83117 “A Good Start (AGS): Scaling-Up Access to Quality Services for Young Roma Children” Case Study of the Roma Education Fund Pilot Project The World Bank Europe and Central Asia (ECA) June 11, 2013 1 This report was prepared, in alphabetical order, by Rosen Asenov (World Bank Consultant), Marijana Jasarevic (Operations Analyst), Joost de Laat (Senior Economist), and Barbara Kits (World Bank Consultant), and prepared in collaboration with Judit Kontsekova (Research Associate at the Slovak Governance Institute), Martina Kubánová (Partner and Research Fellow at the Slovak Governance Institute), and Andrej Salner (Partner and Research Fellow at the Slovak Governance Institute). Sujani Eli (ECSHD Program Assistant) provided support. Detailed comments by Roberta Gatti, Sector Manager and Lead Economist, ECSH4, and peer reviewers Ms. Aglaia Zafeirakou, Senior Education Specialist (HDGPE) and Ms. Sophia V. Georgieva, Social Development Specialist (ECSSO) are gratefully acknowledged. World Bank ECA Vice President : Phillippe H. Le Houerou World Bank ECCU5 Country Director : Mamta Murthi World Bank HD Sector Director : Ana Revenga World Bank HDE Sector Manager : Roberta v. Gatti World Bank Task Team Leader : Joost de Laat This document has been prepared for the European Commission; however, it reflects the views only of the authors, and the Commission cannot be held responsible for any use which may be made of the information contained therein. This report draws in some parts on the results of the UNDP/World Bank/EC regional Roma 2011 survey. However, the report was prepared by the World Bank and does not necessarily represent the views of UNDP. 2 TABLE OF CONTENTS List of Figures ............................................................................................................................................... 4 List of Tables ................................................................................................................................................. 5 List of Boxes ................................................................................................................................................. 5 Abbreviations ................................................................................................................................................ 5 1 Introduction ........................................................................................................................................... 6 2 Methodology ......................................................................................................................................... 7 2.1 Introduction to the AGS Project ....................................................................................................... 7 2.2 Sources of Information for the Case Study ...................................................................................... 9 3 Background Characteristics of AGS Beneficiaries ............................................................................. 13 3.1 Community and Household Conditions of AGS Beneficiaries ...................................................... 13 3.2 Learning Opportunities Provided in the Home Environment ......................................................... 18 4 AGS Activities and Child Outcomes................................................................................................... 20 4.1 AGS Activities: Participation Rates ............................................................................................... 20 4.1.1 Awareness Raising / Information Campaigns ..................................................................... 23 4.1.2 Removing Cost barriers and/or providing additional incentives ......................................... 27 4.1.3 Making preschools more welcoming and Accessible.......................................................... 28 4.1.4 Supporting Home Parenting ................................................................................................ 33 4.2 Parental Feedback on AGS Activities ............................................................................................ 34 4.3 Measuring Project Outcomes ......................................................................................................... 36 4.3.1 Changes in Preschool Participation ..................................................................................... 36 4.3.2 Changes in Home Learning Activities................................................................................. 42 4.3.3 Learning and Socio-Emotional outcomes............................................................................ 43 5 Discussion ........................................................................................................................................... 45 6 Annex – Additional ECEC Resources................................................................................................. 53 3 LIST OF FIGURES Figure 1: The 16 AGS Localities................................................................................................................... 8 Figure 2: AGS Project Timeline .................................................................................................................. 11 Figure 3: Share of Rural Households per Country: AGS- and Comparison Data ....................................... 14 Figure 4: Segregation among AGS Beneficiary Households (%) ............................................................... 16 Figure 5: Average Number of Household Members & Children (0-7) per Household ............................... 17 Figure 6: Running Water & Electricity Access among AGS Beneficiary Households (%) ........................ 17 Figure 7: Employment Rates among AGS Parents ..................................................................................... 18 Figure 8: Parents’ Secondary School Completion Rate .............................................................................. 18 Figure 9: Language Spoken at Home in AGS Families .............................................................................. 19 Figure 10: Activities that Parents Do with their Children ........................................................................... 20 Figure 11: Country-level AGS participation graphs ................................................................................... 22 Figure 12: Share of Parents Reporting that Preschool is ‘Very Important’: Baseline vs. Endline .............. 26 Figure 13: Parents’ Expectations regarding their Children’s Educational Attainments; Sons and Daughters ..................................................................................................................................................................... 27 Figure 14: Baseline Enrollment Rates ......................................................................................................... 37 Figure 15: Changes in Enrolment during AGS ........................................................................................... 38 Figure 16: Reasons for Sending Children to Preschool (age 3-5) ............................................................... 39 Figure 17: Reasons for Not Enrolling the Child in Preschool ..................................................................... 40 Figure 18: Reasons for Not Enrolling the Child in Preschool: Baseline vs. Endline .................................. 41 Figure 19: Share of Children (3-5 years old) having missed one or more class days at preschool in the week prior to the interview .......................................................................................................................... 42 Figure 20: Share of Children Receiving Teacher Feedback on Performance ............................................. 42 Figure 21: Activities Parents Do with their Children: Baseline vs. Endline ............................................... 43 Figure 22: Children living in families which have no Children’s books at home ....................................... 43 Figure 23: Improvement in Learning Outcomes among AGS Participants (Aged 3-5) between Start and End of AGS. ................................................................................................................................................ 44 4 LIST OF TABLES Table 1: The AGS Household Survey: Baseline and FOllow-up ................................................................ 10 Table 2: The UNDP/World Bank/EC Regional Roma Survey.................................................................... 13 Table 3: Characteristics of Targeted Localities According to Community Assessments ........................... 15 Table 4: Satisfaction of Parents with Information at Community Motivation Events ................................ 35 Table 5: Evaluation of AGS Activities by Parents ...................................................................................... 35 LIST OF BOXES Box 1: Hungary’s Home-School Liaison Program (HSCL) ........................................................................ 31 ABBREVIATIONS AGS - “A Good Start”, EU Roma pilot initiated by the European Parliament and supported by Directorate General for Regional Policy BVL – Bernard van Leer Foundation ECEC – Early childhood education and care ISSA – International Step by Step Association PHF – Partners Hungary Foundation REF – Roma Education Fund SGI – Slovak Governance Institute UiDF – Unity in diversity Foundation UNDP – United Nations Development Programme 5 1 INTRODUCTION This report is a case study of the Early Childhood Education and Care (ECEC) project “A Good Start: Scaling-Up Access to Quality Services for Young Roma Children” (hereafter, ‘A Good Start’ or ‘AGS’), implemented by local partner organizations of the Roma Education Fund. It follows the companion report “Toward an Equal Start: Closing the Early Learning Gap for Roma Children in Eastern Europe” (World Bank, 2012), which provided an assessment of the preschool and early learning gap between Roma and their non-Roma neighboring children using the UNDP/World Bank/EC regional Roma survey (2012) data. The companion report identified four policy measures to reduce the large preschool participation gap and to improve early learning opportunities more broadly for Roma children in Eastern Europe: (1) Awareness raising / information campaigns. For example, providing basic information on the returns to preschool and information on registration procedures and available subsidies. (2) Removing cost barriers and/or providing additional (financial) support. For example, basic material needs support (e.g. clothes), removing fees and paying for school lunches, as well as providing small subsidies conditional on regular attendance. (3) Making preschools more inclusive for Roma children and their parents. For example, removing barriers related to enrolment procedures, creating a more welcoming environment for Roma children in the classroom, and – related – greater involvement of Roma parents. (4) Supporting home parenting aimed at promoting early learning in the home environment, as an area for intervention that complements the preschool-related interventions The companion report also highlights distance and lack of infrastructure as an important barrier affecting a minority of Roma children in some localities in Eastern Europe. The objective of this report is to learn from the practical experiences by the AGS project and thus to provide practical insights for policy makers and program managers designing Early Childhood and Care Services promoting inclusion of Roma children and their families . The approach combines quantitative information obtained from project monitoring data and qualitative information from interviews with project stakeholders. These are described in more detail below. As discussed in more detail below, the AGS was explicitly developed as a pilot project, giving local partner organizations across the 16 localities the flexibility to design and pilot a variety of different approaches to promoting better access to early childhood education and care services. Importantly for the purposes of this case study, for a 6 number of reasons it was not possible to include into this project design a rigorous prospective impact evaluation. First, the self-selection of communities by the local partners was based on a unique set of considerations by local partners that would have rendered an ex-post selection of genuine comparison localities implausible. Furthermore, in several of the localities, local partners were piloting new approaches, modifying these over the course the implementation, And finally, there was great heterogeneity across the different localities, which effectively meant that even if plausible comparison localities could have been found, a rigorous counterfactual impact evaluation would not have been able to link specific outcomes to a specific intervention approach. The Annex at the end of this case study provides an overview of some other practical resources. By no means a comprehensive list, it includes information on Early Childhood Education and Care training materials and projects, which have been designed to benefit young Roma children, as well as examples of information on tools and methodologies to improve the quality of ECEC for minority or other disadvantaged children more generally. REF, in collaboration with its international partners, has also developed several useful resources based specifically on the AGS experience1. These include (i) case studies that were developed to explain the different methods used to help the children and their families, (ii) “A Teacher’s Guide to Good Practices in Inclusive Early Childhood Services”, and (iii) a “Good Practise Guide to Data Collection.” 2 METHODOLOGY This chapter introduces the ‘A Good Start’ (AGS) project and summarizes the case study methodology. 2.1 INTRODUCTION TO THE AGS PROJECT “A Good Start: Scaling-Up Access to Quality Services for Young Roma Children” (hereafter, ‘A Good Start’ or ‘AGS’) took place from June 2010 to June 2012, and included projects in 16 different localities, spread out over four countries: Hungary, FYR Macedonia, Romania and Slovakia (see Figure 1). The project was financed by the European Commission Directorate General for Regional Policy (EUR 2,046,104). To supplement this grant budget, the Roma Education Fund involved a number of other donors who financed complementary activities to the AGS project, totaling EUR 437,712 1 For details, see http://www.romaeducationfund.hu/publications/studies-and-researches-0 7 as follows: Bernard van Leer Foundation (EUR 260,918), LEGO Foundation (EUR 107,660), and the Network of European Foundations (EUR 69,134). It was implemented by REF in cooperation with three international partners: the International Step by Step Association (ISSA), the Spanish Fundacion Secretariado Gitano, and the Slovak Governance Institute (SGI). REF contracted local partner non- governmental organizations (NGOs), which implemented the project at the local level. Although the main project target group consisted of vulnerable Roma, the project also supported poor non-Roma children and families in the same localities. FIGURE 1: THE 16 AGS LOCALITIES Source: SGI, 2012. A The project localities are: Nyíregyháza, Hodász, Kántorjánosi, Nagydobos, Nagyecsed, Nyírkáta (Hungary); Skopje/Šuto Orizari (Šuto Orizari is part of Skopje), Trabotivište, Crnik/Delčevo (data for both neighboring villages are reported separately by REF), Vinica (FYR Macedonia); Telechiu, Craiova-Mofleni (Romania); and Abranovce, Zborov, Banská Bystrica, Martin (Slovakia). The local partners of the Roma Education Fund selected the AGS localities in consultation with REF. Poor localities were selected that REF and/or the local partners was already familiar with. In addition, at least one urban and one rural locality were selected in each country. Apart from these general guidelines, no explicit criteria were used in the selection process. Partner NGO’s were selected based on their experience in dealing with local stakeholders, including local governments, and Early Childhood Education and Care needs in the locality. In addition, partners always had previous experience in promoting the social inclusion and education of Roma 8 children. REF decided to replace two of the initially selected local partners after the first year of implementation due to unsatisfactory performance (in Nyíregyháza – Hungary and in Craiova - Romania). A menu of activities addressing various potential barriers was designed with the choice of specific AGS activities decentralized to the local NGOs, as was the targeting of children, parents, and other key stakeholders. The different types of activities facilitated through AGS included community motivation events on education and health issues for parents, parenting education, home visits, and assistance to families in enrolling their children to preschools. These also included various forms of support for preschool attendance of children: material support such as clothes, shoes, school supplies and hygiene packages, the facilitation of transport to and from preschool, accompaniment of children to and from schools, and tutoring for pupils attending the first grade of primary schools. As such, AGS activities were targeted at a wide range of actors with local partner NGOs having considerable flexibility in choosing the specific local approach. The choice of activities by the local NGOs, described in more detail below, reflected the local conditions as well as the actual capacities of the NGOs. The complementary expertise of various international partners was brought into the project. The International Step by Step Association (ISSA) is specialized in promotion of quality early childhood education and care (ECEC), and provided trainings. The Spanish Fundación Secretariado Gitano was in charge of policy dissemination. The Slovak Governance Institute´s (SGI) responsibilities included monitoring and evaluation support. SGI worked in close collaboration with the World Bank and UNDP to implement a common locally managed monitoring framework (described below). The latter were complemented with regular REF assessments. 2.2 SOURCES OF INFORMATION FOR THE CASE STUDY The first source of information on AGS consists of two rounds of survey interviews that were carried out with each of the ‘recurring’ project beneficiaries in the 16 project localities. One survey was carried out after2 the start of the program and a second survey at the end of the project. The questionnaire was designed jointly between REF, SGI, and the World Bank. The surveys included information on preschool enrolment, learning outcomes, parental expectations regarding preschool, and home parenting techniques. From the 1,022 families whose children and/or parents were ‘recurrent beneficiaries’ of AGS, 919 were also interviewed in the follow-up survey.3 The term ‘recurrent 2 For project timing reasons, it was not feasible to carry the first survey out before the start of activities. 3 For an additional 103 families, the second round survey could not be carried out, mostly because the family had moved to a different location (59%) or because respondents refused to participate in the second round (18%). The 9 beneficiaries’ refers to individuals and families who either participated in AGS activities for an extended period of time, or who participated in several AGS activities during the lifespan of the project. 4 All beneficiary children were included in the survey. Data was gathered about children aged 0-6, as well as children who had recently had their seventh birthday.5 Within the families surveyed, a total of 1,621 children were within this age bracket during both surveys. TABLE 1: THE AGS HOUSEHOLD SURVEY: BASELINE AND FOLLOW-UP Hungary FYR Romania Slovakia Total Macedonia Completed Baseline Households 355 341 90 236 1,022 Survey Children 573 590 165 453 1,781 Completed Both Households 340 301 74 204 919 Surveys Children 551 530 139 401 1,621 ‘Missing’ as compared Households 4% 12% 18% 14% 10% to baseline survey Children 4% 10% 16% 11% 9% Source: AGS Data. A In the baseline survey, most of the respondents were mothers (91%), although sometimes fathers (5%), grandparents (3%), and other household members (1%) were interviewed instead. In the final survey, respondents were again mostly mothers (94%). In both surveys, the respondent was almost always the person responsible for taking care of the children: 95% in Survey 1, and 99% in Survey 2. B The Baseline Survey was completed from Nov. 2010 to Apr. 2011; the final survey was completed from Apr. 2012 to Jul. 2012. Survey interviews with the project beneficiaries were conducted by local AGS project staff. The reason for having the staff of the local AGS partners carry out the interviews was deliberate as the interviews were an integral part of the needs based assessment that the AGS staff carried out. This may have had two opposing effects on the quality of the data: on the one hand, some responses regarding the AGS project itself may be biased because subjects were aware that the interviewer was also an AGS staff member. On the other hand, through their role as mediators, the interviewers may have gained the trust of community members, possibly allowing for more honest and open responses. Prior to undertaking the survey, each of the interviewers received in-depth training by a SGI-WB-UNDP team. relative sample size per locality is smallest in Romania, which is partly explained by the high number of ‘dropout’ households for the second survey in this country. 4 Other beneficiaries were excluded because the nature of their exposure to AGS tends to be relatively superficial. For example, parents who have attended one community-based motivation event, without any follow-up, may have benefited from the project to some extent, but cannot be considered ‘core beneficiaries’ of AGS. 5 Those who were born after 1 September 2003 were included, whereas those born before this date were left out of the surveyed group of children. 10 Because the first survey data were collected several months after the start of the project activities (v. Figure 3) certain comparisons between the two survey rounds must be interpreted with caution. One example is enrolment rates: in many AGS localities, enrolment rates may have increased due to AGS activities by the time that the first survey was implemented. This issue of timing should be taken into account when interpreting the information. FIGURE 2: AGS PROJECT TIMELINE June-November 2010: Start June 2012: End of AGS of AGS activities AGS Project Nov. 2010 – Apr. Apr. 2011 – Jul. 2012: 2011: Collection of Collection of Endline Baseline Survey Data Survey Data A team from the Slovak Governance Institute (SGI) and the World Bank also collected qualitative information through field visits to each of the localities toward the end of the program. Qualitative interviews and focus groups covered project implementers, parents of children, community mediators (or school/health mediators), family assistants, kindergarten (head) teachers, Roma teacher´s assistants, trainers and facilitators, and local government representatives. In most localities, the team visited all of the project facilities, such as formal and informal preschools and schools. In addition, they conducted focus group discussions and individual interviews. Focus group discussions were conducted mainly with mothers (in a few cases with fathers), kindergarten (head) teachers and college students. Beneficiary parents were (randomly) selected for individual interviews and the implementing organizations were asked to suggest interviewees who could be of interest such as successful examples. Finally, a limitation of the current case study is that it lacks detailed qualitative information on how integration operated within the classroom. More systematic evidence in this area can guide inclusion policies. This case study also takes advantage of reports prepared by REF and its (local) partners. This includes narrative reports written by local partners and community assessments carried out before and after the project. The 16 Community Assessments provide data on the preschool infrastructure present in 11 each project locality, both at the start- and at the end of the project.6 Community assessments were conducted by the local partner organizations. The sources of information for the Community Assessment included local authorities, school inspectors, school directors and staff, kindergarten and crèche directors and staff, local Roma leaders and/or leaders of local Roma NGOs, and Roma parents. Where possible, these community assessments are used below to complement information obtained from the survey data. Finally, data from the AGS surveys can be compared to data on the ‘general Roma populations’ from the UNDP/World Bank/EC regional Roma survey (2011) that formed the basis of the companion report. This survey (henceforth: the regional Roma survey (2011)) was designed by the World Bank and UNDP in partnership, financed by the European Commission DG Regional Policy, and implemented by UNDP through the IPSOS polling agency in May-July 2011. Interviews were held with a random sample of Roma in (among others) Hungary, FYR Macedonia, Romania and Slovakia, and represents 78% - 90% of the entire Roma population in these countries. The sample was purposefully not representative of all Roma in these countries, but rather focused on those communities where the share of the Roma population equals or is higher than the national share of the Roma population. In each of the countries, approximately 750 Roma households and approx. 350 non-Roma households living in the same neighborhoods or vicinity were interviewed. Here, we limit the regional survey sample to those households which have at least one child aged younger than 8 – i.e. born after 1 September 2003, to enable making valid comparisons with the group of households sampled in the AGS survey. 6 The Community Assessments include information on the numbers of Roma- and majority population in each locality, the numbers of Roma children of eligible age to attend pre-school and primary school and those that are attending in practice. Information is provided across each age group and grade. They also give information about the number and types of education facilities in an area, who maintains them (e.g. state/ religious/ private) and what capacity they have. Information is also included on the numbers of segregated classes and schools, and the number of Roma children attending these (compared to official figures for children who have been assessed as having special needs). The Community Assessments also include information on the number of children commuting to school or to access care services. They also capture data on the availability of health care in the community – such as pre-natal, anti-natal, immunizations and logopaedic services. Lastly, they list how many parents are involved in school/parents' associations and on the number of children targeted through 'A Good Start'. 12 TABLE 2: THE UNDP/WORLD BANK/EC REGIONAL ROMA SURVEY Hungary FYR Macedonia Romania Slovakia Total ECD Total ECD Total ECD Total ECD Sample Households Sample Households Sample Households Sample Households Roma households 753 352 788 369 757 401 756 422 Non-Roma neighboring 354 61 358 96 350 57 350 121 households No. of children aged <8 in Roma 18% 586 (31%) 16% 598 (28%) 20% 697 (31%) 20% 719 (33%) sample Urban Households in 35% 34% 88% 85% 39% 38% 27% 27% Roma sample Source: UNDP/World Bank/EC regional Roma survey (2011). A The label ‘ECD households’ refers to households with at least one child in the agegroup 0-7, i.e. born after 1 September 2003. 3 BACKGROUND CHARACTERISTICS OF AGS BENEFICIARIES This chapter sheds light on the ‘baseline’ – i.e. pre-project – conditions in the 16 project localities. The information presented here is based on the community assessments, the country-specific AGS reports, the focus group discussions, and the baseline round of the AGS household survey. 3.1 COMMUNITY AND HOUSEHOLD CONDITIONS OF AGS BENEFICIARIES In each of the four countries, both urban and rural Roma communities were targeted. The graph of Figure 4 shows that according to the AGS survey data, most of the project beneficiaries resided in urban localities in Slovakia and especially FYR Macedonia, whereas in Hungary and Romania, they were predominantly rural. Overall, most of the beneficiary households are urban (59%). Compared to the national level comparison data on Roma households with young children, the urban-rural divide in the AGS target communities is relatively similar in Hungary, FYR Macedonia and Romania, whereas in Slovakia, rural households are under-represented in the AGS sample. Most AGS beneficiary families’ houses are within a 3 km distance from a local kindergarten. The one exception is Slovakia, where this is the case for only about half of the households. 13 FIGURE 3: SHARE OF RURAL HOUSEHOLDS PER COUNTRY: AGS- AND COMPARISON DATA 80 71 70 73 66 Percent (Households) 62 60 37 40 20 15 7 0 Hungary Macedonia Romania Slovakia AGS National Sources: AGS data (AGS); UNDP/World Bank/EC regional Roma survey (2011) (National) In Hungary, the localities were mostly located in the Mátészalka microregion, which is one of the poorest areas of the country. In total, the region counts 9 segregated Roma settlements, among which were two AGS localities: Nagyecsed and Kántorjánosi. In all Hungarian localities, AGS was run in the first year by the National Charity Association for a Better Future and in the second year by Romano Trajo – Gypsy Cultural and Public Association, that worked in co-operation with Unity in Diversity Foundation, Partners Hungary Foundation, and the College of Nyíregyháza. In FYR Macedonia, conditions between the AGS communities were very diverse. Šuto Orizari was the largest locality to participate in the whole AGS project. This municipality is part of FYR Macedonia’s capital Skopje and is one of the largest Roma settlements in Europe, with poor living conditions. On the other hand, Delčevo is an example of a municipality where Roma are relatively well integrated. Unemployment remains very high, but school enrolment and political participation among Roma are high in this locality. In FYR Macedonia, two local partners were responsible for the implementation of AGS: in Šuto Orizari, Skopje, the Roma Education Centre Ambrela was contracted, and in the other localities, the responsible NGO was the Humanitarian and Charitable Association of the Roma KHAM – Delčevo. In Romania, the two AGS localities have about half Roma and half non-Roma inhabitants. Segregation is again widely prevalent, and the main language spoken by Roma in these localities is Romanes. In Telechiu in particular, living conditions are very poor, with hardly any infrastructure, electricity, or clean water provisions in the Roma settlement. Each locality had its own implementing agency: in Mofleni, this was Romani Criss in co-operation with the NGO Trust, and in Telechiu this was the Ruhama Foundation. 14 In Slovakia, conditions between the AGS communities were diverse. In the AGS locality Banská Bystrica, one of the larger cities of Slovakia, both segregated and integrated Roma households can be found, and living conditions among the Roma are diverse. In Martin, one of the two settlements targeted by AGS called Bambusky, living conditions are poor. In the two rural locations in Slovakia, most Roma are segregated from the rest of the village and live in poor conditions. In the Slovak Republic, the implementing agency for Abranovce and Zborov was Equal Chances. In Banská Bystrica, this was the County Association of Roma Initiatives. In Martin, the Cultural Association of Roma in Slovakia was responsible for implementation. The table below provides an overview of the localities based on the community assessment reports prepared by the participating NGOs. These data are rough estimates in some cases but give a sense of differences across the localities. TABLE 3: CHARACTERISTICS OF TARGETED LOCALITIES ACCORDING TO COMMUNITY ASSESSMENTS Coun- Locality Urban / Rural Overall Roma Roma Preschool try Population Population living in Enrolment of (estimate) AGS target Roma Children settlement (ages 3-6) (NGO Est.) 7 Nyíregyháza Urban, norteast 120,000 4800 2,000 missing Hodász Rural, northeast 3503 A 3423 B 3423 B ? 107/200 (54%) A B B? Hungary Nyírkáta Rural, northeast 1729 450 300 76/76 (100%) Kántorjánosi Rural, norteast 2230 A 580 B 580 B ? 71/119 (60%) A B B? Nagyecsed Rural, norteast 6797 1600 1316 137/160 (86%) A B B? Nagydobos Rural, norteast 2315 620 138 72/73 (99%) Šuto Orizari- Urban 22,017 13,342C 100/1500 Skopje FYR Macedonia Vinica Urban, northeast 19.938 1230C missing 15/67 C D Delčevo Urban, northeast 17,726 651 // 1,000 missing 19/72 Trabotivište Rural, northeast 542 112 missing 0/5 Crnik-Pehcevo Rural, northeast 1,030 350 missing 7/11 Mofleni Urban, south 1,745 800 59/72 Craiova (incl. 20,000 Romania Urban, south 298,928 400/2300 Mofleni) (Craiova total) Telechiu- Rural, northwest 800 450 375 5/37 (14%) Tetchea 7 The two target localities of AGS in the city are called Huszár lakótelep and Keleti lakótelep. 15 commune Abranovce Rural, north-east 624 210 est. 210 12/23 Zborov Rural, north-east 3104 1480 missing 9/160 Slovakia Banská Urban, north 80446 missing Missing 10/43 or 87? Bystrica Martin Urban, north 57500 1500 Missing 20/74 (27%) A B Census estimates: HCSO, 2001. Information obtained from the municipality (2010). Census estimates (2002). D C Information obtained from local NGO’s. The AGS localities in each of the countries are characterized by high numbers of Roma inhabitants and segregation is higher among AGS beneficiaries than for the average Roma, except FYR Macedonia. Segregation is confirmed by both the community assessments – in which segregated settlements are identified in many of the AGS localities – and the AGS survey data. Spatial segregation is low in FYR Macedonia (2% of all households), but much higher in the other three countries (48% - 60%). Overall, 78% of all recurring beneficiary households declare that all or almost all of their neighbors are Roma, which is higher than the national comparison data on Roma in three of the countries. FIGURE 4: SEGREGATION AMONG AGS BENEFICIARY HOUSEHOLDS (%) 100 100 91 87 84 87 90 76 Percent (Households) 80 70 60 63 63 58 60 50 43 43 40 31 30 24 19 20 10 0 0 Rural(71%) Urban (29%) Rural (7%) Urban (93%) Rural (70%) Urban (30%) Rural (37%) Urban (63%) Hungary Macedonia Romania Slovakia Household located in separate settlement outside the town/village (Almost) all neighbors are Roma Source: AGS Data The average beneficiary household has five household members, out of which two children aged 0-7. Household size and household composition among AGS beneficiaries are very similar to national data on Roma families with young children. 16 FIGURE 5: AVERAGE NUMBER OF HOUSEHOLD MEMBERS & CHILDREN (0-7) PER HOUSEHOLD 8 7 No. of Household Members 7 6.2 5.9 6.4 6 5.9 5.4 5.6 5.4 5.4 5.3 6 5.2 5.3 4.5 4.7 5 4.3 4 3 2 1 0 Rural(71%) Urban (29%) Rural (7%) Urban (93%) Rural (70%) Urban (30%) Rural (37%) Urban (63%) Hungary Macedonia Romania Slovakia AGS National Source: AGS Data (AGS); UNDP/World Bank/EC regional Roma survey (2011) (National) In terms of living conditions, most households have access to electricity, but only a slight majority usually has running water inside the dwelling. Water is generally slightly less accessible to the average Roma family with young children in these countries. Overall, however, the situation of an average Roma family is comparable to that of an AGS family in this respect. FIGURE 6: RUNNING WATER & ELECTRICITY ACCESS AMONG AGS BENEFICIARY HOUSEHOLDS (%) 100 94 98 94 100 96 92 100 95 100 Percent (Households) 80 68 70 63 62 56 50 60 42 40 20 0 Rural(71%) Urban Rural (7%) Urban Rural (70%) Urban Rural (37%) Urban (29%) (93%) (30%) (63%) Hungary Macedonia Romania Slovakia Running water inside? Electricity supply? Source: AGS Data The vast majority of parents in AGS beneficiary families are jobless. These levels of joblessness are even higher than among the (national) average Roma family with young children. Among mothers, only 5% has a job, whereas among fathers, this is 19%. A comparison with national survey data shows that employment levels in AGS families are even lower than nationally among Roma families with young children, especially for fathers, suggesting that local AGS partners successfully targeted particularly 17 vulnerable communities and/or households. In Hungary, FYR Macedonia and Romania, employment among Roma fathers in families with young children never reaches below 42% in these comparison data. FIGURE 7: EMPLOYMENT RATES AMONG AGS PARENTS 60 50 50 41 40 Percent 30 22 23 23 23 15 15 17 20 10 7 10 5 3 6 5 1 0 Rural(71%) Urban (29%) Rural (7%) Urban (93%) Rural (70%) Urban (30%) Rural (37%) Urban (63%) Hungary Macedonia Romania Slovakia Father Mother Source: AGS Data All in all, the data presented above show that AGS has targeted families which are even more vulnerable than the average Roma family with young children in these countries. 3.2 LEARNING OPPORTUNITIES PROVIDED IN THE HOME ENVIRONMENT Parents in AGS families generally have low education levels, similar to national averages for Roma. On notable exception is found in Romania, where the rate is much higher in the urban part of the sample. These data are very similar to nationally representative data, except for the outlier in Romania. FIGURE 8 : PARENTS’ SECONDARY SCHOOL COMPLETION RATE 40 35 35 30 25 22 Percent 19 20 15 16 14 14 15 11 9 9 9 8 10 7 6 6 5 5 0 Rural(71%) Urban (29%) Rural (7%) Urban (93%) Rural (70%) Urban (30%) Rural (37%) Urban (63%) Hungary Macedonia Romania Slovakia Father Mother Source: AGS Data 18 The number of children’s books owned by AGS beneficiary families is low, especially in FYR Macedonia, Romania, and Slovakia. This reflects the general situation for Roma in these countries In Hungary, 50% of all beneficiary families own 4 or more children’s books. However, this relatively high number may be influenced by the AGS project itself, given that in Hungary, books were handed out to families at the start of the project – i.e. before the baseline survey was conducted. In the other three countries, at least half of all beneficiary families do not own more than two books. In FYR Macedonia, at least 50% of the surveyed families in fact own no books at all. The language spoken at home among most AGS beneficiaries is the majority language rather than Romanes. This differs from the general situation where a much larger share of families speaks Romanes at home. The exception is Šuto Orizari (Skopje, FYR Macedonia). Hungary is the only country where the majority language is used universally among AGS beneficiary households. In FYR Macedonia, there is also a significant group of beneficiary households that speaks a third language at home (not shown in the graph), instead of either Romanes or the majority language. In this sense, AGS families differ from the average Roma family in these countries: for example, the regional Roma survey (2011) finds that in Romania, the overall share of families that speaks Romanes at home is 44%, and in Slovakia, this share is as high as 73%. FIGURE 9: LANGUAGE SPOKEN AT HOME IN AGS FAMILIES 100% 0 0 0 0 0 0 14 80% 29 6 62 60% 82 81 100 100 67 100 40% 65 20% 38 19 18 19 0% 0 0 0 Rural(71%) Urban (29%) Rural (7%) Urban (93%) Rural (70%) Urban (30%) Rural (37%) Urban (63%) Hungary Macedonia Romania Slovakia Romanes Majority Language Other A Source: AGS Baseline Data. In FYR Macedonia, a relatively large share of beneficiary households speaks a third language. This relates to the specific region in which the targeted communities are located. There was considerable cross-country variation in the learning activities that parents do with their children. On the whole, AGS families devoted more attention to early learning at home than average 19 Roma households in these countries. This selection effect may reflect that participation in the project was voluntary, with those families being more interested in early childhood education and care also being more interested to become project beneficiaries. For all of the activities measured, FYR Macedonia shows the worst outcomes. In general, playing with children is the most common activity, followed by telling stories and drawing. Reading books with children is done less often, which relates to the fact that many Roma parents in the project localities do not have books at home and may not be functionally literate. FIGURE 10: ACTIVITIES THAT PARENTS DO WITH THEIR CHILDREN 100 99 94 92 89 88 85 78 78 75 80 70 68 71 67 66 60 53 47 39 39 40 34 29 20 0 Read books/looked at Told stories to children Played with children Drew things with Taught picture books with children alphabet/counting to children children Hungary Macedonia Romania Slovakia Source: AGS Baseline Survey Data. 4 AGS ACTIVITIES AND C HILD OUTCOMES This chapter elaborates on the specific AGS project activities. 4.1 AGS ACTIVITIES: PARTICIPATION RATES In each of the four countries, AGS included a number of different project activities targeted at parents and/or their children. Participation in these activities among recurring beneficiaries was recorded by the local NGOs in an electronic database. The activities can be grouped as follows:  Generic assistance: Family was visited by AGS staff, who helped them to solve various problems  Parenting classes: Parent(s) attended parenting classes  Child activities outside school: Child attended activities beyond school instruction: e.g. non- formal education at home, activities in community center, extracurricular life skills programme, reading campaigns, after-school tutoring 20  Activities for parents: Parent(s) attended activities organized for them: e.g. community motivation events, Meséd - Your Tale and Home School Community Liaison in Hungary, Mother and child storytelling club in Romania  Material support: Family received material support: e.g. school stationary, hygiene supplies, clothes  Attendance assistance: Child/children received support related school attendance: e.g. (accompanied) transport, tuition fees, school lunches paid by AGS  ID Assistance: Family received assistance with obtaining ID's, as well as registration at medical services and immunization.  Enrolment assistance: Assistance with enrollment into kindergarten/school In addition, AGS sought to strengthen the quality of the education and care services in several localilties, and targeted key local stakeholders in many of the communities. For example, it supported the development of greater preschool capacities for local Roma children in FYR Macedonia and in Slovakia, or providing alternative services (Slovakia). It sought to improve the quality of pre-school services through training of teachers and caretakers, and employing and/or training Roma preschool assistants was also included in the AGS project in FYR Macedonia, Hungary, and Slovakia. It also developed “Toy Libraries” in four localities in the four countries. In addition, local partners sought to work closely with local decision makers and institutions having a stake in education service provision such as the inspection, pedagogical methodological services, and teacher training college (Hungary). As shown in the figures below, the menu of activities (as well as the intensity) aimed at parents differed across the four countries. For example, since ID assistance was not needed in Hungary, the figure below reports “0%” participation for this activity in Hungary. AGS had the broadest scope of activities in Romania, followed by Hungary and Slovakia. 21 FIGURE 11: COUNTRY-LEVEL AGS PARTICIPATION GRAPHS Hungary: Generic Assistance 93 Parenting Classes 0 Child Activities Outside School 66 Activities for Parents 86 Material Support 89 Attendance Assistance 0 ID Assistance 0 Enrollment Assistance 74 0 20 40 Percent 60 80 100 FYR Macedonia: Generic Assistance 55 Parenting Classes 0 Child Activities Outside School 0 Activities for Parents 65 Material Support 23 Attendance Assistance 33 ID Assistance 44 Enrollment Assistance 38 0 20 40 Percent 60 80 100 Romania: Generic Assistance 88 Parenting Classes 76 Child Activities Outside School 91 Activities for Parents 77 Material Support 88 Attendance Assistance 31 ID Assistance 50 Enrollment Assistance 78 0 20 40 Percent 60 80 100 Slovakia: Generic Assistance 78 Parenting Classes 0 Child Activities Outside School 54 Activities for Parents 69 Material Support 100 Attendance Assistance 30 ID Assistance 0 Enrollment Assistance 60 0 20 40 60 80 100 Percent Source: AGS Data. 22 When activities were offered to parents but parents opted out, the most commonly reported reason by parents for not participating was a lack of time. Regarding activities beyond school instruction, such as activities related to learning at home, activities in the community center, reading campaigns or after- school tutoring, participation was particularly low in Romania, where two thirds of the families that did not participate reported that they were away or had no time to let their children participate. Another 17% reported that they had no information on these activities. The same holds for attending parenting classes. In all countries except Slovakia, a lack of time was also the most prevalent reported reason for parents not to participate in other types of activities organized for them – such as community motivation events or story telling clubs. The activities can be further grouped into four categories according to the main recommendations from the companion report: (1) Awareness raising / information campaigns (2) Removing cost barriers and/or providing additional incentives (3) Making preschools more welcoming and accessible for Roma children and parents (4) Supporting home parenting 4.1.1 AWARENESS RAISING / INFORMATION CAMPAIGNS Raising awareness has been an important component of AGS in all four countries. Parents and pre- schools were targeted through home visits, community motivation events, and by providing leaflets and handouts. Home visits were aimed, among other things, at informing parents about the importance of education. During these visits, the need for regular medical check-ups and immunization was also emphasized in Romania and FYR Macedonia, where up to 1000 children were reported to have been vaccinated mainly in Shuto Orizari. Home visits were not only executed directly by staff members of the local NGO’s responsible for implementing the AGS project, but were also conducted by local community leaders (Romania) who were encouraged by AGS staff to reach out to individual families, and by community mediators (Hungary, FYR Macedonia and Slovakia). Home visits also reportedly encouraged regular preschool attendance of children following registration, especially during the wet season (autumn and winter). In addition, home visits allowed mediators to provide information on hygiene and health problems, which were often sensitive topics that could not be treated in more public gatherings such as community motivation events or group meetings at a community center. Furthermore, home visits were 23 reportedly useful for addressing conflicts between parents and preschool teachers, and to help improve the overall relations between parents and teachers. In addition, group sessions were sometimes organized, by inviting the parents of several families to one of the households in the community, thus informing all parents simultaneously. In Romania, in the period prior to enrollment of the children into preschool, the frequency of home visits was increased up to once a day, which, according to various stakeholders had a positive impact on subsequent preschool enrolment rates. In Hungary, FYR Macedonia and Slovakia, mediators were often confronted with a complex demand for information and problem-solving related to the daily challenges faced by AGS beneficiaries. Nevertheless – or perhaps even because of their elaborate assistance, and the trust that was built between mediators and families as a consequence – mediators were reported to have been very successful in raising awareness on the importance of preschool, which, in turn, helped kindergartens to further motivate families for enrolment and regular attendance. The provision of leaflets and hand-outs was reported to be a useful method to target both literate and illiterate parents. In Romania, leaflets, calendars, and other printed objects were distributed among parents, containing not only written text, but also pictures conveying basic messages related to ECD. Since a large share of Roma parents in the two targeted localities in Romania is illiterate, this has proven to be a useful way of making information on ECD more widely accessible. In Romania, to enroll children in kindergarten and school, the local AGS partner also took care of the necessary documentation required for enrolment. Some AGS parents initially did not have identity documents, or the documents had gotten damaged or lost. In such situations, parents were assisted to obtain duplicates or to renew their identity documents. In addition, assistance in acquiring legal documents such as IDs, and birth certificates, was also provided, and registration with a general practitioner was supported as well. Assistance to apply for official documentation and birth certificate and vaccination was also provided in FYR Macedonia Suto Orizari locality. In Slovakia, one of the preschool directors reported that ‘the open form of enrolment’ advocated by AGS was a very effective innovation. Before the start of the AGS project, parents were already being invited to open days, but many did not show up. Measures such as transport provision and awareness raising reportedly helped to solve this problem. Parents were contacted by AGS staff and invited to join an open day at the preschool where they received application forms and where the procedures for enrolling their child were explained. A pre-school director summarized the experience with Open Days at the 24 kindergarten in a following way: “During the enrolment [local NGO representative staff] organizes the parents, we agree on a date, the parents come, we hold an open day so that the parents can see the kindergarten, premises where the child will be, he/she can spend a little bit of time to see the activities taking place at school (…), we hand out applications to the parents and explain what they need to do. Sometimes they are surprised what we have here for the children. They know this preschool exists but many do not know what it looks like, how it is equipped.” In Slovakia, Roma families sometimes faced additional informal barriers, where preschools are reluctant to enroll Roma children. For example, some preschools reportedly charge arbitrary voluntary fees, without clearly communicating to Roma parents that these payments are voluntary. Formal enrolment dates set for a single period in the year by law are sometimes problematic for migrating families. Barriers also exist on the side of the families who mistrust the institutions or are unaware of benefits of preschool attendance for their children. The use of mediators, usually local Roma, sought to effectively overcome many of these information barriers. Finally, preschool facilities were constructed to create spaces. For example, in FYR Macedonia, the locality of Crnik did not have a kindergarten before the start of AGS. In Vinica (FYR Macedonia), new classrooms were built to accommodate Roma children who were not enrolled before the start of the project and basic equipment was also provided. Similarly in Crnik, with project and Pehčevo municipality financial support, the 1st floor of an old building was renovated and adapted to the children ’s need. A parent from Crnik stated: We did not have a kindergarten in the village... the municipality did not help us all these years and our children did not have opportunity to go to kindergarten until now...". In Zborov in Slovakia AGS financed the creation of a new kindergarten classroom and all of the needed equipment. This was not feasible in each location with such needs. For example, in the Slovak AGS localities there were insufficient places in preschools to enroll some of the children younger than five. In one of the urban localities, municipal officials accommodated all Roma children who wished to enroll, including those below the age of five, but this concerned a small number of children. AGS also supported the development and provision of alternative early child development services (e.g. home based service facilitated by Roma mediators). By the end of the AGS project, almost all (100%) parents of recurrent beneficiary families stated that preschool is ‘important’ or ‘very important’ for children aged 0-6. The share of parents reporting that preschool is ‘very important’ increased during the project, as shown in Figure 15. While these results can be biased since the feedback was collected by AGS staff members, the qualitative 25 focus group discussions at the end of the project supports the notion that AGS awareness raising activities increased Roma parents’ knowledge of and interest in preschool. FIGURE 12: SHARE OF PARENTS REPORTING THAT PRESCH OOL IS ‘VERY IMPORTANT’: BASELINE VS. ENDLINE 100 80 60 40 20 0 Hungary Macedonia Romania Slovakia Baseline Endline Source: AGS Data Although parents’ desire for their children to complete secondary school was already high at the outset of the AGS project, this increased further during the project itself. Compared to national data, AGS parents had higher aspirations for their children by the end of t he project. Parents’ desire for both sons and daughters to complete secondary education increased the most in Romania, the country with the lowest baseline levels, and hardly increased in FYR Macedonia. Overall, the share of parents expressing a desire for their children to complete secondary school increased during the AGS project. Whereas at the start of the project, these shares were similar among AGS parents as among nationally representative Roma families, by the end of the project, AGS parents wanted their children to succeed in school considerably more often than parents interviewed in the national comparison data. 26 FIGURE 13 : PARENTS’ EXPECTATI ONS REGARDING THEIR CHILDREN’S EDUCATION AL ATTAINMENTS; SONS AND DAUGHTERS 100 100 90 90 80 80 87 82 83 84 84 85 83 85 83 70 80 82 79 70 60 70 60 71 73 Percent Percent 63 50 50 40 40 30 30 20 20 10 10 0 0 Hungary Macedonia Romania Slovakia Hungary Macedonia Romania Slovakia AGS National AGS National A Source: AGS Baseline Survey Data. Blue bars represent baseline levels for AGS families. The extended lines above the blue bars represent the increase in the share of parents wanting their children to complete secondary education from baseline to endline: i.e. the tops of these lines indicate the share of parents that wanted their children to complete secondary education by the time that the AGS project was concluded. 4.1.2 REMOVING COST BARRIERS AND/OR PROVIDING ADDITIONAL INCENTIVES In each country, cost barriers were addressed by providing children attending preschool with material support, including clothing kits, one hot meal per day, requisites, footwear, diapers for nursery children (0-3 years), and school supply-kits for children in higher age-groups. Offering these at enrolment in kindergarten or in the first grade, at the beginning of the school year, exempts families from otherwise "hidden costs" of kindergarten and school attendance. During field visits conducted by AGS staff, Roma parents often commented that without this support, a lack of money would have prevented them from purchasing the goods required for the education of their children, and thus, from sending their children to preschool. Similar comments made by parents were documented in FYR Macedonia. In Slovakia, mediators sometimes recommended providing aids to the school rather than directly to families. Assistance was organized to directly address missing items required for regular preschool attendance. Some families received shoes, tracksuits, pyjamas and other small items. Many parents stressed that they would not have been able to take their children to kindergarten without the transportation support. In Romania, transportation to and from the preschool was only provided in Molfeini. After the kindergarten would close, children would spend a few hours undertaking activities at the Community Centre and, thereafter, would be provided with transport back home. Transport was also provided in FYR Macedonia. Indeed, during winter, a lack of transportation provisions 27 in some of the AGS localities decreased attendance rates and the dropout rate reportedly increased significantly during this season. In countries where there is a fee for attending preschool, coverage of this fee removed another cost barrier to many parents. In FYR Macedonia, the coverage of kindergarten participation fees secured through the project was reportedly among the most valuable incentives that the AGS project provided. Without such support, many of the targeted children would not have been able to attend preschool: almost all parents said that they cannot continue sending their children to the kindergarten if the project support for covering the participation fee is not continued. In some localities, however, Roma parents were misinformed about the nature and quantity of school fees. In one of the Slovakian localities, pre-schools collected a parental contribution (in some cases up to EUR 17 per month). In formal terms, this contribution was voluntary, but it was understood by parents to be a mandatory fee. In one of the urban localities, the preschool director reported that parents had to pay a fee to the pediatrician for filling out the paperwork for the child to be able to enroll in preschool. Mothers said they sometimes did not have money to cover various fees and that there was often no infrastructure in place to deal with payment arrears. The coverage of payments through the AGS project hence assisted them in financing their children’s access to preschool. 4.1.3 MAKING PRESCHOOLS MORE WELCOMING AND ACCESSIBLE The evidence from AGS localities underscores that interethnic relations and sense of community can be improved with the support of an active kindergarten (or other educational institution) that actively targets and involves all parents without exception. The most successful cases of positive relations between kindergarten representatives and parents were reported in rural Crnik (Macedonia), Abranovce8 (Slovakia) and in particular in Kántorjánosi but also in some other localities in Hungary. “There where I live the situation is much worse [a neighbouring locality not part of AGS], because there is no pre-school institution. I think there are around 60 children in pre-school age and there is no kindergarten for them. Compared to this village it is heaven and earth. (…) Here I see the potential that we achieved during our era, I teach here more than 30 years.” (Kindergarten head teacher, rural locality, Slovakia) 8 In Abranovce (Slovakia) a Roma mother was included into the School board based on the initiative of the kindergarten head teacher with the aim to represent the interests and the opinions of the parents, who otherwise were reluctant to communicate openly in case of some specific problems. The pre-school institution organized also community events where Roma and non-Roma parents participated. 28 “Here [Abranovce] when we have any celebrations both the kindergarten and the primary school holds a presentation. And non-Roma and Roma parents come and participate. In the other village there is no pre- school the primary school is not doing anything and generally the culture of the village is deteriorating.” (Kindergarten head teacher, rural locality, Slovakia) An example of good cooperation among local stakeholders was Telechiu in Romania, where a successful cooperation was built prior to the AGS project. The local NGO was successful in creating a local support group for supporting the Roma community. As a result of this initiative, the AGS project received financial contributions also from local funding and managed to involve several key actors, such as teachers from several institutions, local government representatives as well as religious leaders. “Our collaboration, as outreach workers, returned very good result, between the Director of the school, myself, our school mediator, the healthcare mediator and others. The positive effects do not refer solely to the Roma community, but to the entire village.” (priest, Reformed Church, Telechiu) Another example of good cooperation was the partnership with local municipalities in Eastern Macedonia where some of the municipalities were willing to take over expenses associated with the project implementation. For instance Delchevo municipality paid for the transportation of children from a neighbouring locality, financial contribution was made when building an extension of 100m2 in Vinica, where also the work of a caretaker was financed. In adttion in Šuto Orizari, Macedonia, cooperation was established with several Ministries, including Ministry of Labour and Social Policy that targets the problems of individuals lacking Identification Documents. More generally, qualitative evidence from the AGS underscores that providing integrated education can be one of the important contributors to building better inter-ethnic relations on the local level. One of the head teachers summarized her long-term experiences and compared two localities; one with an existing pre-school institution and one, where it is missing. She concluded that the level of socialization of the children, their skills, social habits and attitudes are on a much higher level in the community, where a pre-school is functioning as opposed to the other municipality, where the children arrive to primary education without any formal preparation, development and insufficient command of Slovak language. A similar case was observed in the Hungarian locality Kántorjánosi, where Roma children represent about 60% of all children attending kindergarten. 29 However in some localities the cooperation with the local authorities was not successful. In Mofleni, Romania, cooperation with local institutions was reportedly weak and the local Roma community did not receive substantial support from local institutional actors, aside from specific activities carried out by the Roma healthcare mediator, the AGS employed the Roma school mediator and the staff of the local kindergarten. To make preschools more inclusive, various interventions targeted at teachers, local government officials and other relevant non-Roma stakeholders and Roma parents. This included providing activities to raise awareness among key Roma and non-Roma stakeholders toward the importance of participation of all children, including Roma children, in early childhood education and care activities. . Activities with Roma parents included assisting parents with enrolment procedures, inviting parents to come to schools and to participate in school-related activities, and renovate some local kindergartens. For both groups, inclusive activities were also aimed at providing a more accessible route for voicing concerns. Community motivation events reportedly successfully removed some of the communication barriers between the Roma and non-Roma population of the village or town. These events brought together Roma and non-Roma parents and children, the local AGS team and local institutions, thus providing a platform for direct interaction and for sharing experiences. In general, respondents in the qualitative interviews reported that local community centers often played an important role in involving parents in the activities of the center, helping to build closer relationships between parents and teachers, and ensuring a high attendance rate at kindergarten and at school. In Hungary, the community motivation events usually had a ‘professional’ and a ‘cultural’ element, which allowed Roma to present aspects of their traditional culture to the audience. The aim of these events was to provide information on ECEC related topics, but also to support community development. In many community events, parents volunteered to contribute to the preparation, by helping to prepare food or by preparing a small dramatic improvisation or other type of performance. The interviewees appreciated specifically that a space for communication was created between representatives of ECEC institutions, local governments and families. In spite of the positive feedback received on this approach, it remains difficult to attribute actual changes in attitudes to this particular activity. Educators, caretakers and the staff from community centers also received training from the International Step by Step Association (ISSA) on ‘inclusive early childhood education’. These 30 training sessions emphasized diversity – e.g. ‘recognizing, naming and overcoming prejudices and stereotypes’ – and methods for working with children from disadvantaged families. In FYR Macedonia, specific training sessions were organized for community mediators, providing an introduction to kindergarten, its role, aims and programs, and elaborating on expectations from the children, communication with children and parents, how to behave when going to home visits and how to overcome conflicts. For example, in Šuto Orizari, the ISSA representative held mentoring meetings with kindergarten teachers. Topics covered in these meetings were based on pedagogical principles and elaborated on both theory and practice, including: communication with families, creating a stimulating environment for learning, planning and evaluation, inclusion, strategies and methods for learning and interaction between teachers and children, among teachers, and between teachers and parents. In Slovakia, a three-day training was organized and in Hungary, multi-stakeholder workshops were organized on ways of cooperating, speech development for children with different mother tongue, and support to disadvantaged children. BOX 1: HUNGARY’S HOME-SCHOOL LIAISON PROGRAM (HSCL) In Hungary, the AGS introduced the Home School Liaison Program (HSCL) was introduced. The main rationale behind this program was to support cooperation and positive interaction between kindergartens and Roma parents, by allowing the parents to be involved in the educational process. Altogether, 98 Roma parents participated in 64 HSCL sessions in 10 kindergartens. Half-day trainings were organized to support the preparations, in which parents – mostly mothers – and kindergarten teachers participated jointly. The trainers aimed at establishing a common understanding about why it is worthwhile to organize HSCL sessions. They facilitated discussion and the exchange of ideas via interactive small group work. They also made participants aware of the diverse competencies and skills of each and every child. The HSCL trainers highlighted the preschool teachers´ pedagogical skills and encouraged them to use their competencies to provide methodological support for the parents. On the other hand, parents were motivated to bring in themes related to Roma culture. Usually one kindergarten teacher along with one or two parents worked out a common idea in these trainings and these were implemented later in the kindergarten classroom in front of the children. Thus, the parents had the chance to present in the classes that were attended by their own children. The respondents, including parents themselves, also repeatedly highlighted that the parents learned to appreciate and to understand the importance and the seriousness of the work done by the institutions 31 thanks to the HSCL program. One of the kindergarten teachers remarked that “the parents felt the trust“ and “It´s good because it enables the parents to get an insight. (…) They have the chance to see that our work is not easy at all.” (Kindergarten teacher, urban locality, Hungary). And, referring to the Roma parent participation in the preschools, a local government representative in a rural locality in Hungary reported: “It´s a very good element of the project, as the source of all of our problems is lack of communication. And this problem is not ethnical, but essentially human. And if we approach our fellow citizens from the Roma community, and they can show their values, I think that the social acceptance will be better.” Community mediators and teaching assistants reportedly often played an important role in familiarizing parents with preschool, and in facilitating communication between kindergarten teachers and parents. As part of AGS in some localities, kindergartens employed assistants of Roma ethnic origin, usually co-financed by local governments. The Roma nurses and assistants helped to overcome the language barriers, improve communication and cooperation between teachers and parents, and facilitate children’s inclusion. They also acted as a role model for the rest of the community, and conducted home visits, organized community events, and communicated with different institutions and local authorities – including social, welfare and health services. Together with the teaching assistants, mediators reportedly played an important role in ensuring that children regularly attended preschool classes. In Slovakia, parents themselves served as project mentors and assistants in most of the project localities. Roma women from the local community were often employed as family assistants or community mediators. Although they were usually paid relatively low wages, many of them saw their role as mediators as a great opportunity to learn and achieve personal growth and self-confidence. Refurnishing activities also aimed to make the preschool environment more welcoming and accessible. For example, in Romania, the adequate refurnishing of classrooms, the building of a lavatory, as well as the purchase of educational materials, games and toys, made the preschool environment more attractive for children. Although language barriers proved to be an obstacle in communicating with teachers, some parents reported that children started asking their parents to be taken to the kindergarten, where they had access to games and toys they could not enjoy at home and had made friends of the same age. In Slovakia, all preschools participating in the project received Lego toys as part of the project extension financed by the Lego Foundation. In addition, it included training for pre-school teachers and 32 caretakers by Roma trainers. This received uniformly positive responses from teachers, mediators and parents. The Lego was welcomed as a benefit for all children rather than just for the children participating in the AGS project. 4.1.4 SUPPORTING HOME PARENTING Several story-telling and reading activities supported home parenting practices. Many parents did not benefit from reading with their own parents during their youth. In Hungary, one of the AGS initiatives to overcome this hurdle was the Your Tale program, an initiative that invited Roma parents to read stories to young children at the kindergarten. According to some AGS REF and local staff, this component was one of the most successful AGS initiatives. In total, 240 mothers with small children participated in this program, out of 408 recurrent beneficiary families approached at the beginning of the AGS project. Your Tale targeted only female participants who met every week under the guidance of a facilitator. The Your Tale program comprised three phases; emphasis on 1) the reading of tales, 2) writing exercises, and 3) role plays/drama pedagogy when the participants tried to dramatize some of the tales, or their own stories. The mothers were given children’s books every week, which they could take home and read with their children. The Your Tale Program aimed to improve parenting skills, increase demand for further education among parents, increase cohesion in local communities and, to a limited extent, improve interethnic and inter-community relations. A very important side impact of Your Tale activities was the reported increase in self-confidence of participating mothers. This confidence was gained not only in reading and learning together with their children, but also in other aspects of their adult life. For example, participating mothers were reported to have returned to school and successfully looked for employment. For this reason, AGS stressed the term “parental empowerment” in the local discussions.In Romania, story reading sessions were conducted in the Community Center. Participants were mostly mothers, sometimes accompanied by their children. Short thematic stories accompanied by pictures were read, followed by a discussion on the lessons learned from the story. One of the story tellers stated that most mothers had never been told stories during childhood, and that this activity was highly appreciated by parents. Parents and children also provided positive feedback. Literacy classes aimed to empower Roma parents to actively support children in their education, and allowed parents to become more involved in the ECD process. In Romania, a large number of parents were found to be illiterate, which severely limited these parents’ ability to support their children in their education. In addition, these parents were unable to communicate with institutions that provide public services, thus limiting the access of children to these services as well. AGS responded to this 33 situation by providing parents with free literacy classes. These courses aimed to develop adults’ reading and writing skills, especially for parents of the children that were targeted by AGS. The adult literacy courses organized in Telechiu reportedly were welcomed by parents even if these courses were not enough to improve adult literacy. Parents valued learning the alphabet, signing their name, and reading simple one-syllable words. Illiterate parents reported feeling helpless in supporting their children's education and excitedly recounted that basic literacy skills helped them better understand and relate to what their children learn in preschool. AGS also organized schooling specifically for parents, aiming to teach about parenting techniques that support the learning environment at home. In Romania, FYR Macedonia, and Slovakia, sessions were held in the community center, and were attended mostly by mothers. Mothers who attended these training courses reportedly shared the lessons learned with other mothers in the community. However, dissemination of these lessons among fathers or other adults in the household was often found to be defective, mainly due to the perception that children’s education is the exclusive dut y of the mother and not the responsibility of other family members. There were exceptions. The ISSA representative in FYR Macedonia, responsible for the educational workshops with parents, recalled one case when a father told her "...now, when I go to take my child I'm asking if they learned something new, what did you do today...before I was just asking about his behavior and if he ate...". At the same time, the need for these workshops was larger in some localities than in others. For example, in Delčevo, records imply that the parenting workshops were in fact redundant, " due to the previous work done in this municipality through other projects". Similarly, the interest among parents in participation was modest in some localities. The most active were parents of children enrolled in the kindergarten and parents who were assisted by the local partners with some other form of AGS support. In Slovakia, mothers reported that they did not have a babysitter or neighbor who could take care of the children, which impacted their ability to participate. 4.2 PARENTAL FEEDBACK ON AGS ACTIVITIES There is considerable variation in satisfaction by parents that varies by both activity and by country. First, satisfaction with the information provided at community events and home visits is high across the board. At least three quarters of all parents, and in some countries almost 90% or 100% of all parents, attended community motivation events. Parents were asked to rate their satisfaction with the information provided at these events on the various topics listed in the table below. The table reports the share of parents that reports to be ‘very satisfied’ with the information provided. In most cases, this share is larger 34 than 75%, with two exceptions in Romania, where information on ‘further support’ for preschool attendance and on activities organized for parents were not perceived to be as useful. Parents were generally also very satisfied with the support provided by AGS staff during home visits. TABLE 4: SATISFACTION OF PARENTS WITH INFORMATION AT COMMUNITY MOTIVATION EVENTS Hungary FYR Romania Slovakia Macedonia Attended at least one community motivation event 99 73 88 74 Satisfaction with information about enrolment 95 78 84 67 Satisfaction with information about further support for 92 80 42 59 your children related to school attendance Satisfaction with information about activities organized 96 79 59 58 for parents Satisfaction with information about vaccination and 94 82 75 58 health issues Satisfaction with the support provided by the staff 97 80 73 60 during home visits Source: AGS data. Second, satisfaction with the different activities varied by country and by activity. This is shown in the table below. The beneficiary responses provide little guidance, if any, which activities were most useful. In part, this may reflect that there was simply not always a clear delineation between different activities. TABLE 5: EVALUATION OF AGS ACTIVITIES BY PARENTS % of parents rating the activity as ‘most beneficial’ Hungary FYR Romania Slovakia Macedonia Assistance with enrollment into kindergarten 12 74 33 42 Child received support related school attendance: e.g. . 79 48 3 (accompanied) transport, tuition fees, lunches paid by AGS Child attended activities beyond school instruction: e.g. home 17 . 63 30 education, community center activities, extracurricular life skills program, reading campaigns, after-school tutoring. Parent(s) attended parenting classes. . . 8 . Parent(s) attended other activities: e.g. community motivation 45 77 26 28 events, Meséd - Your Tale and HSCL in Hungary, Mother and 35 child storytelling club in Romania. Family got assistance with obtaining ID's. . 98 12 . Family received material support: e.g. school stationary, 45 35 57 44 hygiene supplies, clothes. Family was visited by AGS staff, who helped them to solve 26 29 10 21 certain problems: Generic Source: AGS data. The table reports the share of parents that rated an activity among their top 3 of ‘most beneficial’ AGS activities, conditional on having received this activity – e.g. out of all families having received material support, the table reports the share that reported this material support to be among the most beneficial AGS activities they benefited from. 4.3 MEASURING PROJECT OUTCOMES The surveys measured various intermediate and final outcomes, including enrolment and attendance, receiving feedback by teachers, home learning activities, and learning- and socio-emotional outcomes. The data suggest that enrolment is likely to have increased substantially, while ensuring regular attendance remained a challenge, in part due to external factors influencing attendance. 4.3.1 CHANGES IN PRESCHOOL PARTICIPATION At the start of the AGS project, enrollment among beneficiary children in the age group 3-6 was low in FYR Macedonia and Slovakia, while it was higher in Hungary and Romania. In most countries, these figures were similar to nationally representative data. For example, Roma children in FYR Macedonia and Slovakia also had the lowest enrolment rates in national data – with percentages close to those reported in the graph below. Enrolment rates were much lower in Romania in the national data, however: around 30% rather than over 80% as reported below for AGS communities9. 9 In some cases there may be differences with the enrolment data presented above in the community assessments. The data reported here are those reported by parents of recurring beneficiary children through individual interviews. The community assessments were based on initial estimates by the local partners of all target children. 36 FIGURE 14: BASELINE ENROLLMENT RATES 100 81 82 80 Percent 60 40 15 20 20 0 Hungary Macedonia Romania Slovakia Source: AGS Baseline Community Assessments (2010). Country-averages were calculated using weighted enrollment rates from the individual community assessments. Enrolment rates presented refer to age group 3-6. Community assessment data may capture a broader group of children than only the targeted AGS beneficiaries. As such, differences between these community assessment data and baseline survey data may be the result of either a) a difference in sample composition, or b) the time lag between the community assessments – which were conducted in June 2010 – and the baseline survey – which was conducted starting November 2011. Enrolment among the cohort of 3-5 year olds increased substantially in each of the four AGS countries over the project period, although it is not clear how much of this is due to natural increase as children get older. Even though the baseline survey data presented were collected a few months after the start of the AGS project, enrolment is higher in the endline data. In particular, enrolment increased from 92 to 100 percent in the AGS localities in both Hungary and Romania, from 20 to 52 percent in FYR Macedonia, and from 41 to 66 percent in Slovakia. Of course, enrolment is expected to increase regardless of the project as enrolment increases with children’s ages, which makes it impossible to identify the contribution to enrolment by the AGS project. However, despite the fact that beneficiary children are even more vulnerable than national average Roma, enrolment rates among project beneficiary children compare favorably at the end of the project with national average Roma children of the same age group suggesting that AGS helped boost enrolment. A REF FYR Macedonia country evaluation reports: “Without exception, all interviewed stated that the biggest result from the project was "the inclusion of the children in the kindergarten". Based on the judgments of the kindergarten teachers and directors, observations from the local partners and especially the parents, many of the children have improved their Macedonian language skills, gained hygienic habits and socialized with the other children at acceptable level.” 37 FIGURE 15: CHANGES IN ENROLMENT DURING AGS 100 92 100 92 100 84 80 66 60 52 46 41 40 35 20 20 15 0 Hungary Macedonia Romania Slovakia Baseline (ages 3-5) Endline (Children who were aged 3-5 at baseline) National (Agegroup comparable to Endline) Source: AGS data (AGS); UNDP/World Bank/EC regional Roma survey (2011) (National). A Sample restricted to ages 3-5 at baseline; the same age cohort at endline (now aged 4-7); and ages 4-7 for nationally representative data (comparable to the agegroup captured in the endline survey). At baseline, AGS parents reported sending their children to preschool in order to improve later-in- life opportunities and learning, rather than being motivated by factors such parents’ time management, parents’ view of the teachers, or kindergarten being compulsory. In all of the four countries, the most frequently mentioned reasons for sending young children (aged 3-5) to preschool related to ‘learning opportunities’ and a positive preschool experience on the part of the child. In Romania, the share of children for which parents reported that the child ‘likes to go to preschool’ as a reason for sending the child was much lower than in the other countries. In all countries except Hungary, opportunities at preschool to learn the majority language were also important. 38 FIGURE 16: REASONS FOR SENDING CHILDREN TO PRESCHOOL (AGE 3-5) 90 79 80 70 67 69 70 60 52 53 50 52 48 4948 50 43 36 39 40 34 33 29 26 30 24 21 24 23 14 15 17 20 10 10 12 12 10 10 5 5 2 35 302 0 0 0 Kindergarten is compulsory for meal / other material support My child can play there My child likes to go to I like/trust the teachers My child can learn there chance to succeed in primary My child can establish new It gives me time to work on My child can learn the official Kindergarten provides free My child will have a better state language/regionally dominant language there kindergarten friendships there other matters. school and later my child Hungary Macedonia Romania Slovakia Source: AGS Baseline Survey Data. Among those families where children were not enrolled in preschool, the reasons at baseline for not enrolling the child differed more between countries than the reasons for sending the child to preschool reported above. In Hungary, the most commonly cited reason by AGS parents for not sending a child to preschool was a lack of preschool places (nearby). In FYR Macedonia, a major reason for not enrolling children was cost: 78% of all respondents state that preschool is too expensive, and a slightly lower 57% respond that they do not have money to buy proper clothing and/or shoes for their children. About one fifth (19%) of all children would not be able to understand the lessons due to language barriers. In Slovakia, the most commonly stated reason for not enrolling the child in kindergarten was a desire to stay at home to be with the child (47%). Cost was almost equally important, with 45% of respondents saying that enrollment was too expensive and 23% saying they did not have enough funds to equip their children with clothing or shoes for pre-school. Another oft-cited reason for non-enrollment (31%) was the lack of places in pre-school. Importantly, only 1% of respondents said their children did not like kindergarten. 39 FIGURE 17: REASONS FOR NOT ENROLLING THE CHILD IN PRESCHOOL 100 93 80 Percent 60 47 48 49 42 40 40 27 32 19 24 20 20 11 12 14 13 7 0 0 0 0 0 Cost barriers Preschool is not No places Prefer to keep child Other reasons inclusive available/nearby at home Hungary Macedonia Romania Slovakia Source: AGS Baseline Survey Data. A It should be noted that in Romania, only five parents provided information on reasons why their child was not attending preschool. By the time the AGS project ended, reasons for not sending children to preschool had changed substantially, but the direction of change differs per country. In FYR Macedonia, cost barriers seem to have gone down, and the issue of space constraints seems to have been tackled effectively. On the other hand, a lack of ‘inclusiveness’ is cited as a more prevalent cause of non-enrolment at the end of AGS as compared to the initial situation, suggesting that once cost and space constraints are tackled, this becomes a next critical issue at least for some parents. In Slovakia, on the other hand, cost barriers remained an obstacle, and space constraints actually increased rather than decreased, possibly pointing to effective awareness-raising campaigns but a lack of capacity to respond to increased demand for preschool places. In the other two countries, no data were available at endline as all beneficiary children aged 3-5 in these countries were enrolled in preschool at this time. 40 FIGURE 18: REASONS FOR NOT ENROLLING THE CHILD IN PRESCHOOL: BASELINE VS. ENDLINE 100 93 80 Percent 60 47 48 48 48 49 48 43 42 40 33 32 27 11 12 13 14 20 4 7 11 0 0 Cost barriers Preschool is not No places Prefer to keep child Other reasons inclusive available/nearby at home Macedonia Baseline Macedonia Endline Slovakia Baseline Slovakia Endline Source: AGS data. A Sample restricted to children aged 3-5 at baseline. The change in attendance was also mixed, in part, because the project may have been successful in enrolling more vulnerable children. As can be seen below, the percentage of children having missed one or more class days in the week prior – conditional on enrolment –decreased in the AGS localities in Hungary and Romania. These are the localities which had very high (92%) preschool enrolment at baseline suggesting favorable conditions. In Hungary, the share missing one or more class days reduced from 22 to 18 percent, and in Romania it reduced from 57 to 46 percent. In the AGS localities in FYR Macedonia and Slovakia, which experienced a substantial increase in enrolment from a low baseline, the percentage of children having missed one or more class days in the week prior increased from 46 to 54 percent in FYR Macedonia and from 40 to 55 percent in Slovakia. A possible explanation for these findings is that the AGS project was successful in enrolling more vulnerable children, but the activities were not sufficient to ensure regular attendance. Note that the endline survey was conducted in the late Spring, meaning that attendance was likely lower during the winter months. Basic everyday problems around the Roma settlements are likely to have contributed to the attendance challenge. For example, REF staff reported that in FYR Macedonia attendance was very low during the winter period due to external factors such as heating problems in Šuto Orizari, and irregular transportation in Delčevo and Vinica because of the very bad roads, as well as recurring health problems of the children because of very bad housing conditions. In Slovakia, the distance was a reason why children missed days in a number of localities where there were practical difficulties in reaching the preschools as families reportedly had to walk with small children several kilometres. 41 FIGURE 19: SHARE OF CHILDREN (3-5 YEARS OLD) HAVING MISSED ONE OR MORE CLASS DAYS AT PRESCHOOL IN THE WEEK PRIOR TO THE INTERVIEW 100 80 69 54 57 55 Percent 60 46 46 42 40 40 27 22 18 20 13 0 Hungary Macedonia Romania Slovakia Baseline Endline National Source: AGS data (baseline and endline) and UNDP/World Bank/EC regional Roma survey (2011) for ‘National’. Finally, there was a modest increase in the share of children receiving feedback from their teachers at preschool on their performance in class. At the time of the baseline survey, most, but not all children received feedback from their teachers on their performance in class. In each of the four countries, levels increased during the project, reaching 100% in Romania, and above 90% in the other three. FIGURE 20: SHARE OF CHILDREN RECEIVING TEACHER FEEDBACK ON PERFORMANCE 100 80 91 94 94 81 Percent 60 40 20 0 Hungary Macedonia Romania Slovakia Source: AGS data 4.3.2 CHANGES IN HOME LEARNING ACTIVITIES Learning activities at home increased somewhat, except in Slovakia. Since the beginning of the project there were modest increases in reading books to children, drawing with children, and teaching the alphabet or counting. For example, in Hungary and Romania – the localities where the reading clubs were introduced, experienced an increase in book reading from 78 to 91 percent (Hungary) and from 47 to 58 percent (Romania), while it did not increase in FYR Macedonia and Slovakia. In fact, Slovakia, experienced somewhat of a decrease in home parenting activities done by the end of the project, as 42 compared to the initial situation. Finally, compared to national figures for Roma, AGS communities in some countries perform better, whereas in others, they perform worse. FIGURE 21: ACTIVITIES PARENTS DO WITH THEIR CHILDREN: BASELINE VS. ENDLINE Read Books / Looked at Drew Things with Child Taught Alphabet / Picture Books with Child Counting 91 100 8995 100 78 71 71 78 100 8585 70 66 68 75 80 80 67 65 66 80 66 58 55 55 53 56 60 44 47 60 45 60 44 55 48 39 32 34 42 40 29 26 40 40 29 23 20 20 20 0 0 0 Baseline Endline National Baseline Endline National Baseline Endline National Source: AGS data (AGS Baseline and Endline); UNDP/World Bank/EC regional Roma survey (2011) (National). A Sample restricted to children aged 3-5 at baseline (i.e. these children are approx. 4-7 years old by the time the endline survey was conducted), and a group comparable to the endline survey for the national data (i.e. also agegroup 4-7). The share of children living in a family that has no books for young children at home decreased during the AGS project, except for Romania. Overall, this share went down from over 40% to approximately 25% of all 3-5 year old children. The results are counter-intuitive in Romania, where the share of children living in families without books increased rather than decreased. FIGURE 22: CHILDREN LIVING IN FAMILIES WHICH HAVE N O CHILDREN’S BOOKS A T HOME 100 81 80 64 58 53 Percent 60 48 44 42 41 32 34 40 25 18 13 16 20 2 0 Hungary Macedonia Romania Slovakia Total Baseline Endline National Source: AGS data (AGS Baseline and Endline); UNDP/World Bank/EC regional Roma survey (2011) (National). A Sample restricted to children aged 3-5 at baseline (i.e. these children are approx. 4-7 years old by the time the endline survey was conducted), and a group comparable to the endline survey for the national data (i.e. also agegroup 4-7). 4.3.3 LEARNING AND SOCIO-EMOTIONAL OUTCOMES Learning outcomes among the 3-5 year old age cohort improved. However, since this is to be expected with time, it remains particularly difficult to evaluate to what extent these improvements 43 are a consequence of AGS interventions. As shown below, all the learning outcomes that were measured show improved performance at the end of the AGS project as compared to the baseline. However, since the data presented here reflect one and the same age cohort, alternative factors may (partly) explain this improvement in performance: generally, children learn as they get older, and some children may already have been in preschool before AGS started, allowing them to learn new skills without relying on AGS. FIGURE 23: IMPROVEMENT IN LEARNING OUTCOMES AMONG AGS PARTICIPANTS (AGED 3-5) BETWEEN START AND END OF AGS. 100 94 90 84 83 80 84 86 75 77 80 72 70 67 70 Percent 60 54 47 50 42 40 31 33 30 17 20 14 10 0 Baseline Endline Source: AGS data. A Sample restricted to children aged 3-5 at baseline (i.e. these children are approx. 4-7 years old by the time the endline survey was conducted). 44 5 DISCUSSION This report is a case study of the “A Good Start: Scaling-Up Access to Quality Services for Young Roma Children” (AGS). The AGS was explicitly developed as a pilot project, giving local partner organizations across the 16 localities in Slovakia, Hungary, Romania, and FYR Macedonia the flexibility to design and pilot a variety of different approaches to promoting better access to early childhood education and care services. More than 1,000 families with more than 1,600 children of pre-school age were reached by the project as ‘recurring beneficiaries’, referring to those who either participated in AGS activities for an extended period of time, or who participated in several AGS activities during the lifespan of the project. This case study complements a wealth of international experiences on ECD interventions regarding the integration of most vulnerable communities/minorities, including some developed by REF itself based on the AGS experience. The Annex provides some examples, including references to the other AGS case study materials produced by REF in collaboration with AGS partners10. This AGS case study also complements the companion report “Toward an Equal Start: Closing the Early Learning Gap for Roma Children in Eastern Europe” (World Bank, 2012) , which finds that the inequities start early. Using the UNDP/World Bank/EC regional Roma survey (2011) data, it finds that the preference for education among Roma do not differ significantly from non-Roma of similar socio-economic background: the vast majority of Roma parents wish at least upper secondary for their children. However, the inequities start early: 1 in 5 Roma children 3-5 years old in the Czech Republic and Slovakia enroll in preschool compared with three quarters of the general population. In Romania and Bulgaria, 30-40% of children are enrolled. In Hungary, which has seen greater preschool awareness raising campaigns, including by REF itself, and has put in place a policy to address preschool affordability barriers for poor and vulnerable parents, more than 66% of Roma children are in preschool. Consistent with the international evidence on the benefits to preschool, the report finds that Roma children having attended preschool are much more likely to complete secondary education that Roma children living in the same community with similar socio-economic background characteristics. The qualitative evidence from the AGS case study underscores the findings from the regional survey data and international experiences with preschool returns elsewhere: experience from several of the AGS localities highlights that well designed Early Childhood and Care interventions can plays a pivotal role in reinforcing social links and social capital by reducing segregation and prejudice, while helping 10 http://www.romaeducationfund.hu/publications/studies-and-researches-0 45 young children of disadvantaged communities to grow. In the words of a kindergarten head teacher in rural Slovakia: “There where I live the situation is much worse [a neighbouring locality not part of AGS], because there is no pre-school institution. I think there are around 60 children in pre-school age and there is no kindergarten for them. Compared to this village it is heaven and earth. (…) Here I see the potential that we achieved during our era, I teach here more than 30 years.” This case study provides several valuable lessons for governments/organizations in Europe considering improving Early Childhood Education and Care Services for poor and vulnerable children, Roma children especially. With regards to AGS targeting of recurring beneficiaries, the case study finds that:  The project successfully targeted vulnerable families. The local partner NGOs were responsible for selection of communities in consultation with the Roma Education Fund. Socio- economic information collected on the recurring beneficiaries shows that the AGS targeted communities and families were generally poor and vulnerable, even more so than the average Roma family with young children in these countries as measured by, for example, joblessness.  The project targeted both recurring beneficiaries with substantial preschool enrolment challenges and some without substantial preschool enrolment challenges. Preschool enrolment among recurring AGS children at baseline was comparable to the (low) national average among Roma children for FYR Macedonia and Slovakia, while slightly better than the national average in Hungary, and much better than the national average in Romania. In some cases, the local NGOs may have explicitly targeted families with children already in preschool to provide, for example, complementary home parenting support. With regards to AGS project outcomes, this case study finds that:  Enrolment among the cohort of 3-5 year olds increased substantially in each of the four AGS countries over the project period. In particular, enrolment increased from 92 to 100 percent in the AGS localities in both Hungary and Romania, from 20 to 52 percent in FYR Macedonia, and from 41 to 66 percent in Slovakia, and reached levels well above the national averages for Roma children of the same age. Further, by the time the AGS project ended, reasons for not sending children to preschool had changed substantially, suggesting that AGS was able to address several key barriers for non-enrolment such as lack of spaces or cost barriers.  The change in attendance was mixed. In part, this may reflect a combination of AGS having successfully enrolled more vulnerable children who were particularly susceptible to 46 challenges such as distance. The percentage of children having missed one or more class days in the week prior – conditional on enrolment –decreased in the AGS localities in Hungary and Romania. These are the localities which had very high (92%) preschool enrolment at baseline suggesting favorable access conditions. In the Hungary localities, the share missing one or more class days (during late Spring when the interviews were conducted) reduced from 22 to 18 percent, and in Romania it reduced from 57 to 46 percent. In the AGS localities in FYR Macedonia and Slovakia, the percentage of children having missed one or more class days in the week prior increased from 46 to 54 percent in FYR Macedonia and from 40 to 55 percent in Slovakia. A possible explanation for these findings is that the AGS project was successful in enrolling more vulnerable children, but the activities were not sufficient to ensure regular attendance, in part, because every day external challenges may have been a constraint for vulnerable children.  Learning activities at home increased, except in FYR Macedonia and Slovakia. Since the beginning of the project there were modest increases in reading books to children, drawing with children, and teaching the alphabet or counting. For example, in Hungary and Romania – the localities with reading clubs - experienced an increase in book reading from 78 to 91 percent (Hungary) and from 47 to 58 percent (Romania), while it did not increase in FYR Macedonia and Slovakia where there was a stronger focus on increasing low preschool enrolment rates. With regards to AGS project implementation, this case study finds that:  Strong partnerships with local stakeholders – municipalities, school inspectorates, preschools themselves etc. – were a key determinant for project success and sustainability. For example, in Telechiu in Romania and in some of the municipalities in Eastern Macedonia, local authorities were willing to provide financial support to the project implementation. In Slovakia, one of the preschool directors reported that ‘the open form of enrolment’ advocated by AGS was a very effective innovation whereby parents were contacted by AGS staff and invited to join an open day at the preschool. Fostering greater interaction also limits opportunities for creating perception barriers such as charging arbitrary fees without clearly communicating that these fees are voluntary. In Hungary, referring to the Roma parent participation in the preschools promoted by the Home School Liaison project, a local government representative in a rural locality in Hungary reported: “It´s a very good element of the project, as the source of all of our problems is lack of communication. […]”  Under the guidance of REF, the local NGOs piloted a variety of different activities with varying degrees of intensity from one locality to the next, and varied ways of implementing 47 the activities. Activities included home parenting support, such as the Your Tale reading and story-telling program in Hungary; preschool awareness raising (through pamphlets, home visits, community events); helping with enrolment procedures; providing material support in the form of school stationary, hygiene supplies, clothes, etc.; attendance assistance such as (accompanied) transport, coverage of tuition fees, school lunches paid by AGS; to reaching out to teachers (for example, educators, caretakers and the staff from community centers received training from the International Step by Step Association (ISSA) on ‘inclusive early childhood education’) and promoting parental involvement in the preschool (e.g. the home school liaison program). With regards to AGS project monitoring, this case study finds that:  Local implementing organizations can collect high quality project monitoring data, provided there is sufficient technical assistance provided. The local NGO project partners were asked to fill out the Community Assessment both before project implementation and near the end of the project implementation phase. It focused on the relevant characteristics of the localities, such as the provision of ECEC services and children’s participation in them. After the project start and toward the project end, they were also asked to carry out interviews with each recurring beneficiary family, the so-called Household Survey. The information from the Household Survey was locally entered by the local partners in an electronic database that linked to the internet and accessible by REF project staff. Specialized training support (a few days) and regular follow-up support for the Household Survey by data experts from the Slovak Governance Institute were sufficient to ensure systematic good quality data. The quality of the Community Assessments, which did not receive the same targeted expert support, varied more from one locality to the next.11 What next? The wealth of evidence on the benefits of quality early childhood education and care services, especially for vulnerable children, including Roma, underscores the urgent need to close the gaps that are documented in the UNDP/World Bank/EC regional Roma survey (2011). Two countries, Hungary and Spain, have successfully reduced a considerable part of the gap in access to preschool focusing in part on removing affordability barriers. The AGS experiences in Slovakia, Romania, and FYR Macedonia underscore that addressing this barrier will also likely substantially raise preschool participation for Roma children in other countries. The AGS experience also underscores that building partnerships with local stakeholders at the municipal and preschool level will foster an enabling 11 For additional information, see the “Good Practice Guide to Data Collection” prepared by the Slovak Governance Institute in the context of the AGS project. http://www.romaeducationfund.hu/sites/default/files/publications/ags_datacollectionguide_screen-singlepages.pdf 48 environment that can simultaneously benefit participation in ECEC services and promote sustainability, including by crowding in local government co-financing. And, the AGS experience, especially in Hungary, highlights that families welcome home parenting support. Lastly, the data collection and data entry experience highlights how local NGOs are perfectly capable of implementing a systematic system of beneficiary monitoring when provided the necessary – but limited - technical assistance and support. On the other hand, while the information on enrolment, attendance, and child learning outcomes was encouraging, it also underscores that significant ECEC challenges remained in many localities. Since the goal of the project was to pilot different approaches and to leave a considerable degree of local flexibility, it is not possible to identify which specific intervention, or combination of interventions, is most cost- effective. A natural next step is to choose a subset of the piloted AGS interventions, systematically implement these at much larger scale, rigorously evaluate the impact on key learning and inclusion outcomes, and measure their cost-effectiveness. Going from cost to cost-efficiency is unfortunately not possible in the context of this case study as this additionally requires an impact evaluation linking project outcomes to the specific interventions; the multiple interventions are confounding this, as does the absence of a clear counterfactual comparison group. However, being able to clearly identify the impact of different approaches and measuring cost-effectiveness is very valuable. This approach is common in many parts of the world (see for example the ‘Strategic Impact Evaluation Fund’ (SIEF) www.worldbank.org/sief ), and in Europe too, such impact evaluations are increasingly becoming institutionalized. An example of this is the ‘evidence’ program by the Danish Labor Market Authority (see box below). The Danish Labor Market Authority (LMA) is an example of an institution in Europe that has taken a very proactive approach towards building up and disseminating rigorous evidence in its core area, namely active labor market policies, including for vulnerable groups. Its strategy consists of three complementary activities: (1) collecting existing evidence from research reviews on comparable active labor market programs; (2) developing new evidence through randomized control trials of selected LMA projects; and, (3) disseminate evidence to its affiliated job centers, the Ministry of Employment, and the public at large. Information about job center output is available for everyone on the internet12. 12 www.ams.dk and www.jobindsats.dk 49 Programs aimed at Roma inclusion – including around early childhood education and care - could similarly benefit from systematically implementing on much larger scale well-defined programs whose impact is rigorously evaluated to demonstrate which approach works best. The four country AGS experience raise a number of questions for future evaluation, including:  Which (sub-)set of interventions is most cost-effective in boosting preschool participation? While most local partners implemented a comparable menu of project interventions, the intensity of the different interventions varied considerable from one partner to the other. Does this reflect local conditions or might some activity have a greater impact than another?  How can regular attendance be most cost-effectively promoted? For example, does this require more intense follow up by community mediators or preschool staff; can extra financial support to poor families conditional on regular attendance of the child help; or, some combination?  What is/are the right pre-school curricula to promote inclusion and early learning for vulnerable Roma children? While the above questions focus on participation, there are of course many questions around quality conditional on participation.  What is the most cost-effective way to strengthen home parenting support and how does its impact compare vis-à-vis activities targeted strictly at boosting preschool participation? The Hungary and Romania localities that organized reading clubs experienced an increase in reading at home. What is the best way to organize the reading clubs? For example, should fathers be actively encouraged to participate? Should reading clubs meet more frequently? Meet at the local preschool? Encourage Roma and non-Roma participation? These are important design questions. 50 Furthermore, there are increasingly readily available survey resources to measure child learning and socio-emotional outcomes. For example, the AGS household survey included a child assessment (results highlighted above) which drew on UNICEF’s Multiple Indicator Cluster Surveys (MICS4) as well as the International Step by Step Association child profile.13 A detailed overview of some of the most commonly used tools to measure child development is “Examining Early Child Development in Low- Income Countries: A Toolkit for the Assessment of Children in the First Five Years of Life” (World Bank, 2009). While rigorous impact evaluation around Roma inclusion programs are still rare, fortunately there are some early initiatives doing exactly that: in Slovakia, the Government Office of the Plenipotentiary is partnering with the Slovak Governance Institute, REF, the World Bank, and several academics from Central European University and the Poverty Action Lab to implement and rigorously evaluate the impact of a home parenting support program in 20 communities with funding from the EC DG Employment PROGRESS facility on social policy experimentation; in Romania, the Ministry of Education, in partnership with OvidiuRo NGO, REF, the World Bank, and several academics have designed a preschool participation program and impact evaluation that aims to measure the cost-effectiveness of different approaches involving (combinations of) awareness raising, creating inclusive preschools, and financial assistance support upon regular attendance; and, in Bulgaria, the Trust for Social Achievement is collaborating with the World Bank and Poverty Action Lab to similarly implement and evaluate a large scale program boosting participation through a variety of approaches in more than 150 settlements. Such systematic demonstration programs can show what works and build the necessary public support around proven programs. Finally, while rigorous impact evaluations can be selectively incorporated, the universal use of clear results frameworks can benefit the design of inclusive ECEC programs that seek to scale up the AGS efforts. The AGS experience underscores the diversity of pilot approaches both across and within localities that can be implemented. This is not surprising of course. The objectives by local partners were often defined broadly to strengthen access and quality of ECEC services, covering one or more implicit outcomes such as raising enrolment, attendance, promoting greater cognitive skill development in the preschool, promoting greater socio-emotional development, promoting greater stimulation at home, reducing stereotypes and promoting greater mutual understanding, greater local government participation, 13 The full questionnaire is available at: http://www.romaeducationfund.hu/sites/default/files/publications/ags_datacollectionguide_screen-singlepages.pdf 51 etc. The European Commission14 is currently promoting a greater results focus across Europe whereby programs seek to be much more explicit about the specific outcomes that they intend to impact, including a clear articulation of the theory of change explaining how project inputs (human-, financial- and other resources) will translate into activities, measurable outputs, and ultimately outcomes. This European Commission focus is motivated by the experience that EC financed programs frequently focus on the inputs and the activities, without clearly articulating the ultimate outcomes that the programs seek to impact. The Roma Education Fund is similarly institutionalizing a clearly articulated results focus across REF financed programming. These are important efforts that all governments/organizations seeking to implement quality inclusive ECEC services can adopt. 14 For Example, the EC DG Regional Policy document: The Programming Period 2014-2020. Monitoring and Evaluation of European cohesion policy - European Regional Development Fund and Cohesion Fund - Concepts and Recommendations 52 6 ANNEX – ADDITIONAL ECEC RESOURCES This Annex provides an overview of some additional practical resources15. It includes information on Early Childhood Education and Care training materials and projects, which have been designed to benefit young Roma children, as well as examples of information on tools and methodologies to improve the quality of ECEC for minority or other disadvantaged children more generally. The selected resources appear in alphabetical order. Their selection here in no way implies endorsement by the World Bank. Bernard van Leer Foundation 'Early Childhood Matters' and other publications. Early Childhood Matters is a journal published biannually by the Bernard van Leer Foundation (BVL). The June 2007 issue (number 108), focuses on diversity drawing from BVL and other organisations' experiences. It covers the importance of early childhood development for promoting social inclusion and includes experiences from ECD programme around the world. One example given is from Albania, regarding Partner p r F mij t1 s (“Partners for Children”) setting up a community-based early childhood education and care centres from 2003, called the Gardens of Mothers and Children Centre ('Gardens'). Located in rural North-eastern Albania in areas with limited or no ECEC facilities, these centres are open to children from all religious and ethnic backgrounds, including disabled children and other disadvantaged and marginalised children. Home visits are provided for children who cannot access the Gardens, with their mothers given guidance to support their children's educational development. The article describes how parents and communities were involved in setting up and making the projects a reality, with mothers selected to run the Gardens out of their homes, and fathers also involved in decision-making, and how the centres stress play and experiential learning. Mothers and children are also provided consultations and advice on health. The article also details some of the challenges the NGO faces, and how communities, local authorities and Regional Directorates of Education got involved in supporting the centres, promoting sustainability. In another article, in the same issue of the journal, experiences of ECEC delivery are compared between Northern Ireland and Israel, highlighting how inclusiveness and openness can be encouraged amongst children in the early years even in conflict situations. The authors found a number of themes evident in both locations, suggesting that these could be applied elsewhere. Essentially they are: pre-school curricula that respect diversity; strong partners across a range of stakeholders working towards improved ECEC and inclusion; good engagement and communication with parents; strong evaluative research on the project; and a solid media campaign. 15 Support by Zoe Gray (Consultant) in preparing this Annex is gratefully acknowledged. 53 In addition to this the Bernard van Leer Foundation has published many other reports and publications relevant to ECEC for young Roma children. For example, the Early Childhood Matters issue from June 2008 focuses on quality in Early Childhood Education. Also of interest for practitioners planning to work with parents, is the report Practice and Reflections No 23. Making the road as we go: Parents and professionals as partners managing diversity in early childhood education , from 2008, a case study on the Parents and Diversity project , led by Bureau MUTANT in the Netherlands. See http://www.bernardvanleer.org/English/Home/Publications.html Council of Europe 'Pedagogical fact-sheets on Roma history and culture' According to the website these fact-sheets target teachers, decision-makers, practitioners and students with an interest or working on Roma issues so they can gain knowledge and better understanding of Roma history, culture and language. They cover the origins of the Roma, migrations, the holocaust and policies under communism amongst other topics. They are available in Albanian, English, French, German, Romani, Serbian and Swedish, downloadable from the website. See http://www.coe.int/t/dg4/education/roma/histoculture_EN.asp Council of Europe 'Teaching Kit for Children at Preschool Level 'This educational aid is aimed at Roma children aged 5-7 years who have not attended a nursery school or have fallen behind due to absences or other problems. Part of the Education Program for Roma Children in Europe , it is intended to prepare the children for primary school. This includes supporting the transition from the home to the school environment. It can be used in the family or in informal settings such as community centres. According to the website, the kit focuses on development of intellect such as analytical and reasoning skills, learning how to study, and children’s social and emotional development. There are activities for use by the child on his or her own, and activities requiring an adult s guidance or supervision. It includes literary, scientific and creative activities, and a skills record to assess competencies. It is designed to be culturally relevant so activities and examples are familiar to the children. Resources needed to carry out the activities are inexpensive and easy to get hold of by the families. There is considerable use of pictures and drawings to describe activities, so that they can be used by parents who are illiterate. The Kit is available on CD Rom with fact sheets for the different activities and a handbook to guide use. It can also be downloaded from the website in six parts. See http://www.coe.int/t/dg4/education/roma/preschool_en.asp 54 Diversity in Early Childhood Education and Training (DECET) DECET is a network for organisations and projects valuing diversity in early childhood education and training. The majority of its members are based in Europe. It facilitates knowledge exchange between trainers and practitioners and researches and advocates to promote attention to diversity in early childhood education, emphasising the multiple identities of children and families. It holds conferences and produces publications. Its website includes toys and activities for sale which promote inclusion and diversity for young children. See www.decet.orghttp://www.decet.org/ Education Support UK, 'Together towards inclusion: Pre-school Toolkit for Diversity', Inclusion and Diversity Service, 2010, Northern Ireland. This toolkit has been developed by early years experts to aid pre-schools in Northern Ireland to promote inclusion and respect for diversity for children from different cultures and ethnicities including those whose mother tongue differs from the language used in the pre-schools (English). It is aimed at ensuring equality of access for all children. It focuses on preparing the pre-school for the addition of new students from diverse backgrounds, and helping the teachers in their day-to-day work with the children. There are suggestions on presenting the pre-school setting and visual aids, language development games, methods to monitor language progress, and activities to celebrate differences. There are observation tools with ratings to monitor progress on emotional, social and educational development. It includes practical ideas, such as sample forms to help ease communication between the parents with language differences and the teachers. The final section, 'what next?' focuses on review and action planning to improve attention to inclusion. See http://www.education-support.org.uk/teachers/ids/toolkit-for-diversity-pre- school/http://www.education-support.org.uk/ 'ISSA Principles: Competent Educators of the 21st Century: Principles of Quality Pedagogy ' The International Step by Step Association (ISSA) is a membership organization that promotes equal access to quality education for all children, especially in the early years. ISSA's Principles: Competent Educators of the 21st Century: Principles of Quality Pedagogy, was developed within the network and set out a framework for quality in teaching practices. The principles are provided under seven key themes, and according to ISSA's website are intended as ideals to be followed by professionals, administrators, and policymakers. The themes include planning and assessment, and teaching strategies, and of particular relevance to practitioners working with young Roma: inclusion, diversity, and values of democracy; and family and community. There are indicators accompanying each of the principles to promote measurement. The inclusion principles include promoting equal opportunities irrespective of ethnicity or 55 background, and acceptance and appreciation of diversity. The Principles are downloadable on ISSA's website in English, French and Russian. Further, ISSA has developed a number of accompanying resources on quality pedagogy which are available to its members, including a guidebook on the principles (which includes examples of practical applications), and according to its website is also open to new collaborations on projects, using these resources. See www.issa.nl/resources.html ISSA 'Handbook of ECD Experiences, Innovations, and Lessons from CEE/CIS', 2009 The handbook includes examples from the ISSA network and also from other actors, intended to promote learning and positive practices in early childhood development. Regional-focused and country-specific examples are featured. Examples include community-based early childhood education, parenting skills training, provision of counselling to families, mobile preschools, and policy and advocacy work. Brief information is provided on the projects, such as their aims, main activities, and some highlights or lessons learned. In some cases, personal case stories are given about the child or from the perspective of the educator. Contact details are provided so that it is possible to approach the project implementer for further information. There are examples which are relevant for rural and remote communities, the very poor, the marginalised, and the disabled. One example is the Alternative Social Service Solve the Serious Problem of Roma Drop-out Project in Bulgaria. The Health and Social Development Foundation developed this project to deliver a high quality, culturally-tailored Early Childhood Development programme, particularly to respond to high drop outs rates and infant mortality amongst Roma. In another example, a pre-school teacher in Valmiera in Latvia, relates her experiences working with Roma, Russian and Latvian children, supported by the ISSA partner programme, NGO Center for Education Initiatives. The teacher explains about the different approaches used to promote the culture, language and identity of the minority communities within the pre-school setting. In Javorník in Czech Republic, Ester NGO s project provided accompaniment to pre-school for young children, volunteers and staff acted as teaching assistants, and the NGO promoted enrolment of children to kindergartens. See http://www.issa.nl/docs_pdfs/Handbook%20of%20ECD%20Experiences,%20Innovations%20and%20Les sons%20from%20CEE_CIS%20.pdf OECD 'Starting Strong III: A Quality Toolbox for Early Childhood Education and Care', 2012 This toolbox focuses on quality of early childhood education and care, stating that the benefits of ECEC can not be reaped with access alone, and that if quality is low the effect can be detrimental. The toolbox is aimed at policy makers and other actors working to improve quality of ECEC. It describes five levers to 56 improve quality of early education and care. These are: designing and implementing curriculum and standards, improving qualifications, training and working conditions, setting out quality goals and regulations, engaging families and community, and lastly, advancing data collection and monitoring and research. The toolbox provides possible actions and tools to promote each lever, including useful policy materials, international comparisons and statistics and examples from different country projects. Of particular interest to practitioners working to promote inclusion in ECEC is the section 'Engaging families and communities'. It describes different strategies for working to engage parents from diverse backgrounds. These include providing home visits, assisting parents to provide qualitative home learning environments, and providing support in different languages. In the section 'increasing inequity' it provides useful country examples from around the world, on methods being used to promote engagement with families from minority backgrounds – such as priority access, financial support, home-based parenting, mobile nurseries or 'baby buses', singing and story-telling parent-child groups, adult education and parenting skills, and home visits with parenting information on education, community and social services. There is an online version or the toolbox can be downloaded. See: www.oecd.org/edu/earlychildhood/toolbox Open Society Foundations and International Step-by-Step Association the Romani Early Years Network (REYN) Launched in October 2012, this network is for experts and practitioners working on early childhood education and care issues. REYN is a joint partnership of ISSA and the Roma 'Kopaçi' Initiatives at the Open Society Foundations (OSF) Early Childhood Programme (ECP). The aim of this network is to support and help ECD professionals and other relevant staff working with Roma, such as sharing positive experiences and knowledge, study visits and networking. According to promotional material on the network, the emphasis is on promoting access and equity of care for all young children and quality and professionalism in early years education . There is a blog and a Facebook group for REYN. See for example http://www.facebook.com/groups/Romani.Early.Years.Network/ Roma Education Fund: A Good Start Project research reports and case studies The Roma Education Fund in collaboration with its international partners has developed several useful resources based on the AGS experience. These include case studies that were developed to explain the 57 different methods used to help the children and their families. Authored by independent consultants, one focuses on experiences of mother-child reading and ECD groups, delivered in a number of localities in Hungary, known as MESED or Your Tale , managed by Unity in Diversity Foundation. Another presents lessons from the experience of Ruhama Foundation in Romania, where the key activities of the ECEC project are delivered from a Community Centre. Also useful to programme staff working on ECEC targeting Roma is the “A Teacher’s Guide to Good Practices in Inclusive Early Childhood Services”, authored by ISSA in collaboration with the Roma Education Fund within the framework of the AGS, and a “Good Practise Guide to Data Collection,” developed by the Slovak Governance Institute (SGI) in collaboration with the Roma Education Fund, World Bank, and UNDP in the framework of the AGS. See http://www.romaeducationfund.hu/publications/studies-and-researches-0 Save the Children 'Working Towards Inclusive Practice: Play and Learning Activities' , 2006 This pack formed part of Save the Children’s Early Years Gypsy/Roma and Traveller project, developed to help combat the very low uptake of early years services amongst young children in these communities. It provides information to promote the practitioners' understanding of Roma/ traveller culture and play and learning activities which are culturally appropriate and inclusive of Roma culture. The activities can be used in any early years setting whether they are formal or informal. The pack starts with a useful 'myths' and 'facts' section explaining basic information about the Roma and Travellers, including their way of living, and addresses some of the stereotypes held. Six areas of learning are focused on, in keeping with the UK foundation stage curriculum. These include social and emotional development, communication, mathematics, and physical and creative development. Activities are set out with a clear aim (such as to promote discussion of feelings/ understand differences and areas in common across cultures/ develop counting skills using culturally diverse images), and information on what is needed, what play area the activity addresses and how to plan it. They can be adapted for use easily with children from different backgrounds. Finally there is a play resource section which includes images to be used to support the activities, which are culturally relevant, to help promote a sense of belonging for the Roma/ traveller children and understanding from the other children. See: http://www.savethechildren.org.uk/resources/online-library/working-towards-inclusive-practice-play- and-learning-activities Save the Children 'Working Towards Inclusive Practice: Training', 2006 This is a toolkit was designed to support those involved in delivering training in Gypsy/Roma and Traveller culture to teachers and practitioners involved in early childhood development. According to the toolkit, it aims to examine existing prejudices, promote a better understanding of the communities culture 58 and lifestyle, and provide a better understanding of the difficulties the children of these communities in getting quality early childhood education and care. It also provides methods of representing diversity appropriately in the pre-school setting and information on relevant legislation and government duties. The pack provides information to help the trainer to prepare and deliver the training. As with the publication above each activity is presented with a description of the aim and what is needed. Although this pack is designed for use in the UK many of the exercises are transferable, or could be modified for use in a different location – such as the wall of prejudice exercise, examination of books and storytelling and the home corner activity. See http://www.savethechildren.org.uk/sites/default/files/docs/earlyyears_traintoolkit%281%29_1.pdf Save the Children, 'Steps Towards Learning: A guide to overcoming language barriers in children's education', March 2009 This publication provides explanation of the difficulties that children experience when dealing with a school language which they are not familiar with and provides practical steps to overcome such challenges. The guide is relevant for all levels of schooling including pre-school. It illustrates the extent of the disadvantage that children such as Roma face who are attempting to learn and think through a language which is not their first language, and how much this can prohibit their learning. The guide provides a list of questions to help teachers gain a better understanding on whether a child's difficulties are due to language differences. Then it provides strategies and practical actions to help teachers to deal with this and promote the use of the first language. According to the publication, what can be done will depend on the context but a range of possibilities are put forward with examples from the field. Where there are no resources for a teacher or teaching assistant to provide any first language teaching, there are suggestions for getting members of the community involved in the teaching. The guide recognises the need to learn the second language (for exams and integration issues) and provides information on alternatives for doing this. It provides advocacy arguments and examples of moving to mother tongue education. There is a practical resource section at the end of the publication including details of language programmes and good practice guides. See http://www.savethechildren.org.uk/resources/online-library/steps-towards-learning-guide- overcoming-language-barriers-children’s UNICEF 'Early Child Development Kit Activity Guide' 59 The UNICEF ECD Early Child Development Kit (Activity Guide and Treasure box of Activities) was designed for use in emergencies, however can be used in everyday situations. Many of the games detailed in the guide which is downloadable on the ECD section of the UNICEF website would be suitable or can be modified for use with children from diverse backgrounds, especially the games promoting interactions and friendships. The guide details games and activities for babies, 0-3 year olds and 4-6 year olds which can be carried out with inexpensive easily sourced materials, such as cards, balls, building blocks, puppets, and some activities don't require any materials. The kit and thus the guide can be used with up to 50 children (mainly divided into sub-groups). This can be carried in a school or community centre or similar setting, alternatively parents could use many of the games with a child or children in the home. The focus is on learning through play, and the activities are designed to promote emotional, educational, social, and motor development skills, in particular talking and thinking, interacting with others, and movement. In each activity it describes what kind of learning the caregiver/ person doing the activity should look out for in the child, and in some activities tips are given to ensure the child feels safe and secure and help them to raise any worries or concerns they might have. Children are also encouraged to help each other, problem solve, and learn how to resolve situations through role play and other games. See http://www.unicef.org/earlychildhood/index_52596.html UNICEF 'ECD Resource Pack' This Early Childhood Resource Pack is a training resource for project and programme staff working to deliver quality ECD programmes. It includes advocacy arguments and practical resources. The resource pack is targeted at all communities and is a global document, as such some of the examples and topics are more relevant for Roma and others less so, but many still include useful general information on child development care and project planning and delivery. As part of the resource pack there is an introduction on how to use the materials with specific details on what is needed to take it forward and a powerpoint with tips on facilitating the workshops. There are five modules, according to the authors the appropriate modules and sessions can be delivered based on identified need. For example, module 1 focuses on the child rights approach and where it fits within UNICEF's work, and other key rights issues for children such as including nutrition, water and sanitation and birth registration including what they mean for the children. The birth registration module (1.5), for example, will be useful for practitioners working with Roma children on identity documentation. In the Powerpoint for this session there is an explanation of why birth registration is important, what obstacles there can be in getting a child registered (depending on the context), information needed in developing a birth registration project, and also a case study example from a project promoting birth registration in Africa. In the facilitation document for this birth session there is information on learning objectives, materials required and activities (brainstorming, presentation 60 and group materials). In module 2 topics include investing in ECD and conditional cash transfers, using a case study example from a project, and explanation on early childhood development in relation to the Millennium Development Goals and Education For All. Module 3's focus is on the child, and the role of the family and community in early childhood development. There are three sets of sessions. The first set provides information on marginalised groups, attending to issues in ECD and also minority and exclusion issues. The second set provides information on children’s developmental stages and how they are assessed. The third set focuses on the family. Within this set there are sessions on the role of the father in parenting, what early childhood care is and what it involves including basic information on hygiene and sanitation, maternal care issues relevant to programmes (3.14a). The emphasis of the module is on the role of the parents and family, with sessions for example on supportive family practices (3.15), on positive interaction between the children and their parents (3.16a) and Dialogue with Families and Communities about Early Childhood Development (3.18). Module 4 includes many practical tools for programming. For example there are practical project cycle tools to help programme staff to carry out a situation analysis to establish the particular situation regarding early childhood development in the location or locations they are working (4.1), set priorities in support of early development (4.2), and develop indicators and goals related to Early Childhood Development (4.4). Module 5 focuses mainly on the advocacy and policy context of ECD, providing information and strategies on means to affect change. See http://www.unicef.org/earlychildhood/index_42890.html UNICEF, Open Society Foundation, Roma Education Fund 'Roma Early Childhood Inclusion Overview Report', August 2012 and country reports The RECI reports were developed to provide a better understanding and more detailed picture of the situation for Roma children regarding access and quality of early childhood services in Central and Eastern Europe. In total five country reports have been prepared: the Czech Republic, Hungary, the former Yugoslav Republic of FYR Macedonia, Serbia and Romania. The country reports, which were carried out by local researchers, include available data on the Roma and the perspectives of Roma communities gathered through focus groups and interviews. They also provide examples of positive practices. The RECI Overview Report, is authored by ECEC expert John Bennett who also guided the research of the country reports. The Overview report highlights that although there have been 'remarkable' improvements in legislation in the countries progress in practice is very limited, stating: 'it rarely requires public authorities to take specific actions or to achieve measurable results'. It also stresses the lack of adequate 61 indicators, audit, and evaluation to assess what works and what does not. It emphasises the necessity of governments taking responsibility for policies, and that they should be linked to EU Roma Initiatives. The report calls for higher quality services and greater outreach to Roma parents. Data from the country reports is summarised in the Overview Report with key issues and recommendations, while pointing out that there remains a serious dearth of available disaggregated data to show the true picture for the Roma which must be addressed. See http://www.opensocietyfoundations.org/reports/roma-early-childhood-inclusion-overview- report World Bank, 'Toward an equal start: closing the early learning gap for Roma children in Eastern Europe', Washington D.C, 2012 This World Bank report provides an overview on pre-school participation based on household survey evidence from Bulgaria, Romania, Hungary, Slovakia, and the Czech Republic, all EU member states with large Roma populations. It examines why Roma children aged 3-6 are not accessing pre-school. The findings are based on a new household survey: the UNDP/World Bank/EC regional Roma survey of concentrated Roma communities, conducted in May-July 2011. As such the survey is not representative of Roma across the particular country. The report shows that even at the age of six years, enrolment is no higher than 60%, except for Hungary where the number reaches 90% (where preschool is compulsory, and the government offers support for poor families for school lunches and other expenses, and subsidies for regular pre‐school attendance). Cognitive outcomes for the children included under the survey were considerably poorer than for majority children, however those children who access pre-school have better learning outcomes than their peers who did not. The report highlights four main policy areas to increase enrolment and improve learning at home. These are: inform parents on the benefits of preschool for children’s later‐life outcomes; promote inclusive pre‐schools by reaching out to parents and by involving them more directly in pre‐school with the help of Roma teaching assistants; remove cost barriers possibly coupled with regular attendance subsidies; and finally support parenting at home. Part five of the report highlights some of the projects and actions that are occurring which fit within these themes, drawing lessons from the NGOs and other initiatives experiences, and also expansion of places for children in kindergarten. See http://siteresources.worldbank.org/EXTROMA/Resources/RomaECD_FinalReport.pdf 62