  Use kindergartens to train parents on the importance of providing learning stimulation in the home environment.  Invest in the expansion of quality preschool education in communities with a large share of Roma, with adequate staffing, resources, and curricula that give proper attention to the development of socio-emotional skills. Examples of such programs in Europe and the Americas include Tools of the Mind and PATHS.  Involve parents in early childhood education partnerships, including community-based preschool systems.  Provide financial incentives and remove financial barriers to early childhood education. One approach is to introduce annual subsidies for families, in the context of social assistance reforms, conditional on preschool attendance of children aged 3–4 years, covering the tuition and meal costs of kindergarten for families. BOX 4.1. HUNGARY’S APPROACH TO EARLY CHILDHOOD EDUCATION In Hungary, preschool enrollment among Roma is much higher than in neighboring countries: 49% of Roma children attend preschool in Hungary, a significantly higher percentage than in Romania. Among Roma 5-year-olds in Hungary, 86% attend preschool. According to the World 28 Bank, Hungary’s “new education law, passed in December 2011, calls for compulsory preschool from age 3. Mandatory primary school entry age is 6 years, although children may stay in kindergarten for an extra year, until turning 7. Public-sector kindergartens are free of charge; they charge a compensation for extra services not included in their basic tasks, e.g. for meals, excursions and extracurricular activities.” Preschool is financed through a combination of central government funding (covering about one- third of the cost), contributions from parents (covering 10%), and municipal government subsidies (covering slightly more than half of the total, although this has been problematic in poor municipalities). Fees paid by parents are reduced or waived for low-income families. Municipalities can contract with private providers or recruit volunteers to mitigate capacity issues. To further improve access for poor families, meals are provided free of charge when certain criteria are met. Furthermore, a conditional subsidy program was introduced in 2009, which encourages poor parents to enroll their children as early as possible. This “kindergarten subsidy program” grants families a biannual subsidy of 10,000 Hungarian forints (€35) per child aged 3–4, conditional on preschool attendance. Source: World Bank, Toward an Equal Start: Closing the Early Learning Gap for Roma Children in Eastern Europe (Washington, DC: World Bank, 2012), http://go.worldbank.org/F2QMFQ7FJ0. 4.2.2 INTERVENTION 2: INVEST IN SKILLS AND EMPLOYABILITY OF ROMA YOUTH AND YOUNG ADULTS 56. Employment policies must focus on improving the skills and employability of Roma. The study findings point to the need for labor training programs that target disadvantaged, jobless Roma youth and Roma women. Job search incentives should be improved and linked closely with job creation. When these policies are designed, the transformational effect of employment on the individual and family should be considered, including potential intergenerational effects: that is, training and employment among adults may contribute to improved education outcomes among children. Such initiatives may also allow Roma children to experience the role model of working parents. Considering the crucial role of both cognitive and non-cognitive skills in employment outcomes, the following interventions should be considered. In the area of education:  Take steps to prevent dropouts rather than attempting to accommodate dropouts: o Improve the school climate to encourage engagement, learning, and well-being of all students. o Set up an early warning system of students at risk of dropping out, including counseling and mentorship activities targeted to at-risk students.  Make schools more inclusive, relevant, and welcoming to Roma children: o Train teachers in effective approaches and practices, with a focus on inclusive education and the history, language, and culture of Roma. o Adapt the curriculum to reflect the cultural contexts of Roma communities; adjust school hours to accommodate the life patterns of Roma families; use learning materials relevant to the needs of Roma students. 29 o Give attention to the development of socio-emotional skills, with learning activities that promote Roma role models in order to increase the ethnic self- esteem of children. o Enable children to have a voice in their school’s decision-making process by activating student councils, which should include representatives of Roma children. o Invest in expanding mentoring and tutoring activities for Roma in schools.  Scale up and improve the effectiveness of the school mediator program to gradually include all schools with a high percentage of Roma.  Improve the targeting of affirmative measures (i.e., reserved seats in universities) for Roma children and youth. These measures should be geared to Romani-speaking youth from segregated rural communities to enable the most disadvantaged to progress to higher educational levels.  Facilitate parental involvement in education: for example, organize parent support groups in Roma communities and create opportunities for Roma parents to participate in school activities daily.  Provide incentives for teachers to work with Roma students: for example, consider launching the Teach for Romania program. 40 Consider introducing performance-based pay that would reward teachers who improve the achievement of Roma students.  Revise the teacher curriculum: introduce specific modules aimed at reducing school dropout among vulnerable groups like Roma, based on effective approaches.  Invest in transportation of Roma children who live far away from schools. 30 BOX 4.2: LINKING MONETARY ASSISTANCE TO EDUCATION AND HEALTH: CONDITIONAL CASH TRANSFERS Can social policies go beyond mere assistance and become active tools of socioeconomic transformation? Initiatives in Mexico and Brazil show that this is possible. Brazil’s Bolsa Família Program, like its Mexican predecessor, has been found to be one of the key factors behind the country’s recent substantial progress in reducing poverty and inequality. The program reaches 11 million families, a major portion of the country’s low-income population. It grants these families monthly subsidies on the condition that children attend school regularly and receive regular health check-ups. Poor families with children receive an average of 70 reais (about US$35) in direct transfers. According to the World Bank, “Although relatively modest in terms of resources when compared with other Brazilian social programs, such as Social Security, the Bolsa Família Program may be the one that is having the greatest impact on the lives of millions of low-income Brazilians.” Similar programs have been implemented in other countries in Latin America. Even New York City has recently announced its “Opportunity NYC” conditional cash transfer program, modeled on Bolsa Família and its Mexican equivalent. “This is an example of a developed country adopting and learning from the experiences in the so-called developing world.” Source: World Bank, “Bolsa Família: Changing the Lives of Millions in Brazil,” August 27, 2007, http://go.worldbank.org/3QI1C7B5U0. In the area of employment:  Design and implement a skills training program specifically suited to disadvantaged youth. Such a program should combine training in basic cognitive, socio-emotional (life), and technical skills with internships that ensure that training is aligned with employer skills demand (Box 4.3). The program should also include counseling and job search skills (e.g., the ability to write a good CV, identify potential employment opportunities, write an application letter, perform well in an interview), as well as the services of job search mediators. This can start with a pilot that can be gradually scaled up on the basis of a careful evaluation.  Improve second-chance education and traineeships, including adult literacy programs, and increase access among Roma.  Develop trainee, internship, and placement programs in administrative positions at the central, regional, and local levels.  Invest in labor market opportunities for Roma women. This might include expanding the community health nurse program and training women to become kindergarten teacher assistants.  Increase investment in means-tested social protection programs, providing support for the poorest and supporting and incentivizing transitions to employment.  Strengthen the capacity of the National Agency for Employment through adequate staffing and staff training, increasing awareness and trust among potential beneficiaries, and allocating more budget to active labor market programs for vulnerable populations, including Roma. 31  In the context of the ongoing social assistance reforms, eliminate the bias against large families in social assistance programs to help reduce poverty among all large families, including Roma. At the same time, address potential disincentives to work by lowering the effective marginal tax on earnings in the benefit formula of means-tested programs. This can be accompanied by complementary, active labor market programs and remedial services for adults who are able to work but not employed or in education or training. BOX 4.3: BUILDING SKILLS AMONG DISADVANTAGED YOUTH: SUCCESSFUL EXAMPLES FROM LATIN AMERICA Various programs around the globe have successfully linked disadvantaged youth to jobs. These programs all share two fundamental characteristics: (a) they provide an integrated intervention, addressing multiple challenges faced by young job seekers, and (b) they involve private sector stakeholders. Some use the “training plus” approach: technical and/or life skills training followed by private sector internships and job placement assistance. The Jóvenes programs in Latin America, begun in 1991, have provided a model for comprehensive intervention to improve youth employability. Implemented in various countries in slightly different forms, the model targets disadvantaged, low-skilled youth aged 16–29, offering them vocational training and numerous support services. Classroom training is complemented by actual work experience in private sector internships. Providers compete for the opportunity to provide training, keeping prices low, and selected training providers are required to align their courses with internships, ensuring relevance. Ample attention is paid to the development of “soft skills” such as a good work ethic and proper conduct during a job interview. Counseling and job search information complete the package. Employers receive wage subsidies and beneficiaries receive daily stipends to encourage their participation. A similar program in the Dominican Republic features an important soft skills component that, according to recent evidence, is improving the employability of participants. A rigorous impact evaluation finds that slightly over a year after receiving training, participants showed significant improvements in leadership, behavior in conflict situations, self-esteem, and self-organization. Moreover, a positive impact on employment rates, mainly among women, was found. Participation in the program was also associated with lower rates of teenage pregnancy among the treatment group. With respect to the Jóvenes programs in general, the World Bank wrote, “There have been varying estimates of costs relative to benefits. Early evidence from Peru indicates that the positive earnings effects need to last at least 7 years for PROJoven to yield a positive net gain. A recent longitudinal version of propensity score matching of PROJoven showed a positive internal rate of return, consistently above 4 percent. In Dominican Republic, the investment on training is recuperated after 2 years.” Sources: P. Ibarraran, et al. Life Skills, Employability and Training for Disadvantaged Youth: Evidence from a Randomized Evaluation Design, Discussion Paper 6617 (Bonn: IZA, 2012); O. Puerto, Interventions to Support Young Workers in Latin America and the Caribbean: Regional Report for the Youth Employment Inventory (Washington, DC: World Bank, 2007), http://www.youth-employment-inventory.org/downloads/3.pdf; World Bank, Stepping Up Skills for More Jobs and Higher Productivity (Washington, DC: World Bank, 2010); World Bank, Resilience, Equity, and Opportunity: The World Bank 2012–2022 Social Protection and Labor Strategy (Washington, DC: World Bank, 2012). 32 4.2.3 INTERVENTION 3: INVEST IN IMPROVING HOUSING CONDITIONS AND ACCESS TO BASIC INFRASTRUCTURE 57. Housing and infrastructure needs should be addressed with customized, integrated interventions at local level, combining physical investments with soft measures to enhance impacts and sustainability. These could be delivered in integration with the other interventions recommended in this report. For example:  Extend and upgrade infrastructure (water, sewerage, electricity, schools, clinics, access roads).  Expand quality and coverage of social services (health care, waste collection, community policing, youth activities, mentoring, counseling, recreational activities).  Extend transport services.  Improve community space (street lighting, parks).  Facilitate property registration and simplify the procedure for issuing civic documents, such as identity and residential documents. This is important in order to enable Roma, especially those living in informal settlements, to access basic services and credit and secure tenure rights. For example, when a land title cannot be issued to an informal land holding, the local authority can issue a certificate of possession per the amended Law 7/1996, which enables registration of possession. For Roma who live in geographically remote areas, the Public Community Service for Personal Records could possibly facilitate the preparation of necessary identity documents.  Explore the potential of extending housing finance. Options include (a) small housing mortgage loans for those who can qualify and can afford market-based housing, and (b) loans for building or expanding a house or establishing a business (shop, workshop). Mechanisms could be developed for assessing credit-worthiness of informal sector workers; such methods are distinctly different from those in the formal banking system but have been tested successfully across the globe. Another option would be to establish a guaranty fund to incentivize risk-averse banks and non-bank financial institutions to extend loans. 58. Communities should be involved in the design and delivery of interventions. To support management and operation of interventions by communities, training can be offered to community members, such as in participatory decision making, accounting, and basic financial literacy. Additionally, involving local populations in infrastructure upgrading could also increase ownership, create job opportunities, and develop skills. It is important to ensure that improvements to local infrastructure and housing do not lead to increased concentration or further physical isolation and segregation of marginalized groups.41 BOX 4.4: INTEGRATED URBAN INTERVENTIONS: AN EXAMPLE FROM MAGDOLNA NEIGHBORHOOD, HUNGARY The Magdolna neighborhood has been one of the most crowded and disadvantaged areas in Budapest. In 2007, close to 40% of 5,500 flats lacked a toilet, bath, or piped-in drinking water. 33 About 60% of the population was estimated to be economically inactive. Public security has also been a great concern. About 30% of the population is Roma. A series of measures has been implemented since 2005 to improve the living environment of the neighborhood. The measures were introduced in three phases, financed by multiple sources, including the Municipality of Budapest and the European Regional Development Fund (co- funded by the State budget). A key innovation was the emphasis on improving the living conditions of the existing population of the neighborhood, rather than just on physically remodeling the neighborhood. This was important, since a lack of focus on people could have led to gentrification, pushing the original residents out of the neighborhood and leading to their further marginalization. The interventions included the following activities, among others:  Renovation of the housing stock  Construction of a community center and rehabilitation of the main square  Organization of activities at the community center, including job search clubs, clubs for women, information technology training and access to the Internet, job fairs and exhibitions, summer camps for children, family therapies, and programs for talented children  Special youth programs offered by street social workers  After-school activities, training of teachers, changes to curricula  Vocational training designed for single mothers  Crime prevention and public security activities  Drug prevention programs The local population participated in the design and implementation of the interventions, which improved their effectiveness and also contributed to increasing the cohesion of community members. Source: Rév8, http://rev8.hu/english/. 34 4.2.4 INTERVENTION 4: INVEST IN INCREASING THE CAPACITY OF PUBLIC SERVICE PROVIDERS  Ensure that school and health mediators have a strong mandate to fulfill their duties in the school and health care systems.  Actively use the knowledge and information gained from mediation in the development of mainstream policies and programs, for example, by giving Roma mediators a formal advisory role in policy making.  Develop and introduce cultural competency modules within the mandatory curriculum for the initial and in-service training of public officials (teachers, health care providers, police, mediators, mayors).  Introduce a third-party ombudsperson function to monitor the observance of non- discrimination principles by public service providers (e.g., in health care, social assistance programs, and schools). In case of non-observance, retrain providers through the aforementioned cultural competency modules and consider the use of disciplinary measures, such as fines. 4.2.5 INTERVENTION 5: COMBAT NEGATIVE STEREOTYPES OF ROMA AND FOSTER MUTUAL UNDERSTANDING BETWEEN ROMA AND NON-ROMA  Introduce a surveillance mechanism to detect and seek reduction of media content that perpetuates negative stereotypes of Roma. This mechanism could work in a similar manner as those that currently monitor profanity and the presentation of cruelty in the media.  Invest in projects that require interaction and collaboration between Roma and non-Roma and generate benefits for both groups. Such projects could increase mutual understanding and respect, reducing a key source of discrimination.  Launch an intensive nationwide campaign to make discrimination socially unacceptable in Romania. Role models and opinion leaders could be engaged in the campaign to transmit strong messages that condemn discrimination as wrong and disgraceful conduct that should not be tolerated by society.  Enforce the law against discrimination more effectively and consistently by amending the law and the institutional framework of the National Council for Combating Discrimination (NCCD). The law and institutional framework need to clarify how unlawful acts of discrimination will be sanctioned, make the public more aware of the law, and increase the capacity and resources of the NCCD to consistently apply the law. Such reforms, by bringing more clarity and consistency to the application of the anti- discrimination law, should increase likelihood that unlawful acts of discrimination will be penalized or deterred. This in turn would foster a culture of zero tolerance of discrimination. 35 4.3 POLICY MEASURES TO STRENGTHEN ACCOUNTABILITY AND ENSURE CAPACITY TO DELIVER 59. A prerequisite for developing effective policies, implementing priority interventions, and addressing service delivery gaps is well-coordinated institutional mechanisms. In order to enhance the institutional mechanisms for effective local service delivery and use of EU instruments, the clarification of institutional responsibilities through more uniform legislation and clear working arrangements is necessary. A framework law could be enacted to spell out the functional relations between various structures as well as budgetary sources. Supporting and building capacity of local authorities is essential in accessing EU funds. Partnerships with Roma communities could be strengthened by formulating a methodology for Roma consultations at the local level, while stakeholder feedback could also be gathered through online platforms. Disadvantaged communities could be targeted more accurately by mapping them at subregional and local levels. Additionally, the uptake of EU funds by local communities could be increased by simplifying the procedures and reducing administrative burdens on these communities.42 4.3.1 POLICY MEASURE 1: CLARIFY INSTITUTIONAL RESPONSIBILITIES  Issue a framework law that ensures the uniformity of relevant legislation and describes the competencies and responsibilities of key agencies for Roma inclusion (e.g., National Agency for Roma, National Council for Combating Discrimination).  Create working instruments that spell out how the various players will work together, including clear incentives to adhere to these working instruments and to deliver results. 4.3.2 POLICY MEASURE 2: STRENGTHEN THE LINK BETWEEN STRATEGY AND BUDGET  Produce a budget estimate for NRIS implementation. It is important that the NRIS spell out, for each of its objectives, what outputs are needed to achieve the objective, what activities should be implemented to obtain the outputs, what financing is needed to carry out these activities, and where this financing should come from (including through EU funds). This could then be mainstreamed or translated into clear sectoral strategies and action plans that spell out what each of the ministries will deliver and when, what financing is needed for that, and through what budget items that financing will be made available. 4.3.3 POLICY MEASURE 3: IMPROVE TARGETING OF INVESTMENTS IN DISADVANTAGED COMMUNITIES  Issue a normative act—a Law of Poor Communities—that specifies the methodology for identifying marginalized communities. The methodology could apply the small-area estimation techniques used for targeting geographic areas, which are being drawn on poverty maps created by the World Bank. At the neighborhood level, targeting could be based on city maps that identify the geographic location of marginalized communities using data from the population census that have been verified locally, a technique being developed by the Ministry of Regional Development in collaboration with the World 36 Bank. The identification of disadvantaged communities and their particular disadvantages at the neighborhood level would play an instrumental role in developing projects through the Community-Led Local Development (CLLD) approach, which will be introduced in the 2014-2020 program for EU funds.  Issue specific calls for proposals for integrated programs for disadvantaged communities. 4.3.4 POLICY MEASURE 4: SUPPORT AND BUILD CAPACITIES OF LOCAL COMMUNITIES TO MANAGE PROJECTS AND ACCESS EU FUNDS  Establish a central unit equipped with experts who can facilitate community engagement and project planning, especially by helping communities submit applications for EU Structural Funds. This unit could possibly be built on the structure of the existing Romanian Social Development Fund (RSDF), which has demonstrated a functional institutional model that can foster local-level project development and implementation in marginalized communities. The RSDF’s knowledge of the community-driven development approach would become especially valuable in implementing the CLLD approach. Two options could be considered: o Explore the possibility of expanding the mandate of the Regional Development Agencies (RDAs) to support project design, screening, and implementation of the Operational Programme Human Capital and its programs co-funded by the European Social Fund (to the extent allowed by the timeframe of the Partnership Agreement). A Community Mobilization and Social Inclusion Unit could be established within the RDAs to help local communities design and implement social inclusion proposals. Staff of this unit could be trained by RSDF experts through a training-of-trainers approach, with RSDF experts continuing to play a mentoring role. o Explore the possibility of mobilizing the RSDF experts to provide project planning and community mobilization support directly to local communities across the country. The RSDF could establish regional offices to provide this support to local communities in an effective manner. The assistance could be provided to local authorities, civil society organizations, and local populations to enable them to assume their respective roles in developing and implementing projects. While building on the experience of the RSDF is just an option, and the central unit could be established with other arrangements, in all cases special knowledge and assistance would be required to enable local communities to process projects in alignment with the EU funding mechanisms.  Strengthen the methodology for Roma consultations and establish a guiding framework for a participatory approach. This would provide guidance on how local-level stakeholders and communities can be involved and participate in the planning and implementation of projects. It would include, among other things, tools for conducting needs assessment and identifying gaps and opportunities for synergies. Participation of the local community is essential to the success of interventions because local communities know what is needed, why it is needed, what the bottlenecks are, what can be done, what is affordable and can be maintained, and what opportunities exist. The participation of Roma in project preparation, implementation, and monitoring would increase the sustainability and transparency of projects and also contribute to active 37 citizenship of Roma through empowerment. In this regard, inclusion of men, women, the elderly, and youth will be important. 4.3.5 POLICY MEASURE 5: STRENGTHEN THE SYSTEM FOR MONITORING AND EVALUATION  Strengthen the capacity of local stakeholders to collect and manage basic and generic data related to vulnerable groups.  Establish clear results frameworks to monitor progress and results of all programs and projects in a comprehensive manner. A hierarchical results framework is required for the National Roma Inclusion Strategy. Results frameworks need to be linked to a clear monitoring plan that includes indicators and data sources and spells out the institutional responsibilities regarding data collection, analysis, reporting, and use in decision making. Public disclosure of these data and reports will increase the transparency of decisions and use of resources.  Compile rigorous evidence on “what works” and make it readily available. It would be advisable to undertake selected rigorous impact evaluations of some of the key strategic programs (e.g., early childhood development, skills training), including pilot phases, to learn which are successful and cost-effective for Roma inclusion. The knowledge could also be shared at the Council of Europe’s Good Practice for Roma Integration Portal (http://goodpracticeroma.ppa.coe.int/en).  Involve Roma communities and civil society organizations in monitoring the progress and results of inclusion programs and projects. Communities often know best what is happening to households in their areas. An online platform for citizen feedback, for example, could enable citizens to provide feedback on service quality issues that they encounter. This will provide them with the opportunity to send real-time feedback, for example, by SMS messaging. These platforms can be combined with online maps of interventions and projects that show their location and funding flows. In order to enable people without access to computers or the Internet to provide feedback, the platform could be complemented by a telephone-based interface. BOX 4.5: EVIDENCE FOR WHAT WORKS IN EDU CATION: THE “WHAT WO RKS CLEARINGHOUSE ” In 2002, the US Department of Education’s Institute of Education Sciences created the What Works Clearinghouse (WWC) (http://ies.ed.gov/ncee/wwc). The goal was to have one central source that could provide scientific evidence on the impact of different education initiatives. The WWC reviews the research on the impacts of programs, practices, and policies in education and identifies what works in order to allow educators to make evidence-based decisions. For example, an educator interested in the effectiveness of interventions for dropout prevention can get a quick overview—a score card—and more detailed information from a single Web page. The clearinghouse currently reviews programs in the areas of children and youth with disabilities; dropout prevention; early childhood education; English language learners; literacy; math; science; and student behavior. 38 39 ANNEX A. METHODOLOGY AND DATA SOURCES This report offers evidence-based policy advice. The assessment relies on three main sources of information. First, it takes advantage of the Regional Roma Survey (RRS) carried out in 2011 by the United Nations Development Programme (UNDP), the World Bank, and the European Commission (EC).43 This is the most comprehensive survey effort to date to capture the situation of Roma in Central and Eastern Europe. The assessment also includes information from Household Budget Surveys (HBS) conducted in Romania each year from 2008 through 2012 (Box A). Second, this report relies on qualitative information collected through field visits and interviews with stakeholders, including government of Romania officials, Romanian academics, civil society actors, and representatives of Roma communities. Third, each of the chapters highlights relevant international experiences that can provide useful perspectives and best practices for policy formulation on Roma integration. Many of the experiences with integrating poor and marginalized communities elsewhere provide reasons to be optimistic that Roma integration does not have to be a distant goal for Romania—if discrimination against Roma can be addressed. The UNDP/World Bank/EC Regional Roma Survey is the main data source for this report. This comprehensive survey is representative of approximately 89% of the Romanian Roma population, including Roma living in mixed, separated, and segregated neighborhoods. The survey questionnaire was designed by the World Bank and UNDP in partnership. It was implemented by UNDP through the IPSOS polling agency in May–July 2011 on a random sample of Roma living in communities with concentrated Roma populations in Bulgaria, Romania, Hungary, the Slovak Republic, and the Czech Republic. The European Commission Directorate-General for Regional Policy financed the survey. In each of the countries, approximately 750 Roma households (representing over 3,500 individuals) and approximately 350 non-Roma households living in the same neighborhoods or vicinity were interviewed. The sample was not intended to be representative of all Roma in these countries, but rather focused on those communities where the share of Roma population is equal to or higher than the national share of Roma population. This covers 88% of the Roma population in Bulgaria, 90% in the Czech Republic, 78% in Hungary, 89% in Romania, and 83% in the Slovak Republic. Once these communities were identified, a random sample was drawn, and households were randomly sampled within these enumeration areas. The data provide reliable estimates of the conditions in which the vast majority of the Roma in Romania live, and of the conditions of their non-Roma neighbors. Unless otherwise noted, the analysis in this report is based on a comparison of “Roma” and “non-Roma neighbors” living nearby, as identified by the RRS survey enumerators. Comparisons with non-Roma neighbors provide a crucial frame of reference, since the sampled non-Roma households live in the same municipalities as the Roma households and thus share local labor markets, schools, and health facilities as well as other services and collective infrastructure. Hence, if we observe differences in education, health, housing, and employment outcomes between Roma and non- Roma neighbor households, these must reflect particular disadvantages faced by Roma, differences in preferences between Roma and non-Roma, or both. For comparison with the general population in Romania, the report uses the European Union Statistics on Income and Living Conditions (EU-SILC) survey. The EU-SILC does 40 not distinguish between Roma and non-Roma but provides household survey information that is representative of the general Romanian population. The gap between the Roma population and the general population is generally wider than the gap between the Roma and their non-Roma neighbors, who face the same local conditions. BOX A: ISSUES OF DATA COMPARABILITY BETWEEN THE HOUSEHOLD BUDGET SURVEY AND THE UNDP/WORLDBANK/EC REGIONAL ROMA SURVEY Romania’s Household Budget Survey (HBS) is a nationally representati ve household sample conducted every year between January and December among 30,000 households or 70,000 individuals. The HBS questionnaire includes ethnic affiliation that is used to identify “self- declared” Roma individuals. According to HBS 2012, 3.2% of the total population in Romania declares itself Roma. The UNDP/World Bank/EC Regional Roma Survey (RRS) is a comprehensive survey on the living conditions of the Roma. It was implemented between May and July 2011 in six Eastern European countries, including Romania. The sampling framework includes Roma living in mixed, separated, and segregated neighborhoods. In each country, approximately 750 Roma and 350 non-Roma households living in the same neighborhoods or vicinity were interviewed. The sample is representative of communities where the share of the Roma population is higher than the national share of Roma population (covering 89% of the Roma in Romania). The survey includes third-party identification for ethnicity. The main difference between RRS and HBS Roma households lies in their location. RRS Roma are more rural than HBS Roma, and tend to live in compact neighborhoods where the concentration of Roma households is high. RRS and HBS Roma do not differ much in other respects: RRS households have worse access to sewerage and electric goods (rural bias), but display similar access to bikes, telephones, television, bathroom, toilet, and better access to computers and the Internet. Finally, because the RRS social protection module is less detailed than that of HBS, access to social protection is also likely to be underestimated for RRS Roma households. 41 RRS 2011 HBS 2011 Mean St. Err. Mean St. Err. Urban location 0.39 0.49 0.47 0.50 Household size 4.64 2.52 4.16 2.23 Household-head age 45.7 14.9 45.2 14.2 Household-head male 0.74 0.44 0.78 0.42 Household-head secondary education 0.09 0.13 Radio 0.24 0.43 0.45 0.50 Color TV 0.81 0.39 0.91 0.29 Bicycle/motorbike 0.20 0.40 0.24 0.42 Computer 0.16 0.36 0.08 0.27 Internet connection 0.09 0.29 0.06 0.23 Phone (cell/landline) 0.65 0.48 0.67 0.47 Washing machine 0.33 0.47 0.44 0.50 Number of rooms 2.39 1.26 2.24 1.04 Surface of housing (sqm) 48.7 45.0 36.0 19.3 Sewer connection 0.19 0.39 0.41 0.49 Bathroom inside premise 0.18 0.38 0.17 0.38 Toilet inside 0.18 0.39 0.18 0.38 Electricity connection 0.84 0.37 0.95 0.23 42 1 Given the likelihood that Roma substantially underreport their ethnicity on the national census, an alternative set of expert estimates is commonly reported. According to the Strategy of the Government of Romania for the Inclusion of the Romanian Citizens Belonging to Roma Minority 2012 –2020, estimates range from 535,140 (National Census in 2002), to 730,000–970,000 (Romanian government/World Bank survey, “The Roma Communities Social Map,” 2005), to 1,850,000 ( European Commission, An EU Framework for National Roma Integration Strategies up to 2020, based on data from the Council of Europe). 2 World Bank, Roma Inclusion: An Economic Opportunity for Bulgaria, Czech Republic, Romania and Serbia (Washington, DC: World Bank, 2010). 3 The at-risk-of-poverty rate is a key social inclusion indicator in the European Union which indicates the share of people with an equivalized disposable income below 60% of the national equivalized median income. The share of Roma at risk of poverty in Romania is comparable to that in neighboring countries: statistics compiled by UNDP report rates for Romania, Bulgaria, the Czech Republic, Hungary, and the Slovak Republic ranging from 71% (Czech Republic and Hungary) to 87% (Slovak Republic). The estimate for Romania used in this study, 84%, is slightly higher than the UNDP estimate for Romania, 74%. This is caused by two factors: first, the sample used by UNDP differs from the one used in this study. Second, the income levels reported by Roma households were decomposed into “major sources of income” in order to increase the precision of the estimates. The U NDP dataset and the dataset used in this study each used a different approach to deal with missing values within these “major sources of income” question items, creating some level of discrepancy: whereas the UNDP estimate excluded missing values from the calculation, the estimate reported here replaced missing values with the mean value reported by Roma in Romania for each particular source of income —conditional on receiving any income from that particular source. 4 The severe material deprivation rate, as defined by the European Commission, is the share of population living in households that cannot afford least four of the following nine items: (a) to pay rent or utility bills, (b) to keep home adequately warm, (c) to face unexpected expenses, (d) to eat meat, fish, or a protein equivalent every second day, (e) a week-long holiday away from home; or could not afford (even if wanted to) (f) a car, (g) a washing machine, (h) a color TV, or (i) a telephone. 5 The indicator “persons living in households with low work intensity” is defined as the number of persons living in a household having a work intensity below a threshold set at 0.2. The work intensity of a household is the ratio of the total number of months that all working-age household members have worked during the income reference year and the total number of months the same household members theoretically could have worked in the same period. A working-age person is a person aged 18–59 years, with the exclusion of students aged 18–24 years. Households composed only of children, of students aged less than 25, and/or of people aged 60 or more are completely excluded from the indicator calculation. Due to a lack of data on the number of months worked in the previous year, the indicator used here was simplified, indicating whether the ratio of working-age individuals per household who are actually working to the total number of household members who could have worked is over or below 0.2. 6 http://ec.europa.eu/justice/policies/discrimination/docs/com_2011_173_en.pdf. 7 http://www.consilium.europa.eu/uedocs/cms_data/docs/pressdata/en/lsa/139979.pdf. 8 World Bank, Gender Dimensions of Roma Inclusion: Perspectives from Four Roma Communities in Bulgaria (Washington, DC: World Bank, 2014). 9 Roma Early Childhood Inclusion: The RECI Overview Report (Open Society Foundations, Roma Education Fund, and UNICEF, 2012). UNICEF estimates that among those suffering from malnutrition in Romania, 72% are children under age 3. 10 S. Kendall et al., Narrowing the Gap in Outcomes for Vulnerable Groups: A Review of the Research Evidence (Slough, UK: NFER, 2008), http://www.nfer.ac.uk/nfer/publications/LNG01/LNG01.pdf. 11 These results are based on OLS estimations in which cognitive outcomes are the dependent variables. The estimations control for enumeration area fixed effects, which effectively means that the outcomes of Roma children from the same neighborhoods—some of whom participate in the local preschool, while others do not —are compared. The estimations also control for background characteristics of the child (age, gender, hospital birth, general health status); background characteristics of the child’s primary caretaker (age, gender, whether working, preschool attendance in the past, secondary school completion); and quintiles of per capita household income. 43 12 G. Kezdi and E. Suranyi, “Results from a Successful School Integration Program: Evaluation of the OOIH Program, Hungary, 2005–2007” (Hungary: CEU, 2009). 13 A more detailed assessment of the education gap between Roma and non-Roma can be found in the Education chapter of the full study. 14 The labor force participation rate among a certain population group is defined as the share of the working-age cohort that is either employed or looking for work. The working-age cohort includes all individuals aged 15–64. Those who are looking for work are also referred to as “the unemployed.” 15 Fundación Secretariado Gitano, Health and the Roma Community, Analysis of the Situation in Europe: Bulgaria, Czech Republic, Greece, Portugal, Romania, Slovakia, Spain (Madrid: FSG, 2009). 16 G. Dumincã and A. Ivasiuc, “Romii din Romania: De la ţap ispãşitor la motor de dezvoltare” (Bucharest: Agenţia Impreuna, 2013). The Impreuna study is an output of the Roma Observatory Project conducted by Impreuna with support from Germany. In contrast to the Regional Roma Survey, which represents the 89% of Romanian Roma who live in communities with a concentration of Roma higher than the national average, the Impreuna study aims to be representative of all Roma in Romania. At the same time, the sampling method adopted in the Impreuna study decreases the likelihood that Roma living in concentrated communities will end up in the sample. Since this is precisely the group of Roma that generally has worse welfare outcomes, the numbers cited in the Impreuna study should be interpreted as lower-bound estimates. The Impreuna study is used to complement the RRS findings where the latter lacks details on specific subject areas. 17 Data for 1998 indicate that 18% of adult Roma men and 28% of adult Roma women in Romania could not read. C. Zamfir and M. Preda, eds., Romii în România (Bucharest: Editura Expert, 2002). A 2010 survey shows that the situation had not changed much by 2009, when 25% of adult Roma could not read or write. Research Institute for Quality of Life, Legal and Equal on the Labour Market for the Roma Communities: Diagnosis of the Factors Influencing the Employment rate of the Roma Population in Romania (Bucharest: Editura Expert, 2010, for project L@EGAL 2 – European Investment for the Future of the Roma from Romania). Census data from 2011 show that the illiteracy rate among Romanian Roma over the age of 10 is 14.1%. 18 E. Zamfir and C. Zamfir, eds., Ţiganii între ignorare şi îngrijorare (Bucharest: Editura Alternative, 1993). 19 According to the Impreuna study (see note 16), twice as many Roma as non-Roma live in houses made of poor-quality materials, and 50% of the Roma live in lower-comfort-ranking dwellings. 20 Most microfinance institutions require either collateral in the form of a property title (for an individual loan), or a group guarantee (for a group loan). 21 See, for example, H. de Soto, “The Mystery of Capital,” Finance & Development 38, no. 1 (2001), http://www.imf.org/external/pubs/ft/fandd/2001/03/desoto.htm. 22 Impreuna study (see note 16). 23 Ibid. 24 According to UNICEF, for example, interventions in early child development could yield average returns of four to five times the amount invested (http://www.unicef.org/earlychildhood/index_40748.html). 25 European Commission, Compendium of Practice on Non-Discrimination/Equality Mainstreaming (Luxembourg: European Union, 2011), 9. 26 Two well-known models are Becker’s “taste model,” whereby discrimination arises because employers and workers have a “distaste” for working with people from different ethnic backgrounds, and Arrow’s “ignorance model,” whereby discrimination occurs when the employer, through ignorance or prejudice, assumes that certain groups of workers are less productive than others and thus is less willing to employ them. G. Becker, The Economics of Discrimination (Chicago: University of Chicago Press, 1957); K. Arrow, The Theory of Discrimination (Working Paper 403, Princeton University, Department of Economics, Industrial Relations Section, 1971). 27 A qualitative study was carried out in four localities in Romania: Bucharest, Telechiu village/Bihor County, Măieruş village/Braşov County, and Olteniţa town/Călăraşi County. The locations were selected to represent socioeconomic and spatial (rural, urban, integrated, segregated) differences among the Roma. 28 Qualitative research, September–October 2013. 29 M. Wamseidel, E. Vincze, and I. Ionescu, Roma Health: Perspective of the Actors Involved in the Health System – Doctors, Health Mediators, and Patients (Bucharest: Romani CRISS, 2012). 44 30 In a survey of 402 Roma in Romania, Bulgaria, Slovak Republic, Czech Republic, and Hungary, 64% of working-age Roma reported that they had experienced discrimination. An alarming 49% indicated that employers had openly said that they were treating them differently because they were Roma, and a further 5% heard the same from labor offices. European Roma Rights Centre, The Glass Box: Exclusion of Roma from Employment (Budapest: ERRC, 2005). 31 Interview with representative of Goodbee, a non-banking financial institution specialized in microfinance and social enterprise financing, August 5, 2013. 32 C. Hanganu, “Cutting Edge: The Romanian Press and Roma, 1990–1994,” Roma Rights Journal (ERRC), December 7, 1999. 33 Ibid. See also G. Bernath and V. Messing, Pushed to the Edge: Research Report on the Representation of Roma Communities in the Hungarian Majority Media (Budapest: Central European University, 2011); G. Fleck and C. Rughiniş, eds., Come Closer: Inclusion and Exclusion of Roma in Present-Day Romanian Society (Bucharest: National Agency for Roma, 2008). 34 National Council for Combating Discrimination, Discrimination in Romania: Perceptions and Attitudes (Bucharest, 2009). 35 Impreuna study (see note 16). 36 Information provided by Professor Gheorghe Sar ău in a Yahoo group message on August 17, 2013 (Romania_EU_list@yahoogroups.com). Professor Sarău is active in the Romanian Ministry of Education and is responsible for, among other things, the training of Roma school mediators and Romani teachers. 37 Civil Society Monitoring on the Implementation of the National Roma Integration Strategy and Decade Action Plan in 2012 in Romania (Budapest: Decade of Roma Inclusion Secretariat Foundation), http://www.romadecade.org/cms/upload/file/9270_file24_ro_civil-society-monitoring-report_en.pdf. 38 The European Union’s 10 Common Basic Principles on Roma Inclusion recommend “explicit but not exclusive” targeting of measures, “focusing on Roma people as a target group but not to the exclusion of other people who share similar socio-economic circumstances” (Principle No. 3). “Policies and projects should be geared towards ‘vulnerable groups’, ‘groups at the margins of the labour market’, ‘disadvantaged groups’, or ‘groups living in deprived areas’, etc. with a clear mention that these groups include the Roma” (Principle No. 2). http://www.coe.int/t/dg4/youth/Source/Resources/Documents/2011_10_Common_Basic_Principles_Roma_Inclusio n.pdf. On the other hand, the 2011 Regional Roma Survey findings demonstrate that there are considerable outcome gaps in basically all human development areas between Roma and non-Roma living in similar socioeconomic conditions, which implies that in some areas of intervention such as education, policies and programs would benefit from ethnic targeting (notwithstanding the difficulties of measurement due to the lack of ethnic data collection). 39 A study was conducted by the World Bank in the context of the Assessment of the Communication and Collaboration between the Managing Authority and Intermediate Bodies of the Regional Operational Programmes and Facilitation of Proactive and Direct Support for Beneficiaries. It indicates that while many local communities are capable of submitting funding proposals, they often lack the capacity to design and implement projects. Capacity gaps include defining or verifying the technical specifications of investments and undertaking procurement procedures in line with the regulations of EU funds. 40 Teach for Romania is the Romanian adaptation of Teach for America, which recruits and selects university graduates to teach in the nation’s most challenging schools. Teach for Romania was initiated earlier but stalled due to administrative obstacles. 41 This non-segregation principle is also reflected in Article 7 of the European Regional Development Fund (ERDF) Regulation, which was amended in May 2010 to encourage the use of ERDF for housing interventions, insofar as they promote integration and prevent isolation and exclusion of marginalized communities. 42 More concrete policy recommendations are found in Final Report: Component II of the Assessment of the Communication and Collaboration between the Managing Authority and Intermediate Bodies of the Regional Operational Programmes and Facilitation of Proactive and Direct Support for Beneficiaries, prepared by the World Bank for the Romanian Ministry of Regional Development and Public Administration in December 2013. 43 For a presentation of summary findings, see The Situation of Roma in 11 EU Member States: Survey Results at a Glance, prepared by the European Agency for Fundamental Rights and United Nations Development Programme for the European Commission (Luxembourg: European Union, 2012). 45