72985
v1
Diagnostics and Policy Advice on the Integration
of Roma in the Slovak Republic
Economic Cost of Exclusion, Employment and Social Protection,
Financial Inclusion, Education, Housing, Health, Monitoring and
Evaluation, EU Financing
Europe and Central Asia Vice President : Phillippe H. Le Houerou
ECCU5 Country Director : Peter Harrold
Human Development Sector Director : Ana Revenga
HDE Sector Manager and Lead Economist : Roberta Gatti
Task Team Leaders : Joost de Laat and Maurizia Tovo
9 September 2012
Europe and Central Asia Region
The World Bank
1
ACKNOWLEDGEMENTS
In October 2011, the World Bank was asked by the Slovak Ministry of Labor, Social Affairs and
Family to analyze the situation of the marginalized Roma community and recommend policy solutions
on how to better include them in the mainstream society focusing on employment and social
protection, financial inclusion, education, health, housing, monitoring and evaluation, and use of EU
financing instruments. This report details the findings. A complementary overview report provides a
summary of these findings. It highlights the urgency of better integrating Roma in Slovakia and offers
specific policy recommendations in each of the sectors. In carrying out its analyses, the World Bank
benefitted from a large household survey sponsored by the European Commission in 2011 and
implemented by UNDP in collaboration with the World Bank, which compensated for the chronic
lack of data on how the Roma live.
The analysis also benefited from numerous discussions with officials from the relevant ministries, the
Office of the Plenipotentiary, various Slovak NGOs, and field visits to Eastern Slovakia. In particular,
it follows team visits in: (i) November 2011 (planning workshop with MoLSAF and PPO); (ii)
December 2012 (technical meetings with staff from ministries, international conference on monitoring
and evaluation, and workshop with civil society); (iii) February 2012 (joint conference on social
benefits system in Slovakia and on the economic cost Roma exclusion); (iv) March 2012 (field visits
and regional workshop in Kosice); (v) April 2012 (field visits and regional workshop in Presov); (vi)
May 2012 (Banking for Progress conference on financial inclusion); and, (vii) meetings with the
relevant line ministries. It also benefited from a separate World Bank task for the Ministry of Labor,
Social Affairs, and Family to analyze the Benefit in Material Need (BMN) and benefited from the
inputs by UNDP and the Open Society Foundations (OSF).
This report was prepared under the direction and guidance of Peter Harrold (Country Director,
ECCU5), Markus Repnik (WB Country Manager, Slovakia), Ana Revenga (Sector Director, ECSHD),
Roberta Gatti (HDE Sector Manager and Lead Economist, ECSHD), and Katarina Mathernova
(Senior Advisor). Joost de Laat (Senior Economist; jdelaat@worldbank.org) and Maurizia Tovo (Lead
Social Development Specialist, mtovo@worldbank.org) were task team leaders.
The report and summary overview were prepared, in alphabetical order, by Ms. Rabia Ali (Consultant
and Health Specialist), Ms. Ellen Hamilton (Senior Urban Planner), Mr. Sandor Karacsony
(Consultant and Financial Inclusion Specialist), Ms. Barbara Kits (Consultant and Data Specialist),
Mr. Joost de Laat (Senior Economist), Ms. Katarina Mathernova (Senior Advisor), Ms. Patrizia Poggi
(Social Development Specialist), Ms. Claudia Rokx (Lead Health Specialist), Mr. Federico Torracchi
(Consultant and Data Specialist), Ms. Maurizia Tovo (Lead Social Development Specialist), Ms. Nora
Teller (Consultant and Housing Specialist), Mr. Michal Vasecka (Professor of Sociology), and Mr.
Ethan Yeh (Economist). Ms. Anne Anglio (Senior Team Assistant) provided support.
The team is grateful to the comments received from the peer reviewers. The integrated report was
reviewed by Theresa (Polly) Jones (Lead Operations Officer, LCSHS) and Alexandre Marc (Chief
Technical Specialist, OPCFN). The economic cost of exclusion chapter was peer reviewed by Carolyn
Turk (Lead Social Development Specialist, ECSS4) and Ambar Narayan (Lead Economist, PRM).
The education chapter was peer reviewed by Toby Linden (Lead Education Specialist, Human
Development) and Harry Patrinos (Lead Education Economist, HDNED), and the health chapter was
peer reviewed by Michele Gragnolati (LCSHD) and Homira Nassery (Operations Officer, HDNHE).
2
TABLE OF CONTENTS
Acknowledgements ....................................................................................................................................... 2
Table of Contents .......................................................................................................................................... 3
1 Introduction ........................................................................................................................................... 9
1.1 The Present Situation Calls for Urgent Action ................................................................................ 9
1.2 Methodology and Data Sources ..................................................................................................... 11
1.3 The Present Situation also Offers Unprecedented Opportunities .................................................. 12
1.4 Overall Policy Recommendations.................................................................................................. 14
2 The Economic Cost of Roma Exclusion ............................................................................................. 17
2.1 Background .................................................................................................................................... 17
2.2 Methodology .................................................................................................................................. 19
2.3 Results ............................................................................................................................................ 22
2.4 Conclusions .................................................................................................................................... 28
2.5 Bibliography .................................................................................................................................. 29
3 Employment and Social Protection ..................................................................................................... 31
3.1 Introduction .................................................................................................................................... 31
3.2 Making Ends Meet: Financial Coping Strategies .......................................................................... 33
3.3 Employment and Welfare .............................................................................................................. 39
3.4 Job Patterns .................................................................................................................................... 41
3.5 Distance to the Labor Market among the Unemployed and Inactive ............................................. 45
3.6 Policy Recommendations............................................................................................................... 49
3.7 Bibliography .................................................................................................................................. 58
3.8 Annex: Estimation results .............................................................................................................. 59
4 Financial Inclusion .............................................................................................................................. 61
4.1 Assessment..................................................................................................................................... 61
4.2 Policy Recommendations............................................................................................................... 66
4.3 Bibliography .................................................................................................................................. 70
5 Education ............................................................................................................................................. 71
5.1 Introduction .................................................................................................................................... 71
5.2 Educational Outcomes – Roma vs. Non-Roma.............................................................................. 72
5.3 Getting off to a good start – Early Childhood Education............................................................... 77
5.4 The Special Case of Special Education.......................................................................................... 83
5.5 Paying for Roma Education ........................................................................................................... 89
5.6 Policy Recommendations............................................................................................................... 93
5.7 Bibliography ................................................................................................................................ 108
3
5.8 Annex ........................................................................................................................................... 111
6 Housing ............................................................................................................................................. 116
6.1 Introduction .................................................................................................................................. 116
6.2 Housing Conditions ..................................................................................................................... 117
6.3 Current Policy Framework in Slovakia ........................................................................................ 125
6.4 Policy Recommendations............................................................................................................. 131
6.6 Bibliography ................................................................................................................................ 134
7 Health ................................................................................................................................................ 136
7.1 Introduction .................................................................................................................................. 136
7.2 Methodology and Key Data Sources ........................................................................................... 138
7.3 Population and Health Outcomes among the Slovak Roma ........................................................ 139
7.4 Determinants of Health Outcomes ............................................................................................... 146
7.5 Current Slovakia Roma Health Policies and Programs................................................................ 161
7.6 Policy Recommendations............................................................................................................. 167
7.7 Bibliography ................................................................................................................................ 174
8 Monitoring and Evaluation ................................................................................................................ 176
8.1 Introduction .................................................................................................................................. 176
8.2 Policy Recommendations............................................................................................................. 176
9 EU Financing for Roma Inclusion ..................................................................................................... 182
9.1 EU Framework for Roma Inclusion ............................................................................................. 182
9.2 Slovakia and Structural Funds 2007-2013 ................................................................................... 184
9.3 EU Funding for Roma Inclusion in the Programming Period (2007-2013 )................................ 185
9.4 Next EU Programming Period (2014-2020) ................................................................................ 187
9.5 Key building blocks of the next programming exercise for the funding period 2014-2020 ........ 188
9.6 policy recommendations .............................................................................................................. 191
9.8 Annex 1: Programming Period 2014-2020 – Structural Funds Programming Flow .................... 196
9.9 Annex 2: Common Strategic Framework – Sources of Funding ................................................. 197
9.10 Annex 3: Common Strategic Framework – Timeline .................................................................. 198
9.11 Bibliography ................................................................................................................................ 199
10 Appendix: Summary of Overall- and Sectoral Policy Recommendations .................................... 200
10.1 Overall Policy Recommendations ................................................................................................ 200
10.2 Summary of Sectoral Policy Recommendations .......................................................................... 202
4
LI S T O F T A B L E S
Table 2-1: Labor Tax Rates ......................................................................................................................... 20
Table 2-2: Monthly Output Gap Per Working Age Roma – 15-64 years .................................................... 22
Table 2-3: Breaking the Output Gap Down Into a Productivity and Employment Rate Component ......... 24
Table 2-4: Overview of Slovak Labor Taxes .............................................................................................. 25
Table 2-5: Monthly Fiscal Gap per Working Age Roma Resulting from Lower Income Tax and Health
Insurance Payments ..................................................................................................................................... 26
Table 2-6: Monthly Fiscal Gap per Working Age Roma Resulting from Lower Corporate Tax Revenues 26
Table 2-7: Monthly Fiscal Gap per Working Age Roma Resulting from Higher Social Insurance Payments
..................................................................................................................................................................... 28
Table 3-1: Average Adult Equivalent Income Levels among Roma and Non-Roma households (by income
quintile) ....................................................................................................................................................... 34
Table 3-2: Combinations of Income Sources among Roma Households (Column Percentages) ........ 36
Table 3-3: Structure of BMN Claims .......................................................................................................... 37
Table 3-4: Structure of Family Allowances ................................................................................................ 37
Table 3-5: Background Characteristics of Households With and Without Employment ............................ 41
Table 3-6: Occupation and Industry of Employed Roma ............................................................................ 44
Table 3-7: Distance to the Labor Market: Roma and Non-Roma Neighbors (% of Working Age
Population) .................................................................................................................................................. 46
Table 3-8: Distance to the Labor Market: Working-Age Roma and Non-Roma Neighbors (Population
Estimate)...................................................................................................................................................... 46
Table 3-9: Distance to the Labor Market: Age Cohorts Among Roma (Population Estimate) ................... 47
Table 4-1: Usage of Savings Accounts and Other Banking Services among Roma Households in Slovakia
..................................................................................................................................................................... 61
Table 4-2: Households Living Close to a Bank Branch (%) ....................................................................... 62
Table 4-3: Arrears on Household Mortgages, Loans and Utility Bills ........................................................ 63
Table 5-1: Highest Level of Education Attended ........................................................................................ 74
Table 5-2: Preschool Net Enrolment Rates ................................................................................................. 79
Table 5-3: Monthly Preschool Expenses by Parents with Children in Preschool, and Provision of Food .. 81
Table 5-4: Reasons for Not Sending Child (3-6) to Preschool .................................................................... 82
Table 5-5: Home Environment .................................................................................................................... 83
Table 5-6: Trends in the Number of Schools and Students ......................................................................... 85
Table 5-7: Special School Attendance by Region and Gender ................................................................... 86
Table 5-8: Public Expenditures for Education, 2008 .................................................................................. 90
Table 5-9: Fiscal Returns to Educational Investments ................................................................................ 92
Table 5-10: Teaching Assistants by Region .............................................................................................. 101
Table 6-1: Size of the Dwelling ................................................................................................................ 119
Table 6-2: Methods of Cooking and Heating the Dwelling (% of Households) ....................................... 121
Table 6-3: Arrears for Housing Related Expenses (% of Households) ..................................................... 124
Table 7-1: Life-Expectancy for Roma and Non-Roma in Selected Countries .......................................... 141
Table 7-2: Monitoring Data Roma Health Mediators/Community Health Workers ................................. 164
Table 7-3: RHM Role and the Corresponding Health Impacts ................................................................. 166
5
LI S T O F F I G U R E S
Figure 1-1: Population Pyramids: Roma (Left Panel) and Non-Roma Neighbors (Right Panel)................ 11
Figure 2-1: Employment Rates among Roma are Much Lower than among Non-Roma ........................... 18
Figure 2-2: Employment Rates by Education Level ................................................................................... 18
Figure 2-3: Total Output Gap Amounts to Euro 3.1 Billion Annually........................................................ 23
Figure 2-4: Total Fiscal Gains of Roma Inclusion ...................................................................................... 24
Figure 3-1: Happiness and Life Satisfaction: Employed Roma Are Happier than Others of Working Age 31
Figure 3-2: Employment Rates – 15-64 years ............................................................................................. 32
Figure 3-3: Participation Rates among Roma men and women (%), split up into employed and
unemployed ................................................................................................................................................. 33
Figure 3-4: Comparison of Income Sources for Roma and Non-Roma neighboring households ............... 35
Figure 3-5: State Benefits Received by Roma and non-Roma Neighboring Households ........................... 35
Figure 3-6: Comparison between Roma and non-Roma households reporting the main sources of income
..................................................................................................................................................................... 36
Figure 3-7: Income shares from Social Transfers, Employment and Pensions – Roma vs. Non-Roma
Neighbors .................................................................................................................................................... 40
Figure 3-8: Employment Rates, by Region and Gender, Roma and non-Roma nearby .............................. 42
Figure 3-9: Contract Duration among Employed Roma and Non-Roma Neighbors .................................. 44
Figure 3-10: Types of Employment among Roma and Non-Roma Neighbors ........................................... 45
Figure 3-11: Average Number of Years Work Experience: Roma ............................................................. 47
Figure 3-12: Secondary School Completion and Distance to the Labor Market ......................................... 48
Figure 3-13: Adult Learning Course / Professional Apprenticeship and Computer Literacy ..................... 48
Figure 4-1: Roma Households Have a Much Lower Self-Reported Ability to Face Unexpected Expenses
Without External Financial Support ............................................................................................................ 62
Figure 5-1: Enrolment Rates among Roma and Non-Roma by age group .................................................. 73
Figure 5-2: Secondary School Completion Rates among Roma and Non-Roma Neighbors, age 25-64 .... 73
Figure 5-3: Satisfaction with Preschools among Roma Children and Parents ............................................ 81
Figure 5-4: Segregation: % of Pupils Reporting to be in Class with Mostly Roma Pupils ......................... 83
Figure 5-5: Special School Attendance among Roma and non-Roma children .......................................... 84
Figure 6-1: Percentage of Roma Households living in Predominantly Roma Communities in the Three
Regions of Slovakia................................................................................................................................... 118
Figure 6-2: Presence of Sanitary Provisions in the Household (% of Households) .................................. 120
Figure 6-3: Share of Roma Households Receiving the Basic Material Needs Benefit or the Housing
Benefit by Settlement Type ....................................................................................................................... 122
Figure 6-4: Share of Roma Households Receiving BMN or Housing Benefit by Quality of Living
Conditions by Type of Settlement. ............................................................................................................ 123
Figure 6-5: Trade offs Between Costs per Dwelling and Number of Beneficiaries for Different Types of
Affordable Housing Opportunities ............................................................................................................ 126
Figure 7-1: Population Age Structure for Roma and Non-Roma Neighbors in Slovakia ......................... 139
Figure 7-2: Marriage Rates among Slovak Women .................................................................................. 140
Figure 7-3: Total Numbers of Infant Deaths in Slovakia in 2008, by Region........................................... 142
Figure 7-4: Fraction of Adults Suffering from a Long-Standing Illness ................................................... 143
Figure 7-5: Fraction of Adults Suffering from Chronic Diseases ............................................................. 144
Figure 7-6: Fraction of Youngest and Oldest Adults who Reported that their Health Was Either Good /
Very Good, or Bad / Very Bad .................................................................................................................. 145
Figure 7-7: Waste Collection and Water Sources in Vulnerable Roma communities .............................. 147
Figure 7-8: Facilities within Households .................................................................................................. 147
6
Figure 7-9: Share of Households Restricting themselves when Heating the Dwelling ............................. 148
Figure 7-10: Prevalence of smoking among Roma and Non-Roma, by Gender ....................................... 149
Figure 7-11: Alcohol Use among Roma and Non-Roma (A) ..................................................................... 149
Figure 7-12: Utilization of Reproductive Health Care Services................................................................ 150
Figure 7-13: Source of Information about Childhood Development ........................................................ 151
Figure 7-14: Do Residents of Vulnerable Roma Communities Feel That Their Health Needs Are Being
Adequately Met? ....................................................................................................................................... 152
Figure 7-15: Utilization of Outpatient Health Services ............................................................................. 153
Figure 7-16: Routine Medical Examinations and Check-ups (A) .............................................................. 154
Figure 7-17: Utilization of Inpatient Health Services ............................................................................... 155
Figure 7-18: Reasons for not Seeking Care when Needed ........................................................................ 156
Figure 7-19: Affordability of Medicines ................................................................................................... 156
Figure 7-20: Affordability of Unexpected Expenditures, Nutritious Food and Dental Visits................... 158
Figure 7-21: Satisfaction of Households with Outpatient Services ........................................................... 158
Figure 7-22: Distance of Households from Essential Services ................................................................. 159
Figure 7-23: Health Insurance Coverage among Adults ........................................................................... 160
Figure 7-24: Example of a Mobile Phone (SMS) based Pregnancy Advice System for Mothers ............. 170
Figure 8-1: Poverty Map of Bulgaria ........................................................................................................ 177
Figure 8-2: MoLSAF-WB Pilot Map of Roma Inclusion Projects in Slovakia ......................................... 178
Figure 8-3: Results Framework ................................................................................................................. 179
Figure 9-1: Dedicated Operational Programme On Roma Inclusion ........................................................ 192
Figure 9-2: Roma Inclusion As A Priority Axis In A Broader Operational Programme .......................... 193
7
LI S T O F B O X E S
Box 1-1: Past Use Of Structural Funds for Roma Inclusion ....................................................................... 13
Box 1-2: EC Country Specific Recommendations for Slovakia ................................................................. 13
Box 2-1: Four Potential Criticisms to the Economic Costs Calculation Approach ..................................... 21
Box 2-2: Per capita GDP of Slovak Roma in International Perspective ..................................................... 23
Box 3-1: The Dual Role of the Benefit in Material Need System ............................................................... 38
Box 3-2: Attitudes towards Social Assistance in Slovakia.......................................................................... 39
Box 3-3: A Roma Woman at Work Explains the Pressures She Faces ....................................................... 43
Box 3-4: An International Good Practice: Spain‘s Acceder model............................................................. 51
Box 3-5: Institutionalizing Impact Evaluations: Danish National Labor Market Authority ....................... 53
Box 3-6: Filling the Gaps and Increasing Independence: Chile Solidario .................................................. 56
Box 4-1: Experiences with Microcredit among Starting Roma Entrepreneurs: The KIUT Program in
Hungary ....................................................................................................................................................... 63
Box 4-2: "Graduation approach‖: Using Financial Inclusion to Create Pathways to Graduate out of
Poverty ........................................................................................................................................................ 65
Box 4-3: Bank on San Francisco, United States.......................................................................................... 65
Box 4-4: Financial Literacy Training among Roma in Slovakia................................................................. 67
Box 4-5: Earmarking Savings: United states and Ghana ............................................................................ 67
Box 4-6: ‗Targeting Savings‘ Project among Roma in Slovakia ................................................................ 68
Box 5-1: The NRIS and Improving Education among Roma ..................................................................... 72
Box 5-2: Difficulties in Attending High School.......................................................................................... 76
Box 5-3: Evidence on the Importance of Early Childhood Development ................................................... 77
Box 5-4: Preschool in Hungary ................................................................................................................... 79
Box 5-5: The Special School in Chminianske Jakubovany ......................................................................... 87
Box 5-6: The Meséd Project – ―Your Tale‖ ................................................................................................ 98
Box 5-7: Roma Mediator Training Programme ........................................................................................ 100
Box 5-8: Roma Teaching Assistants ......................................................................................................... 102
Box 5-9: The Experience of Bolsa Familia in Brazil ................................................................................ 104
Box 5-10: Key Areas for Inclusive Education Policy ............................................................................... 107
Box 6-1: Summary of Neighborhood Conditions of Roma Households ................................................... 118
Box 6-2: Summary of Housing Conditions among Roma Households ..................................................... 121
Box 6-3: Pre-paid Metered Utility Provision – Extending the Solution to Manage Debts ....................... 124
Box 6-4: Summary Housing Affordability ................................................................................................ 125
Box 6-5: Housing Goals from the Slovak National Strategy for Roma Inclusion (2012), Section D. 2. 4:
Housing ..................................................................................................................................................... 127
Box 6-6: Stara Lubovna: Community Participation in Integrated (and Incremental) Improvements in
Living Conditions in Spatially Segregated Roma Communities ............................................................... 130
Box 6-7: Local Initiatives to Build Skills and Community Infrastructure at the Same Time. .................. 130
Box 7-1: Health Mediator Programs ......................................................................................................... 163
Box 7-2: Example of Successful Household and Community-based Cash Transfer Program in Indonesia
................................................................................................................................................................... 172
Box 8-1: Social Policy Experimentation and the Danish Labor Market Authority ................................... 180
Box 9-1: High Level Event on Roma Inclusion in 2011. .......................................................................... 184
Box 9-2: Foreseen changes for the European Social Fund ........................................................................ 188
8
1 INTRODUCTION
1.1 THE PRESENT SITUATION CALLS FOR URGENT ACTION
The differences in living standards between Roma and the general Slovak population are stark and
start early in life. Findings from a large scale UNDP/World Bank/EC regional Roma survey (2011),
representative of 83% of Slovak Roma, show that the vast majority of Roma in Slovakia, 87% of those
households interviewed, live in poverty. One-third of Roma children go to bed hungry at least once a
month because there is not enough food. The gaps between the marginalized Roma and the majority
population start early. For example, only 28% of Slovak Roma children aged 3-6 attend preschool,
compared with 58% of non-Roma children living nearby. The comparison is unfavorable also with other
countries: 76% of Roma children in Hungary and 45% of those in Bulgaria attend preschool. Despite
widely held perceptions to the contrary, 82% of Slovak Roma indicate the wish for their children to
achieve at least upper secondary or tertiary education, but only 28% actually reach – but not necessarily
complete – upper secondary education. This compares with 94% in the general population that completes
upper secondary.
The gap among adults is similarly large: only 20% of working age1 Roma men and 9% of Roma women
have jobs, compared with 65% of working age men and 52% of working age women in the general Slovak
population. These rates are low also for regional standards: they are less than half those found in Bulgaria,
the Czech Republic, and Romania. Moreover, wage levels among the Roma who do have jobs are on
average half of those earned by the general population. About half of Roma households receive the
Benefit in Material Need, and approximately two thirds receive child allowances and related family
benefits. Yet, because Roma make up a minority in the overall population, Roma families constitute only
about 35% of all households receiving BMN2. With regards to housing, about one third of Roma families
live in very poor, informal housing conditions, and only about half of them have indoor sanitation or
drinking water in their house. Regarding health outcomes, Roma life expectancy has been estimated to be
about 15 years lower than the national average, which is consistent with the survey finding that only 2% of
Slovak Roma are older than 65. As such, life expectancy for Roma is more comparable with that of
countries in Sub-Saharan Africa or South Asia than with life expectancy in the European Union.
Dire employment conditions among marginalized Roma in Slovakia translate into unusually large
gaps in per capita GDP between Roma and the general population. Among the general Slovak
population, average per capita GDP is approximately Euro 13,000 per year, placing Slovakia among the
richest 25% of countries in the world. At the same time, the average per capita output of Slovak Roma is
only Euro 1,400 per year. After accounting for purchasing power, GDP levels found among Slovak Roma
are, once again, equivalent to levels observed in countries in Sub-Saharan Africa and South Asia
belonging to the poorest 25% worldwide. Another way of seeing this gap is that monthly per capita output
among the general Slovak population is similar in magnitude to the annual per capita output among
Slovak Roma.
The dire situation of the Roma communities has been noticed at the European level and addressing
it has become part of mainstream European policy making. The April 2011 EU Framework for
National Roma Integration Strategies up to 20203 clearly positions Roma inclusion as part of the overall
1
‗Working age‘ refers to the age-group 15-64, inclusive.
2
For details, see the World Bank‘s (2012) report on the Benefit in Material Ne ed (BMN)
3
Communication from the Commission to the Council, the European Parliament, the European Economic and Social
Committee and the Committee of the Regions, Brussels, 5.4.2011 COM(2011) 173 final
9
Europe 2020 strategy of smart, sustainable and inclusive growth. In its May 2012 Communication, the EC
explicitly states that ―[f]or Member States with a larger Roma population making sufficient progress
towards the Europe 2020 employment, social inclusion and education targets will require addressing
explicitly and swiftly the situation of the Roma.‖4 The EU therefore urges the Member States to ensure
that EU funding available makes a tangible difference to Roma communities. In fact, in May and June
2012, the Commission and the European Council even included various countries‘ Roma Integration
Strategies in the new economic governance cycle – called the European Semester – by mentioning the
implementation of these strategies as one of the so-called ‗Country Specific Recommendations‘ (CSR).
This was the case for Bulgaria and Hungary. In the case of Slovakia, however, the Commission/Council
went further by including a detailed set of recommendations on Roma inclusion in the education and
employment fields as one of only seven CSRs (see Box 1 below for details). The Slovakia government
request to the World Bank to provide diagnostics and policy advice is an example of the Government‘s
concern.
Roma integration is indeed in the national economic interest of Slovakia. Slovak GDP would be Euro
3.1 billion higher if Roma would have the same employment opportunities and wage levels as non-Roma.
Considering that there are an estimated 320,000 Roma living in communities with large Roma populations
in Slovakia, the gap in individual level output translates into an aggregate output gap of Euro 3.1 billion,
or 4.4% of Slovak GDP. This means that each year, Slovak output would be 4.4% higher if Roma would
have the opportunity to work as much and to earn the same wages as non-Roma. The extremely poor
employment prospects for Roma, hampered especially by extremely low education levels, also result in
lower income tax and social insurance revenue, lower corporate tax revenue, and higher social assistance
payments, not even counting costs in health, special education, and law enforcement. Altogether, these
differences in fiscal revenues between Roma and the general population amount to nearly Euro 3900 per
year per working age Roma. This means that Slovak Government revenues would be 3.1% higher – about
Euro 725 million annually – if employment conditions for Roma would be the same as for non-Roma.
Furthermore, population dynamics make Roma inclusion indispensable for Slovakia‘s long-term
economic sustainability and the long-term viability of its social protection system. With an officially
used population estimate of more than 320,000 Roma in Slovakia and the survey finding of 40% of the
Slovak Roma population being under 15 years of age, more than 13% of new labor market entrants in
Slovakia today are young, mostly unskilled Roma. By 2050, the Roma population is estimated to nearly
double at current growth rates of approximately 1.7% per year, while Slovakia‘s population as a whole is
expected to shrink. The population pyramids for Slovak Roma and non-Roma neighbors illustrate this (see
Figure 1). Hence, in an otherwise rapidly aging society, paying for future pensions and social services will
increasingly depend on today‘s young Roma. Consequently, the quality and level of education that young
Roma achieve today and in the years to come, and the kind of skills they develop to become productive
citizens, bear significant implications for the future of the Slovak society as a whole. .
4
Communication from the Commission to the Council, the European Parliament, the European Economic and Social
Committee and the Committee of the Regions, National Roma Integration Strategies: a first step in the
implementation of the EU Framework, Brussels, 21.5.2012, COM(2012) 226 final.
10
FIGURE 1-1: POPULATION PYRAMIDS: ROMA (LEFT PANEL) AND NON-ROMA NEIGHBORS (RIGHT
PANEL)
75-79 75-79
60-64 60-64
45-49 45-49
30-34 30-34
15-19 15-19
0-4 0-4
15 10 5 0 5 10 15 15 10 5 0 5 10 15
Percent Percent
Female Male Female Male
Source: UNDP/World Bank/EC regional Roma survey (2011).
Investing in Roma inclusion is not only equitable, but also fiscally smart. Estimates of the fiscal
returns to education investments, measured in net present value, show that the returns to these
investments are substantial, ranging from fiscal revenues that are 50% to 105% higher than the
investment costs. These figures conservatively assume that it is 50%-100% more costly to ensure a Roma
pupil completes the extra school years than a non-Roma pupil. On the revenue side, these calculations
assume that more education translates into better labor market outcomes in the future. For example, the
employment analysis finds that Slovak Roma completing secondary education have employment rates that
are approximately 12 percentage points higher than those who only complete basic education. Better
employment outcomes, in turn, translate into higher government revenue through increased income tax
receipts and lower benefit payments. Finally, because these benefits occur in the future while the
education costs are incurred today, the benefits are discounted assuming a relatively high real interest rate
of 2%.5
1.2 METHODOLOGY AND DATA SOURCES
This report is meant to support the Slovak Government in its efforts to address the exclusion of the
Roma by offering evidence-based policy advice. The assessment relies on three main sources of
information. First, it takes advantage of the 2011 UNDP/World Bank/EC regional Roma survey, described
below, which is the most comprehensive survey effort to date to capture the situation of Roma in Central
and Eastern Europe. It also includes information from the 2010 survey by UNDP, done specifically on
Slovak Roma at the request of the Ministry of Labor, Social Affairs, and Family. The results on
comparable indicators from both surveys are very similar. Second, this report relies on qualitative
information, collected through field visits in Eastern Slovakia and through interviews with key
stakeholders from the Slovak Government and from civil society. And third, each of the chapters
highlights relevant international experiences from which Slovak policy formulation on Roma integration
can benefit. Many of the international examples and best practices from integrating poor and marginalized
5
The current market real interest is well under 1%; over 2011, interest rates on long-term (close to 10 years) Slovak
government bonds were approximately 4.5% and inflation 4.1% (Eurostat, 2012), suggesting a real interest rate of
0.4%. Discounting the revenue stream at 1% real would put the net present value of the education investment returns
between 101% and 167%.
11
communities elsewhere provide reasons to be optimistic that Roma integration does not have to be a
distant goal for Slovakia.
The UNDP/World Bank/EC regional Roma survey – the main data source for this report -- is a
comprehensive survey that is representative of approximately 83% of the Slovak Roma population,
including Roma living in mixed, separated and segregated neighborhoods. The survey questionnaire
was designed by the World Bank and UNDP in partnership, and implemented by UNDP through the
IPSOS polling agency in May-July 2011 on a random sample of Roma living in communities with
concentrated Roma populations in Bulgaria, Romania, Hungary, Slovakia, and the Czech Republic
(henceforth: the ―regional Roma survey‖). The European Commission DG Regional Policy financed the
survey. In each of the countries, approximately 750 Roma households (representing over 3,500
individuals) and approximately 350 non-Roma households living in the same neighborhoods or vicinity
were interviewed. The sample was 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 Roma population. This covers 88% of the Roma population in Bulgaria, 90% in the
Czech Republic, 78% in Hungary, 89% in Romania, and 83% in Slovakia. Once identified, a random
sample of these areas was drawn, and households were randomly sampled within these enumeration areas.
The data provide reliable estimates of the conditions in which the vast majority of the Roma in
Slovakia live, and of the conditions of their non-Roma neighbors. Unless otherwise noted, the analysis
in this report is based on the ‗Roma‘ and the ‗non-Roma nearby‘ sampled households as they were
identified by the survey enumerators. Comparisons with non-Roma living nearby provide a crucial frame
of reference, since the sampled non-Roma households live in the same municipalities and thus share local
labor markets, community, school, and health facilities as well as other services and collective
infrastructure. Hence, if we observe differences in education, health, housing, and employment between
Roma and non-Roma households, these must reflect particular disadvantages faced by Roma, differences
in preferences between Roma and non-Roma, or both.
For comparison with the general population in Slovakia, the report uses the EU Statistics on Income
and Living Conditions (EU SILC) survey. The EU-SILC does not distinguish between Roma and non-
Roma and provides household survey information that is representative of the general Slovak population.
In other words, when gaps are reported between an educational or employment attainment of Roma
compared with non-Roma neighbors in the 2011 regional Roma survey, these gaps will generally be
smaller than between Roma and the general Slovak population in the EU SILC.
1.3 THE PRESENT SITUATION ALSO OFFERS UNPRECEDENTED
OPPORTUNITIES
The forthcoming programming of the next generation of the Structural Funds is a truly historic
opportunity for Member States with large Roma minorities to significantly improve the integration
of their Roma populations. The EU will make large sums available to countries to develop and
implement inclusion programs and projects, and will monitor the progress of Member States. This
provides a unique opportunity, also for Slovakia. In the previous programming periods, it was not possible
to combine the European Social Fund (ESF) and the European Regional Development Fund (ERDF)
investments under one Operational Program, let alone under one Priority Axis. This caused inefficiencies
and led to lesser impact, especially in areas where linking soft (ESF) and hard (ERDF) investments is a
key ingredient for success. Roma inclusion is an area where no infrastructure construction (housing,
12
schools, community centers, etc.) will succeed unless accompanied by outreach, training, mediation,
social work, courses for parents, and other soft investments.
BOX 1-1: PAST USE OF STRUCTURAL FUNDS FOR ROMA INCLUSION
In recent years, significant hopes were pinned on the horizontal priority ―Marginalised Roma
Communities‖ and especially on the ―Local Strategies of Comprehensive Approach‖ in the current
programming period 2007-2013 of the Structural Funds. Despite the generous national allocation of 200
million Euro for the local strategies, little has been invested from this amount in Roma inclusion at the
national or local level. In April 2012, one year before the end of the programming period, only some 16
million out of the allocated 200 million had been contracted. The horizontal priority as currently
implemented does not seem to have resulted in a more systematic push for Roma inclusion, or in scaling
up the piloted approaches. While the new National Roma Integration Strategy has been praised by the EC
for its holistic and data-driven approach, experience with previous strategies shows that real challenge
rests with implementation – a key component of which is the efficient utilization of all resources available.
To make full use of this opportunity, the Slovak Government may consider developing one
Operational Program or one Priority Axis for Roma inclusion in a broader Operational Program
(see Figures 14 and 15). In either case, Roma inclusion can be financed jointly from the European Social
Fund (ESF) and the European Regional Development Fund (ERDF). The financing should also be aligned,
to the extent possible, with the programming under the Rural Development Fund (EAFRD). The existing
set-up in the 2007-2013 programming period of a horizontal priority with the mandate for local
comprehensive strategies has proven too complicated and ineffectual (see Box 1). The structure for the
next programming period should be redesigned, taking advantage of the ability to combine the two
Structural Funds in one OP or Priority Axis. It is also critical to have in place one sufficiently staffed and
resourced coordinating body that leads the design and monitoring and guides the implementation of Roma
integration policies and Structural Funds investments.
The Slovak National Roma Integration Strategy (NRIS) of January 2012 has been adopted at an
opportune moment, as the European Commission is committed to monitor the progress individual
Member States make on Roma inclusion. At the same time, as the current EU financing and
programming period is drawing to a close (end 2013), regulations for the new financing period (2014-
2020) are making their way through the legislative process. Structural Funds in the next programming
period will have to be closely aligned with the Europe 2020 Strategy of smart, sustainable and inclusive
growth. Due to the focus of this last pillar - inclusive growth - Roma inclusion will be a key area in which
the Commission will actively engage Member States during the forthcoming negotiations (expected end
2012 and throughout 2013). The Commission‘s tools during the negotiations will comprise: (i) thematic
concentration on a limited number of priorities, including ―promoting social inclusion and combating
poverty‖; (ii) partnership agreements with Member States that will set out the priority areas and themes;
and (iii) ―ex ante conditionalities‖ for all themes setting the preconditions to receive support from the
Structural Funds.
BOX 1-2: EC COUNTRY SPECIFIC RECOMMENDATIONS FOR SLOVAKIA
The European Commission‘s Country Specific Recommendations to Slovakia, released at the end of May 2012 and
confirmed by the end June 2012 EU Summit, deliver a strong message on the relevance of Roma inclusion. Roma
inclusion is the subject of one of the seven specific recommendations to the Slovak Government. As such, apart from
13
two countries with short references to the need to implement their NRISs, Slovakia is the only EU country receiving
a detailed set of recommendations explicitly targeting Roma inclusion. The recommendation specifically reads:
―Marginalised communities, including the Roma, are largely excluded from the labour market and the mainstream
education system, representing a significant underutilised labour potential in the Slovak economy. In order to tackle
this problem, Slovakia should step up efforts to improve educational outcomes of marginalised groups and reinforce
its reintegration policies for adults.
[THE COUNCIL] HEREBY RECOMMENDS that Slovakia should take action within the period 2012-2013 to:
Take active measures to improve access to and quality of schooling and pre-school education of vulnerable groups,
including Roma. Ensure labour market reintegration of adults through activation measures and targeted employment
services, second- chance education and short-cycle vocational training.‖
1.4 OVERALL POLICY RECOMMENDATIONS
The remainder of this report provides a detailed description of the assessment of the economic cost
of exclusion, the situation in the 5 sectors above and in the 2 cross-cutting areas of monitoring and
evaluation and of EU financing, followed by specific policy recommendations in each. One important
area beyond the scope of this report is the ability of the Slovak legal and judicial system to successfully
protect the rights of all citizens, including the courts, issues of legal aid, as well as law enforcement issues.
Recently, a much discussed court ruling in Eastern Slovakia held school segregation as illegal. More of
such cases may emerge, especially in light of the 2007 ruling against school segregation by the European
Court of Human Rights in Strassbourg. While important, these are all beyond the scope of this report.
Appendix Chapter 10 provides a summary of the policy recommendations for each of the sectors.
Detailed diagnostics and policy options by sector should not overshadow the need for a holistic and
integrated approach to addressing the needs of Roma inclusion, requiring ministries to coordinate
reinforcing actions. In addition to sector specific findings, this report highlights important linkages
between each of these areas and stresses the need for close coordination among different sector ministries.
For example, the linkages that exist between poor housing conditions and poor health, or the linkages that
exist between social benefit reforms and activities by key line ministries; for example, early childhood
education can be promoted by informing parents about its benefits through the use of Roma mediators and
teaching assistants, but also by putting in place targeted social protection benefits that support all poor and
vulnerable parents financially if children attend regularly preschool. Conversely, stimulating demand for
pre-school through the benefit system needs to be complemented by infrastructure investments to expand
pre-school facilities in those communities that are currently lacking these. Or, targeted social protection
benefits can be made available on condition that mothers participate in pre- and post-natal check-ups, fully
vaccinate their children, and participate in counseling on early childhood development and nutrition, thus
also requiring coordination between ministries. Or, changing the eligibility criteria for housing allowances
to include the poorest and restricting use of the funds for home improvements should be complemented by
training activities how to make basic home and energy efficiency improvements. In short, it is important
therefore that policy measures in the different areas are designed to reinforce each other. Many of the
policy measures suggested offer opportunities in this sense.
Most policy recommendations attempt to draw and build on programs and policies already in place
in Slovakia, including good examples and best practices the World Bank team came across during its field
visits which are not known nationally. In some cases, putting in place incremental and reinforcing policy
measures can be achieved by modifying existing policy measures such as reforming the system of
14
allowances –labor activation, kindergarten and housing --that already exist as part of the BMN program of
social assistance; or, by addressing the financial incentives that currently exist for municipalities to invest
in special education, and instead provide incentives to invest in integrated regular schooling; or, by
ensuring that the ongoing expansion of social workers will lead to an increase specifically in counseling
on early childhood development and nutrition. In other cases, the World Bank team recommends
expanding and further improving promising initiatives that are currently small scale such as the Roma
Health Mediator (RHM) program, or the home improvement and financial literacy training courses
provided by, for example, ETP Slovakia.
No policy measure can fully succeed if strong prejudices among parts of the majority population
against the Roma are not addressed simultaneously. To illustrate, in 2010, the World Bank carried out
222 qualitative interviews with government and civil society officials in Bulgaria, Romania, Czech
Republic, and Serbia. Respondents were asked to share their view of commonly held perceptions among
the general public. According to more than three-quarters of these officials in each of these countries, a
commonly held perception by the general public is that Roma are "lazy, lack will power, and prefer to live
off social assistance". Many people with whom the World Bank team met in Slovakia similarly mentioned
that these views are common, consistent with the 2008 European Social Survey finding that two-thirds of
respondents claim that the unemployed (regardless of ethnicity) are not seeking employment and 40%
report that social benefits make people lazy. On the other hand, results from the regional Roma survey
indicate that the vast majority of Roma wish at least a secondary education for their children, and value
lower paid, secure and full time employment over social assistance or over irregular employment with
more freedom. Yet, it is easy to see that a typical employer would not be keen on hiring a Roma, no matter
how successful might have been policy measures to improve Roma education and health outcomes.
Tackling Roma stereotypes is a vital component of inclusion strategies and need to be an integral
aspect of any measure as well as a goal in itself. Beliefs shape behaviors and therefore determine both
political outcomes and practical day-to-day decisions in applying policies and implementing programs.
Even in the absence of the measures advocated in this report, there are examples of mayors who have been
able to transform the lives of Roma in their municipality –at the same time improving the lives of all their
constituents - because they challenged stereotypes, forged partnerships with others also willing to
challenge stereotypes (e.g., social workers, school officials, psychologists), thus creating a virtuous circle.
Communication can have a significant impact on people‘s beliefs and behaviors, in particular through
targeted campaigns of communication to change behavior (CCB), which apply a marketing approach to
address social issues.6 The Slovak Government should join forces with civil society, media experts, and
international organizations in determining a roadmap for a comprehensive and creative communication
campaign aimed at fighting stereotypes and encouraging acceptance. For example, the Open Society
Foundation has a long history of supporting Roma inclusion programs in Slovakia and elsewhere,
including projects to fight stereotypes. The European Roma Grassroots Organization (ERGO) has
implemented a number of programs tackling stereotypes across Europe forging partnerships with
municipalities and creatively using (social) media and sports. And, the Roma Education Fund has been
working closely with teacher training colleges to create greater awareness among college students of the
challenges facing Roma children in school and foster interest in contributing personally to closing the gap
by seeking placement in schools with many Roma children. The Slovak government can take advantage of
these and other such campaigns.
6
CCB has been used widely and successfully to encourage environmentally responsible behaviors, as well as in
preventive health campaigns, for example, to promote basic hygiene and responsible sex. Its use to promote social
cohesion and mutual acceptance (e.g., in ethnically divided contexts) is growing and shows very promising results.
15
Successful implementation will also require strengthening capacity at all levels, from policy
coordination to implementation of projects on the ground, and this implies building stronger
partnerships between regional and municipal bodies, and between public, civil society, and private
bodies. The Office of the Plenipotentiary for Roma Communities has started developing the necessary
capacity. It has prepared the National Roma Integration Strategy in a professional and consultative
manner. This capacity can be captured and built upon – regardless of who will politically oversee the
Office – and substantially expanded. In doing so, Slovakia can also build on international experiences
with integrated approaches to addressing the most vulnerable such as the Chile Solidario program. The
sizable allocation of technical assistance under the structural funds (4% of total ESF) can be deployed to
finance capacity building. For example, knowledge sharing and technical assistance to local actors such as
municipal governments requires strong regional level capacity; or, leveraging the knowledge and
experiences of promising NGO initiatives requires building stronger public-private partnerships whereby,
for example, NGOs with a proven track record are financially supported to work with municipal and
regional bodies to scale up promising initiatives.
16
2 THE ECONOMIC COST OF ROMA EXCLUSION
This chapter highlights how dire employment conditions among Roma in Slovakia translate into unusually
large gaps in per capita output between Roma and the general population. After reviewing the
employment gap and providing an overview of the methodology to calculating the economic costs of
exclusion, the chapter first assesses the gap in economic output as it exists today between Roma and the
general Slovak population. This is followed by an analysis of the fiscal implications of the employment
gap, analyzing the implications for income tax revenues, social insurance revenues, as well as corporate
tax revenues. Finally, it highlights the calculations detailed in the Education chapter that investing in
Roma inclusion is not only equitable, but also fiscally smart, with estimates of the fiscal returns to
education investments, measured in net present value (to account for the fact that investments today
provide fiscal returns in the future), showing that the returns to these investments are substantial and
positive even under restrictive scenarios.
2.1 BACKGROUND
For the purposes of this analysis, the population figure of 320,000 Roma living in concentrated
Roma communities is used. In 2010, the total population in Slovakia was estimated to be 5,431,024
(Statistical Office of the Slovak Republic [SOSR], (2012)). Estimates of the size of the Roma population
vary as highlighted in the 2011 National Roma Integration Strategy of the Slovak Republic: "In the
national census of 2001, 89 920 Slovak citizens, representing 1.7 % of the total population, claimed Roma
national minority. The Atlas of Roma Communities [developed in 2004] quotes the number of Roma
living in Slovakia at 320 000 individuals, or 5.8% of the population. The Center for Demographic
Research estimates with great probability that 440 000 Roma resided in the territory of the Slovak
Republic in 2011, which represents around 8 % of the total population.―
The data from the regional Roma survey (2011) used in this analysis are representative of the
communities that were surveyed by the Slovakia Roma Atlas and is representative of approximately
83% of the Slovak Roma. For this reason, the analysis will use this figure of 320,000, and the analysis
will be representative for the Roma living in these concentrated Roma communities. To compare
monetary amounts between 2009 and 2011, the values for 2009 are converted into 2011 values using the
GDP deflator for Slovakia (Eurostat, 2011).7
For comparison to the general population in Slovakia – and in the other four countries where
relevant, use was made of the EU Statistics on Income and Living Conditions (EU SILC) survey.
The EU-SILC does not distinguish between Roma and non-Roma. Given the low levels of education and
employment among Roma, this means that – to the extent the EU-SILC sample captures Roma – any EU-
SILC statistic presented on the ‗general population‘ will be lower than the corresponding statistic on
Slovak non-Roma only.
The key determinant for the economic cost of exclusion is employment: in the case of Slovakia, very
few Roma work. The employment rates of Roma in Slovakia fall well behind those of the general Slovak
population and also compare unfavorably to the Roma employment rates in other Eastern European
countries with large Roma minorities. Figure 2-1a and 1b show the employment rate for the working age
population (15-64 year old) – general population and Roma living in concentrated Roma areas - in
Bulgaria, the Czech Republic, Hungary, Romania, and Slovakia. Individuals are considered employed
regardless of the nature of employment; i.e. these figures include informal employment. Only 20% of
7
The analysis uses the GDP deflator – rather than the CPI deflator - because the focus is on output and not
expenditure.
17
Roma men and as few as 9% of Roma women are reported to be working, compared with 68% and 52%,
respectively, in the general population. While there is a substantial employment gap between Roma and
the general population in each of the countries, nowhere is the gap as large as in Slovakia.
FIGURE 2-1: EMPLOYMENT RATES AMONG ROMA ARE MUCH LOWER THAN AMONG NON-ROMA
A. Men B. Women
80 74 70 80
63 66 65
70 60 70 58
56 56 52 52
60 60 51
Percent
42 43 42 50
Percent
50
40 34 40 26
30 20 30 19 19
20 13
20 9
10 10
0 0
General Population Roma General Population Roma
Source: Roma: UNDP/World Bank/EC regional Roma survey; General Population: EU SILC (2009).
Roma who achieve higher levels of education show a higher rate of employment, but employment
does not improve with education as much as it does for non-Roma. Moreover, the biggest group of
working age Roma has not completed secondary school. Figure 2-2 shows a breakdown of employment
rates by education level, for Roma and non-Roma neighbors. Employment rates for Roma and non-Roma
are similar at lower education levels, but for those having completed secondary education, the gap in
employment rates is very large: the employment rate for non-Roma neighbors in this group have is almost
double the rate for Roma. Moreover, as indicated by the width of the circles, most Roma never reach this
level of education, whereas by far the biggest group of non-Roma neighbors does.
FIGURE 2-2: EMPLOYMENT RATES BY EDUCATION LEVEL
60
Percentage Employed
50
47
40
30 29
20
6 16
10
14
2
0
0 1 2 3 4
Education Level
Roma Non-Roma Neighbors
Source: UNDP/World Bank/EC regional Roma survey (2011).
Education levels: 1= None or incomplete Basic (up to age 15); 2= Basic or incomplete secondary:
vocational/technical; 3= Secondary or higher. The width of each circle reflects the proportion of the working age
population having achieved each education level.
18
In addition to low employment, labor earnings among Roma with jobs, women especially, are also
significantly lower. The net monthly wage for a male from the general population with a job is Euro 729,
compared with Euro 403 for Roma men. The equivalent figures among women are Euro 566 and Euro 278
per month. This means that employed Roma men earn salaries that are only 55% of the salaries earned by
the general male population. Among women the corresponding figure is even lower: 49%. These
comparisons do not control for potential differences in education or work experience. As such, they only
present the eventual outcome of a complex process of influences and interactions, such as low education
levels leading to low employment- and wage outcomes among Roma, leading to low incomes and
potentially to the inability of Roma to send the next generation of children to school. These are explored in
the chapters on employment and education.
2.2 METHODOLOGY
The analysis used in this chapter rests on a purposefully simple thought experiment: how much
higher would total annual output (and thus incomes) in Slovakia be if the average Roma individual
of working age had the same economic output as the average person from the general population?
This estimate first requires (a) an estimate of annual output among Slovak Roma working age individuals,
and (b) an estimate of annual output among the general population in Slovakia; the difference between
these two, then, is the annual output gap per working age individual. In combination with (c) an estimate
of the number of working age Roma, the total annual output gap across all working age Roma can be
calculated. Note that this analysis rests on the assumption that the labor market does not contain a fixed
number of jobs – at least not across different points in time, but rather, that new jobs are endogenously
being created over the course of say, a generation. Through this process, skill levels and employment
opportunities more generally are being equalized, new skilled labor market entrants – Roma and non-
Roma – get absorbed, and new economic opportunities are being generated. Even in the absence of this,
population growth estimates for Slovakia suggest that skilled Roma entering the labor market are unlikely
to cause a net increase of the overall Slovak working age population participating in the labor market: the
United Nations (2010) predicts the total population in Slovakia to decrease by 4.0% and 11.6% between
2010 and, respectively, 2050 and 2070. The Roma population estimate used in the current analysis –
320,000 – reflects 5.8% of the Slovak population.
To calculate estimates of annual output among Slovak Roma working age individuals, and among
working age individuals of the general population in Slovakia, we first focus on workers. To
calculate output per worker (which is essentially the GDP per worker measure), we need to account for the
fact that one share of the economic output by workers translates into labor compensation, while the other
part translates into compensation to the owners of capital; i.e. to cover the payments for buildings,
computers, machinery, corporate taxes, and firm profits. As described below, we can infer the total
amount of labor compensation by combining individual level information on net compensation (obtained
from household survey data) with information on labor taxation (income tax and social insurance
payments) in Slovakia. In this way we can calculate the ‗labor part‘ of total output.
Once we have calculated the ‗labor part‘ of output, we need to add the ‗capital part‘ to arrive at
total output. We do not have individual level information on the total amount of compensation per
worker that companies spend on capital. However, the OECD maintains a database with country-specific
averages on labor and capital shares, using national accounts information (OECD, 2011); in Slovakia, the
labor share reflects slightly more than half (50.8%) of total output and the capital share (49.2%) the
19
remainder.8 Hence, total output per worker is nearly double the value of total labor compensation per
worker. By adding the capital part to the amount of labor compensation received by Slovakian workers,
we arrive at total output.
To calculate the size of the ‗capital part‘, we need to combine information on total labor
compensation per worker, with information on labor and capital shares. We can then estimate the
value of capital compensation per average worker. Note that our calculation assumes that the capital-labor
ratio is the same for the general population as it is for the Roma population. If the actual capital share is
lower among Roma, then we are under-estimating the total output gap (and vice versa). Finally, once the
difference in total output per worker between the general Slovak population and the Roma population is
known, the calculation has to account for the fact that not every working age individual also actually
works; the total output per worker must be multiplied by the employment rates to calculate the total output
per working age individual. Table 2-1 highlights how Euro 100 in net income translates into Euro 192.74
total labor compensation after accounting for taxes and (employee, employer) social insurance
contributions. The EU-SILC and the regional Roma survey both provide information on the net monthly
income for individuals of working age (15-64 years). This information can therefore be combined with the
information in the table below to infer the total average monthly labor compensation.
TABLE 2-1: LABOR TAX RATES
Amount Rate Base
Total Labor compensation 192.74
Employer social ins. 50.18 35.2% Gross income
Gross income 142.56
Employee social ins. 19.10 13.4% Gross income
Taxable income 123.46
Income tax 23.46 19.0% Taxable Income
Net income 100.00
Source: World Bank. (2011). Doing Business Survey; www.oecd.org/ctp/taxdatabase;
http://www.taxrates.cc/html/slovakia-tax-rates.html; and, http://www.worldwide-tax.com/slovakia/slovakia_tax.asp.
Note that this calculation assumes that every worker and employer is fully tax compliant, and that
discrimination of Roma – if present – has no effect on wages. If Roma workers are in fact more likely
to be informal than workers from the general population, then their net incomes will be closer to their
gross incomes than suggested by the tax-rate inference; in that case our estimate of the total productivity
of Roma workers is biased upwards, and the estimate of the total economic loss is biased downward. The
presence of discrimination, on the other hand, may lead to an overestimate of the output losses. To see
this, note that in the presence of wage discrimination, the reported wages underestimate the underlying
level of productivity. The actual output gap is then smaller than the estimated one. The extent to which
discrimination reduces the chances that people have jobs does not, however, affect our estimate of
economic losses; in that case, discrimination simply puts people out of jobs, having a direct impact on
economic losses, which are adequately captured by the low employment rate.
8
For example, suppose that (purely for illustrative purposes) labor gains are estimated to be 120 and suppose that the
labor share in national income is known to be 60% and the capital shares 40% (these shares are calculated from the
national accounts data). Then the corresponding increase in capital income equals 120*(40/60) = 80. In Slovakia the
OECD (2012) estimates the labor share (for 2010) to be 50.8% and the capital share to be 49.2%.
20
BOX 2-1: FOUR POTENTIAL CRITICISMS TO THE ECONOMIC COSTS
CALCULATION APPROACH
1. Is full equality of output a meaningful benchmark against which to measure economic losses in
output, when in reality, inequalities are large and difficult to overcome? While bridging the output
gap will require resources, concerted effort, and mutual acceptance by Roma and non-Roma alike, placing
the reference point anywhere below equality of output would incorrectly suggest that Roma children
provided with the same opportunities as non-Roma would somehow not be capable of achieving the same
outcomes as non-Roma. The examples from Israel and the UK shown in Chapter 5 (Education)
demonstrate that even extremely large gaps in educational attainment can be overcome with the right
policy environment.
2. On choice of indicator, is the static output gap an accurate measure of overall output gains that
will be achieved if productivity levels were equalized? Equality of output levels may actually increase
national output by more than the current gap in output alone if we take into account the multiplier effect,
whereby an increase in incomes and subsequent expenditures stimulate further employment creation and
incomes. However, because there is much uncertainty over the size of the economic multiplier, the current
analysis does not take this into account, thus opting to err on the lower bound of the gap in output.
3. To what extent are information on wages, employment rates, and taxes and social insurance
contribution rates sufficient to infer productivity gaps? As discussed above in the more detailed
description of the methodology, our method tends to err on the lower side of the individual output gap.
Hence, if the assumptions used here do in fact not hold, this means that the true output gap is higher, not
lower, than what is estimated below. This, in turn, makes the current analysis one of the ‗best case
scenario‘: the output gap presented here assumes favorable conditions.
4. Improving education levels among Roma is a costly investment. Would this cancel out the gains
from higher employment rates? As shown in the Education Chapter, the investments needed to improve
education levels among Roma are likely to pay off: for various types of investment in education, a
calculation of the expected financial benefits to the state under different assumptions of how much
education translates into employment gains shows that these investments result in extra revenues which
outweigh the total costs.
Most of the literature in which similar analyses are executed aim to relate productivity gaps to
economic growth. This literature is predominantly about the impact of gender inequalities on
economic growth in particular. Although there are some opposing voices (Senguino, 2000), the
literature by and large argues that promoting female education will raise economic growth (e.g. Galor and
Weil, 1996). Empirical work generally supports this hypothesis (e.g. Dollar and Gatti, 1999; Klasen and
Lamanna, 2011). These studies rely on so-called country ‗growth‘ regressions, in which country level data
are used to estimate economic growth rates as a function of (country) measures of education or
employment differences between men and women.
The current analysis differs from the conceptual approach used in the gender gap-economic growth
literature, both empirically and conceptually – aside from the gender versus minority focus. First,
empirically, this analysis does not seek to link differences in economic growth rates (over time and across
countries) with differences in employment and education – and hence, in productivity – between the Roma
on the one hand and the general population on the other. Accurate time series data on Roma do not exist,
but even if they did, such an approach would rely on too small a set of countries with Roma populations,
21
and be subject to considerable country level heterogeneity impacting both inequalities and growth rates
that cannot be captured by the data (and thus biasing the results). Furthermore, it would require within
country variation over time in the education gap; comparisons of education levels across age cohorts
suggest education rates have increased very slowly over time. For these reasons, the goal in the current
paper is to establish how much greater economic output – instead of economic growth – would be if
productivity rates would be the same for Roma as they currently are for non-Roma. The final section
touches upon the question of how to equalize productivity levels – starting with a push for preschool -,
although a detailed analysis is beyond the scope of this paper.
2.3 RESULTS
2.3.1 THE OUTPUT GAP
Output per working age Roma individual is a small fraction of output per individual from the
general population. Table 2-2 summarizes the findings. First, total monthly output per male Roma
worker is Euro 1530, only slightly more than half of total monthly output per male worker from the
general population (Euro 2764). The ratio of total output among female workers is similar, Euro 1054
versus Euro 2149, respectively. However, the differences in total output per average working age
individual (15-64 years) are much larger because men (68%) and women (58%) from the general
population are much more likely to be employed than Roma men (20%) and women (9%). As a result,
average monthly output per working age Roma is a mere Euro 313 for men and Euro 92 for woman, well
below the Euro 1892 and Euro 1239, respectively, among the general population.9
TABLE 2-2: MONTHLY OUTPUT GAP PER WORKING AGE ROMA – 15-64 YEARS
Net Income for Total labor Total output Employment Rate Total output
Primary costs to firmA (labor, capital) (among working (labor, capital,
Occupation (if per workerB age 15-64) profit) averaged
working) over all working
age (15-64)C
Men Women Men Women Men Women Men Women Men Women
General pop. 729 566 1404 1092 2764 2149 68% 58% 1892 1239
Roma 403 278 777 535 1530 1054 20% 9% 313 92
Difference -326 -288 -627 -556 -1234 -1095 -48% -49% -1579 -1147
Source: UNDP/World Bank/EC regional Roma survey (2011), and EU-SILC (Eurostat, 2009). Authors‘ own
calculations. A The total labor costs include net wages, income tax (calculated as a 19% flat tax, paid on taxable
income = gross income – employee social insurance contributions), employee social insurance contributions (13.4%
of gross income), and employer social insurance contributions (35.2% of gross income). B Total output per worker
follows from the ‗total labor costs‘ multiplied by (0.492/0.508). C Total output per working age individual equals
‗total output per worker‘ x ‗employment rate‘ of working age.
This estimation method generates an estimate of per capita GDP that is very similar to official
national account level estimates of per capita GDP. As a robustness check, we compare our GDP
estimate with the IMF (2011) estimate based on official accounts. Among the general population, 71.6% is
of working age (15-64 years). Among the Roma population, the equivalent figure is 58.3%. Averaging
over males and females, total monthly output per working age individual regardless of sex is therefore
9
These estimates are very similar to national account level estimates of output.
22
Euro 1566 and Euro 203, respectively, for the general population and for Roma. This implies that per
capita monthly output is Euro 1121 and Euro 118, respectively, or Euro 13,452 and Euro 1,417 per
annum. If the Roma population is proportionally captured in the EU-SILC, then our estimate of per capita
GDP is equivalent to that of the general population estimate alone: Euro 13,452. If the Roma captured by
the regional Roma survey (2011) are not captured by the EU-SILC, then per capita GDP is equivalent to a
(population) weighted average of the two, equivalent to Euro 12,749 per annum. Both estimates are close
to the official national accounts statistics per capita GDP estimate of Euro 12,661 for 2011 (IMF, 2011).
BOX 2-2: PER CAPITA GDP OF SLOVAK ROMA IN INTERNATIONAL PERSPECTIVE
Per capita output among Roma in Slovakia is comparable to per capita output in sub-Saharan
Africa. Per capita output among Roma is estimated to be a Euro 1107 per annum, a mere 10.5% of the per
capita output estimate of Euro 13,452 for the general population. According to the IMF (2011), Slovakia
ranked 42 out of 181 ranked countries with respect to per capita GDP (in purchasing power parity; i.e.
adjusted for price differences) in 2010; i.e. among the world‘s richest 25% of countries.1 The per capita
output (PPP) among its Roma population1 would place it at rank 143, among the world‘s poorest 25% of
countries.
Total annual economic output gains from equal labor market conditions for Roma and the general
population would amount to Euro 3.1 billion, equivalent to 4.4% of Slovak GDP. Given that
currently, labor market conditions for Roma are much more restrictive than for non-Roma, one can
calculate the ‗gap‘ in output between the current situation and a hypothetical ‗best case scenario‘ in which
wages and employment rates among Roma would equal those among the general population. The monthly
average of this output gap is Euro 1,579 for men and Euro 1,147 for women (v. Table 2-2), or Euro 16,357
per year per working age Roma, regardless of sex. Using the 320,000 total Roma population estimate, and
with approximately 58% of Roma men and women being of working age, this implies that there are
approximately 186,000 working age Roma. Hence, the total gap in GDP amounts to more than Euro 3.1
billion, divided nearly equally between labor income and capital income gains.
FIGURE 2-3: TOTAL OUTPUT GAP AMOUNTS TO EURO 3.1 BILLION ANNUALLY
Productivity Capital
Income Gains; Productivity Labor
€ 1.55 Billion; Income Gains;
49% € 1.60 Billion;
51%
The largest driver behind this gap in total output between Roma and the general population are the
low employment rates. Table 2-3 shows how much economic output per working age Roma would be if
instead of the current situation (row 1), wages were equal (row 2), employment rates were equal (row 3),
or both were equal (row 4). Rows 6 and 7 show that equalization of wages alone would reduce the output
gap by 16% for men and 8% for women, while equalization of employment rates alone would close the
23
output gap by 46% and 45%, respectively. That the two do not sum to 100% is simply a reflection of their
interaction; an increase in wages will lead to a greater reduction in the gap if Roma employment rates also
increase and vice versa.
TABLE 2-3: BREAKING THE OUTPUT GAP DOWN INTO A PRODUCTIVITY AND EMPLOYMENT RATE
COMPONENT
Output (monthly) per working age Roma Men Women Men Women
1 Actual 313 92 1.0 1.0
2 If wages were equal to general population 566 187 1.8 2.0
3 If employment rates were equal to general population 1047 608 3.3 6.6
If wages and employment rates were equal to general
4 population 1892 1239 6.0 13.5
5 Total output difference Roma and general population A 1579 1147
6 % of output gap closed if wages were equal 16% 8%
7 % of output gap closed if employment rates were equal 46% 45%
A
This row displays the difference between scenario 4 and scenario 1 (4-1).
2.3.2 THE FISCAL GAP
In similar fashion, one can calculate the size of a so-called ‗fiscal gap‘: the difference between
current fiscal revenues, and the fiscal revenues that the Slovak government would earn if labor
market conditions among Roma would be similar to those of the general population. This fiscal gap
amounts to approx. Euro 725 million annually, equivalent to 3.1% of Slovak government
expenditures. Per working age Roma, the fiscal gap is nearly Euro 3900 per year. This gap can be divided
into three components: (a) lower income tax and social insurance revenues (approx. Euro 365 million
across all working age Roma, or 50%), (b) lower corporate tax revenues (approx. Euro 285 million or
39%), and (c) higher social insurance payouts (approx. Euro 75 million, or 10%). These will be discussed
in turn.
FIGURE 2-4: TOTAL FISCAL GAINS OF ROMA INCLUSION
Higher Expenditures
on Material Needs
Benefits and Family
Allowance
€ 75 Million, 10%
Lower Income Tax
Revenue and Health
Insurance
Lower Corporate Contributions,
Tax Revenues, € 365 Million, 50%
€ 285 Million,
39%
24
Annual income tax and social insurance revenues would be approximately Euro 365 million higher
if Roma had the same labor market outcomes. With employment rates and wages lower, the first gap in
government revenues stems from lower income taxes and lower social insurance contributions. Income tax
is a flat rate of 19% on taxable income. Social insurance contributions are divided between those paid for
by the employee (altogether 13.4% of gross income) and those paid for by the employer (altogether 35.4%
of gross income). These social insurance contributions cover six areas: (a) health insurance and
hospitalization, (b) retirement insurance, (c) disability, (d) unemployment, (e) accident insurance, and (f)
other. Table 2-4 illustrates, assuming a gross income that is now normalized at 100.
TABLE 2-4: OVERVIEW OF SLOVAK LABOR TAXES
Rate Amount Total
Total labor compensation 135.2
Health Insurance and Hospitalisation 11.4% 11.4
Retirement insurance 14.0% 14.0
Employer paid social Disability 3.0% 3.0
insurance Unemployment 1.0% 1.0
(35.2% on gross income)
Accident insurance 0.8% 0.8
Others 5.0% 5.0
Gross income 100.0
Employee paid social Health Insurance and Hospitalisation 5.4% 5.4
insurance Retirement 4.0% 4.0
(13.4% on gross income) Disability 3.0% 3.0
Unemployment 1.0% 1.0
Taxable income 86.6
Flat Income Tax 19.0% 16.5
Net income 70.1
Source: World Bank. (2011). Doing Business Survey; http://www.taxrates.cc/html/slovakia-tax-rates.html;
http://www.worldwide-tax.com/slovakia/slovakia_tax.asp.
Whether a social insurance tax payment should be included in the fiscal gap calculation depends on
whether these are defined contribution schemes where the benefit amount depends on the years of
contribution and the level of income. For the gap in labor related tax revenues that are part of the fiscal
cost calculations, we conservatively include only the gap in health insurance payment and income tax
payments. Slovakia has a universal healthcare scheme for all residents funded by compulsory insurance
contributions. Since the State pays the contributions on behalf of children, pensioners, unemployed, etc.
(EC, 2011), lower health insurance payments as a result of lower wages and employment rates should be
included. While the State similarly guarantees social (disability) pensions to people without working
histories, pension receipts are importantly dependent on the employment histories. Including these would
therefore bias the overall fiscal gap upwards while excluding these results in a downward bias of the
overall fiscal gap. The monthly fiscal gap resulting from lower income tax and health insurance payments
amounts to Euro 189 and Euro 137 per working age Roma men and woman, respectively. Considering the
approximately 186,000 working age Roma, this translates into a fiscal gap of approximately Euro 365
million per annum.
25
TABLE 2-5: MONTHLY FISCAL GAP PER WORKING AGE ROMA RESULTING FROM LOWER INCOME
TAX AND HEALTH INSURANCE PAYMENTS
Net Wage for Income Tax Health Insurance Employment Average Income
Primary (19% Gross Contribution Rate Tax and Health
Occupation (if Income) (11.4% Ins.
working) Employer, 5.4% Contribution
Employee of (per working age
Gross Income) 15-64)
Men Women Men Women Men Women Men Women Men Women
General pop. 729 566 171 133 160 124 68% 58% 227 148
Roma 403 278 95 65 88 61 20% 9% 38 11
Difference -326 -288 -76 -68 -71 -63 -48% -49% -189 -137
Source: UNDP/World Bank/EC regional Roma survey (2011), and EU-SILC (Eurostat, 2009). Authors‘ own
calculations.
Annual corporate tax revenue would be approximately Euro 294 million higher as a result of equal
labor market outcomes. Slovakia has a corporate tax revenue rate of 19% of capital income. The Table
2-6 shows the estimates for the amount of foregone corporate tax revenues, using the difference between
total output per worker and total labor costs to the firm per worker as the amount of capital income per
worker. Taking into account the differences in employment rates, the table shows that the monthly gap per
working age Roma amounts to Euro 148 for men and Euro 107 for women, which translates into a fiscal
gap of approximately Euro 285 million per annum.
TABLE 2-6: MONTHLY FISCAL GAP PER WORKING AGE ROMA RESULTING FROM LOWER
CORPORATE TAX REVENUES
Total labor costs to Total output Corporate Tax Employment Rate Average Corporate
firm per worker (labor, capital) per Revenue (19% of Tax Revenue per
worker capital income) working age (15-
64)
Male Female Male Female Male Female Male Female Male Female
1404 1092 2764 2149 258 201 68% 58% 177 116
777 535 1530 1054 143 99 20% 9% 29 9
-627 -556 -1234 -1095 -115 -102 -48% -49% -148 -107
Source: UNDP/World Bank/EC regional Roma survey (2011), and EU-SILC (Eurostat, 2009). Authors‘ own
calculations.
Finally, annual social insurance payments by the State would be approximately Euro 75 million
lower, and Euro 125 million in total if family benefits would also equalize. Slovakia has an ―assistance
in material need‖ minimum income social assistance program in place. This is a means tested program,
intended to support the poorest families to provide for basic needs, and the amount of benefit received
varies with family composition10 (EC, 2011). Slovakia also has a number of family benefits in place (EC,
10
The amount is calculated as the difference between household income and the minimum income floor. The
minimum income floor is (EC, 2011): € 60.50 for singles; € 115.10 for single parents with one to four children; €
26
2011): (a) child benefits (a flat rate monthly allowance of Euro 21.99 per child paid until completion of
compulsory education), (b) parental allowance (a flat monthly rate of Euro 164.22 for parents with
children until the age of 3), (c) a child care allowance consisting of a tax bonus providing a reduction in
income tax of Euro 20.02 monthly per child; and, (d) a lump-sum birth grant to pay for purchase of
necessities for a newborn child.11
While the argument for including differences in material needs payments is clear in the fiscal gap
calculations, one can argue against including differences in payments for child benefits, which are a
function of the number and age of children, and do not directly depend on income levels. The
analysis presents both for two reasons: (a) the likelihood that measures to improve labor market outcomes
(such as those aimed at addressing the low education levels of Roma) will reduce family size (and thus
family benefit receipts) is very high; there is a large, global body of evidence documenting the reduction
in family sizes with improved labor market outcomes of women; and, (b) to clarify that any fiscal savings
from reduced family benefit payments is only a fraction of the much larger fiscal benefits stemming from
increased labor tax and corporate tax revenues.
Table 2-7 underscores how poor many of the Roma households are: as many as 40% of Roma
households report receiving social assistance and report the amount, compared with 4% of
households from the general population12. The recipient Roma households also receive more: Euro 235
per month on average compared with Euro 119 for households from the general population. Taking into
account both the likelihood of being in a household that receives social assistance and the average number
of working age individuals per household (2.59 and 2.08, respectively), the table shows that the average
social assistance payout per working age Roma is Euro 36 per month compared with Euro 2 for the
general population. The differences in family benefits are smaller; while the likelihood of being in a
household that receives family benefits is similar (38 and 39%, respectively), the amounts are
considerably higher among Roma households, Euro 234 versus Euro 55, respectively. This most likely
reflects the fact that the average Roma household has 1.81 children under 15 years of age compared with
0.37 for the general population. Altogether, the extra fiscal payments amount to approximately Euro 128
million, with 58% of this – Euro 75 million – as a result of higher social assistance payments, and 42% -
Euro 53 million – from higher family benefit payments.
105.20 for couples without children; € 157.60 for couples with one to four children; € 168.20 for single parents with
five or more children;€ 212.30 for couples with five or more children. In addition, there are several potential
allowances such as a housing benefit and an activation allowance.
11
A more detailed description of Slovakia‘s social protection programs can be found in the chapter on employment
and social protection..
12
The proportion of Slovak Roma households reporting to receive social assistance is higher: 55%. For the
calculations we need to restrict the analysis to households additionally reporting the amount. The implication of this
bias is that the fiscal gap estimated here is lower than the real fiscal gap.
27
TABLE 2-7: MONTHLY FISCAL GAP PER WORKING AGE ROMA RESULTING FROM HIGHER SOCIAL
INSURANCE PAYMENTS
% Households Average Amount (if Average amount per Total Benefit
recipient household) working age (15-64) Receipt (per
household member working age)
Social Family Social Family Social Family
Assistance Benefit Assistance Benefit Assistance Benefit
General pop. 4 39 119 55 2 10 13
Roma 40 38 235 234 36 34 70
Difference 36 -2 116 179 34 24 57
Source: UNDP/World Bank/EC regional Roma survey (2011), and EU-SILC (Eurostat, 2009). Authors‘ own
calculations.
Investing in Roma inclusion is not only equitable, but also fiscally smart. To estimate the fiscal returns
to education investments, the Education Chapter compares 3 scenarios whereby increasingly larger
investments are made to enhance both the length and the quality of education for children in Roma
communities. In each case, the estimates assume investments that exceed the regular per pupil education
expenditures in Slovakia. Each scenario assumes that education completion rates among Roma increase
with increasingly higher investments. To calculate the potential fiscal returns to higher rates of education
completion, the three scenarios use estimated education returns (i.e. how much higher employment rates
and wages are for someone having completed secondary education) and assume education returns
currently experiences among Roma (scenario 1), education returns currently experienced among non-
Roma living nearby (scenario 2), and education returns experienced among the general population
(scenario 3). In each of these cases, the education investments yield substantially greater fiscal returns
from improved employment prospects.
2.4 CONCLUSIONS
The dire conditions of Slovak Roma have macroeconomic consequences, which will grow over time
as more than 13% of new labor market entrants are young, mostly unskilled Roma. The gap in
individual level output translates into an aggregate gap in Slovak GDP of Euro 3.1 billion, equivalent to
4.4% of Slovak GDP. This gap will grow over time as a result of demographics. The extremely poor
employment prospects also result in lower income tax and social insurance revenue, lower corporate tax
revenue, and higher social assistance payments. Altogether, the difference between the current situation
and a hypothetical ‗best case scenario‘ of full Roma inclusion is nearly Euro 3900 per year per working
age Roma. Measured over all working age Roma, this means that Slovak government revenues would be
3.1% higher –Euro 725 million annually – if employment conditions were the same for Roma as non-
Roma. 90% of this gap is driven by lower income tax and social insurance revenue, with higher social
assistance payouts representing only 10% of the total fiscal gap.
28
2.5 BIBLIOGRAPHY
Dollar, David and Roberta Gatti. 1999. ‗‗Gender Inequality, Income and Growth: Are Good Times Good
for Women?‘‘ Mimeograph, World Bank, Washington, DC
Duflo, Esther, and Abhijit Banerjee (2011) ―Poor Economics: A Radical of the Way to Fight Global
Poverty‖. Public Affairs Publishing.
European Commission (2011). ―Your Social Security Rights in Slovakia.‖ DG Employment, Social
Affairs, and Equal Opportunities
Eurostat. (2009). Statistics on Income and Living Conditions – Slovakia sample.
__________ (2011). Labor Force Survey. Retrieved on November 10th, 2011 on
.
__________ (2011). National Account Statistics. Retrieved on November 10th, 2011 on
.
Galor, Oded and David N. Weil. 1996. ―The Gender Gap, Fertility, and Growth‖. American Economic
Review 86(3): 374–87.
Gould, Eric, Victor Lavy, and M. Daniele Paserman. 2004. ―Immigrating to Opportunity: Estimating the
Effect of School Quality Using a Natural Experiment on Ethiopians in Israel.‖ Quarterly Journal of
Economics. 119(2): 489-526.
Gábor Kézdi and Éva Surányi (2009). A Successful School Integration Program. An Evaluation of the
Hungarian National Government‟s School Integration Program. 2005-2007. Roma Education Fund.
Working paper No. 2
Institute for Public Affairs. (2003). ―A Global Report on Roma in Slovakia‖. Eds. VaÅ¡eÄ?ka, M., Juraskova,
M., and Nicholson, T. Bratislava.
International Monetary Fund (2011). World Economic Outlook.
Klasen, Stephan, and Francesca Lamanna. 2011. ―The Impact of Gender Inequality in Education and
Employment on Economic Growth: New Evidence for a Panel of Countries‖. Feminist Economics, 15:3,
91-132.
OECD. www.oecd.org/ctp/taxdatabase
OECD Stats (2012). http://stats.oecd.org/Index.aspx?queryname=345&querytype=view
Slovak Government Office of the Roma Plenipotentiary (2011). 2011 National Roma Integration Strategy
of the Slovak Republic. Slovak Republic.
Statistical Institute of the Slovak Republic. (2001). Population and Housing Census. Retrieved on
November 10th, 2011 on .
__________ (2012) Statistics Portal http://www.statistics.sk/pls/elisw/casovy_Rad.procDlg
29
Senguino, Stephanie. 2000. ‗‗Gender Inequality and Economic Growth: A Cross -Country Analysis.‘‘
World Development 28(7): 1211–30.
UNICEF (2011), TransMONEE database. http://www.unicef-irc.org/databases/transmonee/
United Nations. Population Division of the Department of Economic and Social Affairs of the United
Nations Secretariat, World Population Prospects: The 2010 Revision,
http://esa.un.org/unpd/wpp/index.htm
World Bank (2010). ―Roma Inclusion: An Economic Opportunity for Bulgaria, Czech Republic, Romania,
and Serbia.‖ World Bank Policy Note. World Bank. Washington D.C
__________ (2011). Doing Business Survey. World Bank. Washington D.C
__________ (2012). ―Closing the Early Learning Gap for Roma Children in Eastern Europe ‖. World
Bank. Washington D.C.
30
3 EMPLOYMENT AND SOCIAL PROTECTION
This chapter highlights by how much employment income and different sources of social protection
income contribute to the financial coping strategies of Roma households in Slovakia, as well their non-
Roma neighbors living nearby. The chapter also provides a detailed analysis of employment patterns,
looking at the determinants of which Roma find jobs and which don’t, and the types of jobs that Slovak
Roma have. Among the unemployment and inactive, the chapter explores how far these two groups are
from the labor market in terms of, for example, previous employment experience. The chapter concludes
with examples of policy actions that the Slovak government can consider to boost employment for Roma,
grouped in three types of incremental policy recommendations: (1) improving job search incentives; (2)
improving efficiency of job search – better matching of labor supply and demand; and, (3) improving the
employability of unemployed Roma through skills building. In addition, the chapter calls for investments
in monitoring and rigorous evaluation of employment policy actions, and includes a call to maintain a
strong safety that continues to protect the poor, but replace incentives in the social protection design that
may currently lead to more exclusion with social protection incentives that promote greater investments in
job search, human capital, health, and housing.
3.1 INTRODUCTION
Roma households with employment have lower rates of malnutrition and higher rates of self-
reported happiness and life satisfaction. Among those Roma living in households where nobody is
employed, 46% experience hunger at least once a month compared with 33% among households where at
least one person is employed. In addition, Roma who are employed report that they are ‗happy‘ and
‗satisfied with their life‘ significantly more often than Roma who do not have work.
FIGURE 3-1: HAPPINESS AND LIFE SATISFACTION: EMPLOYED ROMA ARE HAPPIER THAN OTHERS
OF WORKING AGE
A. Happiness B. Life Satisfaction
90 87 90
85 85 82 82 84 83
80 77
Percent
74 80 77
76
75
75
70 68 67
66
70
65
65
60
Not Employed Total 60
employed Not employed Employed Total
Roma Non-Roma Roma Non-Roma
Source: UNDP/World Bank/EC regional Roma survey (2011). Figure 3-1a represents: the proportion of working age
Roma and non-Roma, respectively, who report that overall, they are ‗Quite happy‘ or ‗Very happy‘. Percentages
were calculated separately for those who work and those who do not work. Figure 3-1b represents: the proportion of
working age Roma and non-Roma, respectively, rating their overall life satisfaction as 5 or higher, on a scale from 1-
10, with 1 being the lowest and 10 being the highest. Percentages were calculated separately for those who work and
those who do not work. Sample restricted to one randomly selected adult (16+) individual per household.
31
However, actual employment rates among Roma are extremely low, especially among women, and
lower than among Roma in each of the neighboring countries. Figure 3-2 shows that while 65% of
working age men and 52% of working age women in the general population are employed, only 20% and
9% of Roma men and women are. Among non-Roma neighbors in Slovakia the rates are respectively 66%
and 26%. In total, only 26% of the marginalized Roma households in Slovakia have at least one employed
household member.
FIGURE 3-2: EMPLOYMENT RATES – 15-64 YEARS
A. Men B. Women
80 74 74 80
70 65 66 63 61 66 65 70
60 61 5754
60 60 56 53
52 49 51
50 42 43 50
Percent
Percent
42
37
40 34 40 32
30 30 26 26
20 19 19
20 20 13
9
10 10
0 0
Slovakia Bulgaria Czech Hungary Romania Slovakia Bulgaria Czech Hungary Romania
Republic Republic
General Population Non-Roma Neighbors Roma General Population Non-Roma Neighbors Roma
Source: General Population: Eurostat 2011 Q2; Non-Roma neighbors and Roma: UNDP/World Bank/EC regional
Roma survey (2011). Individuals are considered employed regardless of the nature of employment; i.e. these figures
include informal employment.
These low employment rates do not reflect preferences: the vast majority of Roma express a desire
for stable jobs, similar to the responses by non-Roma neighbors. Consistent with the finding that
Roma with jobs report greater levels of happiness and life satisfaction, 77% of Roma men and 78% of
women report preferring ―Secure employment but low paid‖ instead of ―Having a higher income but
insecure and irregular‖. These responses are similar to the responses by non-Roma neighbors. Comparable
majorities of Roma and non-Roma neighbors similarly prefer ―Having secure employment but having to
be at work 8 hours a day, 5 days a week, and not having the freedom to manage your time‖ compared with
―Having irregular employment but being free to manage your time‖ Data for the age-group 16-24 show a
very similar pattern.
Many Roma men and women are looking for work, but cannot find jobs. 56% of Roma men and 40%
of Roma women of working age participate in the labor market13: only slightly lower than the rates for
non-Roma neighbors: 64% and 49% for men and women, respectively. However, a large majority of labor
market participants is unable to find work; 64% among men and as much as 78% among women14, higher
rates in Slovakia than anywhere else in the region. Using a population estimate of 320,000 Roma living in
13
The ‗Participation Rate‘ among a certain population group is defined as: ‗the share of the working age population
that is either employed or looking for work‘. The working age population includes all individuals aged 15 -64. Those
who are looking for work are also referred to as ‗the unemployed‘.
14
These rates are referred to as ‗unemployment rates‘: the unemployment rate is defined as the share of the
unemployed among those in the labor force (i.e. those aged 15-64 who are either working or looking for work). As
such, the unemployment rates gives indication of the share among those interested in working who are not able to
find work.
32
Slovakia in total, these figures translate into a total of 165,000 Roma without work, divided between
approx. 15,000 jobless in Western Slovakia, 28,000 in Central Slovakia, and as many as 122,000 in
Eastern Slovakia. Specifically among youth – i.e. the age-group 15-24 – unemployment rates are even
higher: 70% for Roma men and 79% for Roma women. Among youth in the general population,
unemployment jumped up from 19% in 2008 to over 33% in 2010 as a consequence of the global financial
crisis (World Bank, 2012: 14).
FIGURE 3-3: PARTICIPATION RATES AMONG ROMA MEN AND WOMEN (%), SPLIT UP INTO
EMPLOYED AND UNEMPLOYED
60 60
22 21
27 16
Percent
Percent
40 36 40 24
18 14
31 20
20 41 42 41 20
33 26
20 19 13 19
9
0 0
Slovakia Bulgaria Czech Hungary Romania Slovakia Bulgaria Czech Hungary Romania
Republic Republic
Roma Men: Employed Roma Men: Unemployed Roma Women: Employed Roma Women: Unemployed
Source: UNDP/World Bank/EC regional Roma survey. Sample restricted to working age population. Exact figures
for participation rates are not displayed. Labels inside the graphs represent the share of the working age population
that is employed (blue) or unemployed (red). In line with the ILO definitions of Labor statistics, a person is
"unemployed" if he/she was did not have a paid job last week and a) did have a job that would start sometime in the
future, or b) did look for a job within the last four weeks. Those in the second category also had to be ready to start a
job within the next two weeks.
Wages among Roma who do work are much lower than among non-Roma (neighbors) with jobs,
making it even more difficult to make ends meet. Monthly earnings among Roma men with jobs are
approx. three quarters of the monthly earnings than a non-Roma neighbor can expect to earn, while these
earnings are only 60% of the earnings among the general Slovak male population with jobs. Among
employed Roma women, the gap with non-Roma neighbors is similar to the gap among men; approx.
three quarters, while monthly earnings are 55% of those among working women in Slovakia as a whole.15
While some of this can be explained by lower education levels, younger age structure etc., the fact
remains is that low wages contribute to the low incomes, and raise the question of how do Roma families
manage to survive?
3.2 MAKING ENDS MEET: FINANCIAL COPING STRATEGIES
The average Roma household lives from approx. €528 per month, compared with €773 for non-
Roma neighbors. When corrected for household size, the differences in per capita incomes are even
larger. Many Roma households are larger in size and its members are younger on average. The ‗adult
15
Source: General Population: EU-SILC. Author‘s calculations. For comparative purposes, values have been
adjusted to 2011 prices using the Harmonized Index of Consumer Prices (Eurostat, 2012). Non-Roma neighbors and
Roma : UNDP/World Bank/EC regional Roma survey (2011). Earnings (taxable monthly income) for the general
population are normalized at 100%. The corresponding rate for non-Roma neighbors and Roma are relative to the
general population. Data is reported for yearly income. Monthly data is calculated by dividing yearly income by the
number of months worked (full- or part-time).
33
equivalent‘ per capita income corrects for these differences, and amounts to €221 and €390 among Roma
and non-Roma, respectively. Put differently, whereas the total average household income for Roma is
68% of that for non-Roma households living nearby, this is only 57% once equivalized income measures
are compared. Table 3-1 further shows the average per capita (adult equivalent) incomes across the five
income quintiles for Roma and non-Roma neighbors. The average per capita income in the bottom quintile
is only €108 among non-Roma neighbors, underscoring the deep levels of poverty among the bottom 20%.
The average income among the next quintile is €249, more than double. Among Roma, poverty is
concentrated among a much larger share: at €252, the average per capita monthly income of the 4 th
quintile among Roma – representing the top 61-80% of the income distribution - is nearly identical to the
bottom 21-40% among non-Roma neighbors.
TABLE 3-1: AVERAGE ADULT EQUIVALENT INCOME LEVELS AMONG ROMA AND NON-ROMA
HOUSEHOLDS (BY INCOME QUINTILE)
Income Quintiles (%) 0-20 21-40 41-60 61-80 81-100
Roma 73 157 206 252 420
Non-Roma Neighbors 108 249 387 511 737
Source: UNDP/World Bank/EC regional Roma survey (2011).
3.2.1 SOURCES OF INCOME: EMPLOYMENT VERSUS SOCIAL BENEFITS
A typical Roma family in Slovakia receives a small share of overall income from employment and
depends importantly on state transfers, with universal family (child) benefits providing a larger
share of income than means-tested social assistance 16 . Among Roma, only 25%, or approx. €130,
comes from employment or other labor activities, compared with 63%, or approximately €488 on average
among non-Roma neighbors (UNDP/World Bank/EC regional Roma survey (2011)). The regional Roma
survey also captured different types of transfers: unemployment benefits, social assistance (including
poverty and local assistance benefits, stipends and scholarships), and universal child allowances. 4 out of
5 Roma households reported one or more of these as major sources of income. Two-thirds report receiving
universal child allowances, and more than half report receiving social assistance (corresponding to Basic
Material Needs), and a total of 80% of households reports at least one type of state benefit as a major
source of income. Figure 3-4 and Figure 3-5 show the breakdown.
16
Reported income figures were calculated from survey information on the major sources of income of the
household. Households were asked to report the amount of income they received from eight possible income sources.
These amounts were only reported if the household classified them as a ‗major source of income‘. The categories
asked for were: income from employment, income from unemployment benefits, income from social assistance,
income from child benefits, income from pensions, income from other labor activities, income from remittances, and
income from any other sources. If a household reported a particular source to be of significant importance for the
household, but refused to report the amount received from that source, the amount was assumed to equal the
conditional average for Roma households in Slovakia – i.e. the average amount of income reported from that
particular source by all Slovakian Roma households who classified this source as a major income source. In total,
between 2% (for pensions) and 29% (for child allowances) of all Slovakian Roma-household observations were
corrected in this way. The same was done for non-Roma households, by taking the conditional average for this group
within Slovakia. In this group, between 1% (for ‗income from other labor activities‘) and 28% (for child allowances)
of the observations were corrected. If a household did not give information about some of the income sources, but
did report other sources, the non-repor ted sources were assumed to be of no- or trivial importance to the household.
34
FIGURE 3-4: COMPARISON OF INCOME SOURCES FOR ROMA AND NON-ROMA NEIGHBORING
HOUSEHOLDS
Roma Households
Average Total Income: €528
Earnings related to employment
1
2 Unemployment benefits
4 1
Pensions
101
130 52 Social assistance
170
111 Child allowance
44
488
46 18 Other labor activities than
132 employment
Remittances
Other
Source: UNDP/World Bank/EC regional Roma survey (2011). Figures are based on unconditional means for Roma
and Non-Roma households in Slovakia. Missings are not included, meaning that 91% of the Roma sample and 93%
of the non-Roma sample were taken into account when calculating averages. Social Assistance includes: Maternity
leave benefits, Poverty and Local assistance benefits, Stipends and Scholarships.
FIGURE 3-5: STATE BENEFITS RECEIVED BY ROMA AND NON-ROMA NEIGHBORING HOUSEHOLDS
A typical Roma household A typical Non-Roma household
receives €346 per month in the vicinity receives €171
from State Benefits per month from State Benefits
Source: UNDP/World Bank/EC regional Roma survey (2011). Figures are based on unconditional means for Roma
and Non-Roma households in Slovakia. Missings are not included, meaning that 91% of the Roma sample and 93%
of the non-Roma sample were taken into account when calculating averages.
The most striking differences between Roma- and neighboring non-Roma households in terms of
income sources occur in the realms of employment and social assistance. Whereas only one in every
five Roma households (21%) reports employment as one of their major sources of income, this figure
reaches 54% for neighboring non-Roma households. The opposite picture appears for social assistance:
55% of Roma households report this to be a major income source, whereas only 23% of non-Roma
households do so.
35
FIGURE 3-6: COMPARISON BETWEEN ROMA AND NON-ROMA HOUSEHOLDS REPORTING THE MAIN
SOURCES OF INCOME
Social assistance, unemployment - and/or child benefits 80 65
Child Benefits 66 53
Social Assistance (all types) 55 23
Employment 21 54
Unemployment Benefits 20 11
Pensions 15 26
Other Labor Activities 6 4
Remittances 2 0
Other Income Sources 0 1
80 60 40 20 0 20 40 60 80
Non-Roma Neighboring Households Roma Households
Source: UNDP/World Bank/EC regional Roma survey (2011). The figure represents the share of Roma- and non-
Roma neighboring households, respectively, reporting each of the listed categories as a major source of income.
Lower bound estimates. Social Assistance includes: Maternity leave benefits, Poverty and Local assistance benefits,
Stipends and Scholarships.
Most households, also those with employment income, depend on a variety of income sources for
their survival. This is shown in TABLE 3-2. For example, among those 1 in 5 Roma households that
receive income from employment, 35% also report receives income from social assistance (column (1)).
Among households receiving social assistance (column (4)), 83% report also receiving child allowance. In
fact, 50% of all households in the sample report receiving both child allowances and social assistance.
Table 3-2: COMBINATIONS OF INCOME SOURCES AMONG ROMA HOUSEHOLDS (COLUMN
PERCENTAGES)
(1) (2) (3) (4) (5) (6) (7) (8)
Other
Employ- Unemploy- Social Child
Pension Remittances Other
ment ment Assistance Allowance
Employment
Emp. . 18% 17% 14% 24% 30% 29% 100%
Unemp. 17% . 21% 17% 24% 37% 36% 0%
Pension 12% 15% . 11% 11% 15% 36% 0%
S.A. 35% 44% 42% . 68% 46% 64% 50%
Child All. 73% 77% 49% 83% . 76% 86% 0%
Other Emp. 9% 11% 6% 5% 7% . 57% 0%
Remit. 2% 3% 5% 2% 2% 17% . 0%
Other 1% 0% 0% 0% 0% 0% 0% .
N 161 154 111 412 502 46 14 2
Source: UNDP/World Bank/EC regional Roma survey (2011). Missings are not included, meaning that 91% of the
Roma sample households were taken into account in the presented frequencies. Social Assistance includes: Maternity
leave benefits, Poverty and Local assistance benefits, Stipends and Scholarships.
36
3.2.2 TYPES OF SOCIAL BENEFITS AND ASSOCIATED RISK REDUCTIONS
In Slovakia, there are three main types of ‗protection-oriented‘ social assistance programs: the
Benefit in Material Need program (BMN), Family Allowances, and Disability Benefits. The Benefit
in Material Need protects families whose income falls below the subsistence minimum, as calculated via
means-testing. It consists of a basic entitlement and supplements. The exact amount that a family receives
is calculated as the difference between the sum of claims and the income of the claimant. Table 3-3
provides an overview of the basic- and additional claims associated to the BMN. For the conditional
housing benefit, conditions include regular payment for housing costs or an agreed scheme of the payment
of housing debt. Like the BMN, Disability benefits are also income-tested, and eligibility criteria are
relatively strict. This benefit includes a caretaker allowance, transportation benefits, and one-off
compensations for larger purchases needed to accommodate the needs of the disabled. The only non-
contributory cash benefit that the disabled person themselves get is the protection allowance, as discussed
under the Benefit in Material Need. (World Bank, 2012)
TABLE 3-3: STRUCTURE OF BMN CLAIMS
Type of Family Basic Type of Additional Allowance: Additional
Claim: Claim:
Individual with no children €60.50 Health care allowance €2 per person
Protection allowance (for people who for health, €63.07
Individual and 1-4 children €115.10 age and other reasons cannot be activated) (€ 34.69)
Individual with 5 or more
children €168.20 Allowance for pregnant woman €13.50
Couple with no children €105.20 Allowance for child aged <1 €13.50
Couple with 1-4 children €157.60 Allowance for school-aged child €17.20
Couple with 5 or more children €212.30 Activation benefit (CONDITIONAL) €63.07
Housing benefit (CONDITIONAL):
- for one person households: €55.80
- for households with several persons: €89.20
Source: World Bank, 2012.
Of these three types, family allowances are the only universally available form of support. Family
allowances are available to any family with children. They may consist of the categories mentioned in
Table 3-4. Since January 2011, a parent is allowed to work while receiving parental allowance benefits.
TABLE 3-4: STRUCTURE OF FAMILY ALLOWANCES
Type of Benefit Conditions Amount
Child benefits Family with children under 18 or studying children €22 per month per child
under 25
Parental allowance Family has at least one child aged <3 (regardless of €190 per month
amount of children in this agegroup)
Benefit for proper care of Alternative to the parental allowance. Reimburses the Up to €230 per month
the child cost of using a surrogate to care for the child
(including child care facilities)
Child birth grants and Universal
adoption benefits
Source: World Bank, 2012
37
Two thirds of the households who receive the benefit in material need are non-Roma: Roma families
constitute only about 35% of all households that receive the benefit in material need in Slovakia. If
one were to assume that the population of Roma in the Slovak Republic is 320,000 individuals, and a total
population of 5,431,024, one obtains that about 38,800 Roma households receive the BMN, while about
72,200 non-Roma households receive the benefit (World Bank (2012f)) (the average household size
among Roma households is about 4.45; while it is about 2.83 among non-Roma households).
Among marginalized Roma households only, about half receive Material Needs Benefits, and
approximately two thirds receive child allowances. In particular, 49% of Roma households living in
‗mixed‘ neighborhoods, receive the Benefit in Material Need (BMN); this proportion increases to 54%,
among Roma households in ‗segregated‘ and ‗separated‘ settlements.17 Among non-Roma households in
the vicinity, only 4% receive the BMN, and only one third receives child allowances. Similarly, about one
third of all surveyed Roma households receive a housing allowance, which is conditional on receipt of the
BMN. Among non-Roma living in the vicinity, only 3% of all households receives housing allowance
(UNDP Slovak Roma survey (2010); Author‘s calculations).
Access to unemployment benefits is linked to very strict conditions in Slovakia. Access to this
contributory benefit is conditioned on contributions for at least two years during the last three years 18 and
the duration of benefits is limited to six months. ―The system is especially restrictive towards labor market
entrants and those working in unstable jobs, typically in low-skilled and low-paid occupations. (…) A
recent comparative study on the strictness of eligibility criteria for unemployment benefits found the
Slovak Republic to have the third strictest eligibility criteria and sanctions (after Portugal and Romania)
among 36 OECD and/or EU member countries covered in the study.‖ (World Bank, 2012: 36). Among the
general population, 12% of all jobseekers receive unemployment benefits, whereas among Roma, this is
9%19 . In OECD countries on average, almost 50% of jobseekers received unemployment benefits in
2007-08 (World Bank, 2012).
BOX 3-1: THE DUAL ROLE OF THE BENEFIT IN MATERIAL NEED SYSTEM
The BMN serves both as a last-resort social assistance program, protecting against chronic poverty,
and as a non-contributory unemployment benefit. Due to the highly restrictive nature of contributory
unemployment benefits in Slovakia, the BMN in fact assumes a dual role: in addition to providing last
resort social assistance to vulnerable and chronically poor families, it also serves as a de facto second tier
non-contributory unemployment allowance. This is caused by long spells of unemployment and the design
of the benefits, as eligibility for the BMN is established through a means-test that relies largely on current
incomes. Consequently, unemployed individuals who are either not initially eligible for unemployment
benefits, or who run out of benefits before they find another job, become eligible for the BMN.
Source: World Bank, 2012: 36.
17
Roma living in ‗mixed‘ neighborhoods are those living among the majority populatio n. Roma living in
‗segregated‘ neighborhoods are those living in a settlement at some distance from the town or village, or separated
from it by some barrier. Roma living in ‗separated‘ neighborhoods are those living in a certain part of the town or
village – either on its outskirts or within it – without a clear barrier, but in a concentrated settlement.
18
Two years in the last four years in the case of temporary employment or for people who voluntarily contribute to
unemployment insurance. Until 2011, the requirement was at least 3 years of unemployment insurance contributions
during the last 4 years.
19
This number is likely an overestimate and may include people reporting the activation allowance.
38
Moreover, the system of taxes and benefits in the Slovak Republic seems to be biased against part-
time employment at low wages: exactly the types of jobs that many Roma are engaged in. For a
single person receiving the BMN and a housing- or activation allowance, the marginal effective tax rate
(METR)20 for taking up a part-time job at the minimum wage is close to 80%. This means that if such an
individual were to move from social assistance to a part-time job at minimum wage level, 80% of the
additional income received from this change in lifestyle would be ‗taxed away‘. Indeed, the World Bank
(2012) study on Slovakia‘s tax-benefit system concludes that ―existing instruments to ‗make work pay‘
are not sufficient and not very effective‖, mainly because benefits aimed at easing the transition – such as
tax credits – are not substantial enough and have a too short maximum duration (World Bank, 2012: 39).
BOX 3-2: ATTITUDES TOWARDS SOCIAL ASSISTANCE IN SLOVAKIA
In Slovakia, many people believe that the unemployed do not really try to find work, and that social
assistance is too costly to the state. ―In the 2008 European Social Survey, one-third of Slovaks reported
that the primary reason people live in need is because they are either lazy or lack willpower. In
comparison, only slightly more than 10% of Norwegians report the same. (…) In the Slovak Republic,
perceptions that the unemployed are not motivated to find work are particularly strong, with two-thirds of
respondents claiming that the unemployed are not seeking employment. Furthermore, 40% report that
social benefits make people lazy.‖ (World Bank, 2012: 31) Many citizens of the Slovak Republic are of
the opinion that social benefits are too costly for the state, thus placing a strain on the economy, and that
the Benefit in Material Need poses work disincentives (World Bank, 2012).
However, unlike what is often assumed, social benefits do not pose a large burden on Slovakia‘s
state budget, and the most costly component is the universally targeted family benefit. As such, the
benefits received more exclusively by the poor, among which poor Roma households, only constitute
a very small proportion of the total state budget. However, spending on social assistance is relatively
low compared to most EU countries: The Slovak Republic spends about 1.8% of GDP on the three non-
contributory benefits outlined above (World Bank, 2012: 19). Family allowances, which are universal,
absorb almost two-thirds of this amount. Moreover, empirical evidence on potential work disincentives
arising from receipt of the BMN is inconclusive (World Bank, 2012: 6).
3.3 EMPLOYMENT AND WELFARE
How do Roma households with and without employment differ? First, employment raises household
incomes. The typical per capita income of a Roma household without employment is €202 (equivalized),
whereas for a Roma household with employment, it is €275, about 37% higher. 21 Roma households
without employment typically earn an income of slightly below €500 a month. For non-Roma households
without employment, this is slightly below €700. When comparing these amounts to the incomes earned
by families with employment, there is a steep increase for both Roma and non-Roma: 46% and 60% on
average, respectively.
20
METR represents ―the fraction of any additional earnings that is ―taxed away‖ by the combined effects of taxes
and benefits withdrawals‖ (World Bank, 2012: 39). For more details, please see
http://www.oecd.org/els/social/workincentives#models.
21
Based on median income levels, equivalized using the OECD adjusted scale.
39
Roma households at low income levels generally depend on social transfers (universal and means
tested) for almost all of their income. Those Roma households that earn a larger share of their
income from employment are also the better off households. Among non-Roma households nearby, the
share of income obtained from social transfers is very high only in the lowest income quintile, i.e. among
a relatively small group of very poor households. Among Roma households this pattern is sustained also
across higher income quintiles: up to and including the fourth income quintile among Roma, the
predominant share of income comes, on average, from social transfers. Only the relatively better off Roma
families show a different income pattern: these families earn about half of their income from employment,
and the other half from social transfers.
FIGURE 3-7: INCOME SHARES FROM SOCIAL TRANSFERS, EMPLOYMENT AND PENSIONS – ROMA VS.
NON-ROMA NEIGHBORS
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
1 2 3 4 5 1 2 3 4 5
Roma Income Quintiles Non-Roma Income Quintiles
Income from Social Assistance
Income from Employment Income from Pensions
Source: UNDP/World Bank/EC regional Roma survey (2011).Adult equivalent income levels were used to calculate
the share of income, for each household, from each of the different income sources. These shares were then averaged
by income quintile, for Roma and non-Roma households separately. Roma and Non-Roma income quintiles were
calculated separately: this means that among Roma, the income levels represented by each quintile are in fact lower
than among non-Roma. Social Transfers includes: Social Assistance, Unemployment Benefits and Child Benefits.
Employment includes: income from Employment and from Other Labor Activities.
A Roma household in which the household head is employed is 24% less likely to depend on social
assistance than a Roma household where the household head has no job, keeping other background
characteristics the same. This finding is based on regression estimates that explore the correlation
between social assistance and household background characteristics, including employment. As shown in
Annex Table 3.1, the association is stronger than among non-Roma neighbors, for which the chance of
depending on social assistance goes down by only 19% if the household head is employed (Model 4).
Overall, Roma households are 21.6 percentage points more likely to depend on social assistance than non-
Roma, after controlling for background characteristics (Model 2).
Household employment is strongly correlated with household size, education, financial security, and
housing conditions, not with age of the household head. As shown in Table 3-5, households without
employed individuals are larger than those with employed individuals: these two groups respectively have
4.5 versus 3.2 members on average, reflecting mostly larger numbers of children – 2.1 versus 1.1 children
40
on average – among households without employment. In Roma households with at least one employed
person, education levels of the household head and the adult household members overall are generally
higher than in households without employment. Still, the most striking differences with regards to
education remain between Roma and non-Roma households rather than between Roma households with
and without employment. There is no difference in the age of the household head for these two groups of
households. Financial security improves; a much larger share of Roma households in which at least one
person is employed saves, and also have less difficulty to pay for expenses such as mortgages, rent and
utility bills. And, housing conditions improve: 79% with employment live in a well maintained dwelling
compared with 51% without employment.
TABLE 3-5: BACKGROUND CHARACTERISTICS OF HOUSEHOLDS WITH AND WITHOUT EMPLOYMENT
Roma Non-Roma Neighbors Total
Without With Without With Roma Non-Roma
Employment Employment Employment Employment Neighbors
Household Head – Av. Age 39 40 46 44 39 45
Share of households – head
has completed at least some 11% 38% 59% 86% 18% 74%
secondary education (%)
Average Household Size 4.5 3.2 5 3.6 4.6 3.4
Average No. Of Children per
2.1 1.1 2.1 1.1 2.1 1.1
household
Av. share of working age
household members who are 0% 55% 0% 66% 14% 38%
employed (%)
Household has savings (%) 5% 20% 34% 49% 9% 43%
Household lives in well
51% 79% 87% 96% 58% 92%
maintained dwelling (%)
Source: UNDP/World Bank/EC regional Roma survey (2011).
3.4 JOB PATTERNS
This section explores in more detail employment patterns among employed Roma, determinants of
employment, and the distance to the labor market among unemployed Roma.
3.4.1 DETERMINANTS OF WHO WORKS
Employment rates among Roma are especially low in the eastern region, both relative to other
regions and relative to non-Roma nearby. Figure 3-8 depicts employment rates for Roma men and
women in the three regions of the Slovak Republic. Using a population estimate of 320,000 Roma living
in Slovakia in total, these figures translate into a total of approximately 14,500 Roma in Western Slovakia
without employment (unemployed and out of the labor force), 28,000 in Central Slovakia, and as many as
122,000 in Eastern Slovakia. Among non-Roma neighbors, the employment rate in Eastern Slovakia is the
41
same as that in Central Slovakia, suggesting that the low employment rates among Roma living in Eastern
Slovakia must be explained, at least in part, by factors other than a regional labor market with a low
supply of jobs.
FIGURE 3-8: EMPLOYMENT RATES, BY REGION AND GENDER, ROMA AND NON-ROMA NEARBY
Percent
80
68
70
60 54 54
50 41 38 37
40
30 26 24
21 19
20
9 7
10
0
Men Women Men Women Men Women
Western Region Central Region Eastern Region
Roma Non-Roma
Source: UNDP/World Bank/EC regional Roma survey (2011). Sample restricted to Roma subjects in the agegroup
25-64.
Even when differences in background characteristics are taken into account, Roma are much less
likely to be employed than non-Roma, and Roma women are much less likely to be employed than
Roma men. The first estimation model predicts ‗being employed‘ based on a number of background
characteristics, including ethnicity. It shows that Roma ethnic origin lowers the chances of being
employed by almost 16 percentage points; given employment rates of 20% among men and 9% among
women, this reduction is extremely large. When restricting the analysis to Roma only (Model 2), women
with similar background characteristics as men are 14 percentage points less likely to work; again, this is a
very large effect considering the low employment rates. It underscores the challenges that Roma women
especially have in accessing the labor market, and is consistent with qualitative experience by a Slovak
Roma community assistant for medical education, described in the box below.22
Education level is also a large and significant determinant of employment among Roma, although
the effect size is larger among non-Roma. Annex Table 3.2 shows the probability of being employed as
a function of personal and community characteristics. The estimation models predict ‗being employed‘
based on a number of background characteristics. What language is spoken at home (i.e. whether this is
Romani or not), or whether the dominant ethnic identity of the settlement in which one lives is Roma, is
not correlated with the employment probability. It also does not matter whether one lives in a rural or
urban setting. Instead, the only factor that has an effect besides gender and the region is education level:
Roma who did not complete primary school are 6 percentage points less likely to find employment than
those who did, whereas those who complete secondary school or higher are 13 percentage points more
likely to find employment than those who have completed only primary education. The correlation
22
The employment gap between men and women among non-Roma is larger than among Roma, also after
controlling for background characteristics (results available upon request). However, this reflects in part the very low
employment levels among both Roma men and Roma women.
42
between education and employment is larger – almost twice as large - among non-Roma as shown in
Model 3 of the table.
BOX 3-3: A ROMA WOMAN AT WORK EXPLAINS THE PRESSURES SHE FACES
Denisa Gáborová is 35 and works as a community assistant for medical education. She is a Roma who
was raised in a predominantly non-Roma neighborhood. Denisa has four children: two are in secondary,
and the other two in primary school. Her husband drives a tractor for the local landscaping department.
It took Denisa years to get used to ‗settlement life‘ once she got married: the lack of infrastructure and
facilities, and the fact that all residents were Roma were things she wasn‘t used to. When she went to
work, other residents of the settlement scolded her children. Her son once said: ―Mama, do you know
what they told me? That you don‘t take care of me, that you go to some training and don‘t care for us.‖
Many family-members also had doubts about Denisa‘s ambitions. ―In the Roma community, people think
that a wife should be at home, cooking, caring for her children and husband.‖ But with time, people
started to view her differently. When Denisa‘s father-in-law had an accident and broke his ribs, they ran
into one of Denisa‘s colleagues at the hospital. ―She said hello to me, and my husband and father -in-law
were proud that such people greet me and know me and value me.‖
Source: MECEM, Roma Press Agency: 2010.
A majority of Roma report labor market discrimination, both when they look for work and on the
workplace. The large employment gap between Roma and non-Roma controlling for background
characteristics can be explained by a number of factors, including discrimination in the labor market.
Qualitative information suggests that discrimination is indeed a substantial barrier to Roma looking for
work and those with jobs. According to the regional survey, among Roma who looked for work
somewhere in the past 5 years, 78% report that they have experienced discrimination because of their
ethnicity. 12% of non-Roma men report the same. Among those who already had a job in the past 5 years,
the corresponding figures are 57% and 4%23.
3.4.2 JOB CHARACTERISTICS
The large majority of Roma who work are in formal employment, but informality is relatively high,
and most work tends to be temporary or seasonal. Approx. 80% of employed Roma has a written
contract, compared with 95% of non-Roma neighbors. However, in 24% of cases for Roma men with jobs
and 17% for Roma women, the employers do not pay social contributions for pension or healthcare. The
corresponding numbers among non-Roma nearby are 7% and 11%, respectively. Furthermore, as shown in
Figure 3-9, only 38% of men and 44% of women report being in ‗permanent‘ employment, with the rest
being in temporary, seasonal, and periodically (from time to time) employment.
23
Source: UNDP/World Bank/EC regional Roma survey (2011). Sample restricted to one randomly selected adult
(16+) individual per household..The share of Roma individuals who have looked for work in the past 5 years is 52%;
for non-Roma neighbors, this is 24%. The share of Roma individuals who have had a job in the past 5 years is 25%;
for non-Roma neighbors, this is 37%.
43
FIGURE 3-9: CONTRACT DURATION AMONG EMPLOYED ROMA AND NON-ROMA NEIGHBORS
A. Men B. Women
100 100
80
80 80 69
Percent
Percent
60 60
44
38 35 36
40 40
21
20 11 14 12 15
5 20 6
3 4 5
0 0
Permanent Temporary Seasonal Periodically Permanent Temporary SeasonalPeriodically
(from time (from time
Roma Non-Roma Neighbors to time) Roma Non-Roma Neighbors to time)
Source: UNDP/World Bank/EC regional Roma survey (2011). Sample restricted to employed subjects (age-group:
15-64).
The vast majority of employed Roma are unskilled workers, and Roma mainly work in
construction, public utilities and mining. These characteristics clearly show that the lack of education
among Roma translates into low-skilled jobs – if Roma find work at all. Hardly any Roma work in social
services such as leisure services (3% of total), healthcare (2%), or education (2%).
TABLE 3-6: OCCUPATION AND INDUSTRY OF EMPLOYED ROMA
Occupation % Industry %
Unskilled worker 58 Construction 25
Skilled worker 20 Public utilities 25
Semi-skilled worker 9 Industry or mining 19
Professional / Office worker 5 Trade / other commercial services 9
Civil servant 3 Agriculture and forestry 8
Owner of own business 2 Transportation 6
Landless worker 2 Leisure services (tourism, etc.) 3
Foreman, technician 1 Healthcare 2
Other 0 Education and science 2
Other 1
Source: UNDP/World Bank/EC regional Roma survey (2011). Sample restricted to employed subjects (age-group:
15-64).
Self-employment rates among Roma and non-Roma are extremely low. The proportion of Roma
which are self-employed is negligible. When translating the rate of self-employment into a numeric
estimate, the approximate number of self-employed Roma is 750, among a total group of 29,000 Roma
with jobs.24 This suggests that there is scope for more self-employment among both groups. At the same
time, it should be noted that realizing higher self-employment rates is only feasible if other preconditions,
such as availability of start-up capital and sufficient levels of entrepreneurial skills and financial literacy,
are met.
24
An estimate of 320,000 Roma living in Slovakia was used.
44
FIGURE 3-10: TYPES OF EMPLOYMENT AMONG ROMA AND NON-ROMA NEIGHBORS
Roma Non-Roma Employee in private
company
2
2 Employee in public sector
3 4 3
Self-employed in own
41 25 business
55
66 Employer in own
business with employees
Other
Source: UNDP/World Bank/EC regional Roma survey (2011). Sample restricted to employed individuals (age-
group: 15-64).
3.5 DISTANCE TO THE LABOR MARKET AMONG THE UNEMPLOYED AND
INACTIVE
In order to design targeted policy-measures, it is useful to make a distinction between those who are
‗close to‘, and those who are ‗far from‘ the labor market. Among the unemployed Roma, some will
find it very difficult to find work, even with the help of the state and civil society organizations. For
example, individuals who have been without a job for long and have little work experience. A similar
argument can be made regarding education: the higher one‘s education level, the more chances one has of
finding work. As such, groups who are at different ‗distances‘ from the labor market are likely to require
different types of policy-responses. For this reason, the following section will analyze distances to the
labor market among the working age Roma in Slovakia; indicators that capture some basic information on
a person‘s chances of finding a job such as the amount of time during which an individual has not worked
(since recent work experience is often seen by potential employers a s an indicator of having ‗up to date‘
knowledge and skills).
Most Roma without a job are either inactive or very long-term unemployed: precisely the groups
that have the worst chances of getting into employment. Among the entire working age population,
shares of ―short-term‖ unemployment (defined here as < 2.5 years25 because few are unemployed between
1 and 2.5 years) are low: 10% of men and 2% of women, with a majority of these having been
unemployed for less than 1 year. These proportions are fairly similar between Roma and non-Roma
neighbors. However, as many as 25% of all Roma men of working age and 29% of Roma women are very
long-term unemployed (>2.5 years). Among non-Roma neighbors, the corresponding rates are 9% for men
and 13% for women. In addition to (very) long-term unemployed, there is also a large group of working
25
The cutoff at 2.5 years rather than more standard cutoffs such as 1 or 2 years was chosen because data were only
available on the year in which a person last worked, i.e. 2009 or 2010. Since the survey was conducted in May/June,
the maximum time duration of unemployment for someone reporting 2009 as the year in which they last worked is
2.5 rather than 2 years. The cutoff was set at 2.5 years rather than 1.5 years because the group pf 1.5-2.5 years
unemployed was very small, as was the group of 0-1.5 years unemployed.
45
age Roma outside the labor force: i.e. inactive. 44% of men and 61% of women compared with 37% and
51% among non-Roma men and women.
TABLE 3-7: DISTANCE TO THE LABOR MARKET: ROMA AND NON-ROMA NEIGHBORS (% OF
WORKING AGE POPULATION)
Men Women Total
Roma Non-Roma Roma Non-Roma Roma Non-Roma
Employed 20% 46% 9% 29% 14% 37%
Short-term Unemployed 10% 9% 2% 6% 6% 7%
Long-term Unemployed 25% 9% 29% 13% 27% 11%
Inactive 44% 37% 61% 51% 52% 44%
Total 100% 100% 100% 100% 100% 100%
N 1,070 408 1,098 439 2,168 847
Source: UNDP/World Bank/EC regional Roma survey (2011). ‗Short-term unemployment‘ is defined as being
unemployed for less than 2.5 years. Within the group of short-term unemployed Roma (6% of the working-age Roma
individuals), the largest group (4%) has only been employed for less than 1.5 years, whereas a minority (2%) has
been unemployed for 1.5-2.5 years. 14% of the Roma in Slovakia who are not employed did not report the year in
which they last worked. These individuals were assumed to belong to the group of long-term unemployed. The same
procedure was followed for non-Roma neighbors.
In total, there are about 12,000 short-term unemployed Roma in Slovakia, and about 53,000 long-
term unemployed. As many as 104,000 Roma do not participate in the labor force (estimated).
Among the short-term unemployed, the biggest group belongs to the age-group 25-40 (approximately
6,500 people). The same holds for the long-term unemployed: an estimated 26,000 long-term unemployed
Roma belong to this age category. Among the inactive, most individuals are younger: the biggest group
here is aged 15-24 (approximately 41,500 individuals).
TABLE 3-8: DISTANCE TO THE LABOR MARKET: WORKING-AGE ROMA AND NON-ROMA NEIGHBORS
(POPULATION ESTIMATE)
Men Women Total
Employed 19,500 9,000 28,500
Short-term UnemployedA 9,700 2,000 11,700
Long-term Unemployed 24,300 29,000 53,400
Inactive 42,900 61,000 103,900
Total 96,500 101,000 197,600
Source: UNDP/World Bank/EC regional Roma survey (2011). Estimates based on a total estimated Roma population
of 320,000 in Slovakia.
46
TABLE 3-9: DISTANCE TO THE LABOR MARKET: AGE COHORTS AMONG ROMA (POPULATION
ESTIMATE)
Agegroups Employed Short-term Unemployed Long-term Unemployed Inactive Total
15-24 4,700 2,100 11,600 4,100 60,000
25-40 14,600 6,500 26,200 35,900 83,300
41+ 8,800 3,700 15,400 26,000 54,100
Total 28,200 12,400 53,300 103,500 197,600
Source: UNDP/World Bank/EC regional Roma survey (2011). Estimates based on a total estimated Roma population
of 320,000 in Slovakia.
Roma who are long-term unemployed or inactive have very little work experience. Among short-term
unemployed Roma, the average duration of men and women‘s life-time work experience is approximately
9 and 6 years, respectively. For those who are long-term unemployed, this drops to 6 and 3 years. For the
inactive, rates are similar to those of the long-term unemployed, at 5 years on average for men and 3 years
for women. As such, those who are at a greater distance from the labor market in terms of the duration of
their unemployment and their activity status are also the ones with the least work experience, on average.
Figure 3-11b corrects these figures for age, showing the proportion of one‘s working life during which the
subject has been employed. This proportion is already low for the employed and short-term unemployed,
at 40%, and drops to 20% among men and 10% among women among the long-term unemployed and
inactive. In part this reflects that among these groups, approx. half has never worked before, compared
with 30% among non-Roma neighbors.
FIGURE 3-11: AVERAGE NUMBER OF YEARS WORK EXPERIENCE: ROMA
A. Av. No. of Years Work Experience B. Share of Working Age Life Spent Working
12 50
9.8 40 40 40
10
Av. no. of Years
8.6 40
7.7
8
Percent
5.7 6 30
6 5 20 20 20
4 3.3 3 20
10 10
2 10
0 0
Employed Short-term Long-term Inactive
Unemployed Unemployed
Men Women Men Women
Source: UNDP/World Bank/EC regional Roma survey (2011).
Roma who are short-term and long-term unemployed or inactive also have very low levels of
secondary education completion. The main difference in this respect occurs between those with, and
those without jobs. Among the employed, secondary school completion rates are about twice as high as
among those without jobs. For women, secondary school completion is again lower among those who are
long-term unemployed or out of the labor force as compared to the short-term unemployed, whereas for
men, rates are fairly similar across these three groups. Among non-Roma who are out of the labor force,
the secondary school completion rate is 60% for men and 62% for women.
47
FIGURE 3-12: SECONDARY SCHOOL COMPLETION AND DISTANCE TO THE LABOR MARKET
45 39
40
33
35
Percent
30
23
25
20 17 17
13 15 14
15
10
5
0
Employed Short-term Long-term Inactive
Unemployed Unemployed
Men Women
Source: UNDP/World Bank/EC regional Roma survey (2011).
Professional skills are best among the employed Roma, but do not get worse as the distance to the
labor market increases. Shares of unemployed Roma who have received adult learning courses or an
apprenticeship are similar for all groups without jobs; approximately 8% compared with approximately
18% among those with jobs. Computer literacy rates are higher, between one-quarter and one-third, with
rates that are not much higher among the employed. The learning course completion- and computer
literacy rates are about twice as high for non-Roma neighbors as for Roma in all categories (not shown).
FIGURE 3-13: ADULT LEARNING COURSE / PROFESSIONAL APPRENTICESHIP AND COMPUTER
LITERACY
A. Apprenticeship B. Computer Literacy
50
45
40 45 41
Percent
Percent
35 40
30 33
35 29
25 19 17 29
20 30 24 23
15 25 20
7 8 8 8 7 7 20 16
10
5 15
0 10
Employed Short-term Long-term Inactive 5
Unemployed Unemployed 0
Employed Short-term Long-term Inactive
Unemployed Unemployed
Men Women Men Women
Source: UNDP/World Bank/EC regional Roma survey (2011). Apprenticeship: ―Has s/he ever received adult
learning courses or professional apprenticeship (formal or informal of any kind?‖. Computer Literacy: ―Is s/he able
to use a computer word processing program?‖
Family size increases with distance from the labor market among women with employed women
having the fewest number of children and inactive the highest number of children. Among Roma
women who are employed, the average number of children in the household is 1.4. As the distance to the
labor market becomes larger, Roma women live in households with more children, on average: 1.5 among
48
short-term unemployed, 2.0 among long-term unemployed, and 2.3 children among inactive women.
Among non-Roma women, this increase by labor market distance is smaller, from 0.7 children among the
employed to 1.2 among the inactive.
3.6 POLICY RECOMMENDATIONS
Given the prevailing low skill levels and the extent of labor market exclusion (mainly among Roma,
but also among very poor non-Roma), the Government may consider a more holistic approach to
addressing unemployment and inactivity by offering complementary employment services . The
following recommendations to improve employment outcomes among the existing working age Roma
population are geared toward three main objectives: (1) improving job search incentives; (2) improving
efficiency of job search by allowing better matching of labor supply and demand; (3) improving skills;
and, (4) invest in monitoring and evaluation of specific activation measures, and systematically share best
practices across municipalities. However, given the extent and depth of the challenge to improve
employment outcomes in the short-run, the analysis also strongly calls for a fifth objective: (5)
maintaining a strong safety net that continues to protect the poor and combines it with targeted measures
that promote health and human capital investments, especially for children.
3.6.1 POLICY MEASURE 1: IMPROVE JOB SEARCH INCENTIVES
The current system of providing activation allowances to Roma to participate in unskilled menial
jobs is not effective because they provide disincentives to search for work and they do not build
skills that improve employability. First, activation programs in Slovakia such as the small municipal
works program have been found to weaken the incentives to take up ‗real‘ jobs (World Bank, 2012). Most
Roma out of work are registered with employment agencies much more often than non-Roma. High rates
of registration are not surprising since this is required to be eligible for the activation allowance of €63.07
per month for a period of 6 months at a time to participate in the (often menial) municipal works. The
incentives provided through these activation allowances discourage job search since income support is
more generous to ‗jobless‘ individuals on activation than people who have recently found work (e.g. the
form of tax incentives or continuation of some social transfers). Hence, beneficiaries who can only access
low wage, insecure jobs may be better off staying on benefits and receiving additional support through
activation allowances (ibid: 39). Furthermore, most of the benefits administered by the Slovakian
government are unconditional.
A combination of more conditionality – to look for work, participate in apprenticeships etc. - and
continued partial coverage upon finding employment for a significant amount of time can aid in
activating jobless Roma. The World Bank (2012) study on Slovakia‘s tax-benefit system concludes that
―existing instruments to ‗make work pay‘ are not sufficient and not very effective‖, mainly because
benefits aimed at easing the transition – such as tax credits – are not substantial enough and have a too
short maximum duration (World Bank, 2012: 39).
Furthermore, job search incentives, especially for Roma women, and addressing the large gap in
pre-school are linked. The situation of Roma women as caretakers for the children warrants particular
attention. The data showed that family size and inactivity are strongly linked. Furthermore, as shown in
the education chapter, enrolment rates of Roma children aged 3-6 in preschool is extremely low, reflecting
in part a self-reported social desirability by Roma mothers to raise children at home. From this
perspective, addressing the gap in preschool participation is not only essential to ensure Roma children get
49
an equal start with regards to early learning and succeed in school later on (see education chapter for
details), but also addresses an important barrier to labor force participation among Roma women.
3.6.2 POLICY MEASURE 2: IMPROVE EFFICIENCY OF JOB SEARCH – BETTER
MATCHING OF LABOR SUPPLY AND DEMAND
Effective skill building requires addressing counseling the unemployed. In addition to regular
counseling, the employment office can consider providing soft job search skills such the ability to write a
good CV, to identify potential employment opportunities, write an application letter, and present oneself
well in an interview. These can be provided through specific and short job counseling programs, while
‗regular‘ intensive job counseling aims to identify job opportunities and encourage the unemployed to
pursue these.
The government can play a key bridging role enabling employers to identify prospective Roma
employees eager and able to work by reaching out to non-State actors and government entities well
known with the local Roma community. Many Roma are eager to work, and while discrimination is a
barrier to employment prospects, there are also employers interested in hiring Roma men and women, not
only for corporate social responsibility reasons, but simply because they need good productive workers to
contribute to their firms. Information about employment – and employee – opportunities can be a major
barrier. To identify prospective employees, the US Steel factory in Kosice, for example, turned to a local
church in Kosice with a strong presence in a large Roma neighborhood in Kosice. In this case church
officials were able to be a bridge between the private company US Steel, which had no particular
knowledge or expertise identifying motivated and skilled (in the formal but more often in the informal
skill sense) Roma; for example, eagerness to learn new skills on the job, or having obtained non-formal
skills through previous work experience that would be relevant for US Steel. Similarly, other non-State
actors, but also government social workers and community mediators can provide a bridging function for
employment services.
Rather than having each individual firm identify a non-State actor or State actor with knowledge of
the local communities, the employment offices could perform a bridge between these two sides of the
market, and develop a jobs platform – in the form of, for example, a database and/or job fairs – where
firms or government institutions looking to hire workers can communicate with entities that have good
knowledge of local communities to get recommendations for specific job applicants. Note that this is
much like the regular job market with job applicants providing references such as former employers or
university faculty where they have trained. The main difference is that the employment offices actively
build a network of local references that employers can turn to. Of course, this requires building in
measures that provide transparency and ensure that corruption is avoided (e.g. local Roma paying ‗job
brokers‘), but this can be addressed, for example, by relying on multiple sources for references and using
entities that have a clear reputation to keep up.
Similarly, establish a knowledge portal of ―good practices‖, including of municipal activation activities
and actively disseminate these among all the relevant actors. As described below, some mayors are
implementing innovative practices, often taking advantage of the municipal activation allowances. Rather
than having Roma sweep streets or do other menial tasks that do not build skills, these mayors are
employing Roma in, for example, the construction of municipal buildings, and are providing them with
the necessary on the job training that enables them to undertake real construction activities. These types of
experiences should be shared much more broadly across the different municipalities.
50
3.6.3 POLICY MEASURE 3: IMPROVE EMPLOYABILITY THROUGH SKILL BUILDING
Improve the employability of excluded Roma by strengthening skills: This could be achieved
through employment attachments and second-chance education. Given the extremely low education
levels and lack of work experience among unemployed and inactive Roma, job search incentives must be
combined with programs that build employable skills. This includes introducing a robust system of second
chance education, as also recommended in the June 2012 European Council Country Specific
Recommendations to Slovakia. To achieve this, apart from designing and developing a network of second
chance education provider, the current legislative restrictions would need to be amended.
Second chance education may also build on creative solutions being undertaken in some
municipalities to build sector specific skills through employment attachments or use of activation
works to meaningfully build construction skills. For example, a few municipalities have set up social
enterprises in which Roma on activation allowances carry out real municipal construction works,
supervised by a qualified person, with a strong emphasis on on-the-job training. But these types of work
arrangements could also be linked with activities undertaken by non-State actors such as private firms and
NGOs, effectively creating leveraging activation allowances to create traineeship opportunities that can
act as a pathway to employment. Establishing certain standards that these traineeships should meet can
also open the door for certification of skills learned through non-formal means.
A special focus on the social sectors is particularly relevant to promote employment opportunities
for women and to improve the perception of working Roma women. Recall a negligible number of
Roma men and women work in education or health. Providing training, especially to women, to become
community health assistants, or (kindergarten) teacher assistants is likely to generate a significantly
positive impact on the perception of the Roma among the general population, and of working Roma
women among the Roma communities themselves. In addition, future government plans to hire social
workers in the areas including marginalized Roma communities can also put an emphasis on hiring from
the communities themselves so as to ensure better linkages and communication with community members.
Finally, the Government of Slovakia may wish to consider developing trainee, internship and placement
programs for Roma youth in central administration, regional and municipal positions. Again, such
programs could be developed in collaboration with NGOs active on the field of Roma education and
employment. In addition, future government plans to hire social workers in the areas including
marginalized Roma communities can put an emphasis on hiring from the communities themselves so as to
ensure better linkages and communication with community members. An example of a program that has
taken a holistic approach toward improving employability is Spain‘s Acceder model described below,
which has drawn on considerable European Social Funds to support employment prospects for Roma. In
Slovakia, such a program could draw on the capacity of the Social Development Fund of the Slovak
Republic, and seek partnerships with non-State actors such as NGOs, churches, private businesses.
BOX 3-4: AN INTERNATIONAL G OOD PRACTICE: SPAIN‘ S ACCEDER MODEL 26
Acceder is a program designed by Fundacion Secretariado Gitano in Spain aimed at enhancing the access
of the Roma population to mainstream training and employment as an alternative to self-employment or
26
This box is based on a presentation about the Acceder program in the Decent work for Roma conference (Skopje
December 1-2, 2011) and background materials (Framework Document by Jose Manuel Fresno and the technical
staff of FSG; and Methodological model for the socio-labor market integration of Roma) prepared for the EURoma
working group meeting on employment (March 11-13, 2009).
51
family business. It comprises components that could be drawn upon in the context of Roma employment
in Slovakia.
A comprehensive program supporting employment: Although conceived in times of a positive
economic outlook and dynamic labor market, Spain‘s Acceder program remains one of the most robust
and comprehensive programs aimed at facilitating Roma employment because of its holistic approach.
Through providing professional qualifications and access to the labour market, raising awareness on
prejudice and discrimination affecting the Roma, and fostering more active policies regarding the Roma
community, it addresses the challenges faced by Roma from several important angles, with features that
could and should be utilized in the Central and Eastern European context as well.
The Acceder methodology is based on a three-tier approach, which is adapted to the specific
characteristics and circumstances of each territory where the program is active:
(1) Individual intervention: curricular inventory, variable analysis, vocational diagnosis, design of
personalized integration pathways;
(2) Environmental intervention: educational and general diagnosis of the family and community context;
(3) Labor market intervention: Seeking possible labor opportunities and partnerships; the establishment of
commercial networks; establishment of collaboration with public and private economic agents committed
to the integration of Roma.
A 2-level approach – Policy and Practice: The Acceder program has been running on two
complementary levels: (1) a grass-roots approach with teams working around Spain primarily in the larger
cities providing guidance, training, assistance in search for employment and placement; (2) a
complementary policy-level approach: awareness-raising, campaigns, research, data collection and
advocacy.
Good program results: The Acceder program has achieved considerable results: until November 2011,
64,365 beneficiaries have participated in the program, 43,279 labor contracts have been established (one-
third of which constitute ―first jobs‖) and more than 16,000 participants are in training. The program has
also reached gender balance through targeted actions.
Funding and investment: Acceder is primarily financed through the EU Structural Funds: between 2000
and 2008, approximately 68% of the program was funded from ESF and ERDF, with the significant
majority of co-funding provided by central government, regional and municipal budgets. In 2009, the
costs per attended beneficiary was estimated at 1,454 Euros, with costs per labor contract amounting to
2,010 Euros.
3.6.4 POLICY MEASURE 4: INVEST IN MONITORING AND EVALUATION OF
SPECIFIC ACTIVATION MEASURES, AND SYSTEMATICALLY SHARING OF
BEST PRACTICES ACROSS MUNICIPALITIES
The Slovak government can improve the effectiveness of employment related interventions and
expand the reach of the most effective ones by systematically piloting and evaluating promising
ideas, and by investing more in knowledge sharing across municipalities. Governments, also in
52
Europe, are increasingly using rigorous impact evaluations to pilot programs and measure their
effectiveness. This is especially important in areas where policy measures are designed to address some of
the greatest social challenges, including improving employment prospects for the long-term unemployed.
Examples include a program of systemic evaluations of labor market measures in Denmark, the
Netherlands, France, and others.
These evaluations are prospective impact evaluations, or so-called social policy experiments, in
which a subset of potential beneficiaries is randomly selected to receive the pilot program. Randomly
selected recipients and non-recipients are then followed over time and employment outcomes are
compared. Projects that are proven to work will have much more public support for scale up than
promising initiatives that have not been subject to a rigorous evaluation. One of the most well-known
examples of this is the Mexican conditional cash transfer program Progresa (now called Opportunidades),
which was criticized at its inception in 1997 until rigorous impact evaluations showed it was an effective
means of using government resources to promote nutrition and school attendance among Me xico‘s poor,
and was subsequently scaled up. By 2007, program expenditures had reached $3.7 billion and covered
over 5 million families. 27 To implement these evaluations, the Slovak evaluation departments in the
Ministries can reach out to academics and policy think tanks in Europe – e.g. the Poverty Action Lab
Europe – founded at MIT University, and partner with local Slovak think tanks. The European
Commission is promoting (and funding) social policy experiments through its PROGRESS facility, as was
highlighted during the December 2012 conference on monitoring and evaluation in Bratislava28.
BOX 3-5: INSTITUTIONALIZING IMPACT EVALUATIONS: DANISH NATIONAL
LABOR MARKET AUTHORITY
The Danish Labor Market Authority (LMA) has taken a very proactive approach towards building
up evidence on its employment 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 internet (www.ams.dk
and www.jobindsats.dk ).
In designing and carrying out these evaluations, the LMA works closely with external evaluators –
Danish academics – and a selection of its affiliated job centers. Denmark has 98 municipalities, with
93 integrated job centers for all job seekers (insured and uninsured). There are also 4 regional employment
councils.
So far it has completed 4 randomized control trials, 2 evaluations are ongoing, and a new one is
planned starting August 2012 serving particularly vulnerable groups. In each evaluation, the
comparison group is offered the regular package of employment services while the treatment group
receives something ‗extra‘. For example, the first evaluation consisted of an intervention whereby job
seekers were offered bi-weekly counseling as opposed to regular counseling every three months. In the
upcoming evaluation, a ‗social mentoring‘ pilot will be evaluated. The target group consists of youth 18-
29 years old far from the labor market. Local job centers will be provided with funding to hire social
27
E.g. see http://www.povertyactionlab.org/evaluation/role-conditional-cash-transfers-early-childhood-development-
mexico: Lia C H Fernald, Paul J Gertler, Lynnette M Neufeld. ―Role of cash in conditional cash transfer programmes
for child health, growth, and development: an analysis of Mexico‘s Oportunidades‖ Lancet 2008; 371: 828–37
28
http://siteresources.worldbank.org/EXTROMA/Resources/615986-1325095562445/Diane_Angermueller.ppt.
53
mentors who will give intensive counseling to vulnerable youth, including advising on accessing social
services and education and training opportunities.
Furthermore, the Government of Slovakia program could establish a knowledge portal of ―good
practices‖ of municipal activation activities. Many municipalities are undertaking innovative ideas in
the area of employment as well as other areas, sometimes in collaboration with the NGO sector. Creating a
platform where mayors and other municipal level authorities can exchange these ideas with one another –
even a virtual platform – and creating a regional task force that systematically compiles these practises
(including impact evaluation findings) and shares them with municipalities can ensure that good but
isolated ideas can benefit all municipalities.
3.6.5 POLICY MEASURE 5: MAINTAIN A STRONG SAFETY NET THAT PROTECTS
THE MOST VULNERABLE, BUT ALSO ACTIVELY SUPPORTS INCREMENTAL
AND SPECIFIC IMPROVEMENTS IN WELFARE
Maintain a strong safety net that protects the most vulnerable. Even with the current levels of
family and social assistance support, poor families struggle to make ends meet. Continuing to provide
these unconditional social protection benefits is therefore essential for protecting the most vulnerable
families – Roma and non-Roma – from falling deeper into poverty. Some of the reforms of the BMN
program that were proposed in 2011 increase the risk that beneficiaries far removed from the labor market
– the largest group among out-of-work Roma - fall into poverty in case they do not succeed to find jobs or
engage in activation programs.
Include additional incentives for poor families to invest in incremental and targeted improvements
in education, health, employment, and housing. The support system of government transfers to the poor
can have a stronger role in promoting specific household investments in health, education, employment,
and housing that reduce long-term dependency. Families enrolled in CCT programs receive cash subsidies
if they fulfill certain conditions, such as attending preventive health clinics (e.g. child vaccinations),
taking nutrition supplementation for young children, and education attendance (Fernald et al. 2008). There
is considerable rigorous international evidence from around the world indicating that targeted conditional
transfers can lead to significant improvements in health- and education outcomes, and that through these
mechanisms, they enhance productivity and reduce long-term dependency. Slovakia has a number of CCT
type programs, including the activation allowances which adults receive conditional on carrying out local
works. Other countries, and especially countries in Latin America have a number of comprehensive
programs that integrate different measures, including Oportunidades in Mexico, Bolsa Alimentação in
Brazil, Red de Protección Social in Nicaragua, Programa de Asignación Familial in Honduras, Familias en
Acción in Columbia, Subsidio Unico Familiar in Chile, and the Program of Advancement through Health
and Education in Jamaica (Fernald et al. 2008). Of course, such demand side initiatives should be matched
with quality and accessible supply side programs. In the context of Roma in Slovakia, the Government
may consider:
Incentivizing targeted investments in maternal health and early childhood development – from
conception to age 8. This period is critical for the future development and prospects of children, since it
bears long-term implications for people‘s capabilities in school and on the labor market. The Government
of Slovakia may consider providing financial incentives to poor pregnant women and to young mothers to
a) attend pre-natal check-ups, b) have their children fully vaccinated, and c) attend information sessions on
infant care, nutrition, the risks of smoking, and the use of health insurance.
54
Strengthening the existing program providing support to poor young families conditional on
children participating in pre-school represents an additional potential measure in this regard.
Disadvantaged families are currently exempted from paying kindergarten fees and are supported through
subsidies for meals. However, qualitative research by the Slovak Governance Institute (2012) revealed
that many mothers with children of preschool age lose their social assistance benefit in material need
(dávka v hmotnejnúdzi) when they become a recipient of the parental allowance (rodiÄ?ovský prÃspevok)
after giving birth to a child. This is because their new income – i.e. including the parental allowance –
exceeds the minimum subsistence level. As a result, they also lose eligibility for the pre-school subsidies.
This, perhaps unintended consequence of the eligibility rules for the benefit being tied to the calculation of
the Basic Material Needs Benefit threshold excludes among the most vulnerable families – the original
target group of this benefit. With total expenses for the meals and fees paid at kindergartens ranging from
approximately 20 to 50 Euros per child, preschool is then no longer affordable for many of these families.
Incentivizing secondary school completion, not early school leaving. Early school leaving rates among
Roma are extremely high. To adequately address this challenge, the Government could consider making
early school leavers under 20 years old ineligible for the current labor activation allowance, and instead
provide a bonus to parents of vulnerable youth if children complete secondary education. This could also
be linked to encouraging savings (see financial inclusion chapter): for example, including a matching
grant component to the bonus if parents have saved a certain amount in a dedicated secondary education
savings account. Again, rigorously evaluated examples from successful programs around the world could
serve as inspiration for policy makers.
Incentivizing job search incentives. This can be used to promote job search while providing stability.
See Policy Measure 1 above.
Incentivizing investments into improved living conditions, especially by the most vulnerable. The
Basic Material Needs Benefit provides an additional payment to cover housing and utility costs. While
approximately half of all households receive the BMN, approximately one-third (or 24,000) of all Roma
households receive housing allowances. Families who do not live in formally accepted buildings with
formal tenure (lease/ownership) are not eligible to receive housing benefits. Hence, the housing allowance
program rules inherently exclude the neediest Roma families. The Slovak Government may consider (a)
delinking eligibility for the housing allowance from formal residence status, or even developing a new
subsidy program; (b) imposing conditions on receiving housing allowances in such a way that they can be
used only for costs related to housing and utilities, and related expenses; and, (c) linking the housing
allowance to household financial planning skills for self improvement of living conditions (i.e., link to
financial literacy training, use of prepaid metered utilities such as gas, electricity and water).
To address multiple needs of Roma families through the social protection system, Slovakia can build
on international experiences such as the Chile Solidario program. As described below, this program
focuses on providing complementary services targeting the poorest families that go beyond the main
social safety nets available in Chile. The program‘s main goal is to help households progr essively sustain
their exit from extreme poverty by improving their human capital assets, housing, and income-generation
capacity. Chile Solidario also has a supply-side component, aimed at ensuring coordination among
different social protection programs. The rationale comes from the recognition that an approach with
isolated and sectoral programs is not able to address the multiple and interrelated causes of extreme
poverty. The long-term objective is to move away from an approach based on single programs toward a
system of social protection in which bundles of programs are tailored to meet the specific needs of
households that are hard to reach. While the specific target groups are obviously different in the Chile
Solidario, the program feature of a dedicated coordinating body providing personalized social protection
55
support to the most vulnerable provides important lessons for the Roma situation, where the social
protection system provides not only a crucial backbone for security but provides also opportunities to
stimulate targeted investments in employment, (early) education, health, housing, and financial literacy.
BOX 3-6: FILLING THE GAPS AND INCREASING INDEPENDENCE: CHILE
SOLIDARIO
The Chile Solidario program provides a personalized support system whereby the poorest families
receive psychological and social services, guaranteed subsidies, and preferential access to public
social programs. A local social worker follows beneficiary families for two years. During this period,
households receive direct cash transfers while the social worker assesses their needs, assists them in
developing a ―family contract‖ outlining ways the family can improve its living conditions, and connects
them to various social programs. Families are provided with support to develop certain competences,
which will enable them to reach the social welfare system independently and which will motivate them to
strive for change, and with advice and opportunities for orientation.
The program‘s main goal is to help households progressively sustain their exit from extreme
poverty by improving their human capital assets, housing, and income-generation capacity. Unlike
other conditional cash transfer programs in Latin America, which have income support as the main
objective, Chile Solidario transfers a comparatively small amount of money. The program‘s direct cash
transfers (complementary to other social protection programs) are provided at a decreasing rate over 24
months, in order to avoid beneficiary dependency on the program and to gradually prepare families to no
longer receive benefits.
Chile Solidario‘s supply-side component, aims at ensuring coordination among different social
protection programs, enable families to navigate through Chile‘s social services institutions
independently, to orientate themselves on specific programs and to manage solutions to their own
needs. The rationale comes from the recognition that an approach with isolated and sectoral programs is
not able to address the multiple and interrelated causes of extreme poverty. The long-term objective is to
move away from an approach based on single programs toward a system of social protection in which
bundles of programs are tailored to meet the specific needs of households that are hard to reach.
There are four specific programs that aim to facilitate a transition from Chile Solidario to other
social protection programs: The Bridge Program, aimed to support extremely poor families. This
program functions at the municipality-level and is administered and technically assisted by the ‗Solidarity
and Social Investment‘ foundation (FOSIS); The Bind Program, aimed to support vulnerable elderly
citizens who live on their own. This program functions at the municipality-level and is technically assisted
by the National Senior Service (SENAMA); The Street Program, aimed to support adults who are
homeless. This program may be performed by municipalities, Provincial Governments and NGO‗s and its
management and technical assistance are the responsibility of the Ministry of Social Development; And,
the Opening Pathways Program, aimed to support children in families where there are situations of forced
separation because of an enforcement order to one of its members. The program is implemented by NGOs
and its design and methodological support is provided by Ministry of Social Development.
Chile Solidario is part of Chile‘s Social Protection System and is represented by the Executive
Secretariat of the Ministery of Social Development, which is in charge of coordinating social welfare
institutions. It operates as a decentralized system closely related to local government, municipalities, and
people responsible for Chile‘s main social security programs. To reach out to its target group, Chile
56
Solidario makes use of the government administration of other social protection systems, but it also uses
its own institutional network. In addition, it links its own beneficiaries back to other social protection
programs, in order to achieve a fully integrated social protection system.
Chile Solidario‘s structure and functioning are laid down in a legal framework, which explains the
operation principles of the system, its scope and modes, and establishes the state benefits that must be
granted to Chilean citizens by right.
Sources: http://www.chilesolidario.gob.cl/en; ―Safety-nets: How To – A Toolkit for Practitioners,‖ The World Bank:
http://siteresources.worldbank.org/SAFETYNETSANDTRANSFERS/Resources/281945-
1291746977764/HowtoCompletePdfs.pdf.
57
3.7 BIBLIOGRAPHY
Fernald, L., P. Gertler, and L. Neufeld. ―Role of cash in conditional cash transfer programmes for child
health, growth, and development: an analysis of Mexico‘s Oportunidades‖ Lancet 2008; 371: 828–37
Ehrbeck, 2011: ―Full Financial Inclusion in Our Lifetime,‖ CGAP: http://www.cgap.org/gm/document-
1.9.53154/Tilman_Ehrbeck_Transcript.pdf, Accessed on: 7 Sep. 2011.
Habitat.org: http://www.habitat.org/hw/june-july03/feature4.html, accessed on: 7 May 2012.
MECEM, Roma Press Agency, 2010: ―Denisa Gáborová: Roma women are proud martyrs,‖ by K.
Magdolenová and J. Vaňová, http://www.mecem.sk/rpa/?id=journalism&lang=english&show=18116
accessed on: 21 May 2012.
World Bank. 2010. ―Safety-nets: How To – A Toolkit for Practitioners,‖ The World Bank. Washington
DC.
__________. 2012: ―Protecting the Poor and Promoting Employability: An assessment of the social
assistance system in the Slovak Republic,‖ unpublished.
58
3.8 ANNEX: ESTIMATION RESULTS
ANNEX TABLE 3.1: PREDICTING SOCIAL ASSISTANCE AMONG ROMA HOUSEHOLDS IN SLOV AKIA
(1) (2) (3) (4)
All: Base All: Full
Roma Non-Roma
Model Model
Roma .181*** .216***
(.043) (.043)
Household Head – Female -.073** -.067** -.031 -.135***
(.030) (.029) (.039) (.047)
Household Head – Age -.006*** -.005*** -.005*** -.007***
(.001) (.001) (.001) (.002)
Household Head – Employed -.228*** -.190*** -.241*** -.196***
(.033) (.034) (.048) (.045)
Number of household members .031*** .024*** .034*** .019
(.007) (.007) (.008) (.019)
Second income quintile -.027
(.062)
Third income quintile .024
(.060)
Fourth income quintile -.106*
(.062)
Fifth income quintile -.336***
(.061)
Household suffers from hunger .074**
(.035)
Household language is Romani .076* .036 .040 .300**
(.039) (.039) (.042) (.149)
Distance to closest Social welfare office: within 3 .095** .068 .075 .111
km (.043) (.042) (.051) (.083)
Dominant ethnicity of settlement is Roma .021 -.004 .028 .037
(.034) (.034) (.038) (.079)
Central regionb -.155*** -.217*** -.212*** -.083
(.053) (.052) (.064) (.099)
Eastern regionb -.080* -.129*** -.085 -.096
(.045) (.042) (.052) (.090)
Rural household .035 -.011 .033 .018
(.042) (.041) (.048) (.085)
Constant .557*** .752*** .748*** .656***
(.088) (.100) (.096) (.166)
Observations 999 999 676 323
R-squared .209 .262 .112 .136
Source: UNDP/World Bank/EC regional Roma survey (2011).
OLS estimations. Robust standard errors in parentheses. Example (Model 2): In Slovakia, Roma households are 21.6
percentage points less likely to complete secondary school than non-Roma, when background characteristics are
taken into account. *** p<.01, ** p<.05, * p<.1 ; In addition to income quintiles, the analysis includes ‗missing
59
income‘ as a separate income category. Observations for which no information on income was available could hence
be included in the estimations. The ‗missing income‘ category is left out of the table. aOnly one subject per
household was included in the estimation sample: the head of the household. This implies that all individual level
background characteristics, such as gender and age, refer to the household head. bSlovakia‘s Western region is
omitted from the table. Estimates for the Central and Eastern region refer to the difference of these two regions with
the Western region.
ANNEX TABLE 3.2: PREDICTING EMPLOYMENT (AGES 25-64)
(1) (2) (3) (4) (5)
All Roma Non- Roma: Roma:
Roma Men Women
Roma -.159***
(.037)
Women -.157*** -.141*** -.204***
(.015) (.017) (.032)
Age -.001 -.001 -.002 -.002 -.000
(.001) (.001) (.002) (.001) (.001)
No / incomplete basic educationa -.052** -.060** -.120 -.059 -.066***
(.025) (.025) (.088) (.042) (.023)
Complete Secondary education or highera b .188*** .132*** .295*** .134*** .131***
(.030) (.034) (.063) (.046) (.043)
Number of household members .003 .006 -.009 .013* -.002
(.005) (.006) (.018) (.008) (.005)
Household language is Romani -.016 .006 -.148 -.003 .016
(.029) (.031) (.128) (.042) (.032)
Distance to employment office: within 3 km -.028 -.039 -.011 -.052 -.025
(.039) (.041) (.083) (.059) (.035)
Distance to nearest city: within 3 km .009 .012 .017 -.003 .028
(.032) (.033) (.073) (.046) (.030)
Dominant ethnicity in settlement is Roma -.010 -.029 .042 -.019 -.038
(.025) (.027) (.067) (.039) (.026)
Central regionc .080* .126** .010 .121 .134***
(.044) (.052) (.088) (.074) (.050)
Eastern regionc -.031 -.044 -.009 -.081 -.007
(.035) (.039) (.074) (.058) (.036)
Rural household .005 -.002 .045 -.041 .039
(.035) (.037) (.073) (.053) (.033)
Constant .385*** .196*** .290* .258** -.013
(.073) (.075) (.148) (.115) (.069)
Observations 2,085 1,450 635 722 728
R-squared .172 .105 .106 .074 .083
Source: UNDP/World Bank/EC regional Roma survey (2011).
OLS estimations. Robust standard errors in parentheses. Example (Model 1): In Slovakia, Roma of the ages 25-64
are 15.9 percentage points less likely to be employed than non-Roma, when background characteristics are taken into
account. *** p<.01, ** p<.05, * p<.1; aBoth schooling categories shown in the table should be compared to ‗primary
or incomplete secondary education‘. bSecondary school completion also includes those respondents who have
incomplete general secondary school. cSlovakia‘s Western region is omitted from the table. Estimates for the Central
and Eastern region refer to the difference of these two regions with the Western region.
60
4 FINANCIAL INCLUSION
Poor households – regardless of ethnicity - need a broad range of financial services just as much as, if not
more than, non-poor households. This chapter investigates access to a broad range of financial services
for Slovak Roma, explores the link between access to financial services and vulnerability, and highlights
examples of policy actions both in Slovakia itself by civil society, as well as in other countries of the world
facing similar challenges with excluded communities. The policy recommendations promote a
comprehensive, incremental approach to financial inclusion, focusing on (1) expanding financial literacy
and debt management training, (2) improving access to financial services with a focus on savings
facilitation, and linking savings activities with human development outcomes; and, (3) taking advantage of
government social protection payment systems to promote financial inclusion.
4.1 ASSESSMENT 29
ROMA IN SLOVAKIA LACK ACCESS TO FINANCIAL SERVICES
Roma in Slovakia lack access to financial services. Only 29% of Roma households in Slovakia have a
current account, compared with more than three quarters (77%) of the general population. The precarious
economic situation faced by most Roma coupled with a lack of access to saving methods helps explain
why only very few Roma households (9%) have any savings at all, and even less households (5%) save in
savings accounts. This situation is comparable to other countries in the region.
TABLE 4-1: USAGE OF SAVINGS ACCOUNTS AND OTHER BANKING SERVICES AMONG ROMA
HOUSEHOLDS IN SLOVAKIA
Banking Services A % Savings %
B
Saving Account 5 Household has Savings (%) 9
C
Term Deposits 2 Number of Months the Household Could Rely on Savings 3
C
Money Transfer Services 2 Household with Savings has a Savings Account (%) 33
Debit/Payment Card 10 Home Ownership (%) D 73
A
Source: UNDP/World Bank/EC regional Roma survey (2011). Figures refer to the percentage of Roma households making
use of the indicated banking services. B ‗Savings‘ include e.g. cash, bank deposits and highly valued commodity items like
gold. C Conditional on household having some form of savings. D Measured as family- rather than household-ownership of the
dwelling.
A likely cause of this low usage of banking services is the physical remoteness of banks for most
Roma households: only about one-third of Roma households in Slovakia live within 3 kilometers of
the nearest bank branch. As shown in Table 4-2, this is lowest proportion across the five countries, and
reflects the absence of bank branches in the rural areas. Not shown in the table is the finding that two
thirds (65%) of Roma households in Slovakia possess either a landline or a mobile phone (used for online
banking services in some countries), while about 1 in 5 reports having a computer, 1 in 5 reports being
able to use word processing, and 1 in 8 has an internet connection.
29
This chapter has been prepared based in part on selections from the upcoming World Bank report on Financial
Inclusion among Roma in five Eastern European countries: Bulgaria, the Czech Republic, Hungary, Romania and
Slovakia. The report is titled: ―Reducing Vulnerability and Promoting the Self-Employment of Roma in Eastern
Europe through Financial Inclusion‖ and is based on the same survey data that were used for the current report: the
data from the regional Roma survey.
61
TABLE 4-2: HOUSEHOLDS LIVING CLOSE TO A BANK BRANCH (%)
Czech
Bulgaria Hungary Romania Slovakia
Republic
General population
91 79 92 85 64
Roma households
47 69 60 44 36
Roma households in
urban neighborhoods 73 70 79 73 81
Roma households in rural
communities 19 50 50 26 18
Source: UNDP/World Bank/EC regional Roma survey (2011) for Roma household data, and EU SILC (Eurostat,
2008) for general population data. Author‘s own calculations.
A
For Roma, the data refers to the proportion of households living less than 3 km away from the nearest bank branch.
For the general population, the data refers to the proportion of households living close to a bank branch, though the
term ‗close‘ is left unspecified in the questionnaire administered to households. Thus, survey de sign differences do
not allow a strict comparison between the data.
Irregular employment incomes and low levels of savings translate into considerable economic
insecurity. Given the crucial importance of savings for households‘ economic security, the fact that few
Roma have savings and even fewer saving accounts, is an alarming finding. In fact, few Slovak Roma – 1
in 10 compared with nearly 6 in 10 among the general population – are able to face unexpected expenses
without external financial support, and this is directly linked to the lack of savings that Roma families
would otherwise be able to draw from.
FIGURE 4-1: ROMA HOUSEHOLDS HAVE A MUCH LOWER SELF-REPORTED ABILITY TO FACE
UNEXPECTED EXPENSES WITHOUT EXTERNAL FINANCIAL SUPPORT
100
80
60 58 59
60
35 33
40
15 10
20 8 6 6
0
Bulgaria Czech Hungary Romania Slovakia
Republic
General Population Roma
Source: General population: EU SILC (2008); Roma: UNDP/World Bank/EC regional Roma survey (2011). The
figures represent the share of households reporting an adequate ―Ability to pay for unexpected expenses without
relying on loans or other external forms of financial support‖.
Related, Roma are also more likely to have arrears on utility bills than the general population. 15%
of Roma households in Slovakia is in arrears on utility bills, compared with 3% of the general population.
This is lower than among Roma in neighboring countries, but the difference with the general population
remains large, and in fact exceeds the gap between Roma and non-Roma registered in some of the other
countries.
62
TABLE 4-3: ARREARS ON HOUSEHOLD MORTGAGES, LOANS AND UTILITY BILLS
Czech EU-
Bulgaria Hungary Romania Slovakia
Republic 27
Roma All Roma All Roma All Roma All Roma All
Utility Bills (%) A 38 32 34 2 46 0 29 23 15 3 1
Source: UNDP/World Bank/EC regional Roma survey (2011) for Roma household data, and EU SILC (Eurostat, 2008) for
general population data; A Bills for any housing related utilities (e.g. electricity and water supply, heating or phone bills.
Most Roma face significant barriers that limit access to credit. As shown in the previous employment
section and in the World Bank report on financial inclusion of Roma in Europe (forthcoming, 2012), the
potential group of Roma that can take advantage of microcredit is small, as most Roma interested in
becoming self-employed face significant barriers that limit access to credit. These barriers include
indebtedness, little employment- and professional experience, and very low levels of education: even
when compared to that segment of the general population that is being refused credit. The recent social
microcredit experience of the Hungarian Kiút program (box below) provides an important example in this
regard. Hence, while microcredit can be an important tool for supporting a small group of existing micro
entrepreneurs and a small group of starting entrepreneurs, a much larger group may be reached through
more basic financial inclusion instruments.
BOX 4-1: EXPERIENCES WITH MICROCREDIT AMONG STARTING ROMA
ENTREPRENEURS: THE KIUT PROGRAM IN HUNGARY
The Kiut Program is a two-year microcredit pilot program in Hungary implemented by the Polgar
Foundation in collaboration with Raiffeisen Bank, with EU financing. The purpose of the project is to
empower poor households, mainly Roma, to become self-employed through microcredit with a goal of
reaching up to 300 loans by 2012. At the project start in mid-2010, its field officers sought out potential
borrowers in poor communities, especially in Eastern Hungary, offering relatively small-sized, unsecured
loans of up to EUR 3,500 for a period of 0.5-1.5 years. The loans are provided in a group setting, so that
peer pressure – but not full joint liability – is used to secure the loans. Loans are generally provided
sequentially 30 and saving is encouraged among members of the group. An ideal group consists of 5
members, but in practice contains a minimum of 3 and maximum of 6 members. The Kiut program
supports selected entrepreneurs in registering their business and the first loan is disbursed.
The program succeeded in reaching out to poor Roma families. Project summary statistics based on
interviews with serious loan applicants 31 showed that a ‗typical‘ applicant is Roma, male, in his late
thirties, either married or living with a spouse, in a household of 5 members, and in a mixed
neighborhood, most likely in a village. The applicant generally lives far (around half an hour or longer)
from (a) a bank office where a loan could be taken and paid; (b) the local employment office, and (c) any
secondary school. The applicant is typically literate but does not have advanced education – usually
primary or some vocational. He has been unemployed for nearly two years although has a formal
employment history spanning more than 10 years. However, he does not expect to be able to take a regular
30
I.e. credit is given to one borrower at a time under the condition that the previous borrower repaid her loan.
31
Kiut field officers carry out a detailed interview with serious loan applicants. This extensive monitoring system
was developed in collaboration with the World Bank and UNDP, and financed by DG Regional Policy.
63
commercial loan of approximately 500 Euros. Furthermore, the typical applicant has low satisfaction with
life in general, and especially with the financial situation of the household.
The Kiut early experience underscores the challenges to lend to this group of borrowers. Staff drew
several early lessons from the program in May 2011. First, despite the initial goal of concluding loan
agreements with 100 customers by the end of the first year, fewer than 50 loans were disbursed, owing to
the difficulty of finding good candidates. Even then, by early Winter 2011, many businesses experienced
difficulty and there was an increase in the number of late- or non-payment of loan installments, and by
May 2011 the program management decided to change the model of the program, most importantly
requiring field agents to put greater emphasis on identifying potential entrepreneurial capabilities, even if
this may mean in practice attracting fewer clients from the most disadvantaged groups. In addition, the
program decided that counseling and training that had been conducted informally by the field workers and
crisis managers so far, should be done in a more organized manner. This approach led to much improved
lending results.
Source: World Bank (2012): ―Reducing Vulnerability and Promoting the Self-Employment of Roma in Eastern
Europe through Financial Inclusion‖
ACCESS TO FINANCIAL SERVICES IS IMPORTANT, ESPECIALLY FOR THE POOR
Access to financial services is important, especially for the poor. Successful financial inclusion can be
evaluated at two different levels: the provision of additional financial security, mitigating the impact of
economic shocks and uncertainty to marginalized businesses owners and households, and the provision of
additional opportunities to participate in a wide variety of markets – including the formal labor market.
Poor households – including many of Roma households – need a broader range of financial services just
as much as, if not more than, non-poor households (Ehrbeck, 2011). For example, poor people need
careful financial planning to make ends meet with the limited resources, they need access to savings
products to protect themselves against income vulnerabilities that arrive as a result of irregular
employment, or they need savings to build up a down payment for a (micro-) loan to make home repairs.32
The importance of financial inclusion for basic welfare outcomes is underscored by a recent increase in
political attention for this topic.
In June 2012 The World Bank published a Reference Framework as a resource for preparing new
financial inclusion strategies. This Framework highlights that: ―Financial inclusion is emerging as a
priority for policymakers and regulators in financial sector development, with an increasing number of
countries introducing comprehensive measures to improve access to and usage of tailored financial
services, informed by a fast-growing body of experience and knowledge. More than 60 countries have
initiated financial inclusion reforms in recent years.‖ The Seoul G20 summit recognized financial
inclusion as one of the main pillars of the global development agenda, and a growing number of
governments are seeking comprehensive policy-level solutions to improve financial inclusion.
For poor households being able to save in a secure way is perhaps the most crucial form of financial
inclusion since this is an important means of increasing economic security. Yet, Roma households
almost never have savings accounts. Savings allow people to smooth consumption and respond to income
shocks, but also to borrow money, increasing their potential for self-employment. For poor households,
32
For a discussion on microcredit vis-Ã -vis financial inclusion broadly, e.g. see an August 19, 2011 interview with
Mr. Tilman Ehrbeck, the CEO of CGAP: http://www.cgap.org/gm/document-
1.9.53154/Tilman_Ehrbeck_Transcript.pdf.
64
financial exclusion often implies a complete lack of secure means to save, or forces these households to
make use of expensive alternatives to regular banking services. Nonetheless, even under these far from
ideal circumstances the poor still can and do save (Karlan, 2011). Yale Professor of Economics Dean
Karlan concludes that: ―The evidence generated thus far on the impact of access to formal savings is
limited but very promising," (Karlan, 2011).
The European Commission (EC) has followed a narrow approach on financial inclusion geared
towards basic bank account access, and focuses on ‗financial education‘, i.e. citizens‘ understanding
of complex financial products. In the case of the Roma, a broader approach toward financial inclusion
would be more appropriate. In fact, given the extremely high Roma poverty levels, low levels of
education, and general degree of marginalization, the challenge of achieving financial inclusion among
Roma in Slovakia is comparable with that of achieving financial inclusion among citizens of lower middle
income and lower income countries, rather than among the citizens of European countries in general. The
‗Graduation Approach‘ provides an example of how poor households facing extremely high economic
insecurity are targeted in low and middle income countries:
BOX 4-2: "GRADUATION APPROACH ‖ : USING FINANCIAL INCLUSION TO CREATE
PATHWAYS TO GRADUATE OUT OF POVERTY
The CGAP-Ford Foundation Graduation Program is a global effort to understand how safety nets,
livelihoods, and microfinance can be sequenced to create pathways for the poorest to graduate out
of extreme poverty. The graduation model targets poor households facing extremely high economic
insecurity. Ten pilots are taking place in Haiti, India in three places (with Bandhan, SKS, and Trickle Up),
Pakistan, Honduras, Peru, Ethiopia, Yemen, and Ghana. The pilots involve diverse institutional forms,
economic contexts, and cultures. The pilots are implemented through partnerships between financial
service providers, nongovernmental organizations, and government units. Several of the pilots are
measuring the program‘s effects on people‘s lives through rigorous randomized impact evaluations and
qualitative research.
The Graduation Approach addresses at first the most obstructing barriers that poor households
face to accessing financial services. The initial focus on savings illustrates the latter‘s crucial role in
facilitating a successful transition to self-employment. The approach is built on five core elements:
targeting, consumption support, savings, skills training and regular coaching, and an asset transfer. Once
the most immediate needs are addressed, households are encouraged and assisted to save, allowing the
household to build assets and instilling financial discipline. Only in the final two steps, ‗business skills
training‘ and the transfer of in-kind assets, does the purpose of self-employment come into play. The asset
transfer is included to jump-start small businesses, increasing the chance of a successful transition to self-
employment.
Sources: GCAP Graduation Program Overview; Hashemi and De Montesquiou, 2011. CGAP: Haiti Chemen Lavi
Miyo Program.
The Bank on San Francisco approach provides an example of how poor households in a rich
country like the United States are being targeted for financial inclusion:
BOX 4-3: BANK ON SAN FRANCISCO, UNITED STATES
65
‗Bank on San Francisco‘ is a collaborative effort to bring 10,000 of the city‘s estimated 50,000 un -banked
individuals into the financial mainstream. The San Francisco Mayor‘s Office, the Treasurer‘s Office of the
City and County of San Francisco, the Federal Reserve Bank of San Francisco, local non-profit EARN
(Earned Assets Resource Network), and the city‘s financial institutions have worked together to:
(1) Increase the supply of starter account products that work for the low-income un-banked market by
developing baseline product criteria that must be offered by all participating financial institutions;
(2) Raise awareness amongst un-banked consumers about the benefits of account ownership and spur
them to open accounts;
(3) Make quality money management education more easily available to low-income San
Franciscans;
(4) Clamp down on the proliferation of check cashers and payday lenders;
(5) Raise city-wide awareness of the un-banked problem and potential solutions.
Such a combination of supply-side networks, political engagement and NGO involvement could also be
effective in efforts to include the Roma in Slovakia into the financial mainstream.
Source: Stuhldreher, A.: Bank on San Francisco – An initiative to bring all residents into the financial
mainstream (draft).
To promote financial inclusion among Roma in Slovakia, the June 2012 draft Financial Inclusion
Action Plan of the Slovak Republic identifies several goals. These have been formulated on the basis of
the financial inclusion chapter of the NRIS for which the World Bank team provided inputs:
Goal No. 1: Improve financial literacy. This entails providing marginalized Roma communities
with information on financial services, and supporting basic and advanced financial education and
training in marginalized Roma communities.
Goal No. 2: Improve access to financial services, with a focus on access to savings and
savings facilitation.
Goal No. 3: Increase the protection of marginalized Roma communities against loan shark
activities and illegal practices of credit companies. This entails using systematic terrain work,
education in financial literacy and a comprehensive revision of criminal law, as well as financial
measures aimed at customer protection.
Goal No. 4: Support the development of microfinance programs in marginalized Roma
communities with a view to supporting micro-, small- and medium enterprises (MSMEs)
and self-employed individuals. This entails supporting the growth of MSMEs using EU tools
such as the ESF and the ERDF.
4.2 POLICY RECOMMENDATIONS
Building on the Action Plan Goals 1-4, there are three main policy recommendations: (1) expanding
financial literacy and debt management training, (2) improving access to financial services with a focus on
savings facilitation, and linking savings activities with human development outcomes; and, (3) taking
advantage of government social protection payment systems to promote financial inclusion.
Policy Measure 1: Expanding basic financial literacy and debt management training. Excellent
examples from around the world are plentiful, but Slovakia can also build on locally implemented
initiatives that can be scaled up. There are many financial literacy training modules – also for school
children. In Slovakia, the NGO most active in the area of small scale financial inclusion activities with
66
disadvantaged Roma is ETP. Together with Autonomia Foundation, it has been working with standardized
financial training modules that have proven successful entry points into microsavings and microcredit
programs for clients living in marginalized communities. Upscaling initiatives such as this one would help
reach the target of improving financial literacy among Roma households. Another possible way of
increasing financial literacy among Roma would be to target teen agers through schools.
BOX 4-4: FINANCIAL LITERACY TRAINING AMONG ROMA IN SLOVAKIA
The financial education course developed by ETP Slovakia aims to teach clients to use their finances
wisely as well as to help them escape from the trap of debts and give them a new beginning. The
financial education course consists of two modules. The first module („Don‘t Be Afraid of Money―) is for
entry-level program participants, children and youth. In this module, the client learns basic knowledge and
information about financial management, household budgeting and not spending money beyond means.
The second module consists of 6 volumes and is for the intermediate level (i.e. graduates of the first
module). This program is conducted in groups in community centers. The participants gain knowledge,
skills and attitudes that are necessary for good financial management of their earnings, expenses, savings
and investments. The clients learn to plan their finances as well as saving for certain goals in the future.
Both courses inform participants about the dangers associated with the excessive interest rates of loans
provided by non-banking entities, and how this becomes a trap for clients by creating a vicious circle of
growing debts. After the completion of both courses, the successful graduates become eligible to
participate in microsavings and microcredit programs administered by ETP Slovakia.
Source: www.etp.sk
Policy Measure 2: Improving access to financial services with a focus on savings facilitation, and
linking savings activities with human development outcomes. For example, the Government could
support households to open up targeted education- or housing savings accounts, in which households are
encouraged to save with an explicit purpose. International evidence suggests that targeted savings are
successful in raising overall savings. Several locally implemented initiatives can be adopted and scaled up,
such as ETP Slovakia‘s "Individual Development Account", which provides incentives to save for housing
improvements. Slovakia can also adopt international models, such as the "Kindergarten-to-College (K2C)"
savings initiative used in the USA, in which parents are provided with incentives to save for children‘s
education. Both are described below. Initiatives like these can be linked to social protection transfers like
the Benefit in Material Need and family benefit programs.
Earmarking savings for specific purposes is an effective way to not only stimulate saving in general,
but to make sure that accumulated savings are spent ‗wisely‘ by poor households. As highlighted in
Box 4-5, there are examples of initiatives in which poor families are encouraged to save for specific
purposes. This not only guides families through the saving process, teaching them through learning by
doing why saving is important and how even small quantities of money can accumulate to a more
substantial sum, but also provides incentives for families to spend their savings on investments which are
valuable in the long run, such as education and housing improvements.
BOX 4-5: EARMARKING SAVINGS: UNITED STATES AND GHANA
67
The Kindergarten-to-College program in San Francisco, USA: In order to promote savings and
economic mobility, children‘s savings accounts have been established as long-term asset-building
accounts that grow over time with additional deposits and earnings. San Francisco‘s K2C initiative
promises to open a college savings account for children entering kindergarten in San Francisco‘s public
schools, with an opening deposit of $50 from the City of San Francisco. Low-income families receive an
additional $50 deposit from the city, and EARN, the program‘s non-profit sponsor, has committed to
provide an additional $100 to match the savings of the first group of students to participate.
Source: www.earn.org
Labeling savings accounts in Ghana: The focus on earmarking savings is not uncommon. In Ghana, the
international organization Innovations for Poverty Action, which carries out randomized counterfactual
impact evaluations of innovative ideas, is working with Mumuadu Rural Bank (MRB) to study the
response to, and impact of, a new account labelling savings product. It recruited 2100 study participants
and found that customers with a labelled (i.e. ‗earmarked‘ for housing improvements, school fees, etc.)
savings account show a 31.2% increase in total deposits after nine months of account operations as
compared to customers who were offered to open a savings account but were not offered the option to
earmark it.
Source: http://www.poverty-action.org/project/0071
A Slovak example is ETP‘s Individual Development Account (Box 4-6). A central element of ETP‘s
Individual Development Account (IDA) project has been the provision of targeted savings toward home
improvements, providing a way to link financial inclusion to improved housing conditions.
BOX 4-6 : ‗TARGETING SAVINGS‘ PROJECT AMONG ROMA IN SLOVAKIA
ETP‘s Individual Development Account33, is a microsavings initiative which has been made available
for 400 individuals in Eastern Slovakia. Since its launch in 2006, it teaches Roma families how to resist
loan sharks, better manage limited funds through the month, and – most importantly – think about the
future prospects for the whole family. While the program is small, ETP has developed a very important
knowledge base of financial inclusion in the marginalized communities through years of fieldwork, which
can be capitalized on. ETP describes its project as follows:
―The IDA project is focused on marginalized populations with inappropriate living and housing
conditions. The savers are saving their money for a certain amount of time, usually 1 or 2 years, for a
certain set goal. Some of them save for roof reconstruction or house expansion, others save for new
windows or bathroom. The goal of this program is to teach the participants to think forward and plan their
future as well as to teach them to save money, despite their modest income.‖
The terms and conditions of the Savings Program are clear but strict. The clients have to open a savings
account and monthly deposit a certain amount of money, usually 10-50 Euros. During their saving period,
the clients are neither allowed to make any cash withdrawals nor miss their monthly deposit. Those who
meet the requirements of the program receive a bonus equal to the amount on their saving account. The
clients can also participate in a Micro-loan Program after they have finished saving.
Source: http://www.etp.sk/en/sporiaci-program-ida/#more-92
33
http://www.etp.sk/en/sporiaci-program-ida/#more-92
68
Policy Measure 3: Taking advantage of government social protection payment systems to promote
financial inclusion. The Government can play a significant role in incentivizing financial inclusion of
Roma by transferring various benefits to accounts that beneficiaries – Roma and non-Roma - open in
commercial banks. Providing stable monthly inflow from the social benefits into these accounts may be an
important tool for the Government to agree with banks on the provision of low cost ―no frills‖ bank
accounts for the poor.
There are many experiences from around the world that Slovakia can draw from to put in place a
good - non-stigmatizing - system of E-payments to replace cash payments to poor households. See,
for example: www.worldbank.org/paymentsystems. This website includes a report titled ―General
Guidelines for the Development of Government Payment Programs‖ (World Bank, 2012), which presents
a set of comprehensive guidelines that can assist governments and other stakeholders in designing and
operating efficient government payment programs. The Slovak system of E-payments should take into
consideration the relatively large distances to nearby bank branches and consider ways to promote low-
cost branchless banking through the mobile phone. Approximately 65% of Slovak Roma households
currently possess a mobile phone, a landline or both.
In the long run, a good system of E-payments can also provide significant savings to governments34.
For example, disbursing grants and funds electronically to beneficiaries with newly established bank
accounts helped reduce the administrative costs of Brazil‘s conditional cash transfer program Bolsa
Familia from 14.7% to 2.6% of the disbursed grant value, and a recent McKinsey report suggests that the
government of India could save $22.4bn annually by switching to electronic transfers35.
34
World Bank presentations at the Banking for Progress workshop in Bratislava on May 24, 2012.
35
Johnson, C., and P. Jaisinghani, We Must Do Better Than Cash. USAID Impact Blog, 2012.
http://blog.usaid.gov/2012/02/we-must-do-better-than-cash/
69
4.3 BIBLIOGRAPHY
Carbo, S. et al, 2005: Financial Exclusion. New York: Palgrave MacMillan.
Caskey, J., C. R. Duran, and T. M. Solo (2006). The Urban Unbanked in Mexico and the United States.
Policy Research Working Paper 3835. World Bank, Washington, DC.
Collard, S. et al., 2001: Tackling financial exclusion: An area-based approach. Bristol: The Policy Press.
EC(2011) 4977/4: Recommendation on Access to a Basic Payment Account.
Ehrbeck, 2011: ―Full Financial Inclusion in Our Lifetime,‖ CGAP: http://www.cgap.org/gm/document-
1.9.53154/Tilman_Ehrbeck_Transcript.pdf, Accessed on: 7 Sep. 2011.
Ford, J. and Rowlingson, K., 1996: ―Low-income households and credit: exclusion, preference, and
inclusion,‖ Environment and Planning A, 28, 1345-1360.
Ivanov, A. and Tursaliev, S. 2006: ‗Microlending to the Roma in Central and Southeastern Europe: Mixed
Results, New Approaches,‘ Comparative Economic Studies, 48, 36–49.
Johnson, C., and P. Jaisinghani, We Must Do Better Than Cash. USAID Impact Blog, 2012.
http://blog.usaid.gov/2012/02/we-must-do-better-than-cash/
Karlan, D. and Morduch, J., 2009: ‗Access to Finance,‘ Chapter 2 in Rodrik, D. and Rosenzweig, M., eds,
Handbook of Development Economics, Volume 5.
Peachey, S. and Roe, A., 2004: ―Access to Finance: A study for the World Savings Banks Institute,‖
Oxford Policy Management.
UNDP (2002), ―The Roma in Central and Eastern Europe: Avoiding the Dependency Trap‖. Regional
Human Development Report. UNDP: Bratislava.
World Bank (2006). ―Banking Services for Everyone? Barriers to Bank Access and Use around the
World.‖ Policy Research Working Paper 4079, World Bank: Washington, DC.
__________ (2008). ―Finance for All? Policies and Pitfalls in Expanding Access,‖ Washington, DC:
World Bank
__________ (2012). ―Protecting the Poor and Promoting Employability: An assessment of the social
assistance system in the Slovak Republic.‖
70
5 EDUCATION
This chapter provides a diagnostic of the gap between Roma and the general population in the area of
education, and policy recommendations to address this gap. In particular, following an overview of main
indicators of educational outcomes, the chapter analyzes in detail the challenges facing Slovak Roma
children accessing pre-schools, followed by an analysis of the discouraging trends in special education,
and a more macro view of education financing in Slovakia broadly as well as estimate of the fiscal returns
to investing in education for Roma children. The section on policy recommendations to improve
educational outcomes groups these in three areas, with a priority focus on investing in infants and young
children: (1) increase access to quality pre-school, moving toward compulsory preschool from age 3
onwards, and improve home parenting; (2) promote integrated regular primary schooling for all; and, (3)
address early (secondary) school leaving. (International) examples for each of these are provided.
5.1 INTRODUCTION
Roma‘s educational achievements continue to be far below those of non-Roma, regardless of how
these are measured. There is a large gap in preschool access, Roma children are much more likely to be
streamed into special schools for mentally disabled children or into special classes in regular schools, and
rates of secondary school completion and performance on standardized tests are much lower. In some
cases, such as streaming children into special schools, the trends are also getting worse over time.
Education is not only a basic human right, but it is arguably the best passport to a better life for
people living in poverty. Research from around the world indicates that education is a strong
determinant of employment and income.36 Women‘s education in particular is also closely linked to other
positive outcomes such as better child health, nutrition and education, and more manageable fertility rates.
Low education therefore tends to cause poverty, because individuals with little or no education are more
likely to be unemployed or to earn very little. But the reverse is also true: children from disadvantaged
households perform less well in school on average than those from more advantaged households, and
leave school earlier. Thus, low education and poverty form a vicious circle that passes poverty from one
generation to the next. Slovakia is no exception to this worldwide pattern -- employment rates for
individuals who completed upper secondary school are 16.8% higher than for those who did not complete
it, and on average, incomes are 30% higher among this group as well. 37
As Roma educational achievements lag behind those of the majority population in the countries
where they live, improving Roma education is a key priority. The Roma Education Fund (REF),
established in 2005 in the framework of the Decade of Roma Inclusion and supported by several bilateral
and multilateral organizations, symbolizes the international commitment to closing the gap in educational
outcomes between Roma and non-Roma.38 Indeed, education is probably the sector that has received the
greatest attention by scholars, activists and governments seeking to promote Roma inclusion. In
particular, Roma school segregation and Roma over-representation in schools for the handicapped have
been the object of much debate.
36
See, for example: Patrinos, Harrys, and George Psacharopoulos, 2010. Returns to Education in Developing
Countries, Oxford, England: Elsevier.
37
Income was measured based on a) the income streams of all household members; b) the composition of the
household, according to the OECD scale of adult equivalent household size. As such, the income levels that were
used here take into account a number of important characteristics of the household in which each individual lives.
38
REF main programs include: (a) project support grants to governments and NGOs, (b) policy development, (c)
communication and cross-country learning, (d) tertiary level scholarships, and (e) reimbursable grants to help NGOs
and local governments access EU funds for the purpose of expanding Roma access to education.
71
BOX 5-1: THE NRIS AND IMPROVING EDUCATION AMONG ROMA
The Roma Integration Strategy produced by the Office of the Plenipotentiary for Roma Communities in
December 2011 represents the latest Government attempt to tackle the educational needs of the Roma.
The 2004 Concept of Integrated Education of Romani Children and Youth, Including the Development of
Secondary and Higher Education remains the central education-specific document targeting Roma. Its key
objectives include improving readiness for school, reducing attendance to special schools, increasing
secondary and tertiary school attendance, mentoring arrangements for Romani students, and introducing a
study program at the university level focusing on Romani language and literature. The 2011 Strategy calls
for a comprehensive reform of the Slovak education system focusing on three objectives: (a) a massive
increase in the schooling of Roma children three years old and above (from 18% in 2010 to 50% in 2020);
(b) the development and implementation of desegregation standards in schooling; and (c) the development
of specific models of school integration and the eventual establishment of a ―school inclusiveness index‖.
Additional specific objectives include: reaching 100% basic school completion, closing the gap in upper
secondary enrollment, increasing the number of teachers and pedagogical specialists fluent in Romani,
exercising the right to education in a Romani language or to learning Romani, and reducing the
overrepresentation of Roma in special education.
However, since a persistent lack of quantitative information has long stumped efforts to get a more
accurate view of the situation, it has thus far remained challenging to engage in well-informed
dialogue. On the one hand, education data are not disaggregated by ethnic group, and on the other, many
initiatives aimed at improving Roma educational outcomes are not well documented, making it difficult to
assess their performance or learn from their experience.39 Unless otherwise noted, information on Roma
comes from the 2011 Regional Roma Survey, representative of 83% of the Slovak Roma population.
Approximately 42% of Roma that were interviewed in Slovakia report living in communities or
neighborhoods where the majority is Roma. Additional information comes from a survey on the living
conditions of Roma in Slovakia carried out in 2010 by UNDP in cooperation with the Ministry of Labor,
Social Affairs and Family. This survey covered 90 Roma settlements (divided equally among segregated,
separated and scattered settlement depending on the level of integration with the majority population) for a
total of 720 households.
5.2 EDUCATIONAL OUTCOMES – ROMA VS. NON-ROMA
Enrolment rates in basic (grades 1-9) compulsory school are very similar for Roma and non-Roma
children living nearby. In the age group 6-15, which corresponds to the years of compulsory education,
the enrolment rate of Roma is 79% and that of non-Roma living nearby is 83% (see Figure 5-1).
According to OECD‘s Education at a Glance reports, the national ne t enrolment rate for 5-14 year olds is
96.1%.40 Hence, there is a small yet noticeable gap in enrolment rates between Roma and non-Roma,
already at primary school ages. Enrolment differences between Roma boys and girls are very small.
39
See, for example, the evaluation of REF-supported projects carried out by Marek HojsÃk in 2010.
72
FIGURE 5-1: ENROLMENT RATES AMONG ROMA AND NON-ROMA BY AGE GROUP
Percent
100
79 83
80 74
60
36
40
20
0
Agegroup 6-15 Agegroup 16-19
Roma Non-Roma
Source: UNDP/World Bank/EC regional Roma survey (2011).
But beyond compulsory schooling, enrolment differences become large. When they reach 16 years of
age, Roma children start abandoning school and results in a significant enrollment gap in the 16-19 age
group, which corresponds to upper secondary school (see Figure 5-1 above). Thus, while only about one
third of Roma 16-19 years old are still enrolled in school, three fourths of their non-Roma neighbors are
still enrolled. Not surprisingly, even though Slovakia has overall higher secondary school completion
rates than most other countries in the region (Figure 5-2), the percentage of Roma completing secondary
school (21% for boys and 15% for girls) is alarmingly low for European standards and dramatically lower
than for non-Roma (approx. 70 percentage points lower). Also not surprisingly, given the patterns across
the region, girls‘ completion rates are even lower than those for boys.
FIGURE 5-2 : SECONDARY SCHOOL COMPLETION RATES AMONG ROMA AND NON-ROMA
NEIGHBORS, AGE 25- 64
A. Roma B. Non-Roma
90 90 87 85 88 85
80 80 75 76
70 70 59 59
60 60 52
Percent
Percent
51
50 50
40 29 40
30 21 21 18 30
20 15 15 12
10 11 20
10 6 10
- 0
Men Women Men Women
Source: UNDP/World Bank/EC regional Roma survey (2011). ‗Secondary school completion‘ is defined as having
completed either a vocational/technical or a general secondary school program, or a higher level of education.
Sample restricted to age group 25-64.
73
As a result, education levels are very low among the working-age Roma. Table 5-1 shows the
distribution of the highest education level attended. It shows that only about one third (32%) of Roma men
and one quarter of Roma women (24%) has attended (not necessarily completed) some form of upper
secondary school. 41 Among the general population, nearly three quarters (74% and 71% of men and
women, respectively) have reached this level, while more than one in five (22% of both men and women)
have reached tertiary level.
TABLE 5-1: HIGHEST LEVEL OF EDUCATION ATTENDED
Highest level of education attended General Population Roma
(%) (%)
Male Female Male Female
Basic education (or none) 0 0 5 7
Lower secondary/Upper basic 4 7 63 70
Upper secondary 74 71 32 24
Tertiary 22 22 0 0
A
Source: Roma: UNDP/World Bank/EC regional Roma survey; General population: EU-SILC (Eurostat, 2009).
Education levels for 25-64 year-old population excluding current students.
Whether the household head has completed secondary education is the most important factor
related to Roma‘s lower educational achievements. When analyzing the causes of low secondary
education completion rates among Roma 42 (see Annex), it turns out that whether the household head
completed secondary school is the factor that has the largest effect: this makes Roma children over 40
percentage points more likely to complete secondary school (significant at 1% level). Other background
characteristics that appear to play a role are43: the size of the household (the larger the household, the less
likely to complete high school), income (the richer the household, the more likely to complete high
school), suffering from hunger (which has a negative impact), living in rural areas (which makes Roma
more likely to complete high school), having attended preschool and proximity to a high school (both of
41
This category includes: (a) Incomplete secondary vocational/technical; (b) Secondary voc/technical (1 or 2years);
(c) Secondary vocational/technical (3 or 4 years); (d) Incomplete secondary general (4 years); (e) Secondary general
(4 years).
42
The model used to analyze the causes of (low) secondary school completion is an Ordinary Least Squares (OLS)
regression model, with Secondary School Completion as the dependent variable. In this case, ‗secondary school
completion‘ is defined broadly, i.e., also including subjects who completed only part of their secondary school.
Independent variables included are: ethnicity (Roma vs. non-Roma living nearby), gender, age, preschool attendance,
gender of the household head, educational background of the household head (i.e., whether the latter has completed
secondary school), age of the household head, the number of household members, income (included as quintiles),
whether the household suffers from hunger, whether Romani is spoken at home, whether the closest secondary
school is within walking distance from the household‘s dwelling (i.e., within 3 km), whether the dominant ethnicit y
of the settlement is Roma, the region (West, Central, East), and whether the household lives in a rural or an urban
location. The model was estimated for those who had completed at least some primary education, among the age
group 18-30, in three iterations: (1) for both Roma and non-Roma living nearby (with the ethnicity independent
variable included), (2) for Roma only, and (3) for non-Roma only (in models 2 and 3 the ethnicity variable was left
out). The Annex at the end of this chapter gives a more elaborate description of the model, including the full set of
model estimates.
43
This list was taken from the second model iteration, i.e. the model including only Roma subjects. Only significant
(at 10 percent level) predictors are listed here.
74
which have a positive impact). It should be noted that for Roma, contrary to what happens for non-Roma
living nearby, when all other background characteristics are the same, being a girl is not a disadvantage. It
is also interesting that for non-Roma living nearby, only three factors are significantly associated to higher
rates of secondary school completion: gender (with women being at a disadvantage), age, and whether the
household head completed secondary school (which has the strongest effect).44
Low educational aspirations by Roma are unlikely a key impediment to educational outcomes.
While the discrepancy in opportunities is clear, one may wonder whether, even if barriers to accessing
quality education were removed, Roma children would be interested in studying more when their parents
have received so little education themselves. This is partly because parents who have not reached a high
level of education may not be aware of the value of being educated. Household members aged 16 years
and older in the regional Roma survey (2011) were asked ―what do you believe is a sufficient level of
education for a Roma child?‖. The responses were similar for girls and boys, and mirrored the actual
educational attainment among the general population: more than 60% wanted an upper secondary school
diploma. In the same vein, the 2010 Slovak Roma survey found that three fourths of the parents of
elementary school students wanted their children to go on to secondary school. On the other hand, less
than half of the parents of high school students (46%) wanted their children to go on to university,
although this may be more the reflection of significantly greater costs of attending university than of lack
of ambition. As such, low aspirations do not seem to be an impediment to educational outcomes, at least
up to secondary school completion, among Roma.
Ethnicity seems to matter regardless of background characteristics and aspirations. When all
characteristics are equal, on average a Roma is 26 percentage points less likely than a non-Roma to
complete secondary education (significant at 1% level – see Annex Table 5.1). This means that Roma‘s
educational attainment is lower, not only because Roma tend to have unfavorable backgrounds --e.g., a
poor and large family, uneducated parents and no preschool-- but also because of their ‗being Roma‘.
With the information available from existing nationally representative surveys it is impossible to
determine in which ways ―being Roma‖ is an obstacle, but qualitative research provides abundant material
to make some educated guesses. On the one hand, it has been noted that the Slovak school system does
not respect traditional Roma culture and socialization patterns 45 Roma children are raised with much
more freedom than their non-Roma counterparts, without stressing delayed gratification and discipline.
They are expected to learn primarily by mimicking and without correction from adults, with special
emphasis on non-verbal communication –something that helps explain their relatively limited spoken
language development when they start school. Also, even though Roma parents have educational
aspirations for their children, they are generally less equipped to help their children with homework. On
the other hand, there have been many accounts of teacher bias and discrimination against Roma students,
or simply of the lower expectations teachers have for the school performance of Roma children, which are
then reflected in lower achievement and low self-esteem/low ambition.46 Indeed, among Roma in Slovakia
who report having been in contact with educational institutions, 32% have experienced discrimination,
almost half of which was based on ethnicity. Among non-Roma, 27% reports to have experienced
discrimination, but none of these discrimination cases were based on ethnicity.
44
The fact that income does not appear to be a good predictor of high school completion for non-Roma in the survey
should not be interpreted as meaning that income no longer matters for educational achievement in Slovakia. It does
matter overall, but non-Roma living near Roma settlements tend to be in the lower income brackets, and it is the
variations within the lower income brackets that make no significant difference.
45
See for example: Marcincin and Marcincinová (2009), and Kosová and Hul‘ová (2006).
46
Sociological research offers convincing evidence of the impact of low expectations and negative labeling on
performance. Self-fulfilling prophecy is one of the main theories that have been used to explain the influence of
teacher expectations and perceptions of students on student achievements. See for example Minstry et al. (2009).
75
The reasons given by young Roma for dropping out of school are both economic and cultural. When
asked why they had not continued schooling all the way to complete secondary school, Roma drop-outs
younger than 23 years of age gave three main answers: the costs of education were too high, they felt they
were already sufficiently educated or, in the case of girls only, they got pregnant (11%). The first answer
can be easily understood, as even in public schools there are considerable costs associated with studying,
such as transport, books and fees for specific activities. Indeed, almost half of the Roma households in the
2010 survey said that they had difficulties in covering educational expenditures for high school students
(see Box 5-2).47 The feeling that lower secondary school represents a sufficient educational achievement,
which was reported to be the main reason for not completing secondary school by 31% of Roma boys and
24% of Roma girls aged 6-23 in the 2011 survey, reflects a misperception in terms of the impact of
education but can be understood considering that employment rates for Roma are a measly 20% for men
and 9% for women --as succinctly put by a Roma youth, if the end result is going to be unemployment,
why bother getting an education? More in general, this apparent limited ambition can be explained as the
expression of the low self-esteem that psychological studies have typically found among adolescents who
perceive themselves as victims of discrimination (see for example Harris-Britt et al, 2007). However,
bullying, which often underlines a discriminatory attitude by classmates, was not mentioned as a reason to
abandon school.
BOX 5-2: DIFFICULTIES IN ATTENDING HIGH SCHOOL
The staff from a high school with a sizable number of students coming from Roma settlements point out
that some of their students may not come to school because they have no good shoes to brave bad weather,
or simply no decent clothes to wear. Transport money is also a problem, even if the round trip on the bus
is just €1. But perhaps the biggest difficulty is the sense of inferiority that eats away their self-confidence.
For example, at an event offering a buffet, none of the Roma students dared to eat in front of the others --
and they certainly were hungry-- conscious of their poor manners and worried about being judged needy
or greedy.
Source: Personal communication, Kosice.
It has also been suggested that Roma youth may abandon secondary school to take advantage of
labor activation programs. Upon turning 16, youth become eligible for labor activation programs that
provide paid employment for up to six months at about € 60-65 per month (after a six-month break,
another twelve-month cycle is possible). This amount is considerably higher than the social scholarships
provided on the basis of school attendance (€ 20-40 per month) and, more importantly, it goes directly into
the pockets of the youngster while the social scholarship goes either to the school or, depending on the
school‘s decision, to the parents. Needless to say, abandoning school for a short -term job is a short-
sighted strategy, but its attraction from the point of view of an adolescent can be easily understood,
especially when there is no counterargument offered by parents and peers. Indeed, several field workers
have pointed out that, contrary to non-Roma parents, the typical Roma parent in a settlement does not
encourage his/her children to pursue an education. As it is typical of people living in poverty and with
47
There are teaching methods that target problems like the ones described in the bottom part of Box 1. For example a
paper by Sharan examines five such methods, and evaluates their effects on academic achievements, interactions
between children of different ethnic backgrounds, etc. See: Sharan, S., 1980: ―Cooperative Learning in Small
Groups: Recent Methods and Effects on Achievement, Attitudes, and Ethnic Relations‖ Review of Educational
Research, 50 (2): 241-271.
76
uncertain futures, the planning horizon is very short and a certain immediate earning is preferable to the
higher but less certain rewards of a high school degree.
5.3 GETTING OFF TO A GOOD START – EARLY CHILDHOOD EDUCATION
The majority of Slovakia‘s children attend preschool from the age of three. Regular attendance of
quality kindergartens is a high return investment, both for the individual in terms of cognitive
development and for society in terms of improved labor productivity in adulthood. In 2010, there were
162,000 children aged 3-5 years old. In the same year, there were approximately 139,000 children (of
varying ages) in 7,126 kindergarten classes (National Statistics Office), reflecting a gross enrolment rate
of 85%. Net enrolment stood at 72 percentage for the ages 3-5,48 meaning that about 45,000 children in
this age group were not enrolled in pre-school during this period.
A large body of international evidence underscores the importance of early intervention –from
conception to age 8– on child development.49 Early childhood development programs are particularly
beneficial to children from disadvantaged backgrounds.50 As Nobel laureate economist James Heckman
argues, ―investing in disadvantaged young children is a rare public policy with no equity-efficiency
tradeoff‖51. A new review of the scientific literature by The Lancet (October 2011) similarly concludes
that ―[…] unless governments allocate more resources to quality early child development programmes for
the poorest people in the population, economic disparities will continue and widen.‖ 52
BOX 5-3: EVIDENCE ON THE IMPORTANCE OF EARLY CHILDHOOD
DEVELOPMENT
In 2007, the international science journal The Lancet published a first series on early child development,
reviewing the evidence from the international scientific literature. In October 2011, a new series of two
review papers documented subsequent progress worldwide.
The first review article in the series1 underscores that inequalities in child development begin prenatally
and in the first years of life. These inequalities include insufficient early intake of micronutrients (certain
minerals and vitamins) and lower levels of cognitive stimulation. The evidence reviewed underscores that
the most effective and cost-efficient time to prevent inequalities is early in life before trajectories have
been firmly established.
The second scientific review article1 assesses the effectiveness of early child development interventions. It
concludes that parenting support and preschool enrolment can improve early child development, ―[…]
with effects greater for programmes of higher quality and for the most vulnerable children. Other
promising interventions for the promotion of early child development include children's educational
media, interventions with children at high risk, and combining the promotion of early child development
with conditional cash transfer programmes. Effective investments in early child development have the
48
UNICEF, TransMONEE database 2011.
49
Nores and Barnett 2010; Burger 2010; UNESCO 2007.
50
Burger 2010; Heckman and Masterov 2007; Young 1996; Open Society Foundations/Roma Education
Fund/UNICEF (2012).
51
Heckman and Masterov 2007.
52
Engle et al. (2011) ―Strategies for reducing inequalities and improving developmental outcomes for young children
in low-income and middle-income countries‖. The Lancet, 378 ( 9799): 1339 – 1353.
77
potential to reduce inequalities perpetuated by poverty, poor nutrition, and restricted learning
opportunities. A simulation model of the potential long-term economic effects of increasing preschool
enrolment to 25% or 50% in every low-income and middle-income country showed a benefit-to-cost ratio
ranging from 6·4 to 17·6, depending on preschool enrolment rate and discount rate.‖
Roma children benefiting from preschool and parental stimulation have significantly higher
cognitive outcomes, consistent with the international evidence. Roma children 4-6 years old attending
preschool in Slovakia (as well as in the Czech Republic, Hungary, Romania and Bulgaria) are more likely
to be able to (a) identify ten letters of the alphabet, (b) read four simple popular words, (c) write their own
name, (d) recognize numbers from 1-10, and (e) know simple sentences in the national language.53 For
example, about 65% of Slovak Roma children aged 4-6 are reported to know simple sentences in Slovak,
but the percentage goes up to approximately 85% among those who are attending pre-school. Similarly,
while 38% of 4-6 year old Roma not attending preschool can identify at least ten letters of the alphabet,
the percentage goes up to 62% if they attend preschool. Likewise, Roma children whose parents are
reported to have taught them letters or how to count in the past three days have significantly higher
cognitive outcomes.
Roma children benefiting from preschool have also significantly better later life outcomes: they are
much less likely to enroll into special school, more likely to complete secondary school, and less
likely to be on social assistance. Again, consistent with the international evidence, when comparing
Roma from the same neighborhood, those who attended kindergarten as children have better outcomes in
later life than those who did not. For example, subsequent enrolment into special school is reduced by 7
percentage points. Given that 12% of Slovak Roma children are streamed into special primary schools,
this reduction of 7 percentage points is equivalent to a 58% drop. Similar sizeable differences are found
for the Czech Republic. Preschool attendance is also strongly linked with subsequent secondary school
completion: 15% secondary school completion among those who didn‘t attend preschool compared with
20% among those who did, which reflects an increase of 33%. Regression estimates suggest that the
increase is even larger when comparing individuals within the same geographical communities. And,
preschool attendance is linked to lower rates of social assistance as adults – by 11 percentage points for
Slovakia.
The vast majority of Slovak Roma children lack access to preschool, with preschool enrolment rates
far behind the national average for children aged 3-5 years. There are about 27,000 young Roma
children aged 3-554 (9,000 children aged 3, 9,000 aged 4, and another 9,000 aged 6). Since approximately
four out of five are not in preschool, this represents around 21,000 children aged 3-5 who do not have
access to preschool. Within this same age bracket, the average enrolment rate among surveyed Roma
children is a mere 18%, well below the national average of 72%. Roma girls especially are unlikely to be
enrolled in kindergarten. This very large gap is comparable to the gap that exists in the Czech Republic.
On the other hand, Hungary has much higher preschool enrolment rates with more than two thirds of
Hungarian Roma children now in kindergarten (see Table 5-2 below).
53
These results are based on OLS estimations whereby cognitive outcomes are the dependent variables. The
estimations control for enumeration area fixed effects, which effectively means that the outcomes of Roma children
from the same neighborhoods –some participating in the local preschool, while others are not– are compared. The
estimations also control for background characteristics such as the child‘s age, gender, hospital birth, general health
states, background characteristics of the child‘s primary caretaker (age, gender, whether s/he works, attended
preschool in the past, and secondary school completion), and quintiles of per capita household income.
54
Based on the population structure in the sample of Roma interviewed as part of the UNDP/WB/EC Regional
Policy Roma survey (2011), and assuming the Slovak Roma population total of 320,000.
78
TABLE 5-2: PRESCHOOL NET ENROLMENT RATES
Bulgaria Czech Republic* Slovakia* Hungary Romania
Roma girls 38 19 14 71 32
Roma boys 42 25 20 61 33
Roma average (2011) 40 22 18 66 32
National average (2009-10) 75 79 72 88 77
Source: UNDP/World Bank/EC regional Roma survey (2011). A To make comparison with national data, we rely on
the UNICEF's TransMONEE database 2011. National data for Slovakia are for the year 2008-09. To make the
estimates consistent with age groups used to report preschool enrollments in the TransMONEE 2011 database,
enrollments for the Czech Republic and Slovakia (*) were estimated for the 3-5 year age group. For the other
countries, the TransMONEE age group is 3-6 years.
In Slovakia pre-primary education remains voluntary, although the newest education strategy
advocates for compulsory early childhood education. Pre-primary education caters for children from 3
to 6 years of age. Attendance is not compulsory. Public-sector authorities partially charge fees, except to
parents of children in the last year before compulsory school attendance.55 Preparatory classes, called the
zero grade, can be set up in primary schools for children who are 6 or older and who are not considered
ready to enter primary school. These children can be placed into zero grade with the agreement of the
legal guardian (Dral et al., 2008). 56
Several new EU Member states are expanding preschool. Since 2011, Poland, for example, requires all
five year olds to complete a year in preschool. In Bulgaria legal amendments from September 2010
provide for two years of compulsory preschool education, which would encompass all 5-year olds.
Municipalities have two years to ensure that they are able to accommodate all children and the
government, with World Bank financing, has embarked on an ambitious Social Inclusion Program to
expand preschool access. In Romania, according to a legal amendment from January 2011, as of academic
year 2012-13, there will be one compulsory year of preschool education that is meant to equalize the level
of preparation of all children for the primary grades. And, finally, Hungary has been able to reach much
higher enrolment rates by making preschool a priority area, as explained in Box 5-4.
BOX 5-4: PRESCHOOL IN HUNGARY
Preschool enrolment among Roma (and the poor more generally) in Hungary is considerably higher than
in Slovakia, and than in other neighboring countries. For example, while 49% of Hungarian Roma
children report attending some form of preschool, the corresponding rate among Slovak Roma is 10%. At
age five, 86% of Hungarian Roma children are enrolled compared with a mere 36% of Slovak Roma and
75% of Slovak non-Roma neighbors. In other words, Hungarian Roma are more likely to be in pre-school
than Slovak non-Roma living close to Slovak Roma. There are some important differences between the
Hungarian and Slovakia preschool systems.
55
Eurydice (2011). National system overview on education systems in Europe; 2011 Edition. Slovakia
56
It has been remarked that zero grades can also be an important pathway for Roma segregation and enrollment in
special schools and classes. Since Roma are more likely to be ‗unprepared‘ for school than other children, they are
put in zero classes disproportionately. Then, since they don‘t enter with their peers they are not ‗ready‘ at the end of
the zero grade and are placed in special schools/classes, supposedly to give them time to catch up. But the testing
ensures they remain in special schools. This is discussed in School as Ghetto amongst other sources.
79
- In Hungary, kindergarten was optional from the age of 3 and compulsory from the age of 5. The new
education law, passed in December 2011, calls for compulsory preschool from age 3. Mandatory primary
school entry age is 6 years, although children may stay in kindergarten for an extra year, until turning 7.
-Public-sector kindergartens charge no tuition, although they charge a compensation for extra services not
included in their basic tasks, such as for meals, excursions and extracurricular activities. Non state
kindergartens may charge fees.
- To improve access for the poor, meals have been free for families receiving a supplemental child
protection allowance. Furthermore, since 2009, parents of multiply disadvantaged children have been
encouraged through subsidies to enroll their children as early as possible. The so-called "kindergarten
subsidy program" grants disadvantaged families a twice a year subsidy of 10 000 HUF (approx 35 Euro)
per child aged 3-4 years conditional on the child attending pre-school regularly. The eligibility criteria are
based on multiple disadvantages (a legal category, it including low education of the mother and means
testing by the local notary).
- Financing of preschool has come from a combination of central government funding (30-40 percent),
parents (10 parents), and municipal governments (the rest). Parents‘ fees are lowered or cancelled
completely for those with low incomes. Municipal financing has been a challenge for poor municipalities.
It is possible for them to contract with private and voluntary sector providers, services are almost entirely
public.
To address the shortage of preschool places, the government issued specific calls for proposals in 2009
partly for the improvement of school education and partly for the development of kindergartens and
kindergarten-related projects in the 33 most disadvantaged micro-regions. These fell under the scheme of
the infrastructure development within the Regional Operational Programme. Nevertheless, lack of
physical space and personnel in preschools was a constraint in many instances.
Slovak Roma parents with children in preschool report spending approximately € 7 per month on
preschool, with lunch being provided for a fee in three fourths of the cases. The survey asked parents
about the expenses on fees, books, transport, clothes (uniforms), and food. The table below shows the low
actual costs in Hungary (on average €1.3 per month) compared with Slovakia, with the highest expenses
reported in the Czech Republic (€ 25.6). The large majority of Roma parents in Slovakia, as well as in
Bulgaria and Czech Republic, report that children receive food that is covered by a fee charged to the
parents, while in Romania the majority of parents report that their children are expected to take their own
lunch with them. Hungary stands out for freely providing food to virtually all children (see Table 5-3 for
details).
80
TABLE 5-3: MONTHLY PRESCHOOL EXPENSES BY PARENTS WITH CHILDREN IN PRESCHOOL, AND
PROVISION OF FOOD
Czech
Bulgaria Hungary Romania Slovakia
Republic
Monthly cost (mean - Euro) 15.4 25.6 1.3 7.5 7.2
Monthly cost (median - Euro) 17.4 24.7 0.0 2.2 4.0
Provision of food (%)
Yes, freely provided 23.1 4.9 97.0 33.0 28.0
Covered by fee 75.6 92.2 2.4 2.7 72.0
Children must bring own lunch 1.3 2.9 0.6 64.3 0.0
Total 100.0 100.0 100.0 100.0 100.0
Source: UNDP/World Bank/EC regional Roma survey (2011).
Preschool in Slovakia is more segregated than in neighboring countries. According to the regional
Roma survey data (2011), Slovakia had the highest fraction of children attending ‗all‘ or ‗nearly all‘ Roma
kindergartens (48%), indicating a high degree of segregation in early education. In Romania the figure is
similar (46%) but in Bulgaria, Czech Republic and Hungary it is much lower (18%, 16% and 21%,
respectively). It should be noted that the high degree of segregation is not the result of preschools being
located in Roma settlements, as only 11% of Roma children in preschool attend a kindergarten located in
the settlement.
While the majority of Slovak Roma parents with children enrolled in preschool feel that their
children are welcomed, about one third of parents feel they are not. This is presented in Figure 5-3.
The rate of dissatisfaction, although still a minority, is higher in Slovakia than in all other countries that
were surveyed.
FIGURE 5-3: SATISFACTION WITH PRESCHOOLS AMONG ROMA CHILDREN AND PARENTS
A. Do Roma children feel welcome in preschools? B. Parental satisfaction
100
100
% of parents satisfied with preschool
80
80
60
60
40
40
20
20
0
0
Bulgaria Czech Republic Slovakia Hungary Romania Bulgaria Czech Republic Slovakia Hungary Romania
Don't agree Agree Very dissatisfied Somewhat dissatisfied
Completely agree Somewhat satisfied Very satisfied
Source: UNDP/World Bank/EC regional Roma survey (2011).
81
Most Roma households report not sending a child to preschool because they thought the child was
too young and/or because home care was available, but many would consider enrolling their child
into preschool if there were no fees involved or if they would receive food coupons or if the school
had a Roma teacher (assistant). A substantial proportion (16%) of parents also reported ‗too far‘ as a
main reason for not sending their children to kindergarten. Table 5-4 shows that the responses around
home care are comparable with the responses in other countries, while fewer parents in Slovakia (and
Hungary) report expenses as being a main reason, and more parents report ‗too far‘ as being a main
reason. On the other hand, when asked if they would reconsider enrolling their child into preschool if
preschool were free or if they would receive a food coupon, more than 40% of Slovak Roma parents
responded ‗yes‘, and approximately 20% responded ‗maybe‘. This is consistent with reports from
kindergarten staff that children tend to stop coming to school if the household loses eligibility to the
cheaper food fees charged to recipients of social aid. Also, almost half of the parents indicated that they
would reconsider sending their children to preschool if there were a Roma teacher or teaching assistant.
TABLE 5-4: REASONS FOR NOT SENDING CHILD (3-6) TO PRESCHOOL
Bulgaria Czech Hungary Romania Slovakia
Republic
Child is too young 28% 26% 49% 46% 23%
No need (have home care) 29% 40% 24% 21% 37%
Too expensive 40% 21% 4% 22% 5%
Child should stay home 8% 13% 3% 2% 7%
Too far 4% 2% 0% 6% 16%
No place 3% 5% 2% 0% 3%
On the waiting list 1% 5% 3% 1% 2%
Don't trust teachers 0% 2% 0% 0% 1%
Language 0% 1% 0% 0% 2%
Child is ill-treated 1% 2% 0% 0% 0%
Source: UNDP/World Bank/EC regional Roma survey (2011).
Many young Roma children could likely benefit from greater cognitive stimulation at home. Child
development depends of course not just on schooling, but importantly on the home environment. Roma
children face multiple disadvantages in this regard. With so many children growing up in deep poverty,
infants are at higher risk of malnutrition, and families lack the means to purchase books and other learning
tools. Furthermore, the very low education levels among men and especially among women, is a barrier to
effective parenting support for cognitive development. The 2011 survey asked about access to books at
home and several questions on parenting techniques. Indeed, few Roma children aged 3-5 have access to
books: on average 2.6 books and a median of only 1 book, indicating that a typical Roma child in this age
group only has 1 book at home. This is comparable to the situation in Bulgaria and Romania, while Roma
families in the Czech Republic and in Hungary have more books (5 and 4, respectively, for the median).
Parental time spent with children in stimulating activities is also an important input to cognitive
development. Less than one quarter of the Roma children aged 3-5 were taught letters or counting by their
caregivers in the past three days, and less than half looked at picture books or read books, or drew or
painted with their caregivers (Table 5-5).
82
TABLE 5-5: HOME ENVIRONMENT
Bulgaria Czech Hungary Romania Slovakia
Republic
Number of books at home:
Mean 1.8 7.0 7.2 1.2 2.6
Median 0 5 4 0 1
Activities with children, past 3 days:
Look at picture books or read books 23% 50% 57% 17% 44%
Draw or paint 21% 51% 42% 19% 45%
Teach letters or count 15% 21% 29% 12% 22%
Source: UNDP/World Bank/EC regional Roma survey (2011). A Limited to households with children aged 3-5 years
old. Source: regional Roma survey (2011). Authors‘ calculations.
5.4 THE SPECIAL CASE OF SPECIAL EDUCATION
Roma children continue to be over-represented in special schools and classes. Official data in 2003-
04 showed that while the share of Roma children in standard primary schools was 0.53%, the figure for
special schools was almost fifteen times higher – 7.6% (Salner, 2005). In the same vein, estimates based
on field research carried out in 2008-09 on a statistically representative sample were that Roma made up
about 60% of the enrolments in special schools (Friedman et al., 2009). In particular, Roma were
estimated to be 76% of the children in the first four years of special primary education, and 86% of the
students in special classes within standard schools. The findings of the 2010 Slovak Roma Survey and the
2011 regional Roma survey confirm Roma over-representation in special education. 91% of special
classes in regular schools and 65% of special primary schools have only or almost only Roma students.
The regional Roma survey (2011) supports a very high level of classroom-level segregation among
Roma. Figure 5-4 below shows that nearly half of Roma children currently attending basic education are
in classes where most of the children are Roma. Among their non-Roma neighbors, who were also
interviewed, only 7% are in classes with mostly Roma children.
FIGURE 5-4: SEGREGATION: % OF PUPILS REPORTING TO BE IN CLASS WITH MOSTLY ROMA PUPILS
60
47
40
Percent
20 14
7
0
0
Roma Non-Roma nearby Roma Non-Roma nearby
Enrolled in Basic Education Enrolled in Secondary
Source: UNDP/World Bank/EC regional Roma survey (2011).
The most rigorous study on the impact of desegregation efforts finds positive results. A study by
Kezdi and Suranyi (2009) exploits quasi experimental variation to analyze the impact on cognitive as well
as socio-emotional outcomes of primary education students attending one of 45 schools in the Hungarian
National Educational Integration Network (OOIH), which aimed to provide quality education for students
83
in an integrated environment. The study finds that students in integrated schools had small improvements
on standardized reading comprehension tests, a positive impact on the development of non-cognitive
skills, both among Roma and non-Roma, disadvantaged and non-disadvantaged, and sees improvements
on overall tolerance. As the authors point out: ―Program schools seem to achieve integrated education
without hurting non-Roma and non-disadvantaged students‘ skills development.‖
Slovakia is a particularly severe case in the matter of special education: enrolment into special
education attendance among Roma is higher than other countries in the region with the exception of
the Czech Republic. In Slovakia, 12% of the Roma aged 7-18 go to special schools compared to just 3%
for non-Roma living nearby (6% for boys and less than 1% for girls). As a comparison, estimates by the
World Health Organization put the proportion of severely disabled children 0-14 year old in European
countries at 0.8% (WHO, 2011).57 Thus, either Slovak children are extraordinarily affected by disability
or they are unfairly placed in special schools. The findings also indicate that Slovakia is the country with
the second highest rates of special school attendance among Roma children and the second largest gap
between Roma and non-Roma, with only the Czech Republic having higher rates and gaps.
FIGURE 5-5: SPECIAL SCHOOL ATTENDANCE AMONG ROMA AND NON-ROMA CHILDREN
Percent
20 19
18 16
16
14 12
11
12 10
10
8 7
6 5 6
6 4 5 5
4 4
4 3 3 3 3
1 2
2
0
Men Women Men Women Men Women Men Women Men Women
Slovakia Bulgaria Czech Republic Hungary Romania
Roma Non-Roma
Source: UNDP/World Bank/EC regional Roma survey (2011). Sample restricted to subjects aged 7-18.
Special school attendance is on the increase, and so is segregation. As a reflection of national
demographic trends, the number of pupils in basic schools and the number of basic schools are decreasing.
The number of children in special education settings (schools and classes) and the number of special
schools, however, have been steadily increasing, although the slight inflection in 2011-12 could signal a
reversal of the trend (Table 5-6). This increase is primarily due to a larger proportion of Romani children
attending special education. In fact only 6.4% of Roma (5% men and 8% women) over 30 attended special
schools, as opposed to 11% now (11% boys and 12% girls). Thus, in the space of approximately a
generation Roma‘s attendance rate of special schools has more or less doubled. Findings on segregation
tell a very similar story. According to the 2010 UNDP survey, just 7% of the Roma who graduated before
57
The proportion of moderately and severely disabled children is estimated at 4.2 percent, but moderately disabled
children are normally schooled in regular classes.
84
1990 were is classes with only or almost only Roma children, while presently over one third of Roma
children (36%) are in classes with only or almost only other Roma children.
TABLE 5-6: TRENDS IN THE NUMBER OF SCHOOLS AND STUDENTS
2008-09 2009-10 2010-11 2011-12
Basic Schools:
No. of basic schools 2,237 2,224 2,216 2,202
Pupils in basic schools 462,715 448,371 439,675 434,477
Special Schools:
No. of special schools 233 236 244 239
Pupils in special education 28,328 28,543 28,948 28,828
Source: UIPS
It seems that persisting discriminatory practices have resulted in increases in special education
enrollment and segregation. Some other factors contribute. First, the faster demographic growth of
Roma means that the proportion of Roma children in the school system is higher than at any time before.
This is especially true in schools near Roma settlements, resulting in more schools becoming
overwhelmingly Roma not because of deliberate segregation but because of population dynamics. For
example, in Jarovnice (Eastern Slovakia) there was only one basic school until 1992 but the growth of the
large Roma settlement made it necessary to open another basic school; now one of the two schools is
exclusively Roma and the other one is two thirds Roma. Another explanation may be Roma‘s greater
school attendance over time, which may have brought into the educational system children with special
needs who before were remaining at home. Whereas one fifth of the Roma aged 31 or older report never
having been to school or only having attended part of primary school, these figures are slightly lower
among those aged 18-30 (16% and 18%, respectively) and lower still now, indicating that education levels
are improving over time. However, this alone cannot explain the increased segregation. The other
explanation, therefore, is likely to be greater discrimination. Despite the stated policy goal of reducing the
number of Roma children attending special schools or special classes in basic schools,58 it would appear
that a larger proportion of Roma children entering the school system has triggered a number of practices to
minimize the inclusion of these children in the school system. These include putting Roma children in
separate schools and classes, and taking non-Roma children out of schools with a larger –or more
integrated—Roma student body. This last practice is evident when looking at the distance students travel
to go to school: over one third of non-Roma children leaving near Roma settlements travel more than 10
km to go to school, compared to less than one in ten Roma students (UNDP 2010).
Special school attendance is particularly high in Eastern Slovakia, confirming findings from 2008-
09.59 In this region, special school attendance is now even relatively frequent among non-Roma children
(8% of the boys and 1% of the girls) in stark contrast with the other regions where less than 1% of non-
Roma children attend special schools. Eastern Slovakia is also the only region where special school
attendance is higher now than in the past (see Table 5-7 below), for both Roma and non-Roma. This is
consistent with the results of a study carried out in 2010 by the Slovak Governance Institute on the
58
See, for example, the Government‘s Revised National Action Plan for a Decade of Roma Inclusion 2005-2015 for
years 2011-15: goal number 5 reads ‖to deal with problematic issues of education and upbringing at special schools
and school facilities including school consultancy and prevention‖. Before then, the Basic Theses of the Concept of
the Government of the Slovak Republic’s Policy in the Integration of Roma Communities, in 2003, called for an
immediate solution to the over-representation of Romani children in special schools.
59
See Friedman et al., op. cit.
85
education of children from socially disadvantaged backgrounds, both Roma and non-Roma, which found
an increasing number of socially disadvantaged children in special schools (Gallová Kriglerová, 2010).
Considering that Eastern Slovakia is the poorest region, and that it has been particularly hard hit by the
recent financial crisis, the increase in the proportion of its children in special education would appear to be
a confirmation that socio-economic characteristics continue to play an important role in determining
whether a child will end up in special education, even though the School Law of 2008 requires that no
child be placed in special education on the basis of social disadvantage or ethnicity. On the other hand,
economic disadvantage does not appear to be the only explanation for the high proportion of children in
special schools. In fact, when comparing children from equally poor households, those living in Eastern
Slovakia will still be more likely to attend a special school only if they are Roma, suggesting that
discrimination is also present. The experience reported by a social worker during a conference is telling:
she realized that the reports on the individual diagnostic tests used to declare several Roma children fit for
special schooling were exactly the same, with the psychologist not even having made the effort to match
the pronouns to the sex of each child.
TABLE 5-7: SPECIAL SCHOOL ATTENDANCE BY REGION AND GENDER
West Slovakia Central Slovakia East Slovakia
Men Women Men Women Men Women
ROMA (%)
Age cohort 7-18 0 2 9 11 14 13
Age cohort 19-30 12 5 13 15 11 9
Age cohort 31+ 0 0 4 8 6 8
NON-ROMA (%)
Age cohort 7-18 0 (0) (0) (0) 8 1
Age cohort 19-30 (0) (0) 10 (0) 0 0
Age cohort 31+ 0 0 3 3 1 1
Source: UNDP/World Bank/EC regional Roma survey (2011). Figures in parentheses represent estimates from a
total of less than 20 respondents for a given age cohort, region and gender. This means that estimates may not reflect
the actual situation at the national level.
Why do parents send their children to special schools? Much has been written to explain the
disproportionate number of Roma children in special schools. The most thorough study carried out to date
focusing on this subject is probably ―School as a Ghetto‖ (Friedman et al. 2009). The survey and
interviews conducted then identified a number of incentives for Roma parents to enroll their children in
special education, including expectation of better grades, geographic proximity and promises of material
benefits (e.g., meals and school aid), as well as a number of factors that would discourage Roma‘s
attendance of standard schools, such as bullying by non-Roma pupils. The study also suggested that while
the School Law of 2008 requires the parents‘ informed consent (defined as written consent with awareness
of the consequences of consenting), in reality many parents are not aware of the long-term consequences
of their decision and simply trust school officials or are afraid of questioning their authority.
While in principle special schools should cater to the special needs of children with disabilities, only
13% of the Roma respondents to the 2011 Regional Survey attending special school said that a
mental or physical handicap was the main reason for attending (10% reported a mental handicap and
3% a physical one). This compares to 55% for the non-Roma. Material advantages, such as free food or
books, also appear to play a relatively small role in the decision of Roma parents to enroll their children in
86
special schools, as only 6% gave it as the main motivation. None of the non-Roma mentioned material
advantages as a reason, though field research suggests that the financial crisis is pushing poor non-Roma
parents to enroll their children in schools dominated by Roma because of the free meals provided when at
least 50% of the students are from a socially disadvantaged background, i.e., receive social assistance.
What seems to matter most for the vast majority of Roma --over three-fourths—is that special
schools are easier and until recently this ensured access to social assistance. Why would this be the
case? Previous field work reports that teachers frequently explain special education to parents in terms of
its advantages over standard education because an easier curriculum and a more individual approach make
it more likely for children to be successful in their studies (Friedman et al., 2009). It is understandable
that parents‘ natural desire to see their children thrive would make them particularly receptive to such
arguments, especially if they themselves had a positive experience in a special school (see Box 5-5 for an
example). In addition, school aid used to be linked to academic performance, hence providing an
incentive to enroll in schools where better grades were easier to obtain. While the law changed in 2009 to
make aid during compulsory schooling linked to school participation rather than academic performance,
only the respondents who started school after 2009 would have been affected by such change (it is also
likely that not all Roma parents would have been promptly informed of the policy change). The incentive
represented by school aid policies would also explain the relatively large proportion (44%) of non-Roma
living nearby who said the easier program was the main reason for special schooling. On the other hand,
this commonly cited reason for attending special schools may also point to a lack of adequate provisions
for young Roma children to prepare them for regular primary school classes.
BOX 5-5: THE SPECIAL SCHOOL IN CHMINIANSKE JAKUBOVANY
The special school in Chminianske Jakubovany is warm and welcoming to its 395 students, all coming
from the nearby Roma settlement. The school staff explains that children are made to feel good about
themselves and parents do not have to fear that they will be looked down upon. Many of them actually
went to the same school, so they know their children will be safe, and it is not unusual that they try to
enroll their children regardless of a psychological diagnosis attesting to the need for special education
(until 2009 it was possible to accept the children without a psychological diagnosis). Attendance is not a
problem either: children are happy to come because they feel accepted, and in any case they would get
bored at home, with nothing to do but watching TV.
The funding structure provides a strong incentive for school officials to place Roma children in
special education. Schools receive funding on the basis of the number of pupils, taking into account
various parameters such as the school type, the language of instruction, personnel demands and the form
of study. Per-pupil funding in special primary schools is about one and a half times that of standard
primary schools, and per-pupil funding for a special class in a standard primary school is about 1.75 times
higher than for a regular class. Given a dwindling basic school population overall because of national
demographic trends, there is inevitably a certain degree of competition among schools to keep up the
number of students. But the principals of standard schools have to reckon with another factor –the
tendency for non-Roma parents to take their children to another school to avoid mixing with Roma
students (European Roma Rights Center, 2007; Friedman et al. 2009). Opening special classes within a
standard school becomes therefore doubly attractive for the school: it helps retain the non-Roma children
and it gives access to the higher per-pupil funding. In addition, attending a special class makes it easier
for children to have good grades, which in turn may attract more parents (the impact of the 2009 change in
requirements will only become evident as the cohorts that started before 2009 leave school).
87
Funding mechanisms provide a perverse incentive for Special Pedagogical Advising Centers as well.
These centers are in charge of diagnosing children with mental and physical disabilities, and of providing
them with expert assistance in cooperation with family, school, physicians and social workers. Because
their funding depends on the number of clients served, there is an obvious incentive to keep up the number
of students diagnosed as needing the special attention provided by special schools and special classes in
standard schools.
Ethnicity per se may not account for the disproportionate representation of Roma in special
education. When comparing children who share the same characteristics generally associated with
schooling outcomes, such as parents‘ education or family income, it turns out that being Roma does not
make a significant difference (see Annex Table 5-2).60 In other words, if two children are identical in
every way but one is Roma and the other one is not, the estimation suggests that the non-Roma child will
be just as likely as the Roma child to be in a special school. This does not mean that discrimination plays
no role. There is consistent and convincing evidence of discriminatory practices and behaviors that
contribute to the over-representation of Roma children in special education, from inappropriate diagnostic
tests to arbitrary interpretations of testing results and failure to reintegrate Roma children inappropriately
placed in special education (e.g., Amnesty International, 2010; Rafael, 2011).61 But it suggests that the
impact of discrimination is much of what being Roma implies.62 For example, the estimations show that
pre-school participation is an important (negative) determinant of special school enrolment alongside have
a male households head, completion of secondary education by the head, whether the head went to special
school him/herself, and household size. All the factors put Roma children in a disadvantaged situation vis-
à -vis their non-Roma neighboring children. Teacher prejudices may make them more likely to recommend
a Roma child for placement in special classes, which will result in lower educational achievement,
60
An OLS model was used to estimate the determinants of special school attendance among Roma and non-Roma
children. Attendance of special classes in regular school was not included in this case. Independent variables
included are: ethnicity (Roma vs. non-Roma living nearby), gender, age cohort (7-18 or 19-30), preschool
attendance, gender of the household head, educational background of the household head (i.e. whether the latter has
completed secondary school), whether the household head attended a special school, age of the household head, the
number of household members, income (included as quintiles), whether the household suffers from hunger, whether
Romani is spoken at home, whether the closest primary school is within walking distance from the household‘s
dwelling (i.e. within 3 km), whether the dominant ethnicity of the settlement is Roma, the region (West, Central,
East), and whether the household lives in a rural or an urban location. The model was estimated for subjects in the
age group 7-30, in 5 iterations: 1) a base model (with only age and ethnicity as independent variables), 2) a ‗regions‘ -
model (with age, ethnicity, regions, urban/rural, and distance to primary school as independent variables), and three
‗full‘ models: one for Roma and non-Roma taken together (including the ethnicity variable as a predictor), and two
for Roma and non-Roma separately (excluding the ethnicity variable). The Annex at the end of this chapter gives a
more elaborate description of the model, including the full set of model estimates.
61
There has been considerable debate about the process followed to determine whether a child should be in special
education, and in particular about diagnosing. As we have no new evidence to inform such debate, this paper will
not get into it. The reader is referred to works by Friedman et al. (2009), Tomatová (2004), and Spotáková (2011).
62
Social science literature on ethnic discrimination typically focuses on investigating whether there are race-specific
effects present for a particular outcome (e.g., years of schooling, hiring) within a particular domain (e.g., education,
labor markets) at a particular point in time. Cumulative discrimination instead concerns discriminatory effects over
time and across domains. The focus is not on the impact of discrimination on a given outcome at a point in time, but
on the dynamic and systematic processes that may perpetuate or reinforce discriminatory effects. One of the reasons
cumulative discrimination is important is the potential for future discrimination to be causally affected by past
discrimination, including cross-generational effects. Discrimination at one decision point may increase the likelihood
of discrimination at future decision points. See: Blank, Rebecca, and National Bureau of Economic Research.
―Tracing the Economic Impact of Cumulative Discrimination‖. American Economic Revie w, Papers and
Proceedings. May 2006.
88
reinforce stereotypes about Roma‘s inferior mental capabilities and increase discriminatory treatment in
the labor market.
The strongest predictor of whether a child will end up in a special school is whether the parents
went to a special school themselves. If the household head went to a special school, there is a 58%
higher chance that his or her children will also attend a special school, regardless of ethnicity (significant
at 1% level). In a way, this is not a particularly surprising finding as parents‘ education is often found to
be a good forecaster of their children‘s educational achievement. At the same time, this particular case is
exceptional, since parents‘ disability is not a good predictor of their children‘s disability. Qualitative
research suggests some explanations. Parents who themselves went to a special school tend to see nothing
wrong with enrolling their children in special education. On the one hand, rejecting special education for
their offspring would be tantamount to belittling their own educational experience or even questioning
their parental standing (―if special education was good enough for me, why should it not be good enough
for my child?‖). On the other, there may be a feeling that special education is ―normal‖ or ―destiny‖ in
their family, and therefore acceptable. Regardless of the explanation, this finding casts an even darker
shadow on special schooling, as its negative impact appears to be transmitted from one generation to the
next in more ways than one. Hence the crucial importance of efforts to reach out to parents, Roma and
non-Roma alike, and ensure that they understand the long-term consequences of special education, as well
as the fact that it is inappropriate for the vast majority of children.
Attending preschool tends to ―protect‖ against special education for both Roma children and their
non-Roma neighbors. When all other individual characteristics are the same, those who benefitted from
preschool education are less likely to be found in special schools than those who did not (significant at 1%
level). This is consistent with a large body of international evidence suggesting that early childhood
education can be a powerful counterbalance against the burden represented by unfavorable background
characteristics. It is also consistent with the reasons given by Roma for attending special schools: since a
majority reports as the main reason that special school programs are easier, a proper preparation for
regular primary schools, in the form of preschool, is a logical factor to increase chances of attending a
regular school.
5.5 PAYING FOR ROMA EDUCATION
Slovakia education expenditures are among the lowest in Europe. Public expenditures in education
are a mere 3.6% of total GDP, lower than neighboring countries (Czech Republic‘s education expenses are
4.1% of GDP and everyone else‘s are higher) and considerably lower than Western European countries or
countries known for their high educational achievements such as Finland or Korea. Slovakia‘s
expenditures are also very low as a percentage of GDP per capita, and appear particularly low for
secondary school: they are 15% of GDP per capita, against an average of more than 20%.
89
TABLE 5-8: PUBLIC EXPENDITURES FOR EDUCATION, 2008
For primary school student For secondary school
Country As % of GDP as % of GDP per capita A student as % of GDP per
capita
Bulgaria 4.4 24.4 24.2
Czech Republic 4.1 13.6 22.8
Hungary 5.1 21.9 22.9
Romania B 4.3 20 16.6
Slovakia 3.6 15.5 15
France 5.6 17.3 26.8
Netherlands 5.5 17.2 24.9
Portugal 4.9 19.9 31.6
Spain 4.6 20.3 25.8
Source: World Bank. A Public expenditure per student is the public current spending on education divided by the
total number of students by level, as a percentage of GDP per capita. Public expenditure (current and capital)
includes government spending on educational institutions (both public and private), education administration as well
as subsidies for private entities (students/households and other privates entities). B Data for 2007. Source: OECD
In terms of overall education quality, Slovakia‘s performance is slightly above the European
average, but slightly below the new Eastern member states . When looking at the results of the
Programme for International Student Assessment (PISA) for Slovak students, they are broadly in line with
what could be expected from a country with that level of GDP. However, countries from the former
soviet bloc have traditionally better outcomes in the education sector, all other things being equal. When
compared to this smaller group of countries, Slovakia is somewhat below the average, even though scores
have improved considerably between 2006 and 2009 (2009 is the last year for which data are available).
Segregated schooling appears to be an ineffective way to spend the limited budget devoted to
education. As noted earlier, segregation has an impact on pupil attainment through the ―peer effect‖.
Although the data available for Slovakia do not allow for more in-depth analysis, international research
suggests that peer effects tend to be most significant for low-achievement students, and the positive results
obtained by these students thanks to a greater mix of abilities exceeds any negative impact on the
achievements of other students.63 There is also a considerable body of literature on the negative social and
psychological impact of segregation on children, who tend to develop feelings of inferiority and enter a
vicious circle of low expectations and low achievements.64 Low educational achievements, in turn, lead to
higher chances of unemployment and low income, thus setting the stage for poverty and dependence on
social assistance.
Special education in particular seems to represent an unwise use of public resources. This is clearly
illustrated by estimates of the payback period for special education, that is, the number of years it takes to
recover the original investment by summing up the future discounted cash flows. 65 This estimate was
calculated by taking into account the average cost of schooling plus the average cost of students from low-
income families, as well as the costs and revenues that result from the individual‘s employment status later
on. The payback period for individuals who attended only special primary school or special primary
63
In fact, some studies show that there is no negative effect on other (better performing) pupils. For example, the
study from Hungary in: Kézdi, Gábor and Éva Surányi. 2007. ‗A successful school integration program‘. Working
Paper no. 2. Roma Education Fund.
64
See for example, Niles and Peck, 2008.
65
For details on the calculations and the assumptions that underlie them, see Friedman et al.,2009, op. cit. (p. 43-48).
90
school plus three years of practical school is calculated to extend beyond age 60, meaning that the state
will never be able to recover the cost of educating these people.
Improving Roma education would also be a smart investment for the Government. Better educated
Roma are more likely to be employed than their less educated counterparts and tend to have higher
incomes. For the Government, this would mean more revenues from the taxes paid by the (now richer)
Roma and less money going to pay for social assistance. But would the extra expenses needed to improve
Roma educational outcomes be worth it? To answer this question, three different scenarios were
simulated (see Table 5-9 for a summary).
ï‚· The first scenario assumes that Roma pupils would succeed on completing upper secondary
school (12 years of education), thus spending on average three more years in school, but this
would require extra help for these students. So it is assumed that an extra investment equivalent
to 50 percent of the regular per-pupil cost would be necessary. On the other hand, survey results
indicate that with three more years of schooling, Roma employment rates and monthly wages
increase. Thus, the Government would spend €10,458 for the additional education provided to a
Roma but get back €17,014 (in the form of fiscal revenues), equivalent to a rate of return on
investment of 63 percent.
ï‚· The second scenario estimates what would happen if, in addition to the extra spending to finish
upper secondary school, the Government were to enable all Roma children to have two years of
high quality preschool as well as extra help during all of the twelve years of regular education. As
a result of the investment in better and longer pre-school preparation and the higher quality
education, labor market returns to upper secondary education by Roma would become equivalent
to the returns experienced by non-Roma living in close proximity. This would bring the additional
annual fiscal revenues to €26,259 (lifetime net present value), largely offsetting the €17,478
estimated increase in education expenditures per pupil. The rate of return on investment would be
50 percent.
ï‚· The third scenario, assumes even higher investments in Roma education, which would result in
Roma experiencing equivalent labor market conditions as the general population. With
employment rates now at 69% for Roma men and 58% for Roma women (against the current 20
and 9%, respectively), the rate of return on the education investment (105%).
91
TABLE 5-9: FISCAL RETURNS TO EDUCATIONAL INVESTMENTS
Current Scenario 1 Scenario 2 Scenario 3
Additional 3 yrs Additional 2 Additional 2
upper yrs preschool years general
secondary and and even more education and
extra expenditures – even more
expenditures – w/ 'non-Roma expenditures -
w/ 'Roma' nearby' labor general
labor returns returns population
labor returns
Employment probability - men 20% 33% 53% 69%
Monthly wages conditional on working - men 403 500 593 729
Employment probability - women 9% 20% 38% 58%
Monthly wages conditional on working - women 278 309 387 566
Average net wage per adult individual (15-64) 53 112 230 413
Relative to scenario ….? Current Scenario 1 Scenario 2
Extra annual fiscal revenue per adult (15-64) 573 1,144 2,139
Total extra fiscal revenue (NPV) per adult (15-64) 17,014 26,259 45,952
Total extra investment (NPV) per pupil 10,458 17,478 22,371
Difference (Euros) 6,556 8,781 23,581
Rate of return (assuming 2% real interest rate) 63% 50% 105%
Source: Authors‘ calculations using UNDP/World Bank/EC Regional Roma Survey (2011)
The impressive returns on Roma education investments suggested by these simulations may be
overly optimistic, but even under less favorable conditions they would remain largely positive.
Research findings offer strong indications of discriminatory practices in the labor market. For example,
among those Roma subject who have looked for work somewhere in the past five years, 78% report that
they have experienced discrimination because of their ethnicity. Among those who already had a job in the
past five years, 57% report that they have suffered from ethnic discrimination at the workplace. In
addition, the labor market may not be able to absorb as many individuals (Roma or non-Roma) regardless
of their qualifications. Therefore it can be expected that improved education would not translate into quite
the employment conditions envisaged in the three scenarios. But the returns on investment are so
substantial that increasing education expenditures to ensure better education for Roma would still be a
fiscally smart decision under more conservative employment outcomes.
Education investments would also pay off for individual Roma. Many Roma express the view that
studying may be a waste of time because in any case they would find no work, or occasional low-paying
work at best. But estimates show that among Slovak Roma with similar background characteristics, those
completing secondary education have employment rates that are approximately 12% (13% for men and
11% for women) higher than for those who only complete basic education (significant at 1% level). 66
66
An OLS model with ‗employment‘ as the dependent variable was used to arrive at this finding. Employment was
defined as ‗having worked in the previous week for at least one hour, or having been absent during this time from a
job to which the subject will return later‘. Independent variables included are: ethnicity (Roma vs. non-Roma living
nearby), gender, age, school level, the number of household members, whether Romani is spoken at home, whether
the closest employment office is within walking distance from the household‘s dwelling (i.e. within 3 km), wh ether
the dominant ethnicity of the settlement is Roma, the region (West, Central, East), and whether the household lives
92
Even if their non-Roma neighbors experience a much larger increase in their employment rates when they
complete secondary as opposed to basic education (32 percentage points for men and 30 for women), for
individual Roma extra schooling would still not be a waste of time. In addition, their monthly wages
would increase by 24% (from €403 to €500) for men and 11% (from €278 to €309) for women (these
increases correspond to the actual increases observed among the Roma with upper secondary relative to
those having completed only upper basic education, holding various background characteristics
constant).67
5.6 POLICY RECOMMENDATIONS
As noted previously, Roma education in Slovakia is a subject that has attracted considerable
attention and as a result policy recommendations are not in short supply. Summarizing the many
recommendations issued over the years is beyond the scope of the present study, and in any case would be
of little additional value. This section, therefore, focuses on a few policy measures considered crucial, but
this should not be interpreted as a lack of endorsement of other measures advocated elsewhere. In
particular, it does not enter into the debate surrounding the process of diagnostics for determining which
children should be placed into special education, and eventually which children should be streamed back
into regular education. The measures proposed here satisfy two main criteria: (a) they have been tried
elsewhere --sometimes also in Slovakia-- and have given encouraging results, (b) they have a high
potential for synergy, meaning that while each measure has value in its own right, their combined
implementation would increase the effectiveness of each one. In addition, most of them can be
implemented at a reasonable cost and speed.
A commitment to educational inclusion is the precondition for the success of any policy measure
aimed at improving Roma education outcomes. This means nothing less than a paradigm shift, and as
such it will require not only strong political leadership but also a concerted effort with all concerned
stakeholders, including education professionals (e.g., teachers, school staff and administrators, employees
of the Ministry of Education, school inspectors, and academics), professionals whose services are used in
connection with education (e.g., social workers, nurses, doctors, psychologists), local and regional
governments, Roma organizations, and, of course, parents and students. A paradigm shift implies a
change in social norms: from considering segregation normal, or at least acceptable, to considering
inclusion normal –and segregation unacceptable. Needless to say, changing social norms is not easy, but
social norms evolve constantly and influencing the direction of their evolution is possible. Social
marketing techniques, for example, have been used successfully to change social norms concerning a
number of behaviors, from smoking to hygienic habits. ―Edutainment‖, that is, the use of entertainment to
educate people, has also been used with some encouraging results, especially with delicate topics such as
in a rural or an urban location. The model was estimated for subjects in the age group 25-64, in 5 iterations: (1) the
full model for all subjects (Roma and non-Roma); (2) the same model for all Roma; (3) the same model for all non-
Roma; (4) the same model for Roma men; and (5) the same model for Roma women. The Annex at the end of this
chapter gives a more elaborate description of the model, including the full set of model estimates.
67
These Roma (and non-Roma) rates of return to upper secondary education are based on OLS estimations of
employment and wages (conditional on working) regressed on age, household size, whether usual language is
Romani or Slovak, distance to employment office and to nearest city, whether dominant ethnicity in settlement is
Roma, rural versus urban, and region dummies; separate estimations for men and women, ages 15-64. Actual rates of
return may be lower (e.g., due to unobserved characteristics that correlate positively with education completion and
employment conditions) or may be higher (e.g., Roma with upper-secondary education may be able to take
advantage of employment possibilities away from the settlements and move out), and are thus not captured by this
particular survey sample).
93
sexual behavior.68 The specific measures proposed below, therefore, can be accompanied by a pervasive
and sustained effort to change the way non-Roma perceive –and therefore treat-- Roma.69 No policy can
be effective if those supposed to implement it boycott it, whether actively or passively.
Compared to other industrialized countries, Slovakia‘s education policies are ―old fashioned‖ in
their non-inclusiveness. In developed countries there are significant variations in the definition of
―special education needs‖ (SEN), and therefore in the proportion of children who are considered in need
of special educational arrangements. A comparison of special education arrangements in 17 OECD
countries, for example, found that the percentage of students classified as having SEN went from 17.8% in
Finland to 0.9 in Greece, while the percentage of segregated students (in special schools or spending most
of the day in a separate classroom) went from 6% in Switzerland to less than 0.5% in Spain, Italy and
Greece (Powell, 2006). Another comparative study of OECD countries found that the percentage of
students classified as actual beneficiaries of special support and services range from one in twenty to one
in three (Richardson and Powell, 2011). But regardless of definitions, the general trend in OECD is
toward integration and special education is increasingly regarded less as a "place" and more as "a range of
services, available in every school‖.
International good practice for educational inclusion can provide inspiration. Below are some
examples from European countries, but useful lessons can also be found in countries further away that
have had to deal with disadvantaged ethnic minorities, such as Canada, the United States or Australia. In-
depth study of the experiences of these (and other) countries, and possibly study tours and exchange visits,
could help Slovakian authorities identify ways to bring about the needed paradigm shift.
ï‚· Finland. Special educational services are provided alongside regular teaching and children are
only segregated as a last resort, usually in part-time segregated classes within mainstream schools.
There are only eight segregated schools in the whole country, intended for students with visual,
hearing or significant physical disabilities. Special education teachers are assigned to classrooms
where there are students entitled to special education services, and remedial teaching is available
for any student who has fallen behind or needs special support. There is no tracking by school
type: all students engage in the same national curriculum for the period of compulsory schooling.
Every school has a ―student welfare group‖ chaired by the headmaster and composed of school
nurses, counselors and teachers, which works in close collaboration with families. The proportion
of children needing special attention is reported to decline as children progress through school.
ï‚· France. While there are segregated schools and classes, the presumption of the law is inclusion.
Students are assessed, sometimes at an early age, to determine their needs and develop a
personalized education plan. A Network of Specialized Aid to Struggling Students (RASED)
comprised of special education teachers, school pathologists and other service providers is
available to provide advice and specialized support for any child in primary schooling, typically in
mainstream classes.
68
Radio and television soap operas in Brazil, Ethiopia, India, Kenya, Mali, Mexico, Niger, Nigeria, Rwanda, St.
Lucia, and Tanzania have been documented by independent researchers in their massive effects on audience attitudes
and behavior with regard to HIV/AIDS avoidance and use of family planning. See: Ryerson, William. (undated).
―The Effectiveness of Entertainment Mass Media in Changing Behavior‖. Population Media Center. A vailable at:
http://www.populationmedia.org/what/effectiveness/
69
Of course, a successful anti-discrimination campaign would also have positive outcomes in other spheres of
Roma‘s life, starting with employment.
94
ï‚· Italy. The Italian education system is regarded as among the most inclusive in the world. The
right for disabled students to compulsory education in regular classes in public schools was
established in 1971, and in 1992 special classes and special schools were abolished. There are
presently only nine segregated schools in the country, seven for the deaf and two for the blind. If a
disabling condition is suspected, students undergo an assessment to draw a Personalized
Education Plan and determine the number of hours of specialized assistance to which they have
right under the law. The assistance is then provided in mainstream classrooms.
ï‚· United Kingdom. Under the Education Act of 1996, schools have a duty to educate children with
special educational needs in mainstream classes. If schools do not have the means to do it, they
may request additional help, which may require a formal and lengthy assessment. But the basic
presumption remains mainstreaming and special schooling is only allowed if it can be proven that
regular schooling would actually be detrimental to the special need student.
The specific policy measures proposed below are based on the following considerations, drawn from
the evidence and analyses examined so far:
1. The Roma predicament in education is well known, and in many ways not unique . Findings
from the most recent research conducted on the situation of Roma in Slovakia (in particular, the
2011 regional Roma survey and the 2010 Slovak Roma Survey) broadly confirm previous
analyses. This indicates that the Roma educational situation and the factors that contribute to its
creation are quite well understood, but also that there has been little change in the past decade or
so. Another general observation is that the obstacles faced by Roma in Slovakia are not very
different from those faced by ethnic minorities throughout the world, and therefore much can be
learned from the experiences of other countries.
2. Roma‘s educational achievements continue to be far below those of non-Roma, no matter
the way in which they are measured. Whether one looks at the highest grade completed, at the
probability of finishing high school, at scores obtained in standardized tests or at the likelihood of
being sidetracked in special schools and classes for the disabled, Roma are invariably the group
faring the worst. This suggests that (a) Slovakian policies for the integration of marginalized
groups have not been effective, or sufficiently aggressive; (b) future efforts to develop and
implement policies and programs for the integration of marginalized groups should have Roma as
the main target group; and (c) the effect of policies and programs for marginalized groups should
be monitored and feedback can spur action.
3. When it comes to Roma education, the present system is a lose-lose proposition. Failing to
invest in Roma education dooms large numbers of Roma to unemployment or extremely low-
paying jobs and deprives the Government of substantial fiscal revenues. In addition to rethinking
the basic premises of Slovakian education to make it more inclusive, this suggests that
incremental improvements could be possible with minimal additional cost by redirecting expenses
toward services that have been shown to be effective in improving educational outcomes for
marginalized groups.
4. Individual background characteristics explain a large part Roma‘s inferior educational
outcomes. Having poor and uneducated parents --or parents who went to a special school--
represents an enormous burden that is not easy to overcome regardless of personal aspirations.
This suggests that policies and programs aiming at improving Roma educational outcomes should
try to compensate for the lack of material and human resources in children‘s families. Of course,
95
policies and programs aiming at actually improving the socioeconomic status of Roma parents can
complement efforts to make it matter less for the educational success of their children.
5. Cultural differences, including discrimination, also play an important role. Some findings,
especially from qualitative research, can only be explained by a difference in beliefs and attitudes,
and eventually behavior. For example, even when everything else is the same, living in Eastern
Slovakia increases a Roma‘s chances of ending up in special education and being Roma increases
the chances of doing poorly in standardized tests. This suggests that in addition to doing
something to compensate for deprivation in the children‘s background, measures to fight
prejudices (on all sides) are necessary because without them any well-intentioned policy may be
doomed to failure.
To address the Roma education gap, Slovakia can build on positive policy experiences from other
countries that have sought to address large education gaps of minority groups, Roma and non-
Roma. Recommendations to improve educational outcomes can be grouped in three areas, with a priority
focus on investing in infants and young children: (1) increase access to quality pre-school, moving toward
compulsory preschool from age 3 onwards, and improve home parenting; (2) promote integrated regular
primary schooling for all; and, (3) address early (secondary) school leaving. The Slovak program to put in
place community centers can – if designed and resourced sufficiently - perform an important role in
reaching these 3 goals.
5.6.1 POLICY MEASURE 1: INCREASE ACCESS TO QUALITY PRE-SCHOOL, MOVING
TOWARD AGE 3, AND IMPROVE HOME PARENTING
Increase access to quality pre-school, moving toward preschool attendance from age 3 onwards, and
improve home parenting. Given the high returns to early childhood development and the very large gap
that exists between Roma and non-Roma in Slovakia, closing this gap should be a high policy priority
area. Achieving this will require two complementary areas of intervention: increasing preschool enrolment
rates and supporting Roma parents in providing early stimulation at home. The measures suggested below
are based on the following considerations about the present situation:
 Out of pocket expenses for preschool are relatively low (approximately € 7 per month),
particularly in comparison to the monthly child benefit (PrÃdavok na dieÅ¥a, € 21.99 per month).
Experience from Hungary indicates that Roma families living in similarly poor environments are
much more likely to enroll their children, but do not have out-of-pocket expenses for preschool
fees or lunch and preschool from the age of 5 was compulsory at the time of the survey.
ï‚· About two-thirds of parents with children in school is satisfied, but one third are not nor feel their
child is welcomed at the preschool.
ï‚· Many Roma parents express a preference for raising their younger children at home, but more
than half of the parents report they would reconsider enrolment if there were no fees, or if food
coupons were provided, and nearly half of the parents report they would reconsider enrolment if
there were a Roma teaching assistant.
ï‚· A lack of nearby preschool facilities, at least for some Roma children (16%) constitutes an
obstacle.
96
 The average annual gross salary for pre-primary teachers in Slovakia is € 7,622.70 Given average
class sizes71, this means an investment of € 391 per year per pupil. Closing the preschool gap for
all 45,000 children therefore requires an annual investment in teaching salaries of € 17.5 million.
To close the gap in preschool enrolment, the evidence suggests a combination of awareness raising,
implementing demand side incentives, improving physical accessibility, and strengthening the
(social) connection between kindergartens and parents.
a) Raise awareness through kindergarten-community liaisons: Many parents may simply not be
aware that quality pre-school education is one of the highest return investment parents can make
into the development of their children. Making sure that poor parents have this information
enables them to make a better informed decision. This information could be disseminated by
kindergarten teaching assistants for example. The teaching assistants could be recent secondary
graduates from the communities where pre-school enrolment is low, hired at minimum wage
levels of € 327 per month or € 3,924 per year. 72 Suppose that there is one (Roma) teaching
assistant per local kindergarten. With the average kindergarten size at 48.5 pupils in 2010,73 this
means an annual cost of € 81 per pupil. If these would reach out to all approximately 50,000
children aged 3-5 at-risk-of-poverty (of which about 23,500 are Roma), the annual expense would
be € 4.04 million. This implies an additional investment equivalent to 23% of the necessary
regular additional teacher salary expenditures of € 17.5 million required to cover the inflow of the
new pupils.74 More generally, kindergarten teaching assistants can be leveraged to become bridges
between the local communities and the kindergartens themselves. For example, they can help
parents with enrolment procedures, or support the kindergartens engage parents into the
preschool, fostering participation and local ownership. The Roma Education Fund (REF) has been
implementing the Home-Community Liaison Program, as part of its ―A Good Start‖ project,
which is modeled after a similar such program in Ireland. Community liaisons could also be
leveraged to help parents strengthen parenting practices.
b) Remove preschool costs barriers for the poorest parents and provide encouragement
subsidies to poor parents to enroll their children into preschool at an early age (as early as 3
years old) conditional on meeting good attendance. Families below the subsistence minimum
receiving the Benefit in Material Need are eligible for support to offset the tuition and meal costs
of kindergarten. However, qualitative fieldworker by the Slovak Governance Institute75 indicates
that may poor families with infants lose out on this subsidy because the receipt of the parental
allowance for families with infants can push some of these families just above the subsistence
minimum, making children of pre-school age ineligible for the subsidies in the current system. In
addition to covering fees, Hungary, using a more complex eligibility to define ―multiple
70
Eurydice (2010). Teachers‘ and School Heads‘ Salaries and Allowances in Europe, 2009/10.
71
In 2010 there were about 45,000 children aged 3-5 years old not enrolled in kindergarten. Among those who were
enrolled, the average class size was 19.5 children per class.71 The maximum class sizes are 20 per class for children
aged 3 to 4 years and 21 per class for children aged 4 to 5 years. Eurydice. (2010). Structures of Education and
Training Systems in Europe. Slovakia 2009/10 Edition.
72
http://epp.eurostat.ec.europa.eu/statistics_explained/index.php/Minimum_wage_statistics
73
139,239 pupils over 2,869 kindergartens (National Statistics Office, 2010).
74
The total Slovak population has approximately 162k children aged 3-5 (National Statistics Office, 2010). Of these,
roughly 30 percent, or 50,000 children, live in households with incomes below 60 percent of the median income –the
‗at-risk-of-poverty‘ threshold. Among Roma, there are about 27,000 children aged 3 -5 (based on total population
estimate of 320k). 87% live at risk-of-poverty (UNDP calculations based on the UNDP/WB/EC Regional Policy
2011 Roma Survey), equivalent to approx. 23.5k Roma children (from the 50k in total – almost half).
75
Personal communities
97
disadvantaged parents,‖ provides households € 70 per year (divided in two tranches) conditional
on the good preschool attendance of their children 3-4 years. A similar subsidy provided to the
approximately 50,000 children aged 3-5 at-risk-of-poverty would cost € 3.5 million per year,
exclusive of operating expenses. This implies an additional investment equivalent to 20% of the
necessary regular additional teacher salary expenditures.
c) Move toward making pre-school compulsory for at least two years. While it would be best if
parents were freely choosing to enroll their children for their own good, making preschool
compulsory would set an important normative standard and, if accompanied by the other measures
suggested here, would likely constitute a crucial policy element.
d) Invest in preschool infrastructure to accommodate the inflow of children using EU
Structural Funds. Slovakia can importantly take advantage of the European Regional
Development Fund (ERDF) to finance construction. The update of the Slovak Roma Atlas
(underway in Fall 2012 by UNDP) can provide the list of communities lacking preschool facilities
nearby.
Support home parenting. For example, parental participation in extracurricular or in-class activities has
been encouraged in the Step-by-Step projects and the projects financed by the Bulgarian Centre for
Educational Integration of Children and Young People from the Minorities. Discussion groups on
parenting techniques have also been a component of the preschool project of Minority Rights Group
International in Slovakia, and an explicit focus of the Hungary's Meséd project, which is a part of REF's
―A Good Start‖ initiative, and described in more detail in Box 5-6 below. Finally, the Council of Europe's
―Teaching Kit for Roma Children‖ is a set of teaching mater ials developed to help aide young Roma
children prepare for school in a home environment (see ISSA, 2009).
BOX 5-6: THE MESÉD PROJECT – ―YOUR TALE‖
The Meséd Project is being implemented in six locations in Hungary as a part of Roma Education Fund's
EC Roma Pilots ―A Good Start Initiative‖, in collaboration with the Unity in Diversity Foundation and
with the help of students from the College of NyÃregyháza. A major project objective is to improve
parenting practices.
As a part of the project mothers meet in small groups (8 to 15) on a weekly basis for two-hour sessions
and take turns to practice reading out high quality children's story books, which they also get to keep. A
trained facilitator, who is also usually a Roma woman, guides the reading and initiates discussions on
certain elements and messages of the story, thus providing the mothers with a teaching technique they can
use with their children at home. Furthermore, parenting problems and techniques are discussed during the
sessions. Mothers report using the reading techniques they learned during the Meséd sessions and their
children taking pride in the books they have received.
There are also many international non-Roma examples of successful approaches to increase access
to early childhood education for disadvantaged children. The Program for the Improvement of
Education, Health and the Environment (PROMSEA) in Colombia is a long-term large-scale project that
trained disadvantaged Columbian mothers of young preschool-aged children on appropriate educational
approaches. ―Promoters‖ taught mothers about play-based and cognitive methods, culturally-appropriate
games and toys, and how older siblings can be involved (Arango et al. in Siraj-Blatchford and Woodhead
98
2009). Another parenting program that has proven successful is the Mother-Child Education Program in
Turkey. On a nation-wide scale mothers are instructed on their children's development needs and on ways
to create a stimulating home environment. Children who participated in the program demonstrated better
cognitive skills and greater school readiness (Bekman in Siraj-Blatchford and Woodhead 2009).
Piloting of new measures should be accompanied by impact evaluations. The analysis points to several
promising, relatively low-cost areas of intervention. Each of these can build on international experiences.
Nevertheless, the most effective implementation will be locally specific; for example, what is the
appropriate subsidy incentive to encourage poor parents to enroll their children? What are effective
methods to support parenting? Piloting these types of programs first, rigorously evaluating their impact,
and then scaling up proven programs will ensure the interventions are cost effective and receive public
buy-in.
5.6.2 POLICY MEASURE 2: PROMOTE INTEGRATED REGULAR PRIMARY
SCHOOLING FOR ALL - CREATE THE POSITION OF ROMA SCHOOL (OR
COMMUNITY) MEDIATOR
In 2000 the Council of Europe recommended using ―mediators from within the Roma/Gypsy
community‖ to ―ease the contact between Roma/Gypsy, the majority population and schools and to
avoid conflicts at school‖.76 The recommendation of using Roma mediators was based on the experience
of a number of countries, including Spain (the first country to use Roma mediators, in the 1980s) but also
Slovakia (with a pilot in 1994). At the same time, many of the countries of Central and Eastern Europe
elaborated national strategies for Roma integration that included the employment of Roma school
mediators among their education measures. During the following years, the number of Roma mediators
increased significantly in pre-accession countries thanks to the support available from EC funding through
national PHARE projects. The role and profile of Roma school mediators, however, differs from one
country to the other and tends to be confused with that of a teaching assistant. Here it is argued that a
Roma school mediator should have a role closer to what the word ―mediator‖ suggests (a neutral third
party who eases communication between two parties thus preventing conflict and facilitating agreement)
and therefore more similar to that of a social worker or school counselor.
Roma school mediators can act as intercultural mediators. Intercultural mediators are employed in
several European countries with significant non-European immigrant populations to facilitate these
populations‘ communication with public authorities and the local population as well as to improve their
access to public services. The main requirements for an intercultural mediator are to originate from the
immigrant group targeted and to be well integrated in the host country. Communicative competence (e.g.,
empathy, active listening) and pleasant manners are often also a consideration, but generally no specific
academic background is required beyond basic education --the idea is that a cultural mediator is there
primarily to facilitate mutual understanding, and this does not depend on academic credentials. In Italy,
for example, intercultural mediators are required to be familiar with the language and culture of the
immigrant group, to have personally undergone the migration experience, and to be well inserted in the
Italian reality with good knowledge of the language, cultural codes and administrative structures. In
addition, they have to follow a specific training, which was initially provided by NGOs but later
formalized through partnerships with vocational training centers. To facilitate certification, competencies
acquired on the job are recognized.
76
Recommendation No. R(2000)4 adopted by the Committee of Ministers in February 2000.
99
The work of Roma school mediators should have two primary purposes: support Roma children‘s
integration in school and facilitate relations with the families . Activities to fulfill the first objective
would include: accompanying children in their first experiences in the school; explaining to children the
school norms (including discipline and hygiene, both of which are often reported to cause problems to
Roma children), possibly starting before school begins; providing translation and interpretation;
contributing to assess children‘s competencies and possible need for extra assistance (this last
responsibility would be crucial for the successful mainstreaming of children who would otherwise be
bound for special education). In terms of facilitating relations with the family, Roma school mediators
would: explain to families the school organization and way of operating; provide information about school
life, including translation of specific messages; facilitate meetings between parents/caregivers and school
staff, and if needed participate; provide information about the process for entering school (and about the
consequences of enrolling in special classes or schools); and, prevent misunderstanding and conflicts
between Roma families and school staff as well as between Roma and non-Roma parents. Ideally, a third
objective of Roma school mediators should be promoting multicultural education by organizing cross-
cultural encounters and education activities, but this may prove too demanding in many cases.
Because of the nature of their responsibilities, Roma school mediators should preferably come from
the local Roma community. This would ensure that they speak the local version of Romany and have a
deep understanding of the local culture. In some countries, employment decisions for Roma mediators are
also discussed with the community or with Roma organizations to ensure that the mediator is trusted and
accepted. While presently the position of Teaching Assistant for Socially Disadvantaged Children
encompasses some of the responsibilities proposed for school mediators, separating the role of school
mediator from that of teaching assistant would probably facilitate the deployment of both types of
professionals because the requirements for the two positions are not quite the same. Roma mediators‘
should be provided with specific training (see Box 5-7 for an example), support materials and guidelines,
not only before they are employed but also to update them regularly about new administrative dispositions
and policies. But formal academic requirements should be minimal because their most important
qualifications would be their cultural knowledge and personality.77 In principle, each Roma community
should have at least one Roma school mediator for each school in proximity –a disposition that would also
represent an employment opportunity for local Roma.
BOX 5-7: ROMA MEDIATOR TRAINING PROGRAMME
The Roma Mediator Training Programme (Romed), was launched at the beginning of 2011 by the Council
of Europe. The training sessions equip people with a Roma background, either from Roma communities or
with a good knowledge of Roma issues, to act as mediators between the Roma and public institutions. It
has so far been implemented in 16 countries, with 427 mediators trained. The mediators are now working
in the field, with a second series of training sessions to review their work organized in the second half of
the year. The mediators‗ job is to act as ―ambassadors of trust‖ between Roma communities and local
public institutions – for example, getting Roma children into local schools, making sure that
families receive proper health care, helping Roma secure decent housing and find jobs that will bring them
out of long-term unemployment and back into salaried work.
77
The ―Healthy Communities‖ Project, presently implemented in 32 communities, employs Roma health mediators
whose profile and role are comparable to what is being advocated here for education. The experience so far is very
encouraging.
100
5.6.3 POLICY MEASURE 3: PROVIDE SCHOOL-BASED EDUCATION SERVICES
BEYOND REGULAR CLASSROOM TEACHING – TEACHING ASSISTANTS AND
TUTORS
Both international and Slovak experience confirms the effectiveness of teaching assistants in
improving educational outcomes for children from disadvantaged backgrounds. Similarly, after-
school tutors who help children with their homework have been successfully employed in low-income
settings to compensate for the assistance that poorly educated parents cannot provide. On the other hand,
teaching assistants and tutors can only be effective if they are appropriately qualified and properly used.
For example, an evaluation conducted in the UK found out that the number of teaching assistants did not
increase academic performance, but this could be explained by the fact that teaching assistants were
routinely used to substitute for absent teachers rather than to complement teachers‘ efforts (Paton, 2011).
In the case of Slovakia, it has been reported that in many classes teaching assistants are used for tasks not
directly related to children‘s learning, such as serving snacks, cleaning up after teachers or watching
students in the playground (Tankersly et al., 2002). Also, Roma teaching assistants have complained
about being treated by teachers as inferior, which of course constitutes the wrong role model and can
actually contribute to keep marginalized children in a subservient position.
The position of teaching assistant was introduced in Slovakia in 2000 with the title of Roma
Teaching Assistant, and given responsibilities for both cultural mediation and academic support.
Over the years the name of the position evolved (losing its Roma connotation) as well as the definition of
its requirements and responsibilities. Presently, the teaching assistant position covers healthcare assistants
and assistants for socially disadvantaged children. Their work encompasses three main areas: (a) support
to the educational process, including cooperation with teachers in the classroom, help to children to
overcome health, social and language barriers, and tutoring; (b) organization of extra-curricular activities
(e.g., sports, arts, cultural visits); and (c) facilitation of communication with parents and the community.
As can be expected, teaching assistants are more numerous in poorer regions (see Table 5-10 below).
TABLE 5-10: TEACHING ASSISTANTS BY REGION
Elementary schools Special elementary schools
Region Assistant for Health Care Assistant Assistant for Health Care Assistant
Disadv. Disadv. Children
Children
Bratislava 6.0 7.6 2.0 33.8
Trnava 3.2 18.6 5.0 31.3
TrenÄ?Ãn 2.3 20.9 0 19.2
Nitra 8.8 33.0 0 38.9
Å¢ilina 8,4 63.0 2.6 40.6
Banská Bystrica 90.6 60.0 19.1 41.2
Prešov 130.4 38.4 9.7 50.3
Košice 165.1 42.7 19.9 28.0
SLOVAKIA 414.8 284.0 58.3 283.2
TOTAL
Source: Ministry of Education (data collected in March 2011).
101
The number of teaching assistants has been growing but can grow faster, and for this it is necessary
to review the criteria for their funding and recruitment. The number of teaching assistants has been
slowly growing over the years, from 923 in 2005 to an estimated 1,040 in 2011. According to the 2008
School Act, a teaching assistant is required when there are at least 100 students from a socially
disadvantaged background (i.e., whose families receive social assistance) and it is to be financed from
50% of the additional funding provided by the Ministry of Education for each disadvantaged student (€90
per student in 2011-12). This system has three main problems. One is that, under the current funding
conditions, it has been estimated that a school would need at least 111 disadvantaged students to have
enough funds for the salary of a teaching assistant (Oláh, 2011). Another is that one teaching assistant for
100 students is not enough. Another still is that all socially disadvantaged students should have the right
to additional support regardless of whether they are in large groups or not. It has also been remarked that
the academic requirements for teaching assistants have become unnecessarily demanding as a high school
diploma is no longer sufficient and a diploma from a pedagogical school is now required. This makes it a
lot more difficult for a Roma to qualify, even though a Roma teaching assistant could be considerably
more effective with the majority of students from a disadvantaged background who are Roma (see Box
5-8 for an example). It could make sense, therefore, to accept teaching assistants with high school
diplomas for pupils in the lower grades, whose academic needs are unlikely to be too demanding for a
high school graduate.
BOX 5-8: ROMA TEACHING ASSISTANTS
Roma teaching assistants often perform very well at work: a conversation with the principal of a big and
largely Roma school in Eastern Slovakia illustrates this: he emphasizes that there is no way he will let go
his two Roma teaching assistants, no matter what the regulations say about the academic requirements.
They are excellent and hard working – they even come to work when they have a fever, they‘re so
devoted. The non-Roma teaching assistants are also good, but not as effective with the students as the
Roma. So he will do whatever it takes to keep them on the school staff. He would even consider hiring
them as cleaning personnel.
Source: personal communication, April 2012.
A way to increase the number of teaching assistants at a limited cost could be through the use of
students. Presently graduate students in social work or social pedagogy do a nine-month practical
training paid by the Ministry of Labor, Social Affairs and Family. Their work, however, is generally of an
administrative nature and assignments as a teaching assistant (or school mediator) are not envisaged.
Allowing teacher trainees to spend their practicum in schools with high poverty/Roma indicators and
graduate students, including those in pedagogy and from the social academy, to do their internships as
teaching assistants, as it is for example done in the USA, would allow students to gain useful practical
experience and diminish the burden on teachers and teaching assistants. Students in secondary school
could also be volunteer tutors. They would have to be approved by their teachers and, in exchange for
their work with struggling students, they could earn extra credits or other desirable things, such as more
frequent access the school computers and internet.
5.6.4 POLICY MEASURE 4: PROVIDE EDUCATIONAL SUPPORT AND
OPPORTUNITIES OUTSIDE REGULAR SCHOOL HOURS -AFTER SCHOOL AND
SUMMER PROGRAMS
102
Slovakian children in general spend very little time in school, and this is to the detriment of
disadvantaged children. On a typical school day, most children spend less than half their non-sleeping
hours in school and over the course of the year they spend fewer than half of all days in school. This short
time in school puts children from marginalized communities at a significant learning disadvantage relative
to their more advantaged peers whose parents are often able to provide them a rich set of opportunities for
learning outside of school, whether that be after school, on weekends or during the summer. These extra-
curricular activities matter not only because they are enriching in their own right, but also because they
provide experiential background useful for learning as children progress through school.
Extended school hours, after school and summer programs represent a way to compensate for lack
of out-of-school learning opportunities. The quality and nature of these programs matter and this,
combined with their very great heterogeneity, means that the evidence on the effectiveness of after school
programs and summer schools is somewhat mixed (Cooper et al., 2000). Not surprisingly, research shows
that marginally expanding the time students spend in school without improving how that time is used does
not improve learning. At the same time, some high intensity summer programs have generated academic
gains in high poverty areas, and some low-cost reading programs have reduced summer reading loss
(Jacob and Lefgren 2004; Allington et al, 2010). Many dynamic Slovak principals have managed to
complement public finances through their own fund-raising efforts to offer a variety of enriching extra-
curricular activities, from sports to arts, and it would appear that they have quite encouraging results. An
extended school day is also presently being tested, and it would be important to document its outcomes.
Several countries have opted for equalizing extra-curricular opportunities by using their schools
increasingly also as community centers. Many Dutch schools, for example, have been converted into
community schools with a variety of enrichment activities after school hours. Similarly, the highest
performing school districts in the USA have public schools that are open outside school hours to offer a
wide range of sports, arts and other activities, often with the help of volunteers. It should be noted that
relying on volunteers for school-based extra-curricular activities is not just a way to save money. In most
cases, the volunteers are students‘ parents or other family members, so these activities become an
important way to reach out to families and encourage them to participate in school life. Especially when
parents are poorly educated, as it is often the case with Roma parents, involving them in non-academic
activities may represent the first step to encourage them to take a more active role in their children‘s
education.78 An interesting example comes from Romania, where in 2009 an NGO piloted a summer
program to prepare Roma children and their families for kindergarten, and also to better equip teachers to
work with Roma children. Results proved encouraging: 85% of the children who attended the first
summer program also attended kindergarten regularly during the following school year, as did 94% of
children who attended the second summer program.79
5.6.5 POLICY MEASURE 5: PROVIDE ADDED INCENTIVES THAT MAKE A
DIFFERENCE - CASH SUBSIDIES CONDITIONAL ON CERTAIN EDUCATIONAL
ACHIEVEMENTS
Conditional cash transfers (CCT) programs give money to poor families in exchange for specified
verifiable actions, such as sending children to school or taking them to regular health check-ups. In
this way they aim at providing immediate relief from poverty while at the same time contributing to its
78
See for example: Cataluña SecretarÃa de Acción Ciudadana. 2007. Good Practice Guide for the Social Inclusion of
Roma in Europe. Barcelona: Secretaria d'Acció Ciutadana.
79
Presentation by Save the Children at a workshop on Early Childhood Education for Roma Inclusion held in
Bucharest on 28 March 2011.
103
longer-term reduction by supporting human capital development. Part of the appeal of CCT for education
thus lies in their simultaneous action to address current poverty while improving household educational
status and future earning potential. Additionally, CCT schemes may be more politically feasible than
unconditional transfers of comparable size because they introduce an element of responsibility on the part
of beneficiaries thus making them ―work‖ for what they get and reassuring the electorate that tax payers‘
money is not just given away.
CCT programs for education can be divided into two broad types - added benefit and added
condition. Added-benefit CCT consist of conditioning transfers on school enrolment and/or attendance,
and in some cases also on achievement (e.g., graduation). Added-condition CCT rely on the threat of
removing benefits within the existing social safety net if children do not attend school. Most education-
related CCT implemented to date are of the added-benefit type and offer per-student cash grants whose
amount is usually calculated to cover not only the direct costs of school attendance (e.g., school fees,
supplies, transportation), but also opportunity costs (e.g., income lost by not sending children to work).
Expert opinions on individual programs vary, but their extensive and continued use suggests that added
education-related CCT are effective means of affecting educational outcomes. CCT schemes for
education have been widely implemented and analyzed in Latin America, where they have obtained
generally encouraging results (see Box 5-9).
BOX 5-9: THE EXPERIENCE OF BOLSA FAMILIA IN BRAZIL
Bolsa FamÃlia (Portuguese for ―family scholarship‖) started in the 1990s and by 2007 mor e than 11
million families (about 46 million people, a quarter of Brazil‘s population) received payments. Glewwe
and Kassouf (2008) examined the impact of Bolsa FamÃlia on children‘s progress in school and estimate
that, after accounting for cumulative effects, Bolsa FamÃlia has increased enrolment rates by about 5.5
percentage points in grades 1–4 and by about 6.5 percentage points in grades 5–8. The program raised
grade promotion rates by about 0.9 percentage points for children in grades 1–4 and by 0.3 percentage
points for those in grades 5–8. The results show that Bolsa FamÃlia is more effective at increasing the
enrolment of blacks, mulattos and indigenous children than it is for whites, and thus it appears to be
equalizing enrolment by race. While these impacts cast a favorable light on Bolsa FamÃlia, simple
calculations based on the enrolment impacts suggest that the likely benefits in terms of increased wages
may not exceed the program‘s costs. Its long-run effect seems to be increasing participants‘ enrolment
rates by about 18 percentage points, which implies that 82 per cent of beneficiaries would have enrolled in
school even without Bolsa FamÃlia.
Source: Glewwe, P. and A. L. Kassouf (2008). The Impact of the Bolsa Escola/Familia Conditional Cash
Transfer Program on Enrollment, Drop Out Rates and Grade Promotion in Brazil. Department of Applied
Economics, University of Minnesota.
CCT programs for education implemented to date in Hungary, Romania and Slovakia have not met
with the well-documented successes of some of their counterparts from outside Central and Eastern
Europe (Friedman et al., 2009). In particular, neither of the two education CCT implemented in Slovakia
has contributed to increase attendance among Roma. 80 Yet, in theory, the low average Roma incomes
80
As mentioned before, the motivation allowance introduced in 2004 was abandoned in 2009 because it actually
introduced a perverse incentive to enroll children in special classes and schools as a way to ensure the good grades
required to obtain the allowance. Good grades, however, are still required to continue receiving the allowance past
compulsory schooling.
104
should make financial incentives an effective way to encourage Romani families to invest in education. A
number of explanations can be offered for this apparent aberration. One is that lack of money is not the
only reason for low educational achievements among Roma, as indicated by findings from the recent
surveys. Another is that the supply of educational opportunities, both qualitatively (schools which are
segregated, of poor quality, and/or are special schools) and quantitatively (not close to where Roma live),
is also a constraint and therefore stimulating demand through a CCT scheme is not enough. Another still
is that the cash transfer amount is simply not enough to make a difference. For example, as mentioned
earlier, if Roma drop out of school when they turn 16 to take advantage of the €60 labor activation grants,
the transfer should be close to €60.
If properly tailored to the dynamics of Roma education in Slovakia, CCT can be a powerful
instrument to help closing the gap between Roma and non-Roma. Because this gap is small during
the years of compulsory education, it would be unwise to use scarce public resources to encourage
attendance during this time. Instead, CCT should focus on upper secondary school, where the gap
between Roma and non-Roma becomes huge and where the risk of segregated schooling is much
reduced. 81 The amount should be set at a level that realistically covers the total cost of attendance,
including all direct costs (books, supplies, transport, etc.) as well as opportunity costs (e.g., the labor
activation grant). In addition to making grants conditional on regular attendance (e.g., 80% of the time), a
―prize‖ should be given for passing each grade and a bigger ―prize‖ for completing upper -secondary
school. This would encourage students not only to remain in school, but also to strive toward a diploma.
It is interesting to note that in Colombia, conditioning cash transfers on school success rather than school
attendance significantly reduced early pregnancies (Cortés et al., 2011) while in Malawi cash transfers
conditional on attendance reduced teenage pregnancy and early marriage (Baird et al., 2010). These are
surely objectives worth pursuing in addition to improved educational achievements.
5.6.6 POLICY MEASURE 6: MOVE RAPIDLY TO CLOSE MOST SPECIAL SCHOOLS
AND ABOLISH SPECIAL CLASSES IN STANDARD SCHOOLS
Keeping children who do not suffer from a severe mental or physical disability in special classes and
schools is a waste of public money and a violation of the children‘s rights. Leaving aside the moral
argument, straightforward economic efficiency calls for this measure. As seen earlier, special education is
expensive to deliver, between 50 and 75% more expensive than regular basic schooling. It is even more
expensive when one considers its long-term consequences: the fiscal revenues lost because of the lower
wages (or complete absence thereof) of those who attended special education, the social benefits that the
Government will have to pay for the same reason, the likelihood that the offspring of special education
students will in turn end up in special school. On the other hand, there is no demonstrable harm done by
integrated education to children without special education needs, so the higher cost of special education
does not appear to be justified by the avoided losses for the bulk ―normal‖ students.
In addition, there is now convincing evidence that Roma children placed in special education can do
well in regular classes if provided with the appropriate support. A recent qualitative study funded by
the Roma Education Fund (Equality, 2011) followed 38 Slovak Roma and 23 Czech Roma students ages
9-15 that had moved to the United Kingdom on average 3.4 years earlier. Of the 61 students, 17 had
attended a special school before moving abroad, but upon arrival in the UK they were placed in regular
classes. In the new schools, their average attainment in numeracy, literacy and science was only just below
average. Thus, children who had been considered unable to follow a standard curriculum in their own
81
This is also the recommendation of the Roma Education Fund study on CCTs. See Friedman et al., 2009.
105
country, were able to perform on a par with their British counterpart despite having to learn a totally new
language in a totally foreign setting.
The money presently spent on special schools can be used to support mainstreaming, but additional
funding would likely be needed. Integrating SEN82 children in regular classes costs more than keeping
them in special schools or in special classes. Special schools are cheaper because, although classes are
smaller than in regular schools, there are still economies of scale that integration would not allow (this is
reflected in funding norms that provide 150% of regular per-pupil cost for special school students and
175% of regular per-pupil cost for SEN students in standard schools). Special classes in regular schools
are not financially cheaper than mainstreaming under the present financing arrangements because the per-
pupil cost of SEN children attending standard schools is the same regardless of whether they are
integrated, as it is often the case with non-Roma children, or kept in separate classes. But the
administrative burden changes, because mainstreamed children are supposed to have an individual
learning plan while those in special classes are not. With the majority of special schools closing or
becoming standard schools, the number of SEN children in basic schools would inevitably increase, and
this would have budget implications. At a minimum there would be a 25% increase in per-pupil cost
(from 150 to 175% of the regular per-pupil cost) for each of the now integrated students. Assuming that at
least 70% of the 17,326 children attending a public special primary school in 2011-12 would be
mainstreamed, this would cost an additional €7.3 million.83 It should be kept in mind, however, that the
number of children in need of special education services should decrease as children progress through
school (see the example of Finland) and therefore the cost of integrating socially disadvantaged children
would slowly go down each year, until it stabilizes at a lower level.
As special schools are gradually closed or are transformed into regular schools, staff can be
redeployed and trained to provide special services in an integrated setting. The five policy measures
advocated above would help standard schools become better equipped to address the needs of children
from disadvantaged backgrounds, and in particularly Roma children. But they can also become better
equipped to address the needs of SEN pupils. This is why those whose employment depends on special
schools (e.g., special education teachers, psychologists and social workers) would not have to worry about
losing their jobs -- To the contrary, it is likely that mainstreaming would actually increase the demand for
special education professionals, as SEN students would receive a more individualized attention. The
nature of the work of these specialized professionals, however, would change and (re)training would be
needed. In any case, training would certainly be needed for all school staff to eliminate negative
stereotypes and prejudices, learn how to work in an inclusive environment/mode, and sensitize them to the
need to combat harassment and bullying among students. The curriculum at the university level can also
be modified in this sense, so that future generations of education professionals would come already
equipped to work effectively in inclusive environments.
The newly acquired membership in the European Agency for the Development of Special Education
would facilitate the transition. The European Agency for Development in Special Needs Education is
an independent and self-governing organization established by European countries to act as their platform
for collaboration regarding the development of provisions for learners with SEN. Its work program
reflects agreed EU policies and promotes the full participation of SEN students within mainstream
education and training. A wealth of useful information, policy and practice advice, and monitoring tools
82
SEN is used here to designate all children considered different, whether it is because of a physical or mental
handicap, a disadvantaged social background or a special talent.
83
The calculation is as follows: 70 percent of 17,326=12,128; additional 25 percent of the €2370 per-pupil cost of
regular schools= €592.5. Therefore 12,280 x €592.5=€7,275,900. Clearly, these are very rudimentary estimates and
much more detailed calculations are needed.
106
are available to members. For example, a group of experts from 23 countries has developed a set of 67
specific indicators to monitor the inclusiveness of education policies in terms of financing, participation
and legislation (see Box 5-10). In the same vein, key principles for promoting quality inclusive education
were elaborated in 2009 to guide the work of policymakers.84
BOX 5-10: KEY AREAS FOR INCLUSIVE EDUCATION POLICY
1. Legislation and balance/consistency between inclusive education and other policy initiatives.
2. Clear national policy on inclusive education: (a) acceptable national position about the educational
concept of tracking; (b) connection between general and special provision; prevention of the
emergence of special needs.
3. Value statements underlying the curriculum as a point of reference: (a) curriculum; (b) certification.
4. Inclusive assessment systems: identification of SEN by using, e.g., formative/on-going assessment
for learning approaches with all learners.
5. Participation of pupils/students and parents in decision-making.
6. Connection between inclusive education and lifelong learning/ early childhood intervention.
7. Incentives in resources and support allocation; pre-resourcing of schools versus resourcing based
upon diagnosis of needs.
8. Financing and processes linked to funding mechanisms.
9. Inter-sectoral cooperation.
10. Inter-disciplinary support systems.
11. Teacher training/training of professionals (including use of information and communication
technology – ICT).
12. Systems/cultures that encourage collaboration and teamwork among teachers.
13. Differentiation, diversity and multi-cultural education in the classroom.
14. Systems for accountability
84
Together with a number of background papers, they can be found on line at: http://www.european-
agency.org/agency-projects/key-principles
107
5.7 BIBLIOGRAPHY
Allington, R.L., et al. 2010. ―Addressing summer reading setback among economically disadvantaged
Elementary Students‖. Reading Psychology, 31: 411-427.
Amnesty International. Steps to End Segregation in Education – Briefing to the Government of Slovakia.
London: Amnesty International Publications (Index: EUR 72/009/2010).
Baird, Sarah, Ephraim Chirwa, Craig McIntosh and Berk Özler. 2010. ―The Short -Term Impacts of a
Schooling Conditional Cash Transfer Program on the Sexual Behavior of Young Women‖. Available at:
http://irps.ucsd.edu/assets/033/10615.pdf
Cahu, Paul. 2012. ―Education quality and equity in the Slovak Republic – Analysis of the PISA 2009
data‖, unpublished manuscript.
Cataluña SecretarÃa de Acción Ciudadana. 2007. Good Practice Guide for the Social Inclusion of Roma in
Europe. Barcelona: Secretaria d'Acció Ciutadana.
Cooper, H., Charlton,K., Valentine, J.C., Muhlenbruck, L. & Borman, G.D. 2000. Making the Most of
Summer School: A Meta-Analytic and Narrative Review. Monographs of the Society for Research in Child
Development, 65: 1-127.
Cortés, Darwin, Juan Gallego and DarÃo Maldonado. 2011. ―On the Design of Education Conditional Cash
Transfer Programs and non Education Outcomes: The Case of Teenage Pregnancy" Working Paper
008828, Universidad del Rosario, Department of Economics. Available at:
http://ideas.repec.org/p/ces/ceswps/_3531.html.
Equality. 2011. ―From Segregation to Inclusion: Roma Pupils in the United Kingdom‖. Roma Education
Fund.
European Agency for Development in Special Needs Education.2011. Development of a set of indicators
for inclusive education in Europe. Available at: http://www.european-agency.org/publications/ereports
European Roma Rights Center. 2007. The Impact of Legislation and Policies on School Segregation of
Romani Children: A Study of Anti-Discrimination Law and Government Measures to Eliminate
Segregation in Education in Bulgaria, Czech Republic, Hungary, Romania and Slovakia. Budapest:
European Roma Rights Centre.
Friedman, Eben, Elena Gallová Kriglerová, Mária Herczog and Laura Surdu. 2009. ―Assessing
Conditional cash Transfers as a Tool for Reducing the gap in Educational outcomes between Roma and
non-Roma‖. Roma Education Fund. Working Paper no. 4.
Friedman, Eben, Elena Gallová, Martina Kubánová and Martin Slosiarik. 2009. ―School as Ghetto –
Systemic Overrepresentation of Roma in Special Education in Slovakia‖. Budapest: Roma Education
Fund.
GALLUP. 2009. ―Analysis of the impact of affirmative action for Roma in high schools, vocational
schools and universities‖. Roma Education Fund. Working Paper no. 3.
Gallová Kriglerová, Elena (ed.). 2010. ― ― Bratislava: Slovak Governance Institute.
108
Harris-Britt, April, Cecelia Valrie, Beth Kurtz-Costes and Stephanie Rowley. 2007. ―Perceived Racial
Discrimination and Self-Esteem in African American Youth: Racial Socialization as a Protective Factor‖.
Journal of Research on Adolescence, 17(4): 669–682.
HojsÃk, Marek F. 2010. ―REF-supported projects in Slovakia for preventing/reversing segregation of
Romani children in special education‖. Roma Education Fund. Available at:
http://www.romaeducationfund.hu/sites/default/files/publications/comparative_evaluation_sk_-
_english.pdf
Jacob, B.A. and L. Lefgren. 2004. ―Remedial Education and Student Achievement: A Regression-
Discontinuity Analysis‖. The Review of Economics and Statistics, 86: 226-244.
Kézdi, Gábor and Éva Surányi. 2007. ―A successful school integration program‖. Working Paper no. 2.
Roma Education Fund. Available at
http://www.romaeducationfund.hu/sites/default/files/publications/a_succesful_school_integration_kezdi_s
uranyi.pdf
Kosová, Beáta, and Zlatica Hul‘ová (eds.). 2006. ―Education of Roma Students‖. Collectio n of Scientific
Surveys. Banska Bystrica: Matej Bel University.
Ladd, Helen. 2012 ―Education and Poverty: Confronting the Evidence‖, Journal of Policy Analysis and
Management, forthcoming.
Marcincin, Anton, and L‘ubica Marcincinová. 2009. ―The Cost of Non-Inclusion‖, Unpublished
manuscript.
Mervis, Jeffrey. 2011. ―Giving children a head start is possible – But it‘s not easy‖. Science, Vol. 333
(6045):. 956-957.
Mistry, Rashmita S., Elizabeth S. White, Aprile D. Benner, and Virginia W. Huynh. 2009. ―A longit udinal
study of the simultaneous influence of mothers and teachers educational expectations on low-income
youths academic achievement.‖ Journal of Youth and Adolescence 38 (6): 826-38.
Niles, Michael, and Laura Peck. 2008. ―How Poverty and Segregation Impact Child Development:
Evidence from the Chicago Longitudinal Study‖. Journal of Poverty, vol. 12 (3): 306-332.
Oláh, Milan. 2011. ―Assistance Teacher as a (de)Segregation Factor in Education‖ in Rafael, Vlado (ed.).
2011. Answers to Questions on (de)Segregation of Roma Students in Slovak Education System. Bratislava:
Open Society Foundation.
Paton, Graeme. 2011. Teaching Assistants ‗fail to improve school results‘. The Telegraph. Education
News, 26 May 2011.
Patrinos, Harrys, and George Psacharopoulos, 2010. Returns to Education in Developing Countries,
Oxford, England: Elsevier.
Powell, Justin. 2006. ―Special Education and the Risk of Becoming Less Educated‖, European Societies,
8(4): 577-599.
109
Rafael, Vlado (ed.). 2011. Answers to Questions on (de)Segregation of Roma Students in Slovak
Education System. Bratislava: Open Society Foundation.
Richardson, John, and Justin Powell. 2011. Comparing Special Education: from Origins to Contemporary
Paradoxes. Stanford, CA: Stanford University Press.
Roma Education Fund. 2011. Country Assessment: Slovakia. Budapest: Roma Education Fund.
Rus, Calin. 2005. ―The Situation of Roma School Mediators and Assistants in Europe‖. Council of
Europe.
Rus, Calin, and Mihaela Zatreanu. 2006. Education of Roma Children in Europe – Guide for Roma School
Mediators. Council of Europe.
Ryerson, William. (undated). ―The Effectiveness of Entertainment Mass Media in Changing Behavior‖.
Population Media Center. Available at: http://www.populationmedia.org/what/effectiveness/
Salner, Andrej. 2004. Roma Children in the Slovak Education System. Bratislava: Slovak Governance
Institute.
Sharan, S., 1980: ―Cooperative Learning in Small Groups: Recent Methods and Effects on Achievement,
Attitudes, and Ethnic Relations‖ Review of Educational Research, 50 (2): 241-271
Spotáková, Magdaléna. 2011. ―The Role of Diagnostics in Education Directed at Roma Children‖ in
Vlado Rafael (ed.). 2011. Answers to Questions on (de)Segregation of Roma Students in Slovak Education
System. Bratislava: Open Society Foundation.
Tankersly, D., E. Konkova and D. Repiski. 2002. ―transforming the Role of Teaching Assistants‖ in
Educating Children for Democracy. Classroom Practices no.3.
Tomatová, Jana. 2004. Sidetracked. Bratislava: Slovak Government Institute.
White, Julia. 2011. Entry Testing as a Factor in Overrepresentation of Romani Children in Special
Education. Budapest: Roma Education Fund.
World Health Organization. 2011. World Report on Disability. Geneva: World Health Organization.
110
5.8 ANNEX
ANNEX TABLE 5.1: PREDICTING SECONDARY SCHOOL COMPLETION
(1) (2) (3)
All Roma Non-Roma
Roma -.262***
(.042)
Girls -.054** -.016 -.170***
(.021) (.023) (.047)
Age .015*** .007* .040***
(.003) (.003) (.006)
Attended preschool .036 .048* -.045
(.024) (.026) (.069)
Household Head - Male -.004 .017 -.044
(.024) (.027) (.052)
Household Head – Completed Secondary .431*** .401*** .447***
Schoolb (.037) (.044) (.072)
Household Head – Age .003*** .004*** .002
(.001) (.001) (.002)
Size of the household -.022*** -.024*** -.029
(.007) (.008) (.020)
Second income quintile -.018 .011 -.116
(.050) (.051) (.130)
Third income quintile .062 .090* -.072
(.050) (.052) (.124)
Fourth income quintile .120* .142** -.092
(.062) (.067) (.148)
Fifth income quintile .154*** .184*** .007
(.052) (.060) (.116)
Household suffers from hunger -.046 -.062** .060
(.029) (.030) (.096)
Household language is Romani -.013 -.007 -.211
(.031) (.031) (.160)
Distance to secondary school: within 3 km .063* .096** .019
(.035) (.039) (.073)
Dominant ethnicity of the settlement is Roma -.042 -.053* .018
(.027) (.028) (.073)
Central regionc -.037 -.035 -.086
(.046) (.050) (.104)
Eastern regionc -.005 -.016 -.045
(.036) (.037) (.078)
Rural .049 .097** -.015
(.034) (.040) (.070)
Constant -.049 -.213* -.265
(.110) (.120) (.227)
Observations 1,071 823 248
R-squared .494 .314 .425
111
Source: UNDP/World Bank/EC regional Roma survey (2011).
OLS estimations. Robust standard errors in parentheses. Example (Model 1): In Slovakia, Roma of the ages 18-30
are 26.2 percentage points less likely to complete secondary school than non-Roma, when background
characteristics are taken into account.
*** p<.01, ** p<.05, * p<.1
In addition to income quintiles, the analysis includes ‗missing income‘ as a separate income category. Observations
for which no information on income was available could hence be included in the estimations. The ‗missing income‘
category is left out of the table.
a
Only those subjects with at least some primary education were included in the estimation sample.
b
Secondary school completion also includes those respondents who have incomplete general secondary school.
c
Slovakia‘s Western region is omitted from the table. Estimates for the Central an d Eastern region refer to the
difference of these two regions with the Western region.
112
ANNEX TABLE 5.2: PREDICTING SPECIAL SCHOOL ATTENDANCE
(1) (2) (3) (4) (5)
All: All: All: Roma Non-Roma
Base Base + Full
Model Regions Model
Roma .089*** .085*** -.019
(.016) (.016) (.016)
Girls .005 .012 -.028
(.010) (.012) (.020)
Agegroup: 19-30a -.012 -.014 -.014 -.011 -.015
(.015) (.015) (.013) (.016) (.013)
Attended preschool -.080*** -.081*** -.066*
(.018) (.020) (.036)
Household Head - Male -.034** -.042** .009
(.016) (.020) (.017)
Household Head - Completed -.023 -.039** .035*
Secondary Educationb (.015) (.019) (.019)
Household Head - Went to Special .578*** .574*** .559*
School (.056) (.058) (.296)
Household Head - Age -.000 -.000 -.001
(.001) (.001) (.001)
Size of the household .013** .012** .022
(.005) (.005) (.017)
Second income quintile -.005 .013 -.126*
(.038) (.044) (.075)
Third income quintile -.021 -.007 -.139*
(.038) (.042) (.081)
Fourth income quintile .020 .042 -.119
(.035) (.039) (.074)
Fifth income quintile -.002 .020 -.123*
(.039) (.045) (.070)
Household suffers from hunger -.020 -.011 -.094*
(.018) (.019) (.050)
Household language is Romani .024 .027
(.018) (.019)
Distance to primary school: within .077*** .007 .005 -.004
3 km (.021) (.015) (.019) (.019)
Dominant ethnicity of the -.005 -.013 .035*
settlement is Roma (.016) (.018) (.021)
Central regionc .075*** .018 .021 .002
(.027) (.023) (.031) (.016)
Eastern regionc .070*** .037*** .044** .014
(.017) (.014) (.019) (.012)
Rural .023 -.016 -.019 -.021*
(.021) (.017) (.021) (.012)
Constant .027** -.122*** .086* .066 .119*
(.014) (.032) (.051) (.057) (.063)
Observations 2,214 2,214 2,214 1,744 470
113
R-squared .016 .027 .348 .349 .306
Source: UNDP/World Bank/EC regional Roma survey (2011). Sample restricted to subjects aged 7-30.
OLS estimations. Robust standard errors in parentheses. Example (Model 1): In Slovakia, Roma of the ages 7-30 are
8.9 percentage points more likely to attend a special school than non-Roma, when only age differences are taken into
account.
*** p<.01, ** p<.05, * p<.1
In addition to income quintiles, the analysis includes ‗missing income‘ as a separate income categ ory. Observations
for which no information on income was available could hence be included in the estimations. The ‗missing income‘
category is left out of the table.
A
A distinction is made between two agegroups: 7-18 and 19-30. Agegroup 19-30 is represented in the table.
Estimates for this variable refer to the difference of subjects belonging to this agegroup as compared to the younger
agegroup (7-18) which is left out of the table.
B
Secondary school completion also includes those respondents who have incomplete general secondary school.
C
Slovakia‘s Western region is omitted from the table. Estimates for the Central and Eastern region refer to the
difference of these two regions with the Western region.
114
ANNEX TABLE 5.3: PREDICTING EMPLOYMENT
(1) (2) (3) (4) (5)
All Roma Non- Roma: Roma:
Roma Men Women
Roma -.159***
(.037)
Women -.157*** -.141*** -.204***
(.015) (.017) (.032)
Age -.001 -.001 -.002 -.002 -.000
(.001) (.001) (.002) (.001) (.001)
No / incomplete basic educationa -.052** -.060** -.120 -.059 -.066***
(.025) (.025) (.088) (.042) (.023)
Complete Secondary education or highera b .188*** .132*** .295*** .134*** .131***
(.030) (.034) (.063) (.046) (.043)
Number of household members .003 .006 -.009 .013* -.002
(.005) (.006) (.018) (.008) (.005)
Household language is Romani -.016 .006 -.148 -.003 .016
(.029) (.031) (.128) (.042) (.032)
Distance to employment office: within 3 km -.028 -.039 -.011 -.052 -.025
(.039) (.041) (.083) (.059) (.035)
Distance to nearest city: within 3 km .009 .012 .017 -.003 .028
(.032) (.033) (.073) (.046) (.030)
Dominant ethnicity in settlement is Roma -.010 -.029 .042 -.019 -.038
(.025) (.027) (.067) (.039) (.026)
Central regionc .080* .126** .010 .121 .134***
(.044) (.052) (.088) (.074) (.050)
Eastern regionc -.031 -.044 -.009 -.081 -.007
(.035) (.039) (.074) (.058) (.036)
Rural household .005 -.002 .045 -.041 .039
(.035) (.037) (.073) (.053) (.033)
Constant .385*** .196*** .290* .258** -.013
(.073) (.075) (.148) (.115) (.069)
Observations 2,085 1,450 635 722 728
R-squared .172 .105 .106 .074 .083
Source: UNDP/World Bank/EC regional Roma survey (2011).
OLS estimations. Robust standard errors in parentheses. Example (Model 1): In Slovakia, Roma of the ages 25-64
are 15.9 percentage points less likely to be employed than non-Roma, when background characteristics are taken
into account.
*** p<.01, ** p<.05, * p<.1
A
Both schooling categories shown in the table should be compared to ‗primary or incomplete secondary education‘.
B
Secondary school completion also includes those respondents who have incomplete general secondary school.
C
Slovakia‘s Western region is omitted from the table. Estimates for the Central and Eastern region refer to the
difference of these two regions with the Western region.
115
6 HOUSING
This chapter provides a detailed assessment of the large gaps that exist in housing conditions between
Slovak Roma and the general Slovak population, and provides recommendations to address these. It first
highlights the linkages between housing conditions and outcomes in other areas such as health,
employment, and education. It then assesses the neighborhood conditions faced by surveyed Roma, before
focusing on the specific housing conditions and housing affordability. Following a review of the current
policy framework in Slovakia, which relies heavily on two policy tools - (1) construction of social housing
units, and (2) the housing benefit allowance - the chapter concludes with specific policy recommendations
that promote a comprehensive, incremental approach to improving living conditions in situ and/or helping
poor families move into better housing elsewhere.
6.1 INTRODUCTION
―A huge drop in the living standard of Roma communities had been registered in Slovakia in the last
twenty years. In the aforementioned period, the Roma population had relocated from integrated town
districts to town ghettos and rural settlements, mostly in the segregated regions. Housing is undoubtedly
one of the areas in which the gap between Roma on the one hand and the majority population on the other
is ever deepening. …[O]nly Roma communities in Slovakia establish settlements within which are various
types of non-standard dwellings that fail to comply with either technical or hygienic standards. Such non-
standard dwellings are more often than not built on land with uncertain land title, without a planning
permission. The construction materials used, such as wood, tin, clay, are also non-standard. Another
serious problem is the lack of basic infrastructure, such as electricity, access to drinking water, access
roads and sidewalks with public lighting, gas, sewage. An extreme problem in this regard is the waste
removal and disposal.‖
-The Slovak National Strategy for Roma Inclusion, as submitted to the European Commission early 2012,
Section D. 2. 4. Housing)
6.1.1 THE IMPORTANCE OF ADEQUATE HOUSING CONDITIONS FOR INCLUSION
This chapter focuses on the housing conditions among Roma in Slovakia, an increasing number of
whom live in segregated, often informal, settlements --or ―slums.‖85 The term ―slum‖ commonly refers
to informal settlements in or near urban areas with inadequate housing, overcrowding, poor living
conditions, lack of access to basic municipal services such as water, sanitation, waste collection, storm
drainage, street lighting, paved sidewalks and roads for emergency access. Slums usually also lack easy
access to schools, hospitals or other basic services. In general slums are informal and residents lack legal
ownership and buildings are not built to code or in accordance with zoning. Finally, slums are often areas
where the social situation is worsening, where crime and unemployment, for example are increasing. 86 In
Slovakia, a large percentage of Roma households live in slum conditions. The chapter investigates how
many Roma households live under what kinds of conditions, reviews current Government policy
approaches to improving conditions for poor Roma households living in segregated and informal
settlements (i.e., slums), and draws from Slovak and world experience with slum upgrading to suggest
how policies might be complemented or changed to more effectively use available resources to reach a
larger number of people.
85
The Slovak National Strategy for Roma Inclusion, as submitted to the European Commission early 2012, Section
D.2.4. Housing.
86
This definition draws from Cities Alliance. See http://www.citiesalliance.org/About-slum-upgrading.
116
Housing is particularly important for social inclusion as living conditions are linked to other socio-
economic outcomes, such as health, labor market participation and education. Slums where poor
households live under substandard conditions are logical places to focus multi-dimensional inclusion
efforts to reintegrate these places and their residents physically, socially, economically and financially
back into the broader Slovakian society. The Raxen Report (2009) 87 on Slovakia points to spatial
segregation as one particularly important housing challenge. Spatial segregation is highly correlated with
early school-leaving, low labor market participation rates and costly access to other services (public
transport, health facilities, etc.).
The informality of most spatially segregated settlements, as is true of slums worldwide, creates
additional barriers. Without clear land title and a formally accepted dwelling built according to codes in
areas planned for residential housing, Roma households are ineligible to obtain a legal address to use for
registration purposes. A permanent registered address is required for access to housing allowance
benefits, as well as for school registration and access to health care among others. Thus, providing for
adequate housing conditions goes hand in hand with social and economic inclusion for marginalized
Roma households.
6.2 HOUSING CONDITIONS
The UNDP/World Bank/EC regional Roma survey (2011) finds that a large proportion of
households, 42% (approximately 30,240 households when considering the entire population), live in
a settlement where the dominant ethnicity is Roma, underscoring the very high level of spatial
segregation. An estimated 21,000 (30%) Roma families live in the worst conditions, i.e., in ruined houses
or slums as identified by the 2011 Regional Survey enumerators.
As highlighted in the introduction, Roma population growth remains much above that of the
majority population, implying that that this group of ‗unhoused‘ and ‗underhoused‘ Roma is
rapidly growing over time. Assuming a continued annual population growth rate of 1.7% (and assuming
that mortality rates and household size remain the same 88), an estimated 1,200 new Roma households are
formed every year. This is considerably greater than the 0.3% annual increase for the country‘s population
as a whole during the 2001 to 2011 intercensal period.89 A 1.7% growth rate for Roma households means
that the number of Roma families living under the worst conditions (i.e., in ruined buildings or slums) is
estimated to grow by about 360 households per year. At the same time, the main policy intervention for
housing, building low standard social housing, has supplied an average of 290 units per year from 2001 to
2010 suggesting that the program is insufficient even to keep up with natural increase for the worst housed
Roma families.90
6.2.1 NEIGHBORHOOD CONDITIONS
Analysis of the Roma Regional Survey data finds little difference between Roma households and the
nearby comparator households in terms of access to basic services (medical, educational, transport,
and financial). For both groups, urban households have much better access to services than rural
households as would be expected. About one in four households in both groups report deterioration in
their neighborhood conditions during the past four years.
87
FRA Raxen Report 2009
88
Reductions in age-adjusted mortality rates or household size would increase the number of dwelling units needed
each year.
89
The Statistical Office of the Slovak Republic: 2011 Census. See http://portal.statistics.sk/files/table-3.pdf
90
Figures provided by Ministry of Construction.
117
However, the problem with spatial segregation for Roma households is particularly acute. About
four in ten Roma households live in a settlement where the dominant ethnicity is Roma. In Central
Slovakia, this ratio is nearly six in ten. Ethnic segregation is stronger among Roma households living in
ruined buildings/slums than among those living in other types of housing. More than half (56%) of Roma
households living in slums report also living in predominantly Roma communities. At the same time,
nearly three in four Roma households surveyed would prefer to live in mixed areas.
FIGURE 6-1: PERCENTAGE OF ROMA HOUSEHOLDS LIVING IN PREDOMINANTLY ROMA
COMMUNITIES IN THE THREE REGIONS OF SLOVAKIA
80
59
60
Percent
39
40 30
20
0
West Central East
Roma
Source: UNDP/World Bank/EC regional Roma survey (2011).
BOX 6-1: SUMMARY OF NEIGHBORHOOD CONDITIONS OF ROMA HOUSEHOLDS
ï‚· Most Roma households in Slovakia are located in villages. In the Eastern region, the share of Roma
households in villages is 66%.
ï‚· There are few differences between Roma and comparator non-Roma households in terms of access to
basic services (medical, educational, government or financial). Both Roma and non-Roma urban
households are similarly better provided with access than are Roma and non-Roma households in
villages.
ï‚· The majority of Roma households in the sample speak Romani at home.
ï‚· Spatial segregation of Roma reaches approximately 42%.
ï‚· Spatial segregation is stronger among Roma households living in slums than among those living
elsewhere.
ï‚· Roma and non-Roma households are similarly satisfied (dissatisfied) with their neighborhoods and
report few improvements in neighborhood conditions during the past five years.
6.2.2 HOUSING CONDITIONS
Although the neighborhood conditions for Roma and non-Roma households are quite similar,
actual housing conditions for Roma households are consistently worse than for the comparator non-
Roma households. The regional Roma survey finds that 30% of Roma households live in a ruined house /
slum, many times more than the 4% of comparator households with similar housing conditions. Ruined
houses or slum conditions are especially common among Roma households living in segregated
communities: 41% of these households live in slum dwellings. Disaggregated by income level, Roma
households in the lowest income tercile are two and a half times as likely to live in slums as Roma
118
households in higher terciles. Analysis of the EU SILC (2008) finds a similar pattern with 42% of Roma
households in Slovakia living in a dwelling that is in (very) bad condition. Among non-Roma in their
vicinity, this is much lower, 12%, and among the general population, the rate is close to 9% (EU SILC,
2008)91. The share of Roma households where exterior and interior walls are judged to be in (very) bad
condition is significantly higher than the comparator households‘ ratio in all parts of the country. About
40% of Roma households report walls in bad condition whereas for the non-Roma the share is much
lower, below 20% (Roma Regional Survey, 2009).
Roma families are much more overcrowded than non-Roma families, and poorer Roma households
tend to live in smaller units with fewer rooms. Although 45% of Roma households surveyed lived in
dwellings with more than 2 people per room, only 9% of nearby non-Roma households experienced this
level of overcrowding. In general, more than 2 people per room is considered very overcrowded in the
European context.92 Moreover, among the Roma, the poorest households tend to live in smaller units with
fewer rooms. For example, Roma households in the lowest income tercile have, on average approximately
half as large a dwelling as nearby non-Roma households (48 sq m vs. 80 sq m). The median number of
rooms available to Roma households in the lowest income tercile category is 1 compared to 2 for the
highest income tercile Roma households. The median number of rooms for non-Roma households is 3.
The overcrowding measures are especially important when taking into account the size of the
household, as Roma households are generally larger: Roma have access to 14 square meters per
household member on average, whereas among non-Roma, each household member has an average of 28
square meters. The share of Roma households in Slovakia with only one or two rooms in the dwelling is
as high as 72%, whereas among the general population, this is only 27% (EU-SILC, 2008).
TABLE 6-1: SIZE OF THE DWELLING
Roma Income Terciles: Aggregates:
1 2 3 Roma Non-Roma
Number of rooms available to the household:
Mean 1.7 2.1 2.3 2.1 3
Median 1 2 2 2 3
Square meters in dwelling:
Mean 48.5 49.9 61 53.4 80
Median 50 45 55 50 72
Source: UNDP/World Bank/EC regional Roma survey (2011).
In addition to overcrowding, Roma households are significantly more likely to lack basic amenities
such as indoor toilets and kitchens. The figure below illustrates this point clearly. Only about half of
Roma households have indoor sanitation (toilet, bathroom, sewage connection) while about 90% of nearby
non-Roma families have these amenities. The poorest Roma families are least likely to have access to
indoor sanitation. For the lowest income tercile Roma households, fewer than 30% have toilets inside
91
Among the general population, the question asked concerning the condition of the dwelling was more specific than
in the Roma Regional Survey: whereas in the latter survey, the interviewer was asked to rate the exteriors of the
welling on a 1-5 scale, the question asked in the EU SILC survey was: ‗Does the dwelling have a leaking roof, damp
walls/floors/foundation, or rot in window frames or floor?‘
92
This can be considered a rough proxy for EU standards, see the following link for more information.
http://epp.eurostat.ec.europa.eu/statistics_explained/index.php/Glossary:Overcrowding_rate
119
their homes, which suggests potential important health risks. Roma households living in predominantly
Roma communities have lower rates of sanitation than those in mixed communities: only 44% of Roma
households in segregated communities have indoor toilets, 39% are connected to sewerage and 47% have
a bathroom. Similarly, about half of Roma households do not have access to drinking water indoors (as
compared with 12% for non-Roma households).
FIGURE 6-2: PRESENCE OF SANITARY PROVISIONS IN THE HOUSEHOLD (% OF HOUSEHOLDS)
97 92 95
100 89 90 86 84
80
60
Percent
60 53 51
47
40
18
20
0
Kitchen inside Toilet inside Latrine / Toilet Connection to public Shower or bathroom Electricity supply
outside sewerage or waste inside
water tank
Roma Non-Roma
Source: UNDP/World Bank/EC regional Roma survey (2011).
Nearly half of Roma households lack access to indoor piped water (46%) while only 21% of the non-
Roma households nearby do not have access to indoor piped water. Again, the situation is the worst
among Roma in segregated settlements where as few as 39% of households have access to piped water.
When comparing data for Roma in Slovakia to those for Roma in other countries in the region, only Roma
in Romania are worse off in terms of access to drinking water inside their dwelling (Regional Roma
Survey, 2011).
Collection of solid waste is particularly problematic for Roma households, especially in segregated
areas. One in four Roma households reports irregular or no collection of solid waste. In segregated
Roma communities this share increases to nearly one in three (29%). Among nearby non-Roma
comparator households only one in twenty report not having access to regular waste collection.
Regionally, only Bulgaria has lower solid waste collection rates than Slovakia (Regional Roma Survey,
2011).
Roma households are far more reliant on dirty fuels for heating and cooking than are the
comparator non-Roma households (see Table below). Half of Roma households use wood for cooking
and 86% use wood or coal for heating (rates that were characteristic for the overall population in the
1960s). 93 For non-Roma comparator households the rates are much lower: only 12% use wood for
cooking while 53% use wood or coal for heating. Use of dirty fuels is especially common among lower
income Roma households. Three-quarters of all Roma households in the lowest income tercile rely on
wood for cooking and an impressive 92% rely on wood for heating.
93
See the Census data for 1961 and 1971 for Slovakia. http://sodb.infostat.sk/sodb/eng/1848-2001/tab.III.2-
2A_en.htm and http://sodb.infostat.sk/sodb/eng/1848-2001/tab.III.2-2G_en.htm
120
TABLE 6-2: METHODS OF COOKING AND HEATING THE DWELLING (% OF HOUSEHOLDS)
Roma Income Terciles: Aggregates:
1 2 3 Roma Non-Roma
On what do you usually cook in your household?
Gas / Electricity 20 34 51 44 83
Wood 77 54 42 50 12
Other 3 12 7 7 5
And how do you usually heat your house?
Gas / Electricity 4 5 22 8 34
Wood 92 92 69 74 34
Other 4 4 10 18 32
Source: UNDP/World Bank/EC regional Roma survey (2011).
The reliance on wood for heating is exacerbated by the overall lack of energy efficiency of the
structures, which means relatively more wood is necessary as much of the heat escapes the dwelling.
The use of wood and the overall energy inefficiency of the dwellings are in stark contrast with the
EU2020 goals on energy efficiency and emissions.
BOX 6-2: SUMMARY OF HOUSING CONDITIONS AMONG ROMA HOUSEHOLDS
ï‚· Nearly one third of the Roma in excluded communities live in ruined housing / slums, while fewer
than 5% of the non-Roma comparator group lived in ruined housing / slums. Ruined houses or slum
conditions are especially common among Roma households living in segregated communities: 41% of
households. Disaggregated by income level, Roma households in the lowest income tercile are two
and a half times as likely to live in slums as Roma households in higher terciles. An estimated 21,000
Roma families suffer from very poor housing conditions.
ï‚· Roma families are much more overcrowded than non-Roma families. Although 45% of Roma
households surveyed lived in dwellings with more than 2 people per room, only 9% of nearby non-
Roma households experienced this level of overcrowding. Poorer Roma households live in smaller
units with fewer rooms. As Roma households are larger, in terms of living space per capita Roma
households average only 14, which is about one half of the 28 square meters per capita found among
the comparator households.
ï‚· Only about half of Roma households have indoor sanitation (toilet, bathroom, sewage connection)
while about 90% of nearby non-Roma families have these amenities. The poorest Roma families are
least likely to have access to indoor sanitation.
ï‚· About half of Roma households lack access to piped water in their dwellings whereas for comparator
households only 12% lack this access.
ï‚· One in four Roma households reports irregular or no collection of solid waste. In segregated Roma
communities this share increases to nearly one in three. Among nearby non-Roma comparator
households only one in twenty report not having access to regular waste collection. Regionally, only
Bulgaria has lower solid waste collection rates than Slovakia (Regional Roma Survey, 2011).
ï‚· Half of Roma households use wood for cooking and 85% use wood or coal for heating, substantially
more than the use by non-Roma comparator households.
121
6.2.3 HOUSING AFFORDABILITY
The economic transition affected all Central and Eastern European countries and has had a long-
term impact on Roma in the whole region. In Slovakia, the employment rate for groups with little
education shrank to 60% of its 1998 value by 2010, resulting in many low educated Roma households‘
exclusion from the labor market. This, in turn, resulted in increasing vulnerability in terms of security of
income, dependency on transfers and general affordability problems. Slovakia witnessed smooth
economic growth throughout the nineties. Employment rates decreased in the first half of the nineties, but
with the restructuring of the economy (increase of the service sector and decrease of the industrial and
agricultural production sector), employment rates began increasing again until the end of the nineties94.
Currently, the national employment rate is 58.8%, approximately 5% points lower than the EU average.
Employment rates among Roma are much lower though; the Regional Roma survey shows that only 20%
of working age Roma men, and 9% of working age Roma women, are formally or informally employed.
Although approximately half of Roma households depend on social assistance, many do not qualify
for the housing allowance. Among Roma, approximately every second household receives the Benefits
for Material Needs (BMN) (about 35,000 households), and every third family receives the additional
Housing Allowance (about 24,000 households) as can be seen in the Figure below. Among non-Roma
households, the shares are only 4% for the BMN and 3% for Housing Allowance.
FIGURE 6-3: SHARE OF ROMA HOUSEHOLDS RECEIVING THE BASIC MATERIAL NEEDS BENEFIT OR
THE HOUSING BENEFIT BY SETTLEMENT TYPE
60
50
40
Percent
30
20
10
0
Mixed Segregated Separated
BMN Housing Allowance
Source: UNDP 2010 Survey of Roma and nearby Non-Roma in Slovakia
As would be expected, housing allowance receipt correlates strongly with type of settlement. In
segregated areas where housing is more likely to be informal and not legally registered, only 27% of
Roma households receive the housing allowance. In mixed neighborhoods where Roma households are
more likely to live in formal dwelling units, this share increases to 40%. The large gap between BMN
recipients (who comprise the poorest layer of Slovak society) and Housing Benefit receipt can be
explained by several factors, among which informality (not living in a formal dwelling or not having
secure tenure (i.e., a lease)) are the most likely reasons. Families must live in formally accepted buildings
with formal tenure (lease/ownership) in order to be eligible to receive housing benefits, whereas BMN
eligibility is only tied to low income levels.
94
http://www.unece.org/fileadmin/DAM/stats/documents/ces/sem.44/wp.16.e.pdf
122
Households with the worst living conditions are particularly likely to not receive the housing benefit.
Figure 6-4 depicts the share of Roma households who receive the BMN or the Housing Benefit by
household housing conditions. Households who described their housing conditions as average or better
were classified as households with good conditions. Those describing their living conditions as worse
than average were categorized as households with bad conditions. The largest gap between the share of
households receiving the BMN and those receiving the Housing Benefit is found in households living in
bad conditions. This is logical as informal households are more likely to live in poor conditions. It is
important to note, however, that the Housing Benefit, one of the two main subsidies for housing, does not
reach poor Roma households living informally in separated or segregated settlements (in other words, the
most disadvantaged Roma households).
FIGURE 6-4: SHARE OF ROMA HOUSEHOLDS RECEIVING BMN OR HOUSING BENEFIT BY QUALITY OF
LIVING CONDITIONS BY TYPE OF SETTLEMENT.
A. Households with ―Bad‖ Housing B. Households with ―Good‖ Housing
70 63 70
60 53 60 55
48 48 46
50 42 50
39
Percent
40 Percent 40 35
27 28
30 30
20
20 20
10 10
0 0
Mixed Segregated Separated Mixed Segregated Separated
Receives Housing Benefit Receives BMN Receives Housing Benefit Receives BMN
Source: UNDP 2010 Survey of Roma and nearby Non-Roma in Slovakia.
Low incomes contribute to poor housing conditions for Roma families. Extremely low employment
levels – 20% of working age Roma men and 9% of working age Roma women – means that most Roma
households – 87% - live in poverty95, and many depend on social assistance; approximately every second
Roma household receives the Benefit for Meeting Material Needs (BMN), and approximately one third
receives the additional housing allowance. Among non-Roma households, on average 4% receive the
BMN and only 3% receive housing allowances. Even though housing expenditures comprise roughly the
same share of household budgets for Roma as for non-Roma comparator households, the lower income
levels for Roma households mean affordability is much more problematic. Roma households often worry
they will be evicted due to financial difficulties-- 40% of Roma households expressed concern about
eviction. Due to the overall low income levels, restructuring household budgets to mitigate this risk is
difficult or impossible.
Nearly half (44%) of Roma households report having difficulty making mortgage, rent or utility
payments. Among non-Roma comparators a much smaller percentage reports having difficulty to pay the
bills (16%). As shown in
95
As defined by equivalized household incomes below 60% of the national median income.
123
Table 6-3, a small but significant proportion of Roma households is in arrears. For example, 11% of Roma
households is in arrears for electricity bills, compared with 2% of non-Roma neighbors. But as shown
above, the lower rates of arrears in part reflects the fact that a majority of households relies on wood for
heating and cooking, not gas or electricity. Importantly, as many as three quarters (75%) report restricting
heating. The latter is also a challenge among non-Roma where 63% reports restricting heating.
TABLE 6-3: ARREARS FOR HOUSING RELATED EXPENSES (% OF HOUSEHOLDS)
Roma Income Aggregates:
Terciles:
Non-
1 2 3 Roma
Roma
Water supply 4 7 5 6 1
Electricity supply 13 17 7 11 2
Other housing related utilities 4 4 5 4 2
Mortgage 1 1 2 1 1
Credit for household appliance or furniture 4 7 5 4 2
Do you have difficulties paying off mortgages, rent or utility
bills? 37 48 35 44 16
Source: UNDP/World Bank/EC regional Roma survey (2011).
BOX 6-3: PRE-PAID METERED UTILITY PROVISION – EXTENDING THE SOLUTION
TO MANAGE DEBTS
In Slovakia, some attempts have been introduced since the mid 2000s to introduce pre-paid meters
for electricity supply. The benefits of this solution, however, can be extended so as to not only permit
control of consumption, but also can be used to manage arrears effectively by combining pre-paid meters
with housing allowance and debt management schemes. The latter approach is completely missing from
the current Slovak context as housing allowances are only provided as a cash-benefit. In Hungary,
however, prepaid meters are linked to both housing allowances and debt repayment.
A study by the Open Society Foundation describes the Hungarian prepaid meter program.: ―One of the
most popular debt-management models applied in Hungary is a combination of in-kind housing
allowance and repayment of debts via pre-paid public utility services. The model needs a minor
infrastructure investment, that is, a meter which is served by a pre-paid card. Typically, gas and
electricity providers have been offering such meters for many years now. The account can be recharged at
the service points of the utility provider96. For the repayment of the debt, a certain ratio of the recharged
credits is written off, based on a case-by-case debt management contract between the household and the
service provider. The housing allowance scheme is offered in-kind in the form of credits to be consumed.
Households reportedly consciously control their consumption after the meters are installed. The
96
Various service providers have extended this opportunity to bank transfers and uploading the credits at ATMs,
placed in houndreds of news shops and post offices.
124
application usage of this technical solution can be easily learned and it can be transferred if the
infrastructure for recharging the pre-paid cards is available.97‖
The lack of affordable housing and low levels of access to housing finance limit housing mobility for
Roma households. The gap in housing quality and the spatial exclusion of a large share of Roma housing
severely limits the pathways out of housing exclusion. For example, a Roma household that would like to
sell its house generally only has a few potential buyers, i.e., other (usually poor) Roma. This means that
the average price of a home in a predominantly Roma settlement is reported to be a fraction of a regional
average housing price. Social housing represents currently approximately 4% of the total housing stock.
About one in ten social housing units built during the past decade has been intended to serve Roma
housing needs. Unfortunately, the low-standard social housing construction program relies on land
provided by the municipalities and in practice this has resulted in further segregation for Roma families
who receive these new units. Despite these challenges, evidence from ETP, a NGO working with poor
Roma households, shows that many households are able to save small amounts toward home
improvements that can make a significant difference in living conditions.
As is the case with other countries in the region, the lack of affordable rental housing, limited access
to housing finance and discrimination combine to limit Roma households from accessing or
improving housing. The challenge posed by Roma slums is growing both because of rapid population
increase but also because downwardly mobile Roma households move to slums because of evictions, other
affordability problems and on occasion due to discrimination and prejudice.
BOX 6-4: SUMMARY HOUSING AFFORDABILITY
ï‚· Approximately half of Roma households depend on social assistance, although many do not qualify
for the housing allowance, especially those who live in informal housing as formal housing is required
for eligibility. This excludes many of the worst off Roma.
ï‚· Nearly half (44%) of Roma households report having difficulties in paying their mortgage, rent or
utility costs.
ï‚· Although few households are actually in arrears on utility payments this primarily reflects that the
majority of Roma households rely on wood for heating and cooking, not gas or electricity.
ï‚· Three quarters (75%) of Roma households report limiting heating.
ï‚· The lack of affordable housing and low levels of access to housing finance limit housing mobility for
Roma households.
ï‚· Distrust toward Roma moving into mixed neighborhoods also impedes upward mobility.
6.3 CURRENT POLICY FRAMEWORK IN SLOVAKIA
6.3.1 INTERNATIONAL APPROACHES TO IMPROVE HOUSING CONDITIONS OF
POOR AND EXCLUDED COMMUNITIES
97
http://lgi.osi.hu/cimg/0/1/3/9/3/vademecum_supplementary_1703.pdf, section 3.3.2.
125
Providing adequate shelter is a high priority for countries worldwide, however the approach
countries take has evolved during the past several decades. The high cost of providing fully finished
dwelling units has meant that governments rarely are able to meet their targets for public units as the costs
are high and management complex. In an effort to control costs and increase the number of beneficiaries,
governments increasingly have diversified their approaches to housing provision during the past several
decades. Claudio Acioly, who is responsible for housing policy at UN-Habitat, characterizes the last
several decades of housing policy worldwide as having passed through five main stages as it moved from
government (direct) provision of actual housing units to market supply of housing supported by
government regulation and enabling policies.98 He characterizes the five stages as:
ï‚· State sponsored housing production and delivery
ï‚· Lowering standards to reach lower income families
ï‚· Involving future beneficiaries in housing production (i.e., slum upgrading)
ï‚· Intervention in housing inputs to facilitate housing production (i.e., sites and services)
ï‚· Enabling policies: less government involvement
As the figure below shows, the different solutions for housing provision bring different costs and benefits
with provision of public housing being the most costly (and thus having the fewest beneficiaries) and
provision of unserviced sites being the least costly (and thus reaching the most beneficiaries).
FIGURE 6-5: TRADE OFFS BETWEEN COSTS PER DWELLING AND NUMBER OF BENEFICIARIES FOR
DIFFERENT TYPES OF AFFORDABLE HOUSING OPPORTUNITIES
Source: Claudio Acioly Jr., Chief Housing Policy UN-HABITAT, Presentation on Housing Sector Profile, July 28,
2011.
In terms of how existing slums are handled, what the above means is that Governments increasingly
look beyond providing new public units as the sole or even the main intervention. Instead, lower cost
community-driven incremental approaches are increasingly prioritized. These include ‗slum upgrading‘
and ‗sites and services‘.
98
Claudio Acioly Jr., Chief Housing Policy UN-HABITAT, Presentation on Housing Sector Profile, July 28, 2011
126
‗Slum upgrading‘ enables slum residents to engage actively in planning and implementing
community improvements in terms of basic infrastructure, in order to formalize/legalize the slum
areas and the land plots, to strengthen community leadership and skills, and to support housing
improvements. It is this combination of tenure security combined with community-driven approaches
to physical, social and economic services within the community that results in the most effective slum
upgrading projects. Community and housing improvements are usually incremental and take place over a
number of years, creating the conditions for longer-term social and housing mobility for residents.
‗Sites and services‘ approaches provide housing inputs (land, infrastructure and services) with the
idea that the low-income recipients will build their houses themselves, which much reduces the
subsidy required. Commonly services provided to beneficiaries include assistance with materials and
plans for building a core or starter dwelling. For both slum upgrading and for sites and services the costs
to the state are far less, enabling these programs to reach many more households and to leverage other
resources including those of the beneficiaries.99
6.3.2 NATIONAL LEVEL POLICY DEVELOPMENTS: NRIS
Current policies for Roma inclusion in Slovakia draw from the Decade for Roma Inclusion 2005-15
and the EU member states‘ obligations to produce National Roma Inclusion Strategies (NRIS) to
frame the Roma inclusion agenda 2012-2020. The goals of the NRIS (see Box 6-5 below) include a
broader range of interventions ranging from tenure security (legalization of land), provision of basic
physical services/infrastructure, small-scale and incremental housing improvements, social housing
construction, staircase housing models (where beneficiaries are supported in moving gradually from the
lowest quality informal housing through different housing levels to fully finished formal housing units)
and extension of the housing allowance scheme.
The broader goals in the NRIS resonate strongly with the evolution of worldwide thinking on slum
upgrading and incremental housing approaches such as sites and services.
BOX 6-5: HOUSING GOALS FROM THE SLOVAK NATIONAL STRATEGY FOR
ROMA INCLUSION (2012), SECTION D. 2. 4: HOUSING
The overall goal of the strategy regarding housing conditions is to improve access to housing, with
special emphasis on social housing and the need to support abolishing segregation in housing, while fully
exploiting the funds that have been made available recently in the context of the European Regional
Development Fund. In addition, the NRIS aims to bridge the gap between the majority population and the
Roma in access to housing utilities (such as water, electricity and gas), and reduce the proportion of
shacks and illegal dwellings by 25 %.
Partial goals:
ï‚· Disposal of shacks and dwellings unsuitable for living (under the Construction Act and applying
minimum housing standards) in marginalized Roma communities, and establishing mechanisms for
supporting legal housing for citizens whose dwellings have been disposed of.
99
Cities Alliance: Cities Without Slums. The case for Incremental Housing. Cities Alliance Policy Research and
Working Papers Series No. 1, June 2011.
127
ï‚· Analyze chances of repairing the existing apartments in cases where the apartments and/or houses
in question are in such a technical condition, which could endanger health or life of their residents.
Find chances for intervention in emergency cases caused by natural calamities (such as storms, floods)
and fire.
ï‚· Find ways of legalizing and/or disposing of illegal constructions, while giving their inhabitants an
opportunity to acquire legal housing.
ï‚· Introduce financial and legislative tools enabling settlement of land title for the purpose of
building rental social apartments.
ï‚· Ensure completing infrastructure and equipment [access to drinking water, sewage/septic, gas and
electricity] of segregated and separated Roma settlements in Slovakia.
ï‚· Ensure the allocation of funds for the Program of Housing Development that serves to channel
subsidies for procuring standard and low-standard housing. Explore the possibility to use EU funds.
ï‚· Indicator: Number of subsidies granted by the Ministry of Transport, Construction and Regional
Benchmark: Number of low-standard apartments constructed prior to 2010 is 2,890.
ï‚· Introduce and implement a program of gradual assisted housing as a social service . Benchmark:
Currently this program is operational in two towns in Slovakia.
 Prepare legislative framework for providing housing benefit in such a way that it would – in
justified cases – be paid directly to the apartment manager or another provider of services
associated with housing, and also that the circle of recipients be broadened to include applicants
not assessed as citizens in material need, although their income is lower than the sum of the subsistence
minimum. Allowance will, however, be strictly limited to the purpose of covering costs associated with
housing.‖
Analysis of the NRIS suggests that it lacks the specific actions and associated budget needed to
realize these goals. The NRIS‘ housing chapter relies heavily on the Roma Decade Action Plan‘s targets
and does not go beyond the Plan either in terms of goals or in targets. The NRIS instead supports the
status quo more than it does the creation of an inclusive housing policy for the Roma for the following
reasons. First, there are no incentives (or demands) to go beyond creating the sympathy and the step-by-
step education of local authorities to understand options and the responsibility for local solutions, whereas
social housing, spatial planning, and social service delivery is delegated (to a large extent decentralized) to
local level. Second, no steps are foreseen to seriously improve housing affordability (and through this
housing consumption and quality of life). And, third, in a few localities, local practice goes beyond the set
of tools mentioned in the action plan, which suggests that building on good practice in the field is still
problematic, and substantial changes in policy design have not been thought through.
Since June 2010, ERDF regulations have made it possible to use EU Structural Funds to improve
marginal groups‘ housing conditions by desegregation within an integrated approach, that is, a
combination of various soft measures and hard infrastructure investments 100 . The organizational
setting delegates tasks relating to the horizontal representation of Roma related issues and needs to the
Plenipotentiary Office. In the field of housing, the Plenipotentiary Office has a coordination role in
designing future pilot interventions for housing, co-funded from EU funds. At the time of drafting this
report, a new pilot scheme that would have targeted social housing provision via new ―regular‖ local and
integrated social housing construction had been stopped. The pilot should have been a demonstration for
the absorption of ERDF Structural Funds for housing for marginalized groups.
100
http://www.euromanet.eu/upload/86/94/Guidance_note_Housing_interventions_art__7_2_ERDF.pdf
128
Thus far, the Government has relied on two main policy tools to promote housing inclusion:
construction of low-standard social housing and the housing allowance.
Construction of low-standard social housing. The Ministry of Transport, Construction and Regional
Development is responsible for designing and implementing public housing construction programs
targeting also Roma households and this program, which is funded by the State Housing Development
Fund, is the result. The objective is to replace some of the most precarious Roma housing with these
units. The units are small and unfinished in order to reduce costs and increase affordability; however,
despite these cost reductions, only 2,900 units have been built during the past ten years. Each unit costs
an average of 20,000 Euros. The average annual rate of construction (290 units per year) is not sufficient
to even keep up with natural increase and new household formation among the Roma population living in
ruined houses / slums. Flats are rented for approximately 100 Euros monthly, which is more than one
quarter of the average Roma household budget, without taking into account the cost of utilities. Although
housing allowances of about 90 Euros cover almost all of the rent, the residual rent and the additional
utility costs are often unaffordable to these very poor households. Another problem is that unfortunately,
the low-standard social housing construction program relies on land provided by the municipalities and in
practice this has resulted in further segregation for Roma families who receive these new units.
The housing allowance. This benefit is linked to the Basic Material Needs Benefit providing an additional
payment to cover housing and utility costs. One-third (approximately 24,000) of all Roma households
receive housing allowances. The housing allowance has much better reach than the program to build
low-standard social housing; however, the program rules inherently exclude the neediest Roma families.
Specifically, potential beneficiaries must legally rent or own a formal dwelling, which means the unit must
be in an area designated for residential use where land tenure is clear and where the building complies
with local building codes. Although data about tenure status are not available, in 2004 about one-third of
Roma households who lived in separated / segregated communities lived in dwellings built without proper
planning permissions (i.e., informal). 101
6.3.3 INCREMENTAL SOLUTIONS AT THE LOCAL LEVEL
A number of local level solutions in Slovakia go beyond low-standard social housing construction
and provision of the housing allowance to needy households. For example, some local initiatives take
on land title regularization while others supplement low-standard social housing and housing allowances
with ―soft‖ support such as education, community social work, health services or assistance with
microfinance. NGOs (such as ETP Slovakia) as well as municipalities (for example Dolny Kubin) have
implemented these kinds of activities using funds from the national government, donors, EU Funds and
other sources singly or in combination.
Some efforts were made to include unregulated (i.e., informal) areas into the spatial planning
documents of the municipalities, which is a necessary step for regularizing these areas and
eventually providing land title to the Roma residents. Funding was provided to cover a part of the
local authorities‘ costs to update their general urban plans or master plans, and also to facilitate
―finalizing‖ clearing the restituted lands‘ titles for example by swapping land in order to unite plots. In
case Roma settlements were positioned on private land, municipalities could initiate swapping these plots
for plots in their ownership, or, to clarify who was the effective owner of such plots (many times there are
multiple owners and very complicated ownership conditions) and launch negotiations among the ―users‖
and the owners of the plots. Due to lack of public funding, however, the program was stopped, and
101
FRA RAXEN Report Slovakia (2009), p. 24.
129
benefited only a few settlements. 102 The Box below profiles one example of integrated incremental
approaches within Stara Lubovna, a small city near Presov.
BOX 6-6: STARA LUBOVNA: COMMUNITY PARTICIPATION IN INTEGRATED (AND
INCREMENTAL) IMPROVEMENTS IN LIVING CONDITIONS IN SPATIALLY
SEGREGATED ROMA COMMUNITIES
Of the 16,000 people living in Stara Lubovna, approximately 2000 are Roma, of whom about half are
housed in the ‗Podsadek‘ part of the city, below the historical castle ruin area. In Podsadek, most homes
are built of solid materials and some are standard apartment buildings; however, wooden shacks and
insecure extensions can be also found. Many houses lack windows and proper roofs. There is no sewerage,
and most families collect water from public wells. More than half of the residents are children.
Unemployment is close to 90%.
In the community center of Podsadek, the NGO ETP has coordinated various programs in close
cooperation with the municipality of Stara Lubovna since 2003, to improve conditions in Podsadek. The
staff of the local community centre, in part employed by the municipality, organizes after-school
activities, early childhood education programs, health consultations (e.g. drug prevention), vocational
trainings, and, above all, intensive social work. The staff reach out to the different age groups from young
children to adults. Linked to ETP‘s micro-finance and micro-loan program, approximately 150 families
(adults) are offered financial education, and access to micro-loans, which they have invested in improving
their housing quality and energy efficiency by installing new windows, doors, and insulation. The
precondition for accessing micro-loans for housing improvements has been regularizing the land, which
the municipality has supported. In some cases, local Roma residents who move have been supported and
followed up with by social workers (in one case a young couple moved to a social dwelling in the
integrated part of the city).103
Other types of local interventions have resulted in incremental changes, such as immediate
improvement in living and health conditions. Examples include the insulation of walls, windows and
roofs to improve energy efficiency, providing for non-wood heating or in-house access to potable water,
launching of the regularization the land title – sometimes with the help of microloans. These types of
programs can frequently be leveraged to provide job training opportunities and actual work experience for
Roma households, as discussed in the Box below. However, there is a lack of institutional support for
systematic peer-learning and scaling-up of promising initiatives. The Box below describes how existing
programs can be used to provide job training opportunities and actual work experience for Roma
households.
BOX 6-7: LOCAL INITIATIVES TO BUILD SKILLS AND COMMUNITY
INFRASTRUCTURE AT THE SAME TIME.
Some mayors in Eastern Slovakia have taken advantage of the municipal works activation scheme, which
provides allowance to people receiving the BMN in return for work. Most often this work consists of low
skill work like street sweeping. Some mayors, however, have realized that with some on the job training in
basic construction skills, the municipalities can cheaply improve municipal infrastructure and at the same
102
http://www.fao.org/DOCREP/007/AD878E/AD878E05.htm
103
Author site visit in March 2012.
130
time provide unemployed Roma receiving the activation allowance an opportunity to learn employable
skills. Inspired by such examples, ETP in collaboration with Habitat for Humanity started th e ―Sharpen
Your Skills‖ course. In Hodejov, for example, a Community Center was renovated whereby most of the
technical labor was provided by local citizens, who were trained by project tutors. The course consists of a
theoretical and a practical part where participants learn to recognize various building materials, use of
construction tools, and working methods. The tutors carefully explained and then demonstrated each
component. 104
6.4 POLICY RECOMMENDATIONS
The recommendations recognize the desirability of incremental approaches that will reach larger
number of poor Roma households sooner with the expectation that improvements will be gradual
and take place over a period of time. This will require broadening the range of tools (projects and
programs) available to the government and thus to go beyond social housing and housing allowances. The
recommendations here are organized in two broad areas: (1) improving conditions – in situ - for the worst
off Roma living in slum areas, and (2) helping poor families move into better housing.
The Government‘s strategy should prioritize the large group of Roma households in the worst living
conditions (i.e., the 27,000 households living in ruined housing and slums) and can be designed so as to
allow for measurable and immediate improvements in basic living conditions (including access to
infrastructure services, incremental housing improvement, access to state benefits, etc.). This is important
to demonstrate positive progress in counteracting the impacts of unhealthy and dangerous housing
conditions and building on the participation, knowledge and resources of the target population.
If the Government is to show measurable improvement during the next several years, it will be
particular important not to have new groups of households join the category of those living in
ruined and slum conditions. For example, in Lunik IX, tearing down heavily deteriorated buildings
without having an acceptable place to move to for the affected families (formal and informal) will create
an even larger policy problem for the Government as these families will now be fully homeless. To
prevent the further growth of impoverished segregated areas and slums, leveraging different public
resources in an integrated manner will be important. Both sets of recommendations go beyond the NRIS,
because they suggest the importance of thinking beyond the existing policy tools. They take into account
the currently limited availability of public finance resources, while keeping in mind the substantial role
that could be played by current and future EU funding in social inclusion and desegregation and
improving living conditions for the most disadvantaged Roma households.
The new programming period 2014-2020 will have strong ex ante conditionalities that focus on
social inclusion. The expectation is that approximately 20% of all Structural Funds will have to be
dedicated to programs that target inclusion. In the case of Slovakia and the other Decade of Roma
Inclusion countries, Roma inclusion is an essential element of any social inclusion process.
Accomplishing this will require strong political commitment and leadership at the highest level of
government.
104
http://www.etp.sk/wp-content/uploads/2011/08/From-our-lives.pdf
131
6.4.1 POLICY MEASURE 1: FOCUS ON IMPROVING LIVING CONDITIONS FOR POOR
HOUSEHOLDS IN SITU; I.E. UPGRADING
Focus on improving living conditions for poor households in situ; i.e. upgrading. The main challenges
for improving living conditions within a poor community include a lack of legal status of land, building
and people, lack of infrastructure, poor quality of housing, limited social capital within the communities,
physical isolation, and lack of economic opportunities. Specific recommendations for improving living
conditions in situ include the following:
POLICY MEASURE 1A: FACILITATE LEGAL INCLUSION
ï‚· Provide technical assistance to municipalities to support regularization of informal areas through
incorporation of informal areas into city/village plans
ï‚· Allow residents to legally register where they live regardless of type of structure
ï‚· Encourage municipalities and NGOs to undertake steps toward titling and tenure security. For
example, help households register in cadaster, formalize a lease, privatize public land, facilitate
land purchase from private owner or land swapping among owners, etc.
POLICY MEASURE 1B: IMPROVE ACCESS TO BASIC SERVICES
ï‚· Access existing programs, such as the State Housing Development Fund for the provision of
public utility infrastructure, to provide basic infrastructure to poor communities (where feasible)
ï‚· Support families in managing their utility (and other) debts and in regaining access to existing
services
ï‚· Involve beneficiaries in infrastructure upgrading, which builds ownership, creates job
opportunities, and develops skills.
POLICY MEASURE 1C: IMPROVE SUPPORT TO POOR RESIDENTS TO MAKE HOME
IMPROVEMENTS
ï‚· Provide financial literacy and home improvement training and materials to permit residents to
plan for and improve their homes including energy efficiency improvements.
ï‚· Delink eligibility for housing allowance from formal residence status or develop a new subsidy
program aimed at the poorest
ï‚· Consider limiting housing allowances so that they can be used only for costs related to housing
and utilities, and related expenses.
ï‚· Increase availability of micro-finance for home improvements, and assess the experience with
using housing allowances to secure repayments of microloans for housing improvements.
6.4.2 POLICY MEASURE 2: HELP POOR FAMILIES MOVE INTO BETTER AND
INTEGRATED HOUSING.
Help poor families move into better and integrated housing. A large number of Roma households live
in very difficult conditions, which in many cases will mean creating the opportunities to move into
different dwellings. For example, households living in formal apartment buildings that become
132
uninhabitable, overcrowded families, or new households that are being formed. Moving into new housing
is challenging since there is a very limited supply of affordable housing: the relatively small number of
private rentals are largely informal, landlords may discriminate against Roma families, and the supply of
affordable social housing remains below the level of new household formation.
POLICY MEASURE 2A: DIVERSIFY GOVERNMENT PROGRAMS TO ENABLE POOR
HOUSEHOLDS MORE CHOICES THAN JUST LOW-COST SOCIAL HOUSING TO
ACCESS BETTER DWELLINGS:
ï‚· Create incentives to use vacant housing in integrated areas (rent or ownership)
ï‚· Expand the availability of private formal rentals (requires revising the legal framework)
ï‚· Provide strong social support for Roma families transitioning to new neighborhoods, for receiving
majority communities, and in many cases anti-discrimination campaigns
ï‚· Redesign the housing allowance so it can be paid directly to landlord
POLICY MEASURE 2B: INVOLVE FAMILIES IN THE CONSTRUCTION OF NEW
HOMES, WHICH CAN BE MORE COST EFFECTIVE THAN BUILDING FINAL
FINISHED UNITS
ï‚· Provide families with well located land sites and building materials
ï‚· Provide technical assistance to the families in construction
6.4.3 CROSS-CUTTING RECOMMENDATIONS
In addition to these specific recommendations, there are also several cross-cutting
recommendations:
ï‚· Make sure different Roma inclusion programs are synchronized horizontally for a common goal
(i.e. across different housing programs but also programs in education, employment, health, etc.,
or commingling ERDF and ESF funds)
ï‚· Recognize up front the need for sustained and long-term interventions to reach long-term goals
using incremental measures
ï‚· Target the poorest communities and work with them on their problems; this will require revisiting
subsidy programs to make sure the poorest households are being reached
ï‚· Seek integrated approaches
ï‚· Support desegregation (i.e., physical, social and economic inclusion)
ï‚· Apply community development type approach with skilled community worker assigned to
targeted communities
Improve community targeting by taking advantage of (a) the poverty map being produced by the National
Statistics Office and the World Bank that will show small area poverty estimates of Slovak communities,
regardless of ethnicity, and (b) the Slovak Roma Atlas being updated by UNDP (see the Monitoring and
Evaluation section for details).
133
6.6 BIBLIOGRAPHY
Acioly Jr. Claudio. UN-HABITAT, Presentation on Housing Sector Profile, July 28, 2011
Atlas of Roma communities 2004: http://www. romovia. vlada. gov. sk/3553/atlas-romskych-komunit-
2004. php
Cities Alliance. Cities Without Slums. The case for Incremental Housing. Cities Alliance Policy
Research and Working Papers Series No. 1, June 2011.
Cities Alliance. http://www.citiesalliance.org/About-slum-upgrading.
ETP. Accessed 2012. http://www.etp.sk/en/moldava-projekt-vystavby-novych-domcekov/#more-3037,
http://www.etp.sk/wp-content/uploads/2011/08/From-our-lives.pdf
European Commission. DG Regional Policy. 2011. Guidance Note on the Implementation of Integrated
Housing interventions in Favour of Marginalised Communities under the ERDF.
http://www.euromanet.eu/upload/86/94/Guidance_note_Housing_interventions_art__7_2_ERDF.pdf
Eurostat. http://epp.eurostat.ec.europa.eu/statistics_explained/index.php/Glossary:Overcrowding_rate
FAO. 2001. Land Reform in Eastern Europe, Western CIS, Transcaucuses, Balkans, and EU Accession
Countries. http://www.fao.org/DOCREP/007/AD878E/AD878E05.htm
FRA- European Union Agency for Fundamental Rights. 2011. The State of Roma and Traveller Housing
in the European Union: Steps towards Equality Summary Report. http://fra.europa.eu
________. 2009. Housing conditions of Roma and Travellers in the European Union Comparative report.
http://fra.europa.eu
________. 2009. Case studies on specific housing initiatives for Roma and Travellers in the following
countries: Czech Republic, Hungary, Ireland, Slovakia, Spain and United Kingdom. http://fra.europa.eu
________. 2009. Case study Roma housing projects in small communities, Slovakia. http://fra.europa.eu
Open Society Foundation. 2009. Supplementary Background Document to the VADEMECUM
Improving housing conditions for marginalized communities, including Roma in Bulgaria, Czech
Republic, Hungary, Romania and Slovakia through the absorption of ERDF.
http://lgi.osi.hu/cimg/0/1/3/9/3/vademecum_supplementary_1703.pdf,
________. 2011. VADEMECUM Improving housing conditions for marginalized communities,
including Roma in Bulgaria, Czech Republic, Hungary, Romania and Slovakia through the absorption of
ERDF. Methodological Guidance. http://lgi.osi.hu/cimg/0/1/4/2/3/housing_vademecum.pdf
RAXEN NFR Slovakia. March, 2009. Thematic Study: Housing Conditions of Roma and Travellers.
http://fra.europa.eu/fraWebsite/attachments/RAXEN-Roma%20Housing-Slovakia_en.pdf
Roma Education Fund. 2009. School as a Ghetto.
http://www.romadecade.org/files/ftp/School%20as%20Ghetto.pdf
134
Slovak National Strategy for Roma Inclusion, as submitted to the European Commission early 2012,
Section D.2.4. Housing. http://ec.europa.eu/justice/discrimination/files/roma_slovakia_strategy_en.pdf
Statistical Office of the Slovak Republic: 2011 Census. http://portal.statistics.sk/files/table-3.pdf.
________: 1961 and 1971 Censuses. http://sodb.infostat.sk/sodb/eng/1848-2001/tab.III.2-2A_en.htm and
http://sodb.infostat.sk/sodb/eng/1848-2001/tab.III.2-2G_en.htm
135
7 HEALTH
This chapter provides a detailed assessment of gaps that exist in health outcomes between Slovak Roma,
their non-Roma neighbors, and the general population, despite the fact that health insurance coverage is
nearly universal, as well as policy recommendations to improve health outcomes and reduce the gaps.
Following a discussion of main data sources, it assesses several key demographic outcomes and disease
burden, before providing an assessment of the determinants of health outcomes among Slovak Roma,
including not only health infrastructure and services, but assessing also prevention and risky behavior,
and using health services effectively. The chapter then reviews the current Slovak health policy
environment, before concluding with specific recommendations that are grouped under those (1)
promoting more effective use of existing health services, and (2) using entry points outside the immediate
health system to promote better health outcomes.
7.1 INTRODUCTION
Health services in Slovakia are free of charge, health insurance coverage is nearly universal, and
most of the population, including Roma, live within a few kilometers from a health facility. Yet
Roma suffer worse health than the non-Roma population. Previous literature on the health status of
Slovak Roma points to a higher burden of infectious and chronic disease. This is generally supported by
self-reported health outcomes in the regional Roma survey and focus group discussions that were
conducted. Furthermore, research estimates have shown that poor health translates into life expectancy
among Roma that is much lower (an estimate of 15 years for Slovakia) than non-Roma. This is consistent
with the regional survey finding that only 2% of Slovak Roma are older than 65. The findings in this
chapter suggest three complementary reasons that help explain this poor health status.
First, the very poor living conditions described in the housing chapter are obvious contributors to
infectious disease, diarrhea and respiratory disease among children. Recall that housing conditions
for Roma households are much worse than for nearby comparable households. About 30% of Roma
households live in a ruined house/slum, while this is true for only 4% of comparator households. Only
about half of Roma households have indoor sanitation (toilet, bathroom, sewage connection) while about
90% of nearby non-Roma families have these amenities. One in four Roma households reports irregular or
no collection of solid waste. And, half of Roma households use wood for cooking and 85% use wood or
coal for heating, substantially more than the use by non-Roma comparator households.
Second, a high burden of chronic disease is consistent with high risk behavior such as high smoking
rates, both for men and women, poor diet and low levels of physical activity, as well as a high rate of
teen pregnancy. For example, the analysis shows that more than 60% of Roma adults smoke regularly,
compared to about 27% of the general population, and 40% among non-Roma neighbors. The smoking
prevalence among Roma women is three times higher than that among Slovak women nationally. Yet,
there are very few examples of campaigns directed at preventing smoking among Roma, health diets etc.
Roma women reportedly also continue smoking during pregnancy, which will also contribute to poor
reproductive health outcomes. Furthermore, many Roma women have children early. According to data
from the UNDP (2010) survey of the Slovak Roma population, among women aged 15-19, 19% has given
birth. Among women aged 20-24, 64% has given birth. At the same time, practically all women give birth
at the hospital, providing an excellent opportunity for awareness creation about infant and child care,
family planning, and reproductive health.
Third, poor health outcomes can be caused by ineffective use by many Roma of the available health
services. The regional Roma survey finds that most of the population, including Roma, live within a few
136
kilometers from a health facility, and most Roma report being satisfied with the health services received.
Still, 45% of Slovak Roma does not seek health care when they actually need it. Half of those that do not
seek needed care say it is unaffordable to them. In follow-up interviews with focus groups, including
health workers and some Roma community members, the financial barrier was explained; money needed
for transport could be prohibitively expensive for a Roma family. Another 20% wants to wait to see if the
affliction improves by itself. Respondents also indicate that medication is often too expensive and forego
treatment as a result. In part this may reflect that many do not understand the benefits of health insurance,
and instead incur high out of pocket expenses or do not see a doctor for relatively minor check-ups. As a
result of misunderstandings and ‗waiting to seek care‘, the use of emergency services and ambulances is
very high, and costly to both the health system and the patient. Roma Health Mediator programs address
these by assisting Roma in getting health insurance cards and explaining the use as well as accompanying
them to seek health services. However, these programs are small scale and their sustainability is uncertain.
Health, nutrition, and population policies play a fundamental role in economic development and
poverty alleviation. Empirical studies have documented a positive correlation between health and
national income around the world, but this relationship is complex and runs both ways. Higher income can
allow for greater access to inputs that improve health, such as food, clean water and sanitation, education,
and medical care. And in turn, improved health and nutrition can contribute to economic growth through
better educational outcomes, increased productivity, and increased investment, creating a virtuous
circle.105
Improving early child health and nutrition has particularly large benefits. A child born from a
healthy well-nourished mother has a much better chance at a healthy childhood, educational achievement,
and greater labor productivity as adults. Recent evidence also shows that children that are breastfed are
less likely to develop chronic disease in later life; for example, a 4% reduction in obesity risk with every
month of breastfeeding [Singal and Lanigan, 2007]. Children born with low birth weight (<2500 grams)
are at higher risk of childhood stunting, more prone to infectious disease, and show lower educational
achievement and adult productivity. They are also at elevated risk of chronic disease later in life due to the
mismatch between in-utero environments and later environments in which food is more abundant.106
The recently approved Strategy of the Slovak Republic for the Integration of Roma (NRIS) by 2020
includes a clear focus on health issues. The draft Strategy highlights the need to improve access to
quality health services for disadvantaged and vulnerable populations. It focuses on the need for more
health promotion and prevention as well as financial protection, with provisions for including essential
drugs to health insurance packages. In addition, it proposes interventions that would disproportionally
benefit the Roma population such as the free Hepatitis A vaccination, dental care, travel allowances, and
increased support to reproductive health. It also foresees more data-collection on socially excluded sub-
populations in order to improve targeting of interventions and resource allocation. At this time, however,
there remains no indication of increases in budget or resources to support these policies or programs.
The purpose of this chapter is to present findings from the recent regional Roma survey on Roma
health status and performance of the health system for the Roma population in Slovakia. A review
of the current policies and programs, as well as the recommendations of the draft Strategy mentioned
105
Several recent reviews have explored the relationship between health and income; see Kremer and Glennerster
(2012), Bleakley (2010a), Spence and Lewis (2009), Currie (2009), Strauss and Thomas (2007).
106
The mismatch is often between what was available to the mother while pregnant and during early life years with
greater abundance of food later in life. The so-called Foetal Original of Adult Disease (FOAD) theory suggests that
part of the Double Burden Malnutrition problem, whereby under- and overnutrition occurs in the same population,
community, family and even individual over a life-time, is due to a mismatch between the environment of the womb
and that of the world outside when the child is born (Barker Hypothesis in Barker 1998).
137
above is carried out in light of these findings. The policy recommendations below are consistent with the
overall Strategy recommendations, and aim to (a) highlight more specific programs and priorities areas,
and (b) provide a number of complementary entry points that are important to consider.
7.2 METHODOLOGY AND KEY DATA SOURCES
The findings in this chapter rely primarily on the regional Roma survey (the UNDP/World
Bank/EC regional Roma survey, 2011), and comparisons with data on the general population. It
should also be noted that throughout this chapter, we present two comparison populations: 1) the "general
population," which is the national population of each country in the EU Eurobarometer data, and 2) ―non -
Roma,‖ who are non-Roma neighbors of the vulnerable Roma sampled in the Regional Survey. The
chapter will use these terms to refer to specific comparison populations.
All estimates related to the general population of Slovakia are based on Eurobarometer survey data
collected in either 2006 or 2009. Eurobarometer (2006) is data from the 66.2 round of the Eurobarometer
surveys. The survey queried respondents about their general health and quality of life, current or past
health problems, and the location of body pain. Respondents were also asked about treatment for chronic
illness, medical tests or health checkups, and recent changes in health behavior. Demographic and other
background information includes age, gender, origin of birth (personal and parental), and marital status. In
all, Eurobarometer 66.2 interviewed 28,585 citizens aged 15 and over of the 25 countries in the European
Union after the 2004 enlargement, remaining Accession Countries (AC) Bulgaria and Romania, Candidate
Country (CC) Croatia, and among the Turkish Cypriote Community (TCC). National estimates on
smoking and alcohol consumption were based on data from Eurobarometer 72.3, conducted in 2009. This
round of the survey interviewed 30,292 citizens in the 27 countries of the European Union.
When comparing adult estimates across these populations, the figures are also adjusted for age in
order to identify the disparities or gaps between Roma and non-Roma. The Roma population
structure is different from that of the general population or non-Roma neighbors – typically, Roma are
younger on average and have fewer elderly. As a result, comparing simple averages across the different
populations would not account for these differences in age structure.
138
FIGURE 7-1: POPULATION AGE STRUCTURE FOR ROMA AND NON-ROMA NEIGHBORS IN SLOVAKIA
A. Roma B. Non-Roma Neighbors
80+ 80+
70-74 70-74
60-64 60-64
50-54 50-54
40-44 40-44
30-34 30-34
20-24 20-24
10-14 okt-14
0-4 0-4
15 10 5 0 5 10 15 15 10 5 0 5 10 15
Percent Percent
Female Male Female Male
Source: UNDP/World Bank/EC regional Roma survey (2011).
7.3 POPULATION AND HEALTH OUTCOMES AMONG THE SLOVAK ROMA
7.3.1 AGE STRUCTURE AND FERTILITY
Among vulnerable Slovak Roma, family formation starts at a very early age, and there is a high
dependency ratio. Patterns of entry into marital unions among women in the most vulnerable Roma
communities differ substantially from those in the general population (Figure 7-2). In the regional survey,
about 15% of Roma women between the ages of 15 and 19 years were married (7% in the general Slovak
population), and about 65% between the ages of 20 and 24 years (compared to only 13% in the general
population). The mean age at first marriage among Roma women who entered into traditional marriages
(20 years of age) was the same as that among those who were married in church or a municipality office.
Households that participated in the Regional Survey also reported preferences on timing of vital events,
and the mean desired age of marriage for girls was 19.5 years. The desired age for men was slightly higher
at 20 years. Interestingly, preferred age of initiation of sexual life for men and women (18 years for both)
was slightly lower than the desired age of entry into marriage. Desired ages of marriage were slightly
higher among non-Roma neighbors (a mean of 22 and 21 years for men and women, respectively), while
desired age of initiation of sexual life was roughly the same as that among the Roma. Lastly, the Roma
population has a higher dependency ratio (whereby more community- or family members that are not in
the labor-force depend on members in the labor-force or gainfully employed), which, coupled with low
employment rates, exacerbates poverty levels.
139
FIGURE 7-2: MARRIAGE RATES AMONG SLOVAK WOMEN
Source: Roma: UNDP/World Bank/EC regional Roma survey (2011); General Population: Eurobarometer (2006).
The figure represents the fraction of women who have ever been married, for vulnerable Roma and the general
population.
Childbearing starts early and fertility remains high. The Regional Survey did not collect information
on reproductive behavior among women. However, According to data from a similar survey one year
earlier – the UNDP (2010) survey of the Slovak Roma population - the average number of children is 3.7
among women aged 45-49 and 3.6 among women aged 35-39. The median age at first birth is stable over
time – 19 or 20 years of age for each of the five-year cohorts from 25 to 49. Data from the 2011 regional
Roma survey on preferences about timing of first birth also reflect this desire to initiate childbearing early:
the mean desired age to start having children was 20 years for females. Among young women aged 15-19,
19% has given birth. Among women aged 20-24, 64% has given birth (note: this is similar to the reported
marriage rates). Among women aged 45-49, 94% has given birth. This is consistent with findings from
older studies. In 2002, Roma women living in ‗partially integrated‘ and ‗segregated‘ communities had a
mean age at first birth of about 20 years (Mézsáros and Vaňo 2004; Šprocha 2006). Roma women living
in ‗partially integrated‘ communities had a total fertility rate of 3 children per woman, and those living in
‗segregated communities‘ had a total fertility rate of 5.2 children per woman, both high in comparison
with an average of 1.19 children per woman in Slovakia as a whole. In another study in ‗segregated‘
communities in Slovakia, 30% of Roma women had their first child by age 18 (Kumanová, DÅ£ambazoviÄ?
2002). In a household survey across Europe, the fraction of women between the ages of 14-16 years that
had given birth for the first time is three times higher among the Roma than among the non-Roma (FSG
2009). Monitoring data from the RHM programs implemented by OSF in Slovakia show that 49 percent of
Roma mothers were under 18 years of age when they had their first child (6 percent were not even 15
years old) (OSF 2012, Internal Memo).
7.3.2 LIFE EXPECTANCY
Life expectancy among the Roma is considerably lower than that among the general population in
Slovakia, although more recent data is needed. Rigorous data on life expectancy gaps between Roma
and the general population are rare because administrative data are not disaggregated by ethnicity in most
countries. Data that does exist is often estimated or extrapolated from household survey data or official
statistics, and may be based on small sample sizes. Table 7-1 presents Roma life expectancy data for
countries in Eastern Europe, available since 2000, including their source and methodology. Although the
140
data are derived from various sources and should be interpreted with caution, the evidence is suggestive of
lower life expectancy among Roma. The gap in Slovakia of an estimated 14 years is particularly large, but
the data concerns and the year (2000) should be kept in mind.
TABLE 7-1: LIFE-EXPECTANCY FOR ROMA AND NON-ROMA IN SELECTED COUNTRIES
Country Year Life Life Exp Life Exp Source Methodology
Exp Roma Gen Pop
Gap
Moldova 2006 3 65 68 UNDP survey Estimate based on infant deaths
(UNDP-Moldova from household survey, and
2006) correlation between life
expectancy and infant
mortality.107
Serbia 2005 10 63 73 108 2005 Mortality No methodology reported. Only
Database, Serbian reported life expectancy gap.
Office of Statistics
(Bogdanovic et al.
2007)
Romania 2003 ~6 64 70 UNDP survey Estimated from life expectancy of
(UNDP 2003) countries with similar infant
mortality rate. IMR estimate from
separate survey (RHS 1999).109
Hungary 2002 10-15 56-61 71 Unknown (Kovac Not methodology reported.
2002)
Bulgaria 2001 5 67 72 1992 and 2001 Not methodology reported. Data
census, Bulgaria only for people who declared
(UNDP 2003) ethnic affiliation.
Slovakia 2000 15 59 73 110 UNDP 2004, Zoon Not methodology reported. Only
2001111 reported life expectancy gap
Source: World Bank
An important factor contributing to the gap in life expectancy are higher rates of infant mortality
among the Roma. Once again, vital statistics disaggregated by ethnicity do not exist, but available data
suggests that most Roma populations in Europe have higher infant death rates than non-Roma. The most
recent information from the Slovakia statistical yearbook shows the regions were the Roma population are
107
The Roma infant mortality rate was calculated as number of infant deaths (40) over total number of live births
(1,386). No reference to the range of years (e.g., births within the last 10 years) over which the births occurred.
108
Non-Roma life expectancy from WHO European Health for All Database (HFA-DB 2010)
109
Unclear which IMR was used for the calculation. UNDP 2003 cites that the Roma infant mor tality rate is ―roughly
three times higher than the national average‖ and shows IMR figures of 27.1 for the Romanian population and 72.8
for the Roma population for children born in the last 5 years, between 1994-1999. The RHS 1999 final report only
presents figures for children born in the last 10 years, between 1989-1999, with IMR figures of 26.9 for the
Romanian population and 50.6 for the Roma population.
110
From WHO European Health for All Database (HFA-DB 2010)
111
Zoon and UNDP only present life expectancy gap for men (13 years) and women (17 years). For overall life
expectancy gap, Roma male-to-female ratio was assumed to be the same as for the general population, 95:100 (UN
Population Division 2000). Source cited was a report by the International Organization for Migration (2000), which
was not found.
141
concentrated have much higher IMR than those regions where few Roma are living. Figure 7-3: Total
numbers of infant deaths in Slovakia in 2008, by region shows the actual number of infant deaths in 2008
in Slovakia. A total of 336 infants died before the age of 1, with over 50% of the deaths from only two
regions: Presovsky and Kosicky. These are also the two regions with the highest concentration of Roma.
While this correlation strongly suggests that disparities in infant mortality exist, more accurate data is
needed.
FIGURE 7-3: TOTAL NUMBERS OF INFANT DEATHS IN SLOVAKIA IN 2008, BY REGION
Source: Slovakia Ministry of Health statistics, 2008
7.3.3 BURDEN OF DISEASE
Evidence from small, local studies indicates that infectious disease prevalence is high among the
Roma. Measles outbreaks have occurred in recent years among Roma communities in Italy, Portugal,
Germany, Greece, Croatia, Bulgaria, Romania, Serbia and Poland (Loewenberg 2006; Orlikova et al.
2010; Seguliev et al. 2007). Polio has been found to be higher among the Roma as well (Kojouharova et
al. 2003). Evidence from Bulgaria also shows high TB prevalence (Schaaf 2007). A high prevalence of
infectious disease can be linked to low vaccination coverage, low utilization of basic health services,
crowded and unsanitary housing conditions, and poor water and sanitary conditions.
Although the Regional Survey collected no data on infectious disease prevalence in vulnerable Roma
communities, interviews conducted with health practitioners indicate that certain infectious diseases
are common in these communities. Health assistants working in Roma settlements reported that scabies
was a common problem among children, and infectious respiratory ailments, urinary tract infections and
jaundice were common among adults. In one settlement, there were a high number of cases of infection of
the bronchi and lungs among children who then had to be hospitalized when the (untreated) condition
worsened. One health coordinator interviewed in the Focus Group Discussions noted that cases of flu and
other infectious diseases tended to rise in autumn. Another interviewed doctor working in three villages
noted that jaundice had been an issue in the past, but a two-year long vaccination project was successful in
eliminating the disease.
Self-reported health outcomes suggest that Slovak Roma suffer disproportionately from chronic
diseases. Figure 7-4 (a) shows self-reported chronic disease prevalence among vulnerable Roma, non-
Roma living nearby, and the general population in Slovakia, adjusted for age.112 After adjusting for the age
structure of the Slovak Roma, 22% of Roma adults between the ages of 15 and 70 years reported suffering
112
Roma and non-Roma estimates are from the Roma Regional Survey (2011); national estimates from
Eurobarometer (2006).
142
from long-standing disease, a few percentage points lower than in the general population of Slovakia
(29%). Similarly, a quarter of Roma adults reported that they experienced restricted mobility due to a
disability in the past 6 months (not shown), compared to 31% of adults in the general population. In both
regards, prevalence rates among the non-Roma were slightly lower than those among their Roma
neighbours, although the difference was not statistically significant.
FIGURE 7-4: FRACTION OF ADULTS SUFFERING FROM A LONG-STANDING ILLNESS
A. Overall B. By Gender
Source: Roma and non-Roma neighbors: UNDP/World Bank/EC regional Roma survey (2011); General population
(referred to as ‗national‘ in the figure): Eurobarometer (2006). The figure refers to percentages among the age group
15-70. Data were based on self-reported question items, i.e. it was not verified whether the individuals reporting so
were indeed suffering from health problems.
Self-reported, age-adjusted prevalence of anxiety and/or depression is significantly higher among
the vulnerable Roma, and they also report somewhat higher rates of diabetes, hypertension and
asthma. Hypertension and rheumatism/arthritis are among the most prevalent chronic illnesses reported
(Figure 7-5), both among the Roma and the general population of Slovakia. After adjusting for the
different age structures of the two populations, self-reported prevalence of hypertension and chronic
joint/muscular pain stands at about 27 and 21% respectively among the Roma, a few percentage points
higher than that estimated for the country population as a whole. Self-reported, age-adjusted prevalence of
asthma, respiratory disease and diabetes are slightly higher among the vulnerable Roma than in both the
general population and among the non-Roma neighbours in these communities.
Interviews conducted among health practitioners to follow up on these findings confirm a high
prevalence of chronic disease among the vulnerable Roma. A doctor in the focus group discussions
confirmed that high blood pressure and obesity were common among young Roma, as well as
musculoskeletal disorders, which were concentrated among women. Cardiovascular problems and heart
attacks were also common. The doctor pointed out that these problems were prevalent even among Roma
living in more ―integrated‖ villages, where the socioeconomic conditions of Roma are similar to those of
their non-Roma neighbours. The doctor linked high prevalence of these diseases to sedentary lifestyles
(the Roma under his care did not work) and lack of preventive health behaviors. He also pointed to
smoking and alcohol consumption, which were common among men and women, old and young. One
health assistant who was interviewed mentioned angina as a common ailment, and linked this to risky
health behaviors and lack of preventive examinations among the Roma.
143
Most notably, age-adjusted prevalence of chronic anxiety and/or depression is three times higher
among the vulnerable Roma (15% of adults aged 15-70 years) than in the general Slovak population
(5% of adults aged 15-70 years – Figure 7-5). This disparity in mental illness has serious implications not
only for those who suffer from it but for the entire community, as it puts additional strain on families and
communities. In general, these chronic diseases, including mental illness, require regular medical care
(and associated expenditures), and their management requires health literacy113 among both patients and
their families.
FIGURE 7-5: FRACTION OF ADULTS SUFFERING FROM CHRONIC DISEASES
Source: Roma and non-Roma neighbors: UNDP/World Bank/EC regional Roma survey (2011); General population
(referred to as ‗national‘ in the figure): Eurobarometer (2006). The figure refers to percentages among the age group
15-70.Data were based on self-reported question items, i.e. it was not verified whether the individuals reporting so
were indeed suffering from the mentioned diseases.
Self-reported, age adjusted prevalence of each chronic disease is significantly higher among Roma
women than among men, and the gender disparity is always larger than that observed in the general
Slovak population. While self-reported prevalence of long-term illness in the general population was
roughly the same among men and women, among the vulnerable Roma and non-Roma it was higher
among women than among men (Figure 7-4b). Interestingly, the gender disparity was larger for Roma (4
percentage point difference) than their non-Roma neighbours. Further, prevalence of each specific disease
was always higher among women than among men, with the female disadvantage ranging from a low of
about 5 percentage points in the case of asthma, other respiratory illness, and rheumatism/arthritis, to a
little over 10 percentage points for anxiety/depression and diabetes. The corresponding gender differences
for each disease are much smaller in the general population – ranging from 2 to 6 percentage points
depending on the disease, with self-reported prevalence of asthma and other respiratory illness 2
percentage points lower among women than men. Similarly, Roma women were about 5 percentage points
113
Health literacy is the ability to understand instructions on prescription drug bottles, appointment slips, medical
brochures, doctor‘s directions and consent forms, as well as the ability to negotiate complex health care systems.
144
more likely than men to report restricted mobility due to a health problem in the past 6 months (the
difference was 2.5 percentage points in the general population (not shown)).
In spite of being in poor health, the majority of vulnerable Slovak Roma are satisfied with their
health, but satisfaction declines sharply with age. In general, satisfaction levels are comparable to those
in the national Slovak population. The Regional Survey asked a standard self-reported health status
question, where the respondents report whether they are in very good, good, fair, bad, or very bad health.
About three-quarters of the vulnerable Roma surveyed in the Regional Survey reported that their health
was either good or very good. Significant differences emerge within the population, however, when
responses were disaggregated by age group. Among young Roma adults aged between 15-25 years, over
80% reported that they were in good or very good health, slightly lower than the estimate for the general
population. However, for the population over age 55 years, Roma self reports of health status are higher
than among the general population or non-Roma neighbors. Over 40% of older Roma over the age of 55
years reported that they were in good or very good health, higher than the 32% in the general population.
About a quarter of old adults reported that their health was bad or very bad, comparable to the national
population.
FIGURE 7-6: FRACTION OF YOUNGEST AND OLDEST ADULTS WHO REPORTED THAT THEIR HEALTH
WAS EITHER GOOD / VERY GOOD, OR BAD / VERY BAD
Source: Roma and non-Roma neighbors: UNDP/World Bank/EC regional Roma survey (2011); General population
(referred to as ‗national‘ in the figure): Eurobarometer (2006).
Happiness in general was high and mirrored satisfaction levels with health, with younger Roma
reporting higher levels of satisfaction with life than older Roma respondents. The Regional Survey
asked respondents whether they were very happy, quite happy, not very happy, or not at all happy.
Roughly three-fourths of Slovak Roma aged 15-25 years reported that they were quite or very happy
(about 81% among non-Roma neighbours). Older Roma – aged 55 years and above – were significantly
less happy – roughly 50% reported that they were not at all, or not very happy. Older adults in
neighbouring non-Roma households, on the other hand, resembled younger non-Roma, with 75%
reporting that they were either quite or very happy. When asked to rate how satisfied they were with life in
general, the differences were not quite so stark. Among the Roma, the average satisfaction score was 6.6
among young adults (with zero denoting complete dissatisfaction and 10 denoting complete satisfaction),
and slightly lower among the elderly. Among the non-Roma neighbours, the mean score was higher at 7.2
among the young adults, and 6.0 among the elderly.
145
7.4 DETERMINANTS OF HEALTH OUTCOMES
Why are Roma people suffering worse health status? There is evidence suggesting that the socio-
economic conditions in which Roma live and grow up expose them to greater risk factors for poor health
outcomes in comparison with their non-Roma neighbors and the general Slovak population.
7.4.1 PUBLIC SERVICES INFRASTRUCTURE
Provision of public infrastructure in vulnerable Roma communities remains poor and inadequate .
Estimates from regional Roma survey show that while the vast majority of vulnerable Roma communities
have regular waste collection in their neighborhood (at least once every two weeks), there is clearly room
for improvement. In Slovakia, one in five Roma households reported that waste was not collected
regularly from the community. Similarly, only half of Slovakian Roma households have piped water
inside their dwelling, and about 40% access water through a public tap or a source at higher risk of
contamination 114 . Compounding the problem, about half of vulnerable Roma households in Slovakia
reported that their residence was not connected to the public sewerage system, and about 40% did not
have showering/bathing facilities inside the dwelling. Each of these factors places the Roma at higher risk
for contracting infectious disease.
Concerns about access to clean water also emerged in a number of follow-up interviews conducted
with health practitioners in Slovakia. Health personnel interviewed in the Focus Group Discussions
reported that Roma had difficulty accessing water in settlements with public wells or taps during the dry
season and the winter (when water freezes). In general, the lack of hygiene became particularly
problematic during winter when it was harder to find water for cleaning clothes and washing up. One
health assistant related the occurrence of jaundice and other infectious diseases to the lack of clean water.
These interviews also linked the incidence of infectious disease to generally poor living conditions in
vulnerable Roma communities. Health assistants interviewed pointed to the lack of appropriate clothing
and exposure to cold during the winter months as one reason for higher incidence of infectious disease
during this period. One health coordinator suggested that dirty bed linen, lack of ventilation, and dirty and
wet living conditions contributed to illnesses during the colder months.
114
Data on the quality of drinking water that Roma households use has never been collected.
146
FIGURE 7-7: WASTE COLLECTION AND WATER SOURCES IN VULNERABLE ROMA COMMUNITIES
A. Waste Collection B. Water Sources
Source: UNDP/World Bank/EC regional Roma survey (2011).
FIGURE 7-8: FACILITIES WITHIN HOUSEHOLDS
Source: UNDP/World Bank/EC regional Roma survey (2011).
Improvements are indeed possible: In the Czech Republic, for example, each of the services above
exists in more than 90% of Roma households in vulnerable communities. Similarly in Macedonia, about
90% of Roma households in similar communities have piped water inside the dwelling. It is clear that
these improvements will come about only with the right mix of public investments, and this has not
occurred in Slovakia: While a quarter of vulnerable Roma households in Slovakia reported improvements
in roads and pavements, very few reported any recent improvement at all in access to sewerage systems
and drinking water, or in the quality of housing.
7.4.2 SOCIOECONOMIC CHALLENGES
High poverty, low education, and low employment rates among the Roma population contribute to
the poor health in vulnerable Roma communities. As discussed in Chapter 3, Roma unemployment
rates are very high, in some communities even hundred percent. Roma also have lower levels of
education, with only 20% of adults having finished secondary school. Many Roma live in very poor living
conditions, as described in Chapter 6. Facing financial constraints, vulnerable households reduce direly
needed investments in human capital such as education and health, and as this report will show, the vast
147
majority of Roma households reports that they are unable to afford basic necessities that often serve as
inputs to good health.
Low socioeconomic status and poor health among the vulnerable Roma go hand in hand with
concerning levels of reported hunger, and other measures of hardship. Over 40% of Roma
households in Slovakia surveyed in the regional Roma survey reported that a household member went to
bed hungry at least once during the past month. Hunger prevalence among non-Roma households in the
same communities was strikingly low in comparison (about 14%). Roma are also unable to maintain
heating through the winter, as three-quarters of Roma households report that they restrict heating during
winter (Figure 7-9). The fraction is significantly lower (63%) among their non-Roma neighbors.
FIGURE 7-9: SHARE OF HOUSEHOLDS RESTRICTING THEMSELVES WHEN HEATING THE DWELLING
80
75
75
Percent
70
65 64
60
55
Roma Non-Roma Neighbors
Source: UNDP/World Bank/EC regional Roma survey (2011).
7.4.3 DIET, SMOKING AND ALCOHOL CONSUMPTION
Across Europe, Roma adults and children are poorly nourished, a result of unhealthy diets and
nutrition practices reinforced by the low socioeconomic status of the population. Regarding child
health, maternal nutrition and early life nutrition, exclusive breastfeeding and adequate complementary
feeding practices are crucial for a healthy start to a productive life. Roma households in vulnerable
communities are at high risk of experiencing hunger, and additional evidence suggests that young Roma
children are particularly vulnerable: Roma infants typically only receive up to 3 months of exclusive
breastfeeding instead of the recommended six months and practice poor weaning and complementary
feeding practices. Poor diet and nutrition contributes to low immunity, a high incidence of infectious
diseases, and poor child growth, increasing the risk for cardiovascular disease in later life (WHO, 2003).
Surveys of the Roma population in Europe indicate that the intake of fast food is high and the intake of
fruits and vegetables is low (FSG 2009). While poor breastfeeding practices and poor diet are of concern
among the non-Roma population as well, it is critical to ensure that the entire population has proper
nutrition to prevent non-communicable diseases. Related to good nutrition is the condition of teeth; Roma
have high incidence of non-treated cavities and missing teeth and have poor dental check up records (FSG
2009), which can negatively impact the ability to consume fresh and healthy foods such as fruits and
vegetables, whole grains and tubers.
Smoking is common among vulnerable Roma, most likely resulting in poor health outcomes . More
than 60% of Roma adults smoke regularly, compared to about 27% of the general population, and
smoking prevalence is about 8 percentage points higher among men than among women. Non-Roma
148
neighbors smoke as well, but prevalence is significantly lower at 40%. The difference in smoking rates
between men and women is significantly smaller among the vulnerable Roma than it is in the general
population, and the prevalence among Roma women is three times higher than that among Slovak women
nationally. The high smoking rate among Roma women likely contributes to poor reproductive health
outcomes in this community, as women reportedly continue smoking during pregnancy115.
FIGURE 7-10: PREVALENCE OF SMOKING AMONG ROMA AND NON-ROMA, BY GENDER
Source: Roma and non-Roma neighbors: UNDP/World Bank/EC regional Roma survey (2011).; General population
(referred to as ‗national‘ in the figure): Eurobarometer (2006).
By contrast, alcohol use is low (Figure 7-11). About 40% of Roma adults in Slovakia report that they
never drink, only slightly lower than the national average, and only a little over 10% report drinking
regularly: once to several times a week. Differences between the Roma and their non-Roma neighbors are
negligible, but Slovak Roma men do drink significantly more frequently than women (Figure 7-11b):
while roughly 5% of Roma women report drinking once to several times a week, 20% of men report doing
so.
FIGURE 7-11: ALCOHOL USE AMONG ROMA AND NON-ROMA ( A )
A. Overall B. Roma Men and Women
Source: Roma and non-Roma neighbors: UNDP/World Bank/EC regional Roma survey (2011); General population
(referred to as ‗national‘ in the figure): Eurobarometer (2006). A The left-hand figure (figure a) presents overall
shares for Roma, non-Roma neighbors and the general population. The right-hand figure (figure b) presents shares
for Roma men and women.
115
Focus Group Discussion – April 2012
149
7.4.4 REPRODUCTIVE HEALTH
While most Roma women had visited a gynecologist at least once and gave birth in a hospital, the
frequency of reproductive health check-ups was low. A very high fraction of Roma women in
vulnerable communities and their non-Roma neighbors (roughly 90%) had visited a gynecologist, at least
once in their life (Figure 7-12). It remains unclear, however, how frequently or regularly these visits were
made. Also, while the fraction of Roma women delivering in hospitals was very high (almost 95%), less
than a third of Roma women had undergone a cervical smear examination in the past year 116. As with
other examinations, the fraction was slightly higher among the non-Roma neighbors (33%), which is also
very low. These results suggest that while women in these vulnerable communities may be receiving care
at the time of birth, access and utilization of pre- and postnatal care may still be very inadequate. A recent
qualitative study in Serbia and Macedonia supports these findings, showing that Roma women have poor
prenatal care due to various factors including lack of health insurance and poor knowledge (Janevic et al.
2011). Poor preventive health among women is not only a Roma issue (although outcomes seem to be
worse), but also a broader issue for all women across the region117.
FIGURE 7-12: UTILIZATION OF REPRODUCTIVE HEALTH CARE SERVICES
Source: UNDP/World Bank/EC regional Roma survey (2011).
Abortion rates are also relatively high. According to a 2002 UNDP/ILO 2002 report, abortion is about
25% higher among Roma women than the non-Roma, and abortion is regularly used as an anti-conception
116
Women of reproductive age are recommended to have annual pap smears. Regular pap smears dramatically
reduce the development of invasive cancer.
117
Fewer than 50 percent of sexually active women in Moldova and Romania reported ever having had a Pap test; in
Azerbaijan and Georgia, fewer than 5 percent of women reported having had the test in 2003 (Population Reference
Bureau).
150
method. Induced abortion seems to be acceptable as a form of birth regulation (mostly among older
women who have a higher number of children). Repeated induced abortion is not uncommon among some
women living in Roma settlements, which poses another risk to their reproductive health. Women are not
well informed about methods of birth control, in part related to low education levels and cultural factors
that lead to tabooisation of sexuality in traditional Roma communities (UNDP/ILO 2002).
7.4.5 BEING INFORMED ABOUT CHILDHOOD DEVELOPMENT
Health practitioners and community health workers are an important source of information about
childhood development, and the latter could play a central role in overcoming information barriers.
Over 60% of Roma and 80% of non-Roma households in vulnerable communities reported in the Regional
Survey that they had received information about childhood development from a doctor, nurse, or health
worker (Figure 7-13), indicating that most households have had some contact with personnel in the health
system, and that health outreach work currently underway has been successful in reaching vulnerable
households – although we do not have any information on the impact of these programs. The fraction of
households reporting that they had received information from a health worker or practitioner stood second
only to the fraction reporting that they had received information from a family member, indicating that
these communities rely significantly on family and social networks for information on childhood health
and development. Outreach work can help raise awareness in these circles.
FIGURE 7-13: SOURCE OF INFORMATION ABOUT CHILDHOOD DEVELOPMENT
Source: UNDP/World Bank/EC regional Roma survey (2011).
7.4.6 HEALTH CARE SERVICES
As shown above, the Slovak Roma population is in much poorer health than the general population of the
country. Is the health system providing adequate services to fulfill the health needs of the Roma
151
population? Do the Roma have access to and are they using these services? Evidence from the regional
Roma survey sheds light on these questions.
OUTPATIENT DETECTION AND TREATMENT CARE
A large fraction of Roma households in vulnerable communities do not access outpatient care when
needed. How do the most vulnerable Slovak Roma feel about the adequacy of the health system and its
ability to protect their health? Overall, the results are not bad: Only 9% of Roma households reported that
they did not have access to a doctor when they needed one in the past year, while a slightly higher fraction
(15%) reported feeling unsafe with regard to their health needs (Figure 7-14). The striking finding is that
while the vast majority of vulnerable Roma feel that they have access to doctors when needed, only 55%
of Roma households report seeing a doctor when they needed one in the previous year. In contrast, 70% of
their non-Roma neighbors report seeing a doctor when needed.
FIGURE 7-14: DO RESIDENTS OF VULNERABLE ROMA COMMUNITIES FEEL THAT THEIR HEALTH
NEEDS ARE BEING ADEQUATELY MET?
Source: UNDP/World Bank/EC regional Roma survey (2011).
Among those who do use health services, the use of outpatient services was reported to be high
among the vulnerable Roma and their non-Roma neighbors; however, Roma reported to
significantly more likely to use emergency services. Roughly a third of Roma adults and their non-
Roma neighbors had accessed outpatient medical services at least once during the month preceding the
survey, with on average six visits per adult among those who used the services (Figure 7-15). a large
fraction of these outpatient episodes involved the use of emergency services: among Roma adults, about
40% of all outpatient visits were to emergency services. Among the non-Roma neighbors, the fraction was
lower with about a quarter of all outpatient visits involving use of emergency services.
152
FIGURE 7-15: UTILIZATION OF OUTPATIENT HEALTH SERVICES
Source: UNDP/World Bank/EC regional Roma survey (2011).
Interviews conducted with health officials in Bratislava as follow-up to these findings support
frequent (and possible inappropriate) use of emergency services among the Roma. People wait ―until
the situation is really bad‖, as social workers interviewed in Bratislava pointed out. One government
official claimed that many Roma use ambulances as ―a taxi service to go to the hospital‖, in order to avoid
having to pay transport costs associated with a medical visit. While there is a fine for inappropriate use of
emergency services, it is waived for recipients of social assistance benefits. Similarly in the rural areas, a
health assistant reported that the Roma pay a visit to the doctor only when they have a serious problem,
and even in these instances, they prefer to wait until the evening and let the health assistant call an
ambulance for transport to the hospital, rather than go to a doctor during the day and pay the transport
cost.
ROUTINE MEDICAL EXAMINATIONS
Dental examinations and checkups for blood pressure and sugar levels were low among the
vulnerable Slovak Roma (Figure 7-16a and b). However, coverage of other examinations was
comparable to that in the general population, after adjusting for the different age structures of the
populations. When asked whether an adult had been tested during the last year, the Roma disparity was
largest for dental check-ups (about 40% of Roma accessing care compared to 80% of the general
population). Surprisingly, with the age structure standardized to that of the general population, the rate of
undergoing heart check-ups was higher among the Roma than in the general population by about 8
percentage points. The use of X-ray or other scans was also higher, though the difference was slight. An
explanation for this difference may be due to the higher use of emergency services by Roma discussed in
the previous section. Blood pressure check-ups (Figure 7-16b) during the past year were significantly
lower among the Roma (41%) than in the general population (66%)118. This low rate of examinations
places the Roma at risk for delayed or wrong diagnosis. Inadequate testing coupled with a low utilization
of health services and lack of resources to afford medication even when tested and diagnosed paves the
way to a high prevalence of undetected and un-treated illness. For all procedures, the rate of use was
always higher among the non-Roma neighbors of the vulnerable Roma, and the difference varied between
15-25 percentage points depending on the type of examination considered.
118
Only a quarter of Roma adults had their blood sugar levels tested during the same period. No comparable estimate
was available for the general Slovak population.
153
FIGURE 7-16: ROUTINE MEDICAL EX AMINATIONS AND CHECK-UPS ( A )
A. Dental, X-ray / scan, Cholesterol, Heart B. BP, Blood Sugar
Source: Roma and non-Roma neighbors: UNDP/World Bank/EC regional Roma survey (2011).; General population
(referred to as ‗national‘ in the figure): Eurobarometer (2006). A The figure refers to percentages among the age
group 15-70.
INPATIENT CARE
Although the Roma and their non-Roma neighbors had similar utilization rates for inpatient care,
the number of hospitalizations was significantly higher among the Roma, conditional on having
accessed care (Figure 7-17). Just over 10% of Roma adults reported having accessed inpatient services at
least once in the past year, and among these adults, an average of 4 visits were conducted per adult. The
number of visits among the non-Roma was significantly lower, at 2 hospital visits per adult in the same
period. As discussed above, there is concern among health practitioners that many Roma choose not to
seek timely medical care and wait until the health concern has deteriorated significantly before accessing
care. It is possible that this leads to the high use of emergency services and the high rate of hospitalization
observed here.
154
FIGURE 7-17: UTILIZATION OF INPATIENT HEALTH SERVICES
Source: UNDP/World Bank/EC regional Roma survey (2011).
7.4.7 BARRIERS TO ACCESSING HEALTH SERVICES
Among vulnerable Slovak Roma who did not access a doctor when they needed to, financial
constraints were the most commonly reported barrier. Why do 45% of Roma choose not to consult a
doctor, even when they believe that the services are available to them? The most commonly reported
reason was that a consultation would be ―too expensive‖ (50% of households), with ―wanted to wait‖, i.e.
to see if situation improves by itself or worsens, being the next most frequently cited reason (about 20% of
households)119. Reasons such as distance from services, fear, availability of time, and knowledge about a
good doctor played a very small role in a Roma households‘ decision about whether to seek a consultation
or not. This is not uncommon in the region; other countries show similar results.
119
The fraction reporting "too expensive" and "wanted to wait" varied with income quintile, with the highest fraction
among the richest Roma quintile. However, the number of observations in each quintile, ranging from 30-50, may be
too small to assign any significance to this. There is no explanation for the positive correlation with income.
Analyzing the fractions of households reporting they couldn't afford medicines in each income quintile, there is, as
expected, a negative relationship with income (see para and figures on medicines).
155
FIGURE 7-18: REASONS FOR NOT SEEKING CARE WHEN NEEDED
Source: UNDP/World Bank/EC regional Roma survey (2011).
Interestingly, the reasons why the neighboring non-Roma households chose not to seek care were
quite different: slightly less than a quarter of households reported that the direct cost was a barrier, and
compared to the Roma, a significantly higher fraction (20%) reported that lack of time was a factor. The
fraction of households that ―wanted to wait‖ before seeking care was higher as well (30%).
A large fraction of Roma households cannot afford to buy medicines when needed. Lack of
affordable health care is reflected not just in the utilization of doctors, but also in the purchase of essential
medicines (Figure 7-19). About 45% of Roma households reported that they could not afford to buy
medicines at least once in the past year. 72 (43)% of households in the lowest (richest) income quintile
could not afford to buy medicines when needed. Again, significantly, only 20% of non-Roma neighbors
reported being similarly constrained.
FIGURE 7-19: AFFORDABILITY OF MEDICINES
A
A
Source: UNDP/World Bank/EC regional Roma survey (2011). The figure refers to the Fraction of households
reporting they couldn’t afford to buy medicines in the past year.
156
Follow-up interviews conducted with health practitioners confirm that medications are often too
expensive for Roma households. Interviews conducted with health officials in Bratislava confirmed that
while some prescription medications are free, insurance plans may cover anywhere between 20 to 80% of
others, depending on the category of medicine. Non-prescription medicines are not covered, and even a
painkiller like aspirin may be too expensive for poor households. While some sympathetic doctors try to
prescribe the cheapest (generic) medication, it is possible that others may not due to incentives from
pharmaceutical companies. Among larger families, more than one child falling sick at the same time
results in a large one-time expense that may not affordable. Officials say that in such times it is not
unusual for households to borrow money from family, although the extent of this has not been quantified.
Health workers in Kosice stated that Roma families do not know how to save money, and even with social
benefits and allowances arranged through the community center nearby, medicines are still so expensive
that many Roma decide not to buy them, or buy them in smaller quantities than are prescribed, and do not
take them regularly as a result. A health coordinator interviewed reported the same financial constraints,
pointing out that most Roma knew about the usefulness of medicines, and many often asked health and
social workers for painkillers such as aspirin. Only a minority of Roma did not trust Western medication
or the instructions of doctors, and preferred to buy juice or collect herbs and brew tea for themselves and
their children. One doctor who was interviewed and who is working in three villages in one of the regions
where Roma reside found that insurance did not adequately cover prescribed medicines, and so he often
prescribed medicines that either required no additional payment, or just a ―symbolic‖ payment. In some
cases though, he found that the Roma would not go to the pharmacy to pick up the medicine, even in cases
where it was free. Since he was aware that a lot of Roma were unlikely to return for prescription refills, or
that they would not come back soon enough, he would typically prescribe a larger quantity of medicine
than was necessary.
Vulnerable Roma households face similar constraints when making other health investments,
including the purchase of nutritious foods. More generally, financial constraints bind Roma and non-
Roma households differently, even though these households are located in the same communities and
experience similar socioeconomic conditions (Figure 7-20). Only a small minority (10%) of Roma
households report that they would be able to pay for a large, unexpected expenditure out of the
household‘s own resources – compared to 40% of their non-Roma neighbors. Similarly, only 1 in every 5
Roma households can afford to eat meat every second day (compared to 2 in every 5 non-Roma
households), and again a mere 10% report that they can afford regular dental visits (compared to just over
50% of all non-Roma neighbors).
157
FIGURE 7-20: AFFORDABILITY OF UNEXPECTED EXPENDITURES, NUTRITIOUS FOOD AND DENTAL
VISITS
Source: UNDP/World Bank/EC regional Roma survey (2011). The figure refers to the Fraction of households
reporting the couldn‘t afford each item.
Dissatisfaction with the quality of outpatient care received was relatively low among Slovak Roma.
Dissatisfaction with quality of services could keep beneficiaries from accessing services. However, about
55% of Roma households reported being either fairly or very satisfied with the quality of outpatient
services used in the past month (Figure 7-21), while dissatisfaction with services seemed to be limited to a
significant minority of 20% of households who reported that they were fairly or very dissatisfied with the
quality of services accessed.
FIGURE 7-21: SATISFACTION OF HOUSEHOLDS WITH OUTPATIENT SERVICES
Source: UNDP/World Bank/EC regional Roma survey (2011).
Most Roma households were located within less than a 1-3 km distance from critical medical service
providers. Distance from points of access to important services such as those provided by a general
practitioner (GP) were not as high as could be expected, given that Roma communities surveyed were
among the most isolated ones in Slovakia (Figure 7-22). Just over a third of Roma households were
resident within less than a kilometer from a GP, while roughly another quarter live within 1 to 3
kilometers. About two-thirds of Roma households were resident within a 1-3 km or smaller radius around
a primary medical center and a pharmacy. However, it should be noted here that even where service
158
providers were situated within a 1-3 km radius of settlements, the health practitioners interviewed for this
report frequently cited transport costs and a reluctance to seek health care outside a village or settlement as
typical reasons why the vulnerable Roma choose to forego seeking medical care.
FIGURE 7-22: DISTANCE OF HOUSEHOLDS FROM ESSENTIAL SERVICES
Source: UNDP/World Bank/EC regional Roma survey (2011).
Direct and indirect financial costs appear to be a significant barrier to accessing medical services
among vulnerable Slovak Roma communities. Monthly per capita household expenditures on health are
indeed high: 8.80 Euro per capita per month among the Roma (4.3% of household income), and slightly
higher at 11.80 Euro (3.3% of household income) among the non-Roma neighbors. Unlike the vulnerable
Roma, for whom financial constraints featured prominently as a factor influencing a households‘ decision
to seek medical consultations or buy medications, opportunity costs of seeking care (e.g. alternative use of
time) seemed to matter more for their non-Roma neighbors.
Interviews conducted with health officials in Bratislava to provide follow-up to some of the findings
from the Regional Survey shed further light on financial barriers to accessing medical services .
Although visits to doctors are free, waiting times for those accessing services without an appointment are
very long. In order to make an appointment, patients have to pay a 10 Euro fee. For many Roma, this is
prohibitively expensive. Another observation pointed to how Roma households are unable to smooth
consumption, even over the course of several weeks: ―Health care is more easily afforded in the days after
payment of social assistance‖. Afterwards, ―the money is gone, so it‘s also a question of timing‖.
159
A health care coordinator who was interviewed also confirmed that Roma who did not want to visit
a doctor or were unable to do so when needed did not have the money to travel to the doctor. She
had worked with cases where a child was ill and urgently needed care, and the child‘s parents were doing
―everything to find money. They borrow from relatives or they try to save money from allowances‖. In
cases where a child was not critically ill, the opportunity costs to seeking medical care might be high: the
coordinator reported that in such cases, many parents would refuse to take their children to the doctor,
giving ―excuses‖ even when the health worker offered to accompany the parent and child to the doctor. In
these cases, threatening the parents that if they didn‘t go to the doctor, the child would be taken away and
placed in an orphanage proved to be an ―efficient method‖ to get the parents to comply with the advice of
the health worker.
It may be however that distance from provider translates into ―costs‖ that vary across regions. One
doctor interviewed reported that distance was not a constraint to accessing a doctor, because the villages
were within a 4 km distance of each other and ―a lot of Roma who are interested in their health just walk
to a doctor‖. It has to be kept in mind though that a distance of 4 km is a r elatively easy walking distance,
except when one is ill and needs assistance. Further research is needed to ascertain whether such
systematic differences exist across regions.
The low rates of utilization of health services and lack of affordability of many health inputs among
the vulnerable Roma in Slovakia may be surprising as health insurance coverage is very high . As is
often the case, health insurance does not equal financial protection. In the vulnerable Roma communities
surveyed in the Regional Survey, coverage of health insurance is widespread (Figure 7-23), although there
is a significant minority of households in which the adult respondent lives with at least one household
member that is not covered under insurance (17%). This fraction is slightly lower among the non-Roma
households, but the difference is not statistically significant. Interviews conducted with health
practitioners working in Roma settlements confirm that insurance coverage itself is not a problem. Rather,
the issue is that most Roma have no understanding or information about insurance programs and what
they cover. One health assistant interviewed reported that in many cases the Roma do not even know the
name of the insurance company they are insured by. Often, the Roma switch insurance providers when a
new company offers them an incentive of 1 to 2 Euros for switching, but they rarely understand the
implications of this, leading to confusion during medical visits. Another health assistant interviewed added
that many Roma do not have valid insurance cards and that in many cases the community center, social
workers and health assistants have to intervene and help them acquire new cards.
FIGURE 7-23: HEALTH INSURANCE COVERAGE AMONG ADULTS
Source: UNDP/World Bank/EC regional Roma survey (2011).
160
7.5 CURRENT SLOVAKIA ROMA HEALTH POLICIES AND PROGRAMS
7.5.1 STRATEGY FOR ROMA INCLUSION
There is positive news regarding policies and programs targeted at Roma in the health sector.
Slovakia recently approved the National Roma Inclusion Strategy (NRIS), which includes a large focus on
improving health conditions. The main goal for the health strategy is to improve access to quality health
care and emphasize preventive care and health education. This goal is very well aligned with the priorities
that have been identified from this analysis which shows there is a lack of awareness about the importance
of seeking timely and preventive care when needed. The strategy‘s focus on improving access to quality
care and preventive care is therefore well founded.
The strategy proposes partial goals or sub-goals for key health factors. They include personal
responsibilities regarding behavior change such as increased awareness of parenthood, reproductive
health, and campaigns on modern contraception. They include local government responsibilities such as
improving the environment, hygiene and waste disposal, rodent removal and reducing pollution levels as
well as ensuring quality of drinking water. Specific health services are addressed: dental hygiene and care,
preventive vaccinations and reproductive health. The strategy does not include however, clear targets or
indicators to be met toward which policies can be developed.
The strategy includes an implementation proposal, which builds on existing programs such as
strengthening the communication between socially excluded populations and service providers
through community workers. It looks at strengthening coordination with NGOs and the Association of
Towns and Municipalities in Slovakia or ZMOS. Finally the Strategy includes an implementation plan and
budget proposals and it identifies potential resources.
There are areas that could be strengthened by adding more details regarding new priorities and
interventions based on the most recent analysis of the health situation of Roma in Slovakia. For
example, population growth has not been mentioned and as shown is a major issue. The Roma population
has higher fertility rates and lower life expectancy than the Non Roma Slovak population. They have
poorer health outcomes and die younger. Infant mortality rates are higher. Population policies such as
improved access to accepted contraceptive methods, inclusion of contraceptives in the health insurance
package, and reproductive health education to young adults and future parents could be considered.
Infant health and maternal health are crucially important and explain, in part, the higher mortality
rates among Roma children. Intervening at an early age, starting at conception, has been shown to have
substantial benefits in later life (Alderman 201,0 Bhutta et al 2008; Black et al 2008; Hoddinot et al 2008).
The strategy remains very general in targeting populations and could be improved by being more specific
with respect to interventions targeted at pregnant women, young children and young adults. These are
population groups that are accessible and whose health and behavior will have positive impact on the
community as a whole. As such they could be prioritized.
Children going to bed hungry is something that should not happen and would not if social assistance
would be targeted to address these issues. Making parents more aware of nutritious food, and what
children‘s needs are is important so parents make more appropriate choices, even within a limited budget.
In this regards, reducing the intake of sweets and better dental care will also help improve nutritional
status and health throughout the lifecycle.
161
Chronic disease prevalence is high and is likely to grow further. The strategy does not yet specify any
interventions for particularly vulnerable groups such as those suffering from chronic disease, mental
disease or disabilities. Unfortunately, chronic disease prevalence will likely continue to grow as shown in
the above analysis. More is needed to encourage preventive screening and early detection and treatment,
especially among women. Inclusion of chronic disease drugs in the health insurance package for
vulnerable groups is a policy that addresses the financial barriers, which impede Roma from getting the
right treatment and following up on doctor prescriptions.
The very high smoking rates are worrisome and do not get sufficient attention. This is not only an
issue for Roma, it is a problem in Slovakia. Among the Roma, women have some of the highest smoking
rates in the world. Communication about the risks, especially during pregnancy and the consequences for
the child needs to be targeted and adapted to the Roma population.
At the same time, much focus is given to access to services, and from the analysis it is clear that
certain aspects of access, financial access and trust are the main issues which need attention . As such
it is important the strategy emphasize these areas. The strategy mentions the Roma Health Mediator
programs as being successful and would support local government and non-governmental organizations to
get involved. Better access to dental care is specifically mentioned and as the analysis shows is much
needed.
Strengthening communication around the various health issues to different population groups is a
large area of potential gain for improving access. As mentioned above there are many areas where
behavior change can have major impact of current and future health status (with respect to children and
chronic disease specifically). These areas are currently not well addressed, early life nutrition, maternal
nutrition, risk of smoking during pregnancy etc. These areas can be included in the training and
communication manuals and given more attention during training of RHM and all health workers.
7.5.2 INSTITUTIONS AND COORDINATION
It is difficult however, from the aforementioned Inclusion Strategy which was prepared by the
Plenipotentiary (PPO), how much commitment there is from the Ministry if Health for this new
strategy. The PPO was mandated to prepare the above strategy however the current institutional set up to
improve the health of the Roma lays for a large part with the Ministry of Health who is responsible for the
health of the entire Slovak population. It also oversees the Health Insurance regulation and coverage
principles. MOH assures universal access to health services and financial protection to all citizens through
the health insurance. It is however clear from the analyses that not all Roma understand what universal
access and free care is.
As a response, the MOH implements one of the existing Roma Health Mediator Programs, which
aims at strengthening the link between the services and the Roma population using mediators . This
program falls under the responsibility of the National Public Health Institute under the MOH. The budget
is set on an annual basis. Unfortunately there is little information available on the coordination between
the different programs, the oversight and impacts.
Moreover, due to the principle of not recording ethnicity there is very little known from the
administrative data about Roma, their burden of disease and utilization of services. This hampers the
targeting of not only groups but also specific messages and interventions. As a result there is also little
information on specific targeting of Roma and how well programs and campaigns work. Certain programs
162
however do prioritize the Roma, vaccination campaigns, de-lousing and certain education programs, due
to the fact that Roma disproportionally suffer these afflictions.
7.5.3 EXISTING PROGRAMS THAT ADDRESS ROMA HEALTH
There are ongoing programs that address Roma Health issues in Slovakia and which show positive
results; best known among these is the Roma Health Mediator program. The Roma Health Mediator
(RHM) program is the most visible and widely implemented Roma Health program. The RHM programs
exist in Romania, Bulgaria, Spain, and Serbia among others. There are also health mediator programs in
Slovakia, but they are small in scale and there are many variations between the three programs, which are
implemented by different institutions, the National Public Health Institute and NGOs.
An assessment of the RHM programs in Bulgaria, Finland and Romania also found that RHMs
have greatly assisted individuals in Roma communities, particularly in overcoming language
barriers during doctor visits and bureaucratic obstacles in accessing health care (OSI 2005).
Importantly, the RHM programs have been found to decrease prejudice among doctors who participate.
However, this assessment notes that some of the mediator programs operate in isolation without being
sufficiently integrated into the overall public health system, and they have yet to be accompanied by
effective legislative changes that address systemic issues affecting Roma health (OSI 2005).120
A more recent assessment 121 of the Slovakia programs lays out in more details the job profiles and
activities the RHM workers carry out. However, this was an assessment not an evaluation and it is
highly recommended to undertake more rigorous evaluations of the existing programs. Nevertheless, the
assessment does provide very valuable information on focus and activities of the RHM in the 3 different
programs that are ongoing, of which two are NGO-led and one is implemented by the government. The
programs differ in emphasis and the NGO led one if more varied and includes more activities. The
training of RHM varies and in all programs is considered to be too little. In the best case, the RHM get a
two-week training and an additional 4 times a year education session. In addition to training challenges,
RHM are contracted on a yearly basis, with little job security.
There is much written about the programs and assessments of how they work; however, to date
there has been no rigorous evaluation of the impacts of mediator programs. Box 7-1 describes more
details regarding the ongoing programs in Slovakia and highlights the health benefits these program can
accomplish if implemented effectively.
BOX 7-1: HEALTH MEDIATOR PROGRAMS
The objective of Roma Health Mediator (RHM) programs is to facilitate interaction and outreach
between the health system and the Roma community. Typically, Roma women that have completed at
least primary education are selected and trained as mediators with various responsibilities including
facilitating health visits, increasing communication between health providers and Roma patients, and
improving knowledge and awareness of healthy behaviors in the community.
120
Noting that these systemic issues (including financial limitations, flawed legislation and inadequate political will)
are beyond the scope of RHM programs, the report calls for other components of Roma health strategies to address
these issues.
121
This is an internal assessment. OSF made the information available but it will not pursue publishing it, it remains
an internal memo.
163
The Slovak RHM program has been operating for nearly six years. There are three health mediator
projects currently operating122:
1. The program for the health promotion of disadvantaged Roma communities. This program is
organized and implemented by the Ministry of Health in collaboration with the Office of the
Representative for Roma Communities, regional offices for Health, Slovak Medical University and Non-
Governmental Organizations. A first stage, pilot, was carried out in 2007-8 and a second phase, after
positive findings of the pilot, expanded the program to 13 districts in about 150 communities. Vaccination
rates have improved, and health awareness was higher among the Roma participating in the program. A
total of 30 health mediators are involved. The second, expanded phase is ongoing Monitoring data from
the program show an increase in variety of activities between 2008-2009, but it is unclear whether this is
solely due to the program or other factors. While there is no control group or data for 2006 and 2007, the
monitoring data provides a good overview of the type of activities that were carried out and the changes in
volume between 2008 and 2009. For example, vaccinations initiated by the community workers are about
the same, about 10,000 in both years. Health education to about 40,000 people. Translating the monitoring
data into % of targeted Roma population reached, as well as more information on time spent per activity
etc will allow for better analysis and can help inform policy decisions regarding scaling up, allocation of
resources, training needs and prioritization of activities.
TABLE 7-2: MONITORING DATA ROMA HEALTH MEDIATORS/COMMUNITY HEALTH WORKERS
Program for the Health Promotion of Disadvantaged Roma Communities (OSF 2012, Internal Memo)
2. The Healthy communities pilot program, carried out by the Non-Governmental Organization
Association for Culture, Education and Communication (ACEC). This program involves 91 field
health assistants and coordinators, is implemented in 67 Roma Settlements. Most of the filed assistants
live in the community itself, the coordinators vary. The program coordinates closely with over 100
General Practitioners (GPs), 26 primary schools and the municipal authorities where the program is
122
The following information on the Roma programs is based on an internal memo by OSF.
164
situated. Union Health Insurance is a partner and finances the program as well as the vaccination programs
against hepatitis and influenza. The project assessment shows that a large focus is on vaccination and there
has been 100% follow up. The excellent collaboration with GPs contributes to the success. The report
includes data on vaccinations, however no data on population reached. It does however report that
preventive medical check-ups increased in the targeted areas by 36% between 2007-8 while nationwide
this only increased by 16%.
3. The Health field officers program carried out in 12 localities between 2006-2008. In addition to
carrying out similar activities as the above two programs, with health workers, the included a Peer
Program targeted at children. Between 2009-2010, the so-called Peer Program, involved 12 different
groups of children between the ages of 12 and 24. There was strong collaboration with schools and among
the results reported are improved hygiene and sanitation practices; peer group members themselves
organized community-wide activities sharing their knowledge; some reported overcoming addictions and
about 1/3 of the group (50) of 145 children reporting interest in continuing with vocational training, while
26 children aimed at becoming a community health worker. Children became actively involved in
community activities regarding scabies prevention, delousing and awareness creation.
Unfortunately, the available information on the activities and benefits from RHM programs do not
provide adequate data to allow for comparisons between the programs or what would have
happened in the absence of these programs. It is therefore difficult to rigorously quantify the impact;
however, based on qualitative evidence from several countries, the programs definitely appear to have had
positive impacts.
What the various assessments provide is more information about the challenges the programs face
and what future directions could be considered. First and foremost not all communities are being
reached. About 200 out of 650 communities in need are covered under the three programs. At present the
largest program, reaching 150 communities with 30 health mediators, is showing signs of stagnation with
inadequate health education, lack of training for mediators and unfavorable contract situations (non-
standard short term contracts and unclear future). Training to health mediators is provided by individual
projects and is not systematic. When asked, mediators would like more training and opportunities for
sharing knowledge. They encounter situations where they are asked to do more than what is in their job-
description and although they are willing they do not feel qualified. More and better coordination with
local authorities is often mentioned in assessments, especially regarding infrastructure and with hospitals.
There have been instances of miscommunications between families of newborns and hospital authorities
on length of stay. Often these can be resolved with the help of mediators, family support and with better
information on all sides.
In the long-term, the RHM programs can potentially have positive economic benefits in addition to
health impacts. Evidence on health interventions in other countries has shown that improving health can
lead to economic benefits, with the strongest evidence for early child health and nutrition programs.
Perhaps the most well-known study in this area is the Institute of Nutrition of Central America and
Panama (INCAP) nutritional experiment conducted in four rural villages in Guatemala from 1969-1977
(see Behrman 2009 for more info). Long-run follow-up studies showed that children in the villages
receiving a high-protein supplement were healthier and obtained more schooling. The effects on adult
income are not statistically significant, but are suggestive of potentially large increases in income (by
~25%) and hourly wages (by ~33%). Another important experiment focusing on child health is a
randomized school-based deworming program in Western Kenya from 1998-2001. Children in schools
receiving the deworming drugs were healthier and had higher school attendance in the short-run (Miguel
165
and Kremer 2004). Through long-run follow-up surveys, researchers showed that those same children as
adults ended up working 17% more hours per week, and that among wage earners, earnings increased by
over 20% (Baird et al. 2011). While the situation among the Roma in Slovakia are substantially different,
the empirical evidence strongly suggests that improving early child health can greatly improve educational
outcomes, cognitive development, and future adult productivity.
TABLE 7-3: RHM ROLE AND THE CORRESPONDING HEALTH IMPACTS
RHM role Health impact
Documentation and insurance information; Manage chronic diseases in adults through regular check-
encouraging doctor visits ups/preventive care
Early detection and treatment of chronic disease
Vaccination coverage Reduce infant and child mortality
Community-wide protection from mass vaccination
Education sessions on ante-natal care and Improved pregnancy outcomes
reproductive health
Reduce LBWs
Family planning
Health and hygiene education Reduce infectious diseases
Inclusion/integration All indicators
There is a variety of other programs and interventions that are being implemented in Slovakia and
in other countries to improve Roma Health, but few are sustainable, and none have been evaluated
rigorously. In Slovakia123, mobile health clinic units were used to reach hard-to-reach communities and
isolated slums to deliver basic health services. In Greece, Spain and Poland mass immunization campaigns
to increase vaccine coverage in Roma communities were carried out (EC 2004, Orlikova et al. 2010).
While these measures were important, they are also regarded as ―stop-gap‖ and not sustainable in the long
term (EC 2004).
NGOs play an important role in implementing intervention and maintaining issues around Roma
health on the political agenda. In Helsinki, Finland the NGO Roma Mission developed health education
programs for Roma adolescent girls to ask questions they cannot ask their mothers (Dimitrijevic 2009). In
Bulgaria, the Foundation for Promotion of Roma Youth and the Initiative for Health Foundation conduct
discussions on HIV prevention and outreach to drug users in Roma communities (Dimitrijevic 2009,
Initiative for Health Foundation 2007). In Romania NGO led interventions included free medical exams
and free medicines for the poor; health care education for pregnant women, mothers and children;
ensuring basic conditions for the adequate nutrition of children; informing Roma women of their rights
and obligations to increase their access to medical services; and training health mediators (Cace et al.
123
Program funded by Phare (2004)
166
2006). Most of these projects, however, were implemented over a limited period and have not been
continued. One exception was the health mediator program, which was initiated by the NGO Romani
Criss, adopted by other NGOs and eventually taken up by the Ministry of Health.
Roma participation in the design and implementation of these interventions has played a key role in
the success of programs implemented in Spain, such as the Cartuja health clinic in Granada, where
programming is based on research conducted by local Roma women (ERRC 2006). 124 In Kiunstendil,
Bulgaria, a medical center was created in a Roma community following a participatory approach where
community members voiced their most urgent needs (Tomova 2009). The establishment of the medical
center has led to over 90% of the community being registered with a GP, improved vaccination rates and a
greater use of preventive health services, resulting in improvements in general health outcomes for adults
and children (including reduced infant morality) in the community.
At the national health planning level, however, the OSCE High Commissioner on National
Minorities, the Council of Europe and the European Monitoring Center on Racism and Xenophobia
highlight the consistent lack of Roma participation (OSI 2005, CoE 2003). To this end, several NGOs,
including the Fundación Secretariado Gitano, recommend that governments ensure the active
participation, and leadership where possible, of Roma in the interventions that affect them, at the stages of
planning, implementation and evaluation, and at both macro- and micro-levels of action (FSG 2009).
―Active participation‖ is also one of the ten Common Basic Principles for Roma inclusion adopted by the
EU‘s European Roma platform in order to guide EU member and candidate states‘ Roma policies.
7.5.4 SLOVAK HEALTH EXPENDITURES
Slovakia spends more than its neighbors on health care. In 2009, Slovakia spent 7.8% of its GDP on
health, which is well above the average of the EU12 countries125 (Szalay et al, 2011). Since 2010, the
Slovak social health insurance system, introduced after the establishment of Slovakia in 1993, provides
universal coverage for a broad range of benefits. It guarantees free choice of one of the three nationally
operating health insurance companies and is based on solidarity. Contributions are collected from
employees and employers, self-employed, voluntary unemployed and state-insured. The benefits package
is broad and most essential pharmaceuticals are included without co-payment. At the same time, out of
pocket health expenditures amount to 1.8% if GDP in 2009 and an average Slovak spends about 200US$
on health out of pocket annually. This is a significant rise from 2002, when it was about half this amount.
Explanations for the increase include rising co-payments on drugs, increase in use of over the counter
(OTC) drugs, increased use of private providers and information payments in the state sector. The data
and information does not allow to analyze the proportion of public funding spent on Roma, but most
Roma do have a health insurance as discussed earlier.
7.6 POLICY RECOMMENDATIONS
Health services in Slovakia are free of charge, health insurance coverage is nearly universal, and
most of the population, including Roma, live within a few kilometers from a health facility. Yet
Roma suffer worse health than the non-Roma population. First, the very poor living conditions
described in the housing chapter are obvious contributors to infectious disease, diarrhea and respiratory
124
This information is based on personal interviews; the report provides no further information on this initiative.
125
EU12: The countries that joined the EU between 2004 and 2007.
167
disease among children. Second, a high burden of chronic disease is consistent with high risk behavior
such as high smoking rates, both for men and women, poor diet and low levels of physical activity, as well
as a high rate of teen pregnancy. Third, poor health outcomes can be caused by ineffective use by many
Roma of the available health services.
The recently approved Slovak Roma Inclusion Strategy includes a strong focus on improving the
health of Roma. Much emphasis is given to improving access through the active involvement of local
governments (ZMOS), the NGO community, and expanding the piloted Roma Health Mediator programs.
The policy recommendations below are consistent with the overall Strategy recommendations, and aim to
(a) highlight more specific programs and priorities areas, and (b) provide a number of complementary
entry points that are important to consider.
7.6.1 POLICY MEASURE 1: MEASURES PROMOTING MORE EFFECTIVE USE OF
EXISTING HEALTH SERVICES, MEASURES TO EXPAND HEALTH
KNOWLEDGE AND AWARENESS, OR A COMBINATION OF BOTH:
As discussed, even though the Slovak health system provides universal access and all citizens are
covered under the health insurance, many Roma still do not use the system and those that do use, do
not use it very effectively. Moreover, many who say they do not use, identify financial barriers as the
main constraint, which reflects they do not understand the full benefits of the health insurance benefits. As
a result of misunderstandings and ‗waiting to seek care‘ the use of emergency servi ces and ambulances is
very high and costly to both the health system and the patient. The RHM programs mentioned do address
these by assisting Roma in getting health insurance cards and explain the use as well as accompanying
them to seek health services. However, these programs are small scale and their sustainability is uncertain.
Therefore it is highly recommended to seek better integration of Roma in the health system by more
effective access and utilization.
POLICY MEASURE 1A: EXPANDING THE RHM PROGRAM AND IMPROVING ITS
QUALITY THROUGH THE USE OF RIGOROUS EVALUATION METHODS.
Available evidence suggests that RHM programs work. Although more evidence is needed, RHM
programs in other countries have been effective in increasing health care utilization and outreach. The
Slovak government recognizes their effectiveness and mentions their scale up in the future directions of
the Slovak Strategy on Roma Integration.
Vaccination coverage was a problem, especially the completion of vaccinations, and RHM programs
have helped improve vaccination rates. Mediators can encourage parents to get children vaccinated and
are also instrumental in supporting vaccination campaigns. Evaluating the actual benefits of this increase
in vaccination - how many more children were vaccinated due to the program - would allow policy makers
to see results of these investments and justify more investment in such programs.
Upgrading the programs to address the challenges that have been identified The existing RHM
programs as described above, address several of the key issues that arise from the new data-analysis and
some other priorities could be considered. For example, from the available documentation RHM currently
appear to spent considerable time facilitating visits to health facilities and vaccinations. Both are
extremely important and should be continued, however, the attention given to preventive care and
awareness-raising, especially regarding the importance of maternal and child health cannot be
underestimated. Most health gains and later life health benefits are made in the early life years. And the
168
risk of later life chronic disease can be considerably reduced with preventive actions earlier in life. Clearly
a large number of people are being reached, with number as high as 40,000 by one program with health
education. These are excellent opportunities to strengthen the messages and raise awareness about the
areas that need to be prioritized such as maternal and child health and smoking. It is beyond the scope of
this paper to go into details, but it is highly recommended to undertake further operational research and
information gathering on the actual training, messages, communications methods to align the training of
the mediators and the education materials to these priorities.
An important part of an upgrade of the program would involve the mediators themselves and their
supervisors. Among the main challenges mentioned in the various assessments of RHM programs not
only in Slovakia but also in other countries where Roma reside, is the issue of the contract situation of the
mediators. Often they are on short term, no extension guarantee contracts. There is little career
development opportunities and most indicate strong interest in more systematic training opportunities. The
available information does show the RHM program is promising and has the potential to grow both in
scale and scope. However, this is only feasible with better trained mediators, who enjoy more and better
supervision and support as well as more systematic coordination and above all who have job security.
There is also a need to systematize the job profile among the different ongoing programs and again,
evaluate what works best. As indicated above, the job of the mediator is varied and it is unclear from the
available data what time is spent in which task and whether that is the best way to spent that time.
The Slovakia RHM NGO led program piloted peer learning with about 145 adolescents and young
adults, aged between 12 and 24. The program saw positive changes in personal hygiene behavior,
motivation to take charge of one‘s life, and recognizing the importance of overcoming addictions (OSF
2012, Internal Memo). Initial monitoring data results show that one third of the children participating
indicated interest in continuing with vocational training and 26 of the 145 showed interest to become a
RHM. Children themselves started monitoring the health problems in their communities and in schools.
This is a promising pilot, which merits more evaluation and possible scaling up.
More sustainable budget is needed. The largest program is already said to be stagnating and a large part
can be explained by the lack of adequate budget especially to pay for salaries. More job secure contracts
and better training will require more funding. Having better data to justify higher levels of investments
will help make the case.
POLICY MEASURE 1B: BETTER USE OF EXISTING ENTRY POINTS WITHIN THE
HEALTH SYSTEM –
In particular ante natal care and birth delivery - to promote greater health knowledge and
awareness, as well as access to services, among Roma. Birth delivery, but also ante natal care visits are
two crucial periods during which Roma women can be reached and given information. Most births already
occur at the hospital, which substantially reduces mortality risks for both mother and child. Together with
antenatal visits, they allow an opportunity to discuss many elements of reproductive health, family
planning, teenage pregnancy, breastfeeding, and post-natal care with the new mother and family-members.
However, these needs to be well implemented with trusted health workers and messages adapted to the
Roma culture and needs, therefore more coordination between the two is important.
POLICY MEASURE 1C: EXPLORING CHANGING HEALTH PROVIDER PAYMENT
SYSTEMS AND LINKING PAYMENTS TO RESULTS.
169
In particular health coverage as measures by several indicators of Roma population and regular
check-ups could be considered. A range of countries is applying results based financing in the health
care sector, and these systems have shown good results.
POLICY MEASURE 1D: PROMOTING THE USE OF MOBILE AND WIRELESS
TECHNOLOGIES
This can open up many opportunities ranging from making appointments, providing reminders, as
well as reminding patients about prescription renewals or vaccinations. Send test results and ensuring
treatment compliance are also among the possibilities. M-health is the use of mobile and wireless
technologies. Cell phone use is becoming more common and opens up many opportunities ranging from
text messaging to make and remind about appointments, remind patients about prescriptions renewals and
can also be used to remind people about vaccinations, send test results, treatment compliance, questions
about prescriptions, complaints and in many countries health call centers are being evaluated as
successful. Mass SMS can be sent out when campaigns take place. And cell phones can be used to
respond to emergencies in a more efficient manner, for example determining the need for ambulance
transport, informing the hospital about in-coming emergencies. Cell phones among Roma are common;
65% of SK Roma households have "mobile phone or landline" (93% of non-Roma neighbors have it) and
this percentage if likely to grow in a more and more connected world. ‗Innovations for Poverty Action‘
(IPA126) is testing mobile phone messaging with software that reads SMS messages out loud, which is
helpful for illiterate people, xx% of Roma adults are illiterate. Considerations would need to be given to
data-security, privacy and network capabilities among others. Good examples can be found on ante-natal
care in Bangladesh, nursing care in Saskatchewan in Canada for aboriginal populations.
FIGURE 7-24: EXAMPLE OF A MOBILE PHONE (SMS) BASED PREGNANCY ADVICE SYSTEM FOR
MOTHERS
Source: Ministry of Health and Family Welfare of Bangladesh
126
An NGO that applies rigorous techniques to develop, test and scale up proven solutions to real-world problems
faced by the poor in developing countries.
170
7.6.2 POLICY MEASURE 2: USING ENTRY POINTS OUTSIDE THE IMMEDIATE
HEALTH SYSTEM TO PROMOTE GREATER HEALTH AMONG ROMA:
IMPROVING HEALTH THROUGH THE EDUCATION SYSTEM
POLICY MEASURE 2A: USE PRESCHOOL TO PROMOTE HEALTH
Preschool offers not only education benefits, but can also offer many opportunities for health,
promoting early childhood development broadly. Educating hand-washing and other basic hygiene can
be part of the curriculum. Preschools can offer an information channel for vaccination campaigns and
follow up, if relevant de-worming, de-lousing and other hygiene-related efforts, and they offer support for
parenting education such as encouraging parents to read to children and instill healthy behavior, including
nutrition.
POLICY MEASURE 2B: SCHOOLS OFFER EXCELLENT OPPORTUNITIES TO
EDUCATE ON HEALTH
Children are excellent channels to get through to parents and experiences from other countries show
that children can influence for example what parents choose to eat, change hygiene behavior
(demonstrated by Slovakia‘s own Peer Program). During the middle school years, it is very important to
reach young adolescents with messages and information on fertility, family planning, HIV/AIDS and STIs
in general. The Roma Peer program is an excellent start and could be expanded through schools and as
encouraged in the education chapter as after-school programs.
POLICY MEASURE 2C: PROMOTING SPORTS, ESPECIALLY AMONG YOUTH
Playing sports not only improves health; it also offers opportunities for social engagements, pride in
community, and again a venue to learn. Especially reaching an often difficult to reach group, adolescent
and young adults with crucial messages about the risks of teenage pregnancies, drugs, risk of chronic
disease and health lifestyles.
IMPROVING HEALTH THROUGH THE HOUSING AND INFRASTRUCTURE
SECTOR
POLICY MEASURE 2D: PROMOTE ACCESS TO CLEAN WATER, AND REDUCE
WASTE AND INDOOR AIR POLLUTION
The large majority of Roma do not have proper access to clean water and even fewer are connected
to adequate sewage and waste collection facilities. The consideration of improving personal hygiene
behaviors through education and awareness through the RHM programs will not be sufficient if there is
not coordination with the local government infrastructure on water and sewage systems. It is also
recommended to look into temporary solutions that worked in other countries such as a successful project
by Innovations for Poverty Action in Kenya and Haiti, which distributed chlorine dispensers to households
as a start, while more permanent connections to infrastructure are being built (www.poverty-
action.org/work/projects/safewater). As mentioned in the chapter on Housing, the poverty map and Slovak
Roma atlas being produced can support proper targeting of these.
171
IMPROVING HEALTH THROUGH THE SOCIAL PROTECTION SYSTEM
POLICY MEASURE 2E : LINKING SOCIAL ASSISTANCE AND FAMILY BENEFITS TO
RESULTS BY MAKING CASH TRANSFERS FOR CERTAIN HEALTH OUTPUTS.
A majority of Roma receive social assistance and family benefits of which a portion or additional
funds could be linked to actual results, so-called Conditional Cash Transfers. Parents or even
communities would receive additional cash for having their children vaccinated on time, attending health
education sessions and getting adequate pre-natal check-ups for example. There are other examples of
results and a number of countries have very positive experiences with these types of programs. The Bolsa
Familia in Brazil and the Opportunidades in Mexico described in the social protection and employment
chapter and in the education chapter are among the first and largest and have been extensively evaluated.
Another large and successful program is being implemented in Indonesia, the PNPM-Generasi program
described in the box below.
BOX 7-2: EXAMPLE OF SUCCESSFUL HOUSEHOLD AND COMMUNITY-BASED
CASH TRANSFER PROGRAM IN INDONESIA
Promising Interventions for Increasing Community Demand for Maternal and Child Nutrition
Services in Indonesia. In 2007, the Government of Indonesia piloted two programs aimed at improving
access to basic health and education services: PNPM Generasi, and the Keluarga Harapan Program
(Hopeful Family Program, or PKH). The two programs were designed to target health and education
indicators, and were piloted in the same six provinces. The programs apply different implementation
mechanisms, however. PKH is a household-based conditional cash transfer program targeting primarily
‗supply ready‘ parts of the country. PNPM Generasi by contrast is an incentivized community block grant
program, which allows communities to address both demand- and small-scale supply-side problems that
restrict access to services. The objective of PNPM Generasi is consistent with GoI priorities and the
Millennium Development Goals (MDGs): to reduce poverty, maternal mortality, and child mortality, and
to ensure universal coverage of basic education.
PNPM-Generasi has been shown to be an effective mechanism for reaching the poor, reducing
poverty, and improving health and education indicators especially among low-baseline
communities. As part of the Generasi project, communities with assistance from trained facilitators,
identify ways to use funding provided by village-level block grants to achieve 12 health and education
targets. These include: (i) Four prenatal care visits for pregnant women; (ii) taking iron tablets during
pregnancy; (iii) Delivery assisted by a trained professional; (iv) Two postnatal care visits; (v) Complete
childhood immunizations; (vi) Ensuring monthly weight increases for infants; (vii) Monthly weighing for
children under three and biannually for under-fives; (viii) Vitamin A twice a year for under-fives; (ix)
Primary school enrollment of all children 6 to 12 years old; (x) Minimum attendance rate of 85% for all
primary school-aged children; (xi) Junior secondary school enrollment of all 13 to 15 years old; (xii)
minimum attendance rate of 85% for all junior secondary school-aged children. Health outcomes such as
birthweight, malnutrition, and the prevalence of diseases have all shown improvements in the project
locations.
In doing this, PNPM Generasi creates spaces in which elected community representatives, local
health and education workers, and trained facilitators interact. These spaces help to raise awareness
of health and education priorities, and promote discussion around barriers preventing villagers from
172
accessing services and the actions needed to overcome these constraints. Block grants provide resources
for communities to carry out activities that allow for greater access to health and education services. Built
in performance incentives encourage communities to focus on activities linked to priority health and
education goals. Finally, village implementations teams and facilitators regularly monitor village
performance towards achieving the 12 indicators. Communities use monitoring results to revise work
plans and propose new activities throughout the implementation cycle where progress is judged to be
unsatisfactory.
173
7.7 BIBLIOGRAPHY
Barker D.J.P. 1998 Mothers, Babies and Diseases in Later Life, Churchill Livingstone, London
Bleakley, Hoyt. 2010a. ―Health, Human Capital, and Development‖ Annual Review of Economics 2: 283-
310.
Bogdanovic, D., D. Nikic, B. Petrovic, B. Kocic, J. Jovanovic, M. Nikolic, and Z. Milosevic. 2007.
―Mortality of Roma population in Serbia, 2002-2005.‖ Croat Med J 48(5): 720-6.
Currie, Janet. 2009. ―Healthy, wealthy, and wise: Socioeconomic status, poor health in childhood, and
human capital development.‖ Journal of Economic Literature 47(1): 87-122.
Fundación Secretariado Gitano (FSG). 2009. Health and the Roma community: Analysis of the situation in
Europe. Madrid, Spain.
Fundación Secretariado Gitano (FSG) and Ministry of Health and Consumer Affairs. 2005. Health and the
Roma community: analysis of action proposals. Madrid, Spain.
Hajioff, S. and M. McKee. 2000. ―The health of the Roma people: a review of the published literature.‖ J
Epidemiol Community Health 54(11): 864-9.
International Organization for Migration. 2000. Social and Economic Situation of Potential Asylum
Seekers from the Slovak Republic, Bratislava.
Janevic, T., P. Sripad, E. Bradley. 2011. ―Access to Prenatal and Maternity Care among Romani Women
in Serbia and Macedonia: Research Brief. Yale Global Health Initiative.
Kremer, Michael and Rachel Glennerster. 2012. ―Improving Health in Developing Countries: Evidence
from Randomized Evaluations. In Handbook of health economics, Mark V. Pauly, Thomas G. Mcguire
and Pedro P. Barros (eds.), Oxford, Elsevier Science BV: 201-315.
Koupilova, I., H. Epstein, J. Holcik, S. Hajioff, and M. McKee. 2001. ―Health needs of the Roma
population in the Czech and Slovak Republics.‖ Soc Sci Med 53(9): 1191-204.
Kovac C. 2002. Overall health of Hungary's Roma worse than average population. British Medical
Journal. 324(7340): 755.
Open Society Institute (OSI) 2005. Mediating Romani health – policy and program opportunities.
Parekh, N. and T. Rose. 2011. ―Health Inequalities of the Roma in Europe: a Literature Review.‖ Central
european journal of public health 19(3): 139-42.
Petrova, Dimitrina. 2011. A Rights-Based Approach to Roma Health. Manuscript, World Bank.
Schaaf M. 2007. Confronting a hidden disease: TB in Roma communities. Open Society Institute (OSI).
Kumanová, Z., DÅ£ambazoviÄ?, R. (2002). Rómska rodina: na rozhranà medzi tradicionalistou a modernitou.
[Roma family: on the edge between tradition and modernity]. In: ÄŒaÄ?ipen pal o Roma. Súhrnná správa o
Rómoch na Slovensku. Bratislava: IVO.
174
Mészáros, J., Vaňo, B. (2004). ReprodukÄ?né správanie obyvateľstva v obciach s nÃzkym Å£ivotným
Å¡tandardom [Reproductive behaviour in municipalities with a low living standard]. Bratislava:
INFOSTAT.
Sepkowitz, K. A. 2006. ―Health of the world's Roma population.‖ Lancet 367(9524): 1707.WHO (2011)
MHealth New Horizons for Health Through Mobile Technologies. WHO, Geneva
http://www.who.int/goe/publications/goe_mhealth_web.pdf
Singhal, A and Lanigan J, 2007 Breastfeeding, early growth and later obesity. Obesity Reviews 8
(Suppl.1):51-54
Spence M. and M. Lewis (2009). Health and Growth. Commission on Growth and Development.
Å procha, B. (2006). PopulaÄ?ný vývoj rómskeho obyvateľstva na Slovensku. [Population development of
Roma population in Slovakia]. Master Thesis. Department of demography and geodemography. Prague:
Charles University.
Strauss, John and Duncan Thomas. 2007. ―Health over the life course.‖ UC Los Angeles: California
Center for Population Research.
Szalay T, PaÅ£itný P, Szalayová A, Frisová S, Morvay K, PetroviÄ? M and van Ginneken E. Slovakia:
Health system review. Health Systems in Transition, 2011; 13(2):1 – 200.United Nations Development
Program (UNDP) Moldova. 2006. Roma in the Republic of Moldova. Chisinau, Moldova.
United Nations Development Program (UNDP). 2004. Social assessment of Roma and HIV/AIDS in
Central East Europe. Bucharest, Romania.
United Nations Development Program (UNDP). 2003. Avoiding the dependency trap: Roma human
development challenges and opportunities.
Zeman, C. L., D. E. Depken, and D. S. Senchina. 2003. ―Roma health issues: a review of the literature and
discussion.‖ Ethn Health 8(3): 223-49.
Zoon, Ina. 2001. On the Margins: Roma and Public Services in Slovakia. Open Society Institute.
175
8 MONITORING AND EVALUATION
This short chapter provides an overview of several key monitoring and evaluation (M&E) tools that the
Slovak government can use to boost results on Roma inclusion. These include M&E tools for better
targeting of resources, to improve monitoring of gaps in human development outcomes, to promote better
planning and transparency of Roma inclusion programs, and M&E tools to promote learning and sharing
of best practices.
8.1 INTRODUCTION
Monitoring and evaluation are critical for results on Roma inclusion. Monitoring and Evaluation
(M&E) provides a toolkit for performance measurement with a view to enhancing the effectiveness of
policy interventions, to ensuring greater transparency and accountability, and to ultimately supporting
governments‘ efforts to deliver on their commitments to the inclusion of Roma. The introduction of
results-based M&E ensures that policy efforts are well managed internally, and that lessons are captured
in time and disseminated to all stakeholders and partners to provide continuous feedback for adjustments
if necessary. Results-based M&E is also the cornerstone of demonstrating the results of funding.
The April 2011 EU Framework for National Roma Integration Strategies127 calls upon EU member
states to include strong monitoring and evaluation components. This was underscored during the
recent December 2011 conference on the topic organized by the government of Slovakia in partnership
with the World Bank, as well as UNDP, the Open Society Foundation, the European Commission, the
Slovak Governance Institute, and the Poverty Action Lab Europe (J-PAL).128 A similar conference was
organized in Bulgaria in January 2012. Both conferences highlighted several well-established monitoring
and evaluation tools, and there are encouraging international collaborative efforts to promote and use these
tools to achieve results for Roma inclusion.
8.2 POLICY RECOMMENDATIONS
There are five main policy recommendations in the area of monitoring and evaluation:
1. Take advantage of the Slovak poverty map being produced and the Slovak Atlas of Roma
Communities to improve targeting of inclusion programs.
2. Bi-annually expand the EU-SILC survey implemented by the Statistical Office of the Slovak
Republic to include extra households from the poorest communities in Slovakia.
3. Ensure that the programs being financed have results frameworks in place that clearly define
inputs, activities, outputs, and impacts.
4. Undertake rigorous impact evaluations to learn what program works best; and,
5. Building an inexpensive ‗municipal best practice‘ online knowledge portal.
127
COM(2011) 173 Final: An EU Framework for National Roma Integration Strategies up to 2020.
http://ec.europa.eu/justice/policies/discrimination/docs/com_2011_173_en.pdf
128
Slovakia December 2011 Conference, http://go.worldbank.org/YAA3HUQT40; Bulgaria January 2012
Conference: http://go.worldbank.org/H715FNC250
176
8.2.1 POLICY MEASURE 1: TAKE ADVANTAGE OF THE SLOVAK POVERTY MAP
BEING PRODUCED AND THE SLOVAK MAP OF ROMA COMMUNITIES TO
IMPROVE TARGETING OF INCLUSION PROGRAMS
Small area poverty estimation is a method to identify the poorest and most vulnerable communities .
It combines household surveys such as the EU-SILC with information from the national censuses to
estimate poverty rates at, for example, municipal level, and show these on spatial maps. The World Bank,
in partnership with statistical offices around the world, has produced small area poverty maps for more
than 70 countries (as an example, see the poverty map of Bulgaria below). Throughout 2012-2013, the
WB will work with National Statistics Office of Slovakia to develop a similar map for Slovakia. The
experience of the recent crisis map of the least developed microregions by the Government of Hungary
and the Open Society Institute shows that such maps can be used to increase the allocation of EU funds
towards the poorest regions, including poor and predominantly Roma communities.
FIGURE 8-1: POVERTY MAP OF BULGARIA
The poverty map can be combined with the map of Slovak Roma communities and be combined
with a map showing where social inclusion projects are being implemented . In collaboration with the
Ministry of Labor, Family, and Social Affairs, the World Bank has produced an internet map of the
Slovak Roma communities derived from the 2004 Roma Atlas file. UNDP will update Roma Atlas in
2012-2013 as part of its work with the Ministry of Labor. In addition, the team developed a pilot map
(below) showing all Roma related projects under the ESF OP on employment and social inclusion. The
Slovak Government can consider systematically mapping the Roma inclusion projects, provide periodic
updates on these projects using standardized results framework reporting (below), and even allowing
Slovak citizens to comment on the projects and add information to the online maps.
177
FIGURE 8-2: MOLSAF-WB PILOT MAP OF ROMA INCLUSION PROJECTS IN SLOVAKIA
8.2.2 POLICY MEASURE 2: BI-ANNUALLY EXPAND THE EU-SILC SURVEY TO
INCLUDE EXTRA HOUSEHOLDS FROM THE POOREST COMMUNITIES IN
SLOVAKIA
Collecting regular and detailed household information on poverty, employment, education, housing,
health, finance, discrimination, etc. essential to monitor progress and build policies on sound
evidence. The Bratislava December 2011 conferences on M&E for Roma inclusion highlighted how the
nationwide sample of the annual EU-SILC survey – which collects similar information - can be expanded.
For example bi-annually, to include extra households from the marginalized communities identified
through the poverty maps. This would have the advantage of using the same EU-SILC questionnaire and
the same implementing agency — the National Statistics Office — to simultaneously measure outcomes
and monitor progress among both the poorest communities and the general population. As such, it would
also allow the Slovak government to perform standardized monitoring of progress on inclusion using
existing instruments for reporting. The government can allocate EU structural funds to implement this
booster. This would cost approximately Euro 250,000 per round.
8.2.3 POLICY MEASURE 3: ENSURE THAT THE PROGRAMS BEING FINANCED
HAVE RESULTS FRAMEWORKS IN PLACE THAT CLEARLY DEFINE INPUTS,
ACTIVITIES, OUTPUTS, AND IMPACTS
The draft guidance document Monitoring and Evaluation of European Cohesion Policy‘ 129
highlights that it is often difficult to demonstrate the value of a policy because programs frequently
focus on spending rather than achieving well-defined results on outcomes, such as improving job
prospects, keeping children in school and learning, etc. Fortunately, there are basic tools — results
frameworks — which clearly articulate the results chains by summarizing how the project envisions that
inputs (financial and human resources) will translate into specific activities that will in turn lead to
129
DG Regional Policy. "The Programming Period 2014-2020. Monitoring and Evaluation of european cohesion
policy - European Regional Development Fund and Cohesion Fund - Concepts and Recommendations."
178
specific — monitorable — outputs (e.g. number of unemployed who have received job training), which in
turn will contribute to achieving the ultimate desired impacts (results):
FIGURE 8-3: RESULTS FRAMEWORK
To institutionalize their use in Roma inclusion projects and programs, the Slovak government may
consider capacity building and requiring their use for funding proposals. For example, the WB is
currently collaborating with the Slovak Government and UNDP to map European Social Fund projects
using a mapping tool developed under the WB‘s Open Aid Partnership program.130
8.2.4 POLICY MEASURE 4: UNDERTAKE RIGOROUS IMPACT EVALUATIONS TO
LEARN WHAT PROGRAM WORKS BEST
The chapters on education, employment, financial inclusion, housing, and health provide policy
recommendations that can benefit from rigorous evaluations. In some cases, there may be different
ways of implementing the same policy – e.g. what is the appropriate Euro level of a conditional cash
transfer that seeks to encourage parents to enroll their children into preschool? Or, is job counseling as
cost effective as a subsidized traineeship? Does earmarking of savings accounts for, say, education
purposes, increase savings? In these, and many other cases, there are different policy options that seek to
achieve similar policy outcomes.
Social policy experimentations – also called prospective randomized impact evaluations - provide
rigorous evidence on project impacts. These impact evaluations involve randomly selected treatment
130
http://wbi.worldbank.org/wbi/open-aid-partnership
179
and comparison groups. Randomization ensures the two groups have identical characteristics at the start of
the program. This generally ensures that any differences that arise between the two groups – for example
in education or labor market outcomes – can be attributed to the intervention and not to pre-existing
differences that are often present if groups self-select into a project. For this reason, these randomized
control trials are often considered the most robust methodology for conducting program impact
evaluations. Promoted by the EU PROGRESS facility, and implemented around the globe by
governments, civil society, and international organizations, these can identify the most cost-effective
interventions and build public support around proven programs for Roma inclusion. As part of the
advisory services to the Slovak Government, the World Bank team provided technical assistance to the
Office of the Plenipotentiary, which submitted a proposal to the EU PROGRESS facility in December
2011 to carry out a social policy experiment on early childhood education. The program was developed in
partnership with the Slovak Governance Institute and the Roma Education Fund. It was selected by the EU
PROGRESS facility in August 2012.
BOX 8-1: SOCIAL POLICY EXPERIMENTATION AND THE DANISH LABOR MARKET
AUTHORITY
The Danish Labor Market Authority (LMA) has taken a very proactive approach towards building
up evidence on its employment 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 internet (www.ams.dk
and www.jobindsats.dk ).
In designing and carrying out these evaluations, the LMA works closely with external evaluators –
Danish academics – and a selection of its affiliated job centers. Denmark has 98 municipalities, with
93 integrated job centers for all job seekers (insured and uninsured). There are also 4 regional employment
councils.
So far it has completed 4 randomized control trials, 2 evaluations are ongoing, and a new one is
planned starting August 2012 serving particularly vulnerable groups. In each evaluation, the
comparison group is offered the regular package of employment services while the treatment group
receives something ‗extra‘. For example, the first evaluation consisted of an intervention whereby job
seekers were offered bi-weekly counseling as opposed to regular counseling every three months. In the
upcoming evaluation, a ‗social mentoring‘ pilot will be evaluated. The target group consists of youth 18-
29 years old far from the labor market. Local job centers will be provided with funding to hire social
mentors who will give intensive counseling to vulnerable youth, including advising on accessing social
services and education and training opportunities.
Additionally, seeking direct community level feedback on interventions can add critical local
perspectives for social innovation. For example, as part of the monitoring and evaluation component of
the ongoing European Parliament Roma Pilots, the WB, UNDP, Roma Education Fund, and the Slovak
Governance Institute designed and implemented a local system of data collection whereby partner
organizations — mainly small NGOs — collect beneficiary level outcome indicators and project feedback
that is entered in an online tool for project monitoring accessible by project partners. Slovakia can
similarly implement such monitoring tools.
180
8.2.5 POLICY MEASURE 5: BUILDING AN INEXPENSIVE ‗MUNICIPAL BEST
PRACTICE‘ ONLINE KNOWLEDGE PORTAL.
The Slovak government can consider building an inexpensive ‗municipal best practice‘ online
knowledge portal. Various mayors and civil society organizations in different municipalities in Slovakia
implement very innovative ideas to promote Roma integration. For example, to build skills by engaging
unemployed Roma on activation in real municipal construction activities supervised by an expert, or to
undertake land titling by setting up a partnership with the NGO ETP providing loans which families pay
back as soon as they become eligible for the housing allowance. Many mayors may be interested in
addressing similar issues in their communities but do not know where to start. A municipal best practice
database, enforced by a small regional team of experts, can help the spread of innovative ideas. This ‗best
practice‘ database can be part of an integrated knowledge portal, which also includes the poverty map, the
Slovak Roma atlas map, and a mapping of the social inclusion projects. To make the knowledge portal
both inexpensive to operate and to keep up-to-date, it may be a dynamic website whereby local user can
upload information.
181
9 EU FINANCING FOR ROMA INCLUSION
This chapter reviews the opportunities that exist to use EU financing for Roma inclusion, particularly
during the next 2014-2020 EU programming period. Part of the chapter is backward looking, reviewing
past experiences on the use of EU finances for Roma inclusion to draw lessons for future programming. It
highlights six key building blocks to consider for the next programming period: a common strategic
framework, thematic objectives, structural funds regulation, partnership agreements between the
Commission and Member States, ex-ante conditionality, and efficient and flexible use of EU funds.
9.1 EU FRAMEWORK FOR ROMA INCLUSION
The issues of Roma inclusion have been actively taken on board by the European Commission in
2009/2010, with the publication of the first Communication on Roma inclusion in April 2010, ahead
of the Second Roma Summit in Cordoba.131 While the document clearly states that Roma inclusion is a
matter for policy intervention of the Member States at national level, it does recognize, for the first time,
that the poverty and exclusion of the largest European minority is a matter of concern for the EU as a
whole. In April 2011, the Commission went a step further in a second Communication, entitled: EU
Framework for National Roma Integration Strategies up to 2020132 Communication. This document sets
out four key aeas for action at national level, in the areas of education, employment, health and housing. It
calls on MSs to prepare or revise national strategies for the integration of the Roma in line with the EU
Framework and reprogram EU Structural Funds to support these policy objectives with financial
commitments. It aims to have tangible improvement in the inclusion of the Roma by the end of the current
decade.
In May 2012, the Commission published the first overview and brief evaluation of the submitted
NRISs in a Communication entitled: National Roma Integration Strategies: a first step in the
implementation of the EU Framework133. The May 2012 Communication on the EU Framework134 makes
the link between Roma inclusion and Europe 2020 explicit: ―[f]or Member States with a la rger Roma
population making sufficient progress towards the Europe 2020 employment, social inclusion and
education targets will require addressing explicitly and swiftly the situation of the Roma.‖ 135 Three out of
the five headline targets of the Europe 2020 strategy are directly related to Roma inclusion: i) promoting
social inclusion and fight against poverty; ii) ii) reducing the number of school dropouts and increasing
atendance in tertiarary education; and iii) raising employment levels. And, as mentioned in the
introeduction, in the case of Slovakia, the Country Specific Recommendations (released by the
Commission end-May 2012) 136 strongly reinforce this message by making the inclusion of Roma children
131
Communication from the Commission to the Council, the European Parliament, the European Economic and
Social Committee and the Committee of the Regions, The social and economic integration of the Roma in Europe ,
7.4.2010 COM/2010/0133 final
132
Communication from the Commission to the Council, the European Parliament, the European Economic and
Social Committee and the Committee of the Regions, Brussels, 5.4.2011 COM(2011) 173 final
133
Communication from the Commission to the Council, the European Parliament, the European Economic and
Social Committee and the Committee of the Regions, National Roma Integration Strategies: a first step in the
implementation of the EU Framework, Brussels, 21.5.2012, COM(2012) 226 final
134
Ibidem, footnote no. 4.
135
Ibidem, p. 13.
136
Recommendation for a COUNCIL RECOMMENDATION on Slovakia‘s 2012 national reform programme and
delivering a Council opinion on Slovakia‘s stability programme for 2012 -2015, {SWD(2012) 326}, Brussels,
30.5.2012 COM(2012) 326 final provisoire
182
and youth in mainstream education and adults in the labour market one of only seven specific
recommendations to the Slovak Government.
The EU Framework urges the Member States to take measures to ensure that EU funding available
for the period 2007-2013 makes a tangible difference to Roma communities. Member States are
encouraged to adapt their existing operational programmes, which are co-financed by the Structural Funds
and the EAFRD, so that they better support projects targeted at disadvantaged Roma communities. It also
calls on Member States to make better use of the sizeable amount of technical assistance available under
the Structural Funds (up to 4% of programme allocation) to better design, manage, monitor and evaluate
projects targeting the Roma communities. Where Member States lack the know-how or capacity to
manage projects effectively, the EU Framework suggests that the management and implementation of part
of their national programmes could be entrusted to intermediary bodies. These actors, who should have
proven experience in Roma inclusion on the ground, could include international organisations, regional
development bodies or non-governmental organisations.
The Communication is accompanied by a Commission Staff Working Document137, which highlights
key elements and remaining gaps of each individual NRIS. The Commission‘s assessment focuses on
the four key areas of intervention set out in the EU Framework -- education, employment, healthcare and
housing – as well as on how structural requirements, such as cooperation with civil society and with
regional and local authorities are addresses in individual NRISs. It also looks at monitoring,
antidiscrimination measures and establishment of a national contact point, as well as the identification of
funding.
Overall, the Slovak strategy is praised for using a holistic approach and for providing an accurate
description of the dire situation of the Roma in the Slovak Republic. The Staff document recognizes
the emphasis on cooperation with Roma civil society and local and regional authorities, as well as the
need to work with the majority population to tackle widespread prejudices and stereotypes. The overall
budget allocation is quantified at €141m for the period 2011-2015. However, individual measures in the
action plan do not always specify budget implications. It should, therefore, be improved by better
quantification of required funding. The document stresses the need to better align Operational Programs
for the programming period 2014-2020 with the NRIS by drawing lessons from the current programming
period concerning the implementation of Roma inclusion projects financed by the Structural Funds.
Finally, the commitment to a strong involvement of civil society and local authorities in planning and
implementation of projects should be ensured.
The success of the Strategy will ultimately depend on its implementation. At the moment, the Slovak
Government has not drafted action plans for the key areas through 2020. This was partly due to the fact
that the revised National Action Plan for the Decade of Roma Inclusion for years 2005 – 2015 were
adopted with a delay – only in June 2011. Thus the action plans on education, employment, housing and
health drafted in the framework of the Decade are used also for the NRIS. Since the Strategy is
considerably more ambitious and far reaching that the Decade Action Plans, it is not clear at the moment
how its commitments will be transformed into actionable items and implemented. Also, the Decade
extends only through 2015, while the Strategy will extend till 2020. It will be up to the new government to
137
Commission Staff Working Document Accompanying the document Communication from the Commission to the
Council, the European Parliament, the European Economic and Social Committee and the Committee of the Regions,
National Roma Integration Strategies: a first step in the implementation of the EU Framework , Brussels, 21.5.2012,
SWD(2012) 133 final.
183
prepare updated action plans for the NRIS. Perhaps not surprisingly, the most problematic area in the
current Action Plans is that of education where all activities are due in 2015.
9.2 SLOVAKIA AND STRUCTURAL FUNDS 2007-2013
Although Slovakia explicitly identified support to marginalized Roma communities in the 2007-2013
programming period as a ―horizontal priority: Marginalized Roma Communities (MRC)�, little
money has been spent. An indicative allocation of about 200 million EUR to support this priority though
a "comprehensive approach" was ear marked. The comprehensive approach has been expected to be
delivered through selected Local Strategies of Comprehensive Approach (about 150 out of over 230 were
selected) which are in turn expected to be financed through six separate Operational Programmes: (1)
Regional OP, (2) OP Employment and Social Inclusion, (3) OP Education, (4) OP Environment, (5) OP
Competitiveness and Economic Growth, and (6) OP Health. However, as of May 2012, only about Euro
11 million out of the earmarked Euro 200 million had been used to finance any projects from the local
strategies, with an additional approx. 5 million Euro of calls outstanding at the time of this writing.
Several factors explain the low number of projects and low absorption of funds under the local
strategies funding for Roma inclusion:
(1) Preparing high-quality projects with potential to achieve real impact and positive outcomes
for the marginalized Roma communities is a key constraint. The Commission provides to the
Member States with the Structural Funds generous technical assistance (TA) resources – four
percent (4%) of the total allocation. These are meant to enable the authorities in the countries to
manage the programmes as well as invest in project preparation, if and when needed. At the
moment little, if any, of the TA resources are used for project preparation on the local level and in
Roma communities.
(2) The architecture of the current Roma integration funding scheme is overly complex. The
MRC priority is a ―horizontal priority‖, with no specific allocation or funding source, and it
covers a number of OPs. The Local Strategies of Comprehensive Approach have a notional
allocation of 200 million Euro, but as this funding is expected to come from six separate OPs,
there are six different Managing Authorities, each with a separate set of incentives, to deal with.
Incidentally, in the Slovak original, the ―comprehensive approach‖ is called ―complex approach‖.
The name is particularly fitting in this case.
(3) The coordinator of Roma integration, the horizontal priority MRC, and the local strategies,
is the Office of the Plenipotentiary. It has neither the official authority, nor direct access to
resources, nor staffing capacity to carry out the coordination task effectively. Due to this lack
of authority or standing vis-a-vis the Managing Authorities in the individual ministries, it is
exceedingly difficult to plan any coherent and integrated actions targeting the Roma communities.
Due to the insufficient staffing and budget, it does not have capacity for carrying out analytical
work, policy development and active outreach.
BOX 9-1: HIGH LEVEL EVENT ON ROMA INCLUSION IN 2011 .
The European Commission organized in May 2011 a High Level Event on the contribution of the
Structural Funds to Roma integration in Slovakia (23-25 May) with the express purpose of bringing the
184
Roma inclusion agenda into stronger focus and fueling a momentum for a more active use of the funds
and implementation of the Local Strategies of Comprehensive Approach. The event was also a very good
platform for sharing information, good practices, projects and lessons learned. The main conclusions of
this HLE reiterated the recognition of the horizontal priority and the comprehensive approach with the
indicative allocation as an innovative approach that Slovakia introduced. At the same time, some
presenters expressed concern over the low usage of funds through the local strategies. In addition, HLE
participants made the following points:
ï‚· The 2010 amendment of Article 7 of the ERDF Regulation has allowed the financing of housing
actions in rural areas, including the replacement of existing housing, for marginalised communities.
The Commission encouraged Slovakia, especially given the very bad housing conditions of many
Roma communities to experiment and pilot housing interventions, combining ERDF and ESF funds.
ï‚· Key to Roma inclusion is the involvement of all relevant stakeholders, in particular representatives
from Roma communities themselves, to ensure the appropriateness of the proposed integration
measures and assure their buy-in and hence improve the chances of smooth implementation. Lessons
learned from the implementation of the comprehensive approach can be analysed and reflected
upon, especially for the next programming exercise.
• Not much progress can be made in Roma inclusion unless and until the majority population is
also targeted with a well prepared information campaign. This is important not only to raise
awareness of the conditions the marginalised Roma live in and fight prejudices and stereotypes, but
also explain the economic benefits and importance of including the Roma population in the mainstream
education and labour market. This point has just been made in the official Country Specific
Recommendations published by the Commission on 30 May, 2012, as part of the new economic
governance cycle called European Semester (see more below).
A strong monitoring system and data collection can be introduced. It will help in developing evidence-
based policies which is very much needed in the area of Roma inclusion. The recently published results of
the household survey carried out in the summer 2011 by UNDP/World Bank and financed by DG
Regional Policy, as well as the earlier 2010 UNDP survey, provide excellent baselines for any future data
collection exercises. They can also prove instrumental to any evaluation of the impact and targeting of
various policies and projects.
9.3 EU FUNDING FOR ROMA INCLUSION IN THE PROGRAMMING PERIOD
(2007-2013 )
Availability of financial resources does not seem to be the primary constraint to increasing the
integration of the Roma population and to a better design and implementation of inclusion policies
in Slovakia. For the current 2007-2013 programming period, the EU has several policies and programmes
with different financial instruments (funds and programmes) that can provide financing for Roma
inclusion related activities. The most important among them is the Cohesion Policy with a budget of 350
billion Euros which is implemented through financial instruments known as the EU Structural Funds.
These cover a range of different areas. The Structural Funds operate under an implementation modality
called ―shared management‖ between the EC and the national or regional authorities. Once the
Commission approves an Operational Programme, all financing and procurement decisions rest with the
national/regional authorities (so called ―Managing authorities‖). The EC retains financial control and
185
oversight. The two main Structural Funds that can target social inclusion are the European Social Fund
(ESF) and the European Regional Development Fund (ERDF). Slovakia‘s total envelope during the curent
programming period is 11.7 billion Euro, of which 6.5 billion is for the ERDF and about 1 billion for the
ESF.
European Social Fund (ESF) The ESF invest primarily in ―soft‖ measures supporting employment and
the improvement of living standards; it helps people enhance their education and skills. The ESF invests to
improve the situation of the Roma population through funding, for example, education and employment
related activities, guidance and counseling, and micro credit to support self-employment. (For changes
proposed to the ESF for the next programming period see Box 9-2.)
European Regional Development Fund (ERDF) The ERDF supports ―hard‖ investments in
infrastructure, environment, regional development, and enhanced competitivenes. For the Roma
population, apart from improving access to basic infranstructure such as roads, water and waste water,
ERDF can invest in a variety of social infrastructure from kindergardens to community centers. In 2010,
Article 7 of the ERDF regulation was amended to allow the ERDF to invest in the building and
refurbishment of housing for marginalised communities, which includes many Roma communities.
European Agricultural Fund for Rural Development (EARDF) The European Agricultural Fund for
Rural Development is implemented, like the Structural Funds, through shared management. It aims to
improve the competitiveness of the EU‘s agriculture and forestry, develop the EU‘s environment and
countryside, boost the quality of life in rural areas, and encourage the diversification of economic
activities. The EAFRD can thus be used to help Roma communities – a great proportion of which live in
rural areas - by funding economic activities and small local infrastructure in rural areas. During the current
financing period, little of the EAFRD resources were aimed at martginalised Roma communities.
Fundamental Rights and Citizenship Programme Part of the General Programme ‗Fundamental Rights
and Justice‘, the Fundamental Rights and Citizenship Programme aims to help protect children‘s rights
and combat racism. The Programme supports actions that can benefit Roma people, such as projects that
increase mutual understanding between Roma and non-Roma people in order to breakdown stereotypes.
Funding is available through the European Commission, via direct calls for proposals.
Other EU sources of funding There is a variety of programmes in the European Commission which are
not specifically targetting the Roma population. Nevertheless, Roma specific projects can benefit from
such funding. Among the most relevant ones are: the PROGRESS program, which supports supporting the
development and coordination of EU policy in five areas (Employment; Social inclusion and social
protection; Working conditions; Anti-discrimination and Gender equality); the EU‘s Employment and
Solidarity Programme; Daphne III – a programme to combat violence against children, young people and
women, and to protect victims and groups at risk; the Lifelong Learning Programme; the Youth in Action
Programme; the Culture Programme; and the Health Programme. Unlike the ESF, ERDF and the
EAFRD, all these other programmes are directly managed by the European Commission that makes the
selection of individual projects and subsequently the funding decisions.
Over the last two decades, Roma projects in Slovakia received support from a variety of public and
private sources, bilateral and multilateral institutions, as well as Slovak government programs. This
support resulted in a number of interesting and worthwhile projects, ranging from education, through
health, housing, to employment, to cultural and linguistic identity. Undoubtedly these projects helped
individual project beneficiaries and/or their communities.
186
Due to their fragmentation and mostly unsustainable nature, these individual projects do not offer
systemic lessons on how to alleviate poverty among the Roma communities and improve their
overall economic and social status. Even though a good number of the projects were conceived as pilots
to test different approaches and bring the successful ones to scale, they invariably stopped after their
funding ended. One example for all is an EU pre-accession PHARE financed project to train and deploy
medical assistants to work in segregated communities as public health workers and cultural connectors of
the Roma community to the mainstream health system. In 2009, the project had about two dozens of such
assistants. Despite the apparent success of this project according to all relevant stakeholders and the
documented need for such interventions in over 700 segregated or separated settlements in Slovakia, the
project ended once the EU resources came to an end. Projects have not become programs and did not even
have a chance to develop into policies. This is true despite the participation of the Slovak Republic in the
Decade of Roma Inclusion, where Slovakia held the rotating Presidency in 2010.
Significant hopes were initially pinned by promoters of Roma inclusion on the horizontal priority
―Marginalised Roma Communities� and especially on the “Local Strategies of Comprehensive
Approach� in the current programming period 2007-2013 of the Structural Funds. Despite the
generous notional allocation of 200 million Euro for the local strategies, little has been invested from it in
Roma inclusion at national or local level. In April 2012, one year before the end of the programming
period, only some 16 million of the 200 million have been contracted. The complex design of the
horizontal priority where six different Operational Programmes contained the actual allocations for Roma
inclusion, the fact that the coordinator – Office of the Plenipotentiary for Roma Communities – does not
have any leverage vis-Ã -vis the managing authorities of the individual OPs, and overly complex
bureaucratic processes have all contributed to the low absorption of the funds so far. More importantly
though, the horizontal priority as currently implemented does not seem to have not resulted in a better
policy design and more systematic push for Roma inclusion, or scaling up the piloted approaches.
9.4 NEXT EU PROGRAMMING PERIOD (2014-2020)
The guiding strategy for the next generation of the Structural Funds is Europe 2020 strategy of
smart, sustainable and inclusive growth.138 The Operational Programmes under the Structural Funds are
expected to be, for the first time in the next programming period, fully aligned with the objectives and
targets of Europe 2020. To achieve this, the Commission has proposed a variety of instruments: the newly
proposed Common Strategic Framework is intented to ensure that Cohesion Policy investments (through
the ERDF, ESF and the Cohesion Fund), and also the rural development fund (EAFRD) and the maritime
and fisheries instrument, will all fall within a subset of thematic objectives that have been developed
around the Europe 2020 goals and headline targets. During negotiations of the next programming period,
Member States and the EC are expected to enter into Partnership Agreements which will set out: i)
Member States‘ priorities and the specific thematic objectives the Member States will pursue with the
Structural Funds financing, and ii) ex-ante conditions that have to be fulfilled in order to receive the
financing. The Partnership Agreements will contain ex-ante conditionality, which will define, by sectors,
the minimum conditions that need to be met for a Member State to receive financing from EU funds.
138
European Commission (2010) ‗Europe 2020. A European strategy for smart, sustainable and inclusive growth’ .
3.3. 2010 COM(2010)2020.
187
9.5 KEY BUILDING BLOCKS OF THE NEXT PROGRAMM ING EXERCISE FOR
THE FUNDING PERIOD 2014-2020
9.5.1 COMMON STRATEGIC FRAMEWORK
The European Commission presented on 14 March 2012 a "Common Strategic Framework" (CSF)
for the use of EU funds (see Annexes 2 and 3). The CSF follows the proposals of Cohesion Policy
regulations and to help Member States prepare for the next programming period in a more strategic
manner, concentrating their investments more consistently on Europe 2020 taagets. In other words,
Member States will have much less freedom to chose where to invest EU funds than has been the case in
the previous programming rounds. The CSF is also intented to encourage combination and coordination of
investments from the individual funds, to maximise the impact of EU investments. On the basis of the
CSF, national and regional authorities will negotiate with the EC their ―Partnership Agreements,‖
committing themselves to meeting Europe's growth and jobs targets for 2020.
BOX 9-2: FORESEEN CHANGES FOR THE EUROPEAN SOCIAL FUND
The European Social Fund is the key EU instrument for investment in people. Target areas are: improving
employment opportunities, promoting education and life-long learning, enhancing social inclusion,
contributing to combating poverty and developing institutional capacity of public administration. Around
10 million final beneficiaries have been supported by ESF every year. With the current jobs and
employment crisis, the new legislative proposal aims at reinforcing the role of the ESF in the future:
 increasing the minimum share of the budget for each category of regions -- 25% for less
developed regions, 40% for transition regions and 52% for more developed ones. This share corresponds
to at least €84 billion for the ESF, compared to the current €75 billion;
 concentrating the ESF on a limited number of objectives and investment priorities in line with the
Europe 2020 Strategy;
 dedicating a minimum share of 20% of the ESF to social inclusion actions and combating poverty;
 combating youth unemployment, promoting active and healthy ageing, and supporting the most
disadvantaged groups and marginalized communities such as Roma;
 increasing support to social innovation, i.e. testing and scaling up innovative solutions to
addressing social needs and providing social services;
 ecouraging the participation of social partners and the civil society in the implementation of the
ESF, through capacity building, promotion of community-led local development strategies and the
simplification of the delivery system. (Rules governing the reimbursement of projects by the ESF are
expected to be simplified, in particular for "small" beneficiaries: NGO's, SMEs, and others who make up
at least 50% of ESF recipients; and
 making equipment linked to investments in social and human capital eligible for support from the
ESF.
9.5.2 THEMATIC OBJECTIVES
Compared to the past, the new programming period will require a fuller alignment of Structural
Funds financing with the objectives of the Europe 2020 Agenda. The Commission has thus fixed
minimum allocations under the Structural Funds for a number of EU priority areas. For example, for more
developed and transition regions, at least 80% of ERDF resources at national level needs to be allocated to
188
energy efficiency and renewable energy sources, innovation and the improvement of the competitiveness
of small and medium-sized enterprises (SMEs). This amount will be 50% in less developed regions,
reflecting their broader development needs. ESF investments need to be fully aligned with EU objectives
and targets on employment, education and poverty reduction. A minimum of 20% of the national ESF
allocation will have to be allocated to investments combating social exclusion and poverty.
To reinforce this strategic programming process and maximize impact of EU funds, the
Commission has proposed to define a list of 11 thematic objectives in the Regulation (below), aligned
with the Europe 2020 Strategy. During the forthcoming negotiations of the programming documents,
Member States will have to choose only a subset of these thematic objectives for EU funds financing.
1. Strengthening research, technological development and innovation;
2. Enhancing access to and use and quality of information and communication technologies;
3. Enhancing the competitiveness of small and medium-sized enterprises, the agricultural sector (for
the EAFRD) and fisheries and aquaculture sector (for the EMFF);
4. Supporting the shift towards a low-carbon economy in all sectors;
5. Promoting climate change adaptation, risk prevention and management;
6. Protecting the environment and promoting resource efficiency;
7. Promoting sustainable transport and removing bottlenecks in key network infrastructures;
8. Promoting employment and supporting labour mobility;
9. Promoting social inclusion and combating poverty;
10. Investing in education, skills and lifelong learning;
11. Enhancing institutional capacity and an efficient public administration.
9.5.3 STRUCTURAL FUNDS REGULATIONS
Legislative proposals for Cohesion Policy for the 2014-2020 period were adopted by the European
Commission in October 2011, and by the Concil (body representing the Member States), a new
negotiated version was adopted in April 2012. At the time of this writing, the legislative package is
being discussed by the European Parliament. The new Regulations need to be adopted in 2012 or early
2013, to be ready for the start of the new programming period in 2014.
The guiding principles of the next programming period are:
ï‚· focusing resources on a few priorities aligned with the Europe 2020 Strategy
ï‚· defining clear and measurable targets for investments
ï‚· introduction of ex-conditionality and performance incentives
ï‚· increasing the leverage effect of investments through private sector finance
ï‚· concentrating on the poorest Member States and regions
Additionally, the new regulations are supposed to further simplify the often overly complex regulatory
regime for the management and use of the funds. ―Simplification‖ should entail the introduction og
simplified reimbursement rules; allow the implementation of funds on the basis of joint action plans
reimbursed on the basis of results; harmonising eligibility rules and the management and control systems
between the different EU funds, etc.
189
9.5.4 PARTNERSHIP AGREEMENTS BETWEEN THE COMMISSION AND THE
MEMBER STATES
In 2013, each Member State will be asked to draw up a Partnership Agreement where they will
assess their development needs and define their national priorities supporting their National
Reform Programmes and concrete actions to achieve their national targets for delivering on the
Europe 2020 strategy. While the Common Strategic Framework contains the EU's top priorities and will
apply to all funds (including rural development and fisheries), the Partnership Agreements will be tailored
for and with each Member States. Member States will be allowed to combine ERDF, ESF and the
Cohesion Fund in "multi-fund" programmes to better suit their growth plans, improve coordination on the
ground and achieve integrated development. The purpose is to allow the biggest impact on the ground.
The plan at the moment is to allow suspension or cancellation of funding for failure to achieve progress
and targets agreed upon the the Partnership Agreement. It remains to be seen whether this intention will
survive the process of negotations of the legislative package.
The Partnership Agreement will contain notably:
ï‚· choice of the above thematic objectives
ï‚· investment priorities for each thematic objective
ï‚· ex-ante conditions as a pre-requisite to EU funding
ï‚· targets that the Member States plan to reach by the end of the programming period
ï‚· performance indicators and milestones
9.5.5 EX-ANTE CONDITIONALITY
The biggest novelty in the new programming period is the introduction of ―ex-ante conditionality‖.
EU funding is seen as a strong incentive to deliver Europe 2020 objectives. Hence in all investment
areas, some "ex-ante" conditions will need to be in place before the funds are disbursed (for instance, the
proper functioning of public procurement systems). For social inclusion and anti-poverty, the ex-ante
conditionality implies the existance of both a national strategic framework for poverty reduction, aiming at
active inclusion, as well as a national Roma inclusion strategy, where applicable. Conditions are also laid
down for reporting on the fulfillment of ex-ante conditionalities, in the absence of which interim payments
could be suspended. At the time of this writing, the ex-ante conditionality provisions are being debated,
with the Commission pushing for stronger provisions and the Member States (through the Council
formation where the regulations are discussed) trying to water the provisions down. One of the key actor
in this process will now be the European Parliament that will have to vote on the whole legislative
package. But all in all, the concept and the need for some ex-ante conditionality has already been accepted
by the Member States. Some iteration of the current provisions is likely to become law.
9.5.6 FLEXIBLE COMBINATION OF EU FUNDS FOR ROMA INCLUSION
The EU Framework specifically recommends making Structural Funds and other EU funds more
accessible for Roma inclusion projects; making full use of the amendment of the ERDF regulations for
housing, as part of integrated actions; making greater use of technical assistance by the EU; developing
results-oriented projects and increasing their duration; ensuring a more effective, flexible and integrated
combination of EU funds; strengthening the partnership principle in the planning, implementation,
monitoring and evaluation in the programmes of the Common Strategic Framework funds and the
190
capacities of NGOs specifically in ESF funded programmes; making access to funds easier, especially to
small beneficiaries, through simplified cost options; and developing robust monitoring mechanisms and
evaluating results.
The Commission‘s draft Structural Funds regulations propose the possibility to combine more
systematically ESF and ERDF programs for social inclusion. This proposal opens up interesting
possibilities for integrated actions, for example in disadvantaged micro-regions, settlements or
neighbourhoods, and for the simplification of the planning, programming and coordination of Roma
inclusion projects. It has been taken on board by France, Greece, Hungary, Italy, Romania, Slovakia and
Spain. In some cases a valuable reflection on existing managing models and institutional makeups and on
finding the most appropriate tools to achieve effective results on the ground is made. Slovakia, in
particular, suggests the establishment of a multi-fund operational programme using ESF and ERDF funds,
which would finance programmes centred on inclusive infrastructural development and public services as
well as specific labour inclusion and educational programmes.
9.6 POLICY RECOMMENDATIONS
The new Slovak National Roma Integration Strategy of January 2012 has been adopted at an
opportune moment: European Commission is fully committed to monitoring progress individual
Member States make on Roma inclusion and will require annual reporting. At the same time, as the end of
the current financing and programming period is drawing to a close (end 2013), regulations for the new
financing period (2014-2020) are making their way through the legislative process.139
For Slovakia, Roma inclusion -- especially in education and the labour market -- has become one of
only seven ―Country Specific Recommendations‖ (CSRs). The CSRs are those priority areas that the
Commission will track and review very closely on an annual basis as part of the new economic
governance process called the European Semester.Structural Funds in the next programming period will
be required to be closely aligned with the Europe 2020 Strategy of smart, sustainable and inclusive
growth. Due to the focus on the third pillar -- inclusive growth, Roma inclusion will be a key area where
the Commission will actively engage the Member States during the forthcoming negotiations (end 2012
and 2013).
It is important that Slovakia takes full advantage of the opportunities offered by Structural Funds
in the next programming period 2014-2020 to reduce the various gaps between the Roma and the
majority population, focusing on integrated actions that make a difference in areas where large Roma
communities experience segregation and severe poverty. To improve on the current system of using the
Structural Funds to take advantage of the new EU financing for Roma inclusion, Slovak Government may
consider the following policy measures:
139
The Commission adopted the legislative package in October 2011. Following negotiations in the relevant Council
working group, the Council adopted a new text in April 2012. At the time of this writing, the European Parliament is
deliberating the Council text. Pursuant to the Lisbon Treaty, all these regulations fall under the so-called ―co-
decision‖ procedure, where both the Council and the Parliament have to approve the legal text. The EP is expected to
vote on the set of regulations in the [Fall 2012].
191
9.6.1 POLICY MEASURE 1A: DEVELOP ONE 2014-2020 OPERATIONAL PROGRAMME
FOR ROMA INCLUSION
Policy Measure 1a: Develop one 2014-2020 Operational Programme for Roma Inclusion and finance
it jointly from the European Social Fund (ESF) and the European Regional Development Fund
(ERDF) Under this option, the current coordinator, the Office of the Plenipotentiary, would become a
Managing Authority (body in charge of an OP), as indicated in Figure 9-1 below.
FIGURE 9-1: DEDICATED OPERATIONAL PROGRAMME ON ROMA INCLUSION
192
9.6.2 POLICY MEASURE 1B: ALTERNATIVELY, ROMA INCLUSION COULD BECOME
A SEPARATE PRIORITY AXIS
Policy Measure 1b: Alternatively, Roma inclusion could become a separate Priority Axis within a
larger Operational Programme, again jointly financed by the ESF and the ERDF (see Figure 15).
Under this option, the Office would become an ―intermediate body‖ (body delegated by a Managing
Authority to be in charge of a specific priority axis) as shown in Figure 9-2.
FIGURE 9-2: ROMA INCLUSION AS A PRIORITY AXIS IN A BROADER OPERATIONAL PROGRAMME
COMMON ESF ERDF EAFRD COHESION EMFF
STRATEGIC FUND EUROPEAN
EUROPEAN EUROPEAN EUROPEAN MARITIME
FRAMEWORK REGIONAL AND
SOCIAL AGRICULTURAL
DEVELOPMENT FUND FOR RURAL FISHERIES
EU FUNDS FUND FUND DEVELOPMENT FUND
MANAGING OPERATIONAL OPERATIONAL OPERATIONAL OPERATIONAL
AUTHORITY PROGRAMME PROGRAMME PROGRAMME PROGRAMME
3
(MA) 1 2 ….
PRIORITY
AXIS
INTERMEDIATE
PRIORITY PRIORITY …
BODIES PRIORITY AXIS
(DELEGATED AXIS AXIS
BY MA TO ROMA
… …
MANAGE) INCLUSION
ANTI-DISCRIMINATION
EMPLOYMENT
EDUCATION
MEASURES
HOUSING
HEALTH
Picture No. 2. Roma Inclusion as a Priority Axis in a broader Operational Programme
193
9.6.3 POLICY MEASURE 2: STRENGTHEN THE CAPACITY OF THE OFFICE OF THE
PLENIPOTENTIARY
Policy Measure 2: Take advantage of the EU Structural Funds to strengthen the capacity of the
Office of the Plenipotentiary to design, promote and manage inclusion policies and manage a Structural
Funds programme. The Structural Funds themselves can contribute to building this capacity through the
substantial amounts available from the EU budget for technical assistance (up to 4% of OP allocation).
9.6.4 POLICY MEASURE 3: ASSIST THE LOCAL COMMUNITIES AND
MUNICIPALITIES DEVELOP HIGH QUALITY PROPOSALS
Policy Measure 3: Take advantage of the EU Structural Funds‘ technical assistance resources to
assist the local communities and municipalities develop high quality proposals with appropriate
evaluation built-in up front. Again, the Structural Funds themselves can contribute to building this
capacity or hiring technical expertise elsewhere through the substantial amounts available from the EU
budget for technical assistance.
9.6.5 POLICY MEASURE 4: SIMPLIFY THE USE OF ESF AND ERDF, ESPECIALLY
FOR LOCALLY LED INITIATIVES AND SMALL PROJECTS
Policy Measure 4: Simplify the use of ESF and ERDF, especially for locally led initiatives and small
projects, often implemented by NGOs. Already in the current programming period, the Commission
greatly simplified use of funds for small projects under 50,000 Euros where the beneficiaries no longer
need to keep certain complicated accounting measures; once approved, all that projects need to
demonstrate is the implementation of the proposal; no other reporting is necessary. Such simplification
will be applied also in the next programming period. Several countries have taken advantage of such a
simplified process, but not all countries. Slovakia would be well advised to apply these leaner procedures
to small local development initiatives that cannot have the necessary capacity to implement full Structural
Fund financing management and reporting system.
9.6.6 POLICY MEASURE 5: DEVELOP A SYSTEM OF PAYMENT/REIMBURSEMENT
FOR PROJECT PROMOTERS OF SMALL PROJECTS THAT DOES NOT UNDULY
PENALIZE
Policy Measure 5: Develop a system of payment/reimbursement for project promoters of small
projects (beneficiaries of small grants) that does not unduly penalize them. Since the Structural
Funds reimburse expenditure ex post, countries often reimburse projects promoters also that way. This can
complicate or completely prevent implementation for smaller NGOs or even municipalities that lack
sufficient cash flow to pay for expenditures up front. Fortunately, the Commission gives Member States
yearly payment advances, and these can be passed on to final beneficiaries. Some countries routinely do
so. For projects implemented in poor parts of Slovakia, including Roma communities, it would be
important to use this advance and provide up front financing.
194
9.6.7 POLICY MEASURE 6: USE THE POVERTY MAPS TO BETTER TARGET THE
FUNDS
Policy Measure 6: Use the poverty map that will be prepared for targeting Structural Funds
financing to target inclusion projects to the most needy communities. The poverty map is described in
more detail in the Monitoring and Evaluation chapter.
195
9.8 ANNEX 1: PROGRAMMING PERIOD 2014-2020 – STRUCTURAL FUNDS
PROGRAMMING FLOW
IMPLEMENTATION OF STRUCTURAL FUNDS
PROGRAMMES
EUROPEAN COMMISSION
DG REGIONAL POLICY & DG
EMPLOYMENT
OPERATIONAL PROGRAMMES
Managing Authorities
INTERMEDIATE BODIES
(OPTIONAL STEP IF DELEGATED)
BY MA IN A SPECIFIC OP)
Call for Proposals
Benef
iciari
es
Tenders
Companies
196
9.9 ANNEX 2: COMMON STRATEGIC FRAMEWORK – SOURCES OF
FUNDING
 CSF establishes provisions for five funds: European Regional Development
Fund (ERDF), European Social Fund (ESF), Cohesion Fund (CF), European
Agricultural Fund for Rural Development (EAFRD) and European Maritime
and Fisheries Fund (EMFF)
 CSF proposes main targets and policy objectives for each of the 11 thematic
objectives and key actions (for the purpose of Roma integration, objective
no. 9 Social Inclusion and Combating Poverty)
 CSF provides integration and synergies between the five funds (for Roma
inclusion purposes, the three more relevant funds are ERDF, ESF and EAFRD)
ERDF ESF EAFRD
 All thematic  Focus on rural
objectives  Four objectives :
areas
 Companies employment,
 Six priorities:
(infrastructure, education, skills,
knowledge
business services, social inclusion
transfer,
innovation, ICT and and
innovation,
RTD) administrative
competitiveness
 Services to citizens capacity
of agriculture,
(energy, on line building
management of
services, education, natural
health, quality of the resources and
environment , on line climate action
services) and inclusive
development
CF EMFF
 Environment, sustainable  Viability and competitiveness
development and TEN-T of fisheries and aquaculture
 Social inclusion and job
creation in fisheries dependent
communities
 The thematic objectives can be implemented through either monofund
programmes or multifund programmes combining ERDF, ESF, CF (for Roma
inclusion purposes, the three more relevant funds are ERDF, ESF and EAFRD)
197
9.10 ANNEX 3: COMMON STRATEGIC FRAMEWORK – TIMELINE
14 March The Common Strategic Framework (CSF)
2012 reflects the 11 priorities of the EUROPE 2020
Strategy 11 Thematic objectives:
1. Research Technology Development Innovation
2. Information and Communication Technologies
3. SMEs
4. Low-carbon economy
5. Climate change mitigation
6. Environment and resource efficiency
7. Transport and Network Infrastructure
8. Employment
9. Social Inclusion and Combating Poverty
10. Education
11. Institutional Capacity Building
End
2012/Beginning
PARTNERSHIP
2013 AGREEMENT
National and regional level: possibility for multi-fund programmes combining ERDF,
ESF and CF (For Roma inclusion purposes, the three more relevant funds are ERDF, ESF
and EAFRD)
End OPERATIONAL
2012/Beginning
2013
PROGRAMMES
st st
• 7 year timeline (1 January 2014 – 31 December 2020)
• Programmes will list priorities with specific objectives, financial
appropriations of support from the CSF Funds and corresponding
national co-financing
198
9.11 BIBLIOGRAPHY
Analysis of references to the Structural Funds in National Roma Integration Strategies (NRIS), 7th March
2012, www.euromanet.eu
Council of the European Union (2011) Council Recommendation on policies to reduce early school
leaving. Brussels, 7 June. Available at: http://ec.europa.eu/education/school-
education/doc/earlyrec_en.pdf
__________ (EU) (2011) Council conclusions on an EU Framework for National Roma Integration
Strategies up to 2020. 3089th Employment, Social Policy, Health and Consumer Affairs Council meeting.
Brussels. 19 May. http://www.consilium.europa.eu/uedocs/cms_data/docs/pressdata/en/lsa/122100.pdf
National Roma Integration Strategies: http://ec.europa.eu/justice/discrimination/roma/national-
strategies/index_en.htm
European Commission (2010) ‗Europe 2020. A European strategy for smart, sustainable and inclusive
growth‘. COM(2010)2020. 3 March. Available at:
http://europa.eu/press_room/pdf/complet_en_barroso___007_-_europe_2020_-_en_version.pdf
__________ (2010) Communication from the Commission. An Agenda for new skills and jobs: A
European contribution towards full employment. COM (2010) 682 Final. Strasbourg, 23 November.
Available at: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=COM:2010:0682:FIN:EN:PDF
__________ (2010) Communication from the Commission to the Council, the European Parliament, the
European Economic and Social Committee and the Committee of the Regions, The social and economic
integration of the Roma in Europe, COM/2010/0133 final
__________ (2011) Communication from the Commission. Early Childhood Education and Care:
Providing all our children with the best start for the world of tomorrow. COM (2011) 66 final. Brussels
17 February. Available at: http://ec.europa.eu/education/school-education/doc/childhoodcom_en.pdf.
__________ (2012) Commission Staff Working Document. Elements for a Common Strategic Framework
2014 to 2020. Brussels, 14 March. Available at:
http://ec.europa.eu/regional_policy/sources/docoffic/working/strategic_framework/csf_part1_en.pdf
European Parliament and European Council (2010) Regulation (EU) No 437/2010 of the European
Parliament and of the Council of 19 May 2010amending Regulation (EC) No 1080/2006 on the European
Regional Development Fund as regards the eligibility of housing interventions in favour of marginalized
communities. Available at: http://eur-
lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2010:132:0001:01:EN:HTML .
Review of EU Framework National Roma Integration Strategies (NRIS) submitted by Bulgaria, the Czech
Republic, Hungary, Romania and Slovakia, Open Society Foundation, 2012.
199
10 APPENDIX: SUMMARY OF OVERALL- AND SECTORAL
POLICY RECOMMENDATIONS
10.1 OVERALL POLICY RECOMMENDATIONS
Detailed diagnostics and policy options by sector should not overshadow the need for a holistic and
integrated approach to addressing the needs of Roma inclusion, requiring ministries to coordinate
reinforcing actions. In addition to sector specific findings, this report highlights important linkages
between each of these areas and stresses the need for close coordination among different sector ministries.
For example, the linkages that exist between poor housing conditions and poor health, or the linkages that
exist between social benefit reforms and activities by key line ministries; for example, early childhood
education can be promoted by informing parents about its benefits through the use of Roma mediators and
teaching assistants, but also by putting in place targeted social protection benefits that support all poor and
vulnerable parents financially if children attend regularly preschool. Conversely, stimulating demand for
pre-school through the benefit system needs to be complemented by infrastructure investments to expand
pre-school facilities in those communities that are currently lacking these. Or, targeted social protection
benefits can be made available on condition that mothers participate in pre- and post-natal check-ups, fully
vaccinate their children, and participate in counseling on early childhood development and nutrition, thus
also requiring coordination between ministries. Or, changing the eligibility criteria for housing allowances
to include the poorest and restricting use of the funds for home improvements should be complemented by
training activities how to make basic home and energy efficiency improvements. In short, it is important
therefore that policy measures in the different areas are designed to reinforce each other. Many of the
policy measures suggested offer opportunities in this sense.
Most policy recommendations attempt to draw and build on programs and policies already in place
in Slovakia, including good examples and best practices the World Bank team came across during its field
visits which are not known nationally. In some cases, putting in place incremental and reinforcing policy
measures can be achieved by modifying existing policy measures such as reforming the system of
allowances –labor activation, kindergarten and housing --that already exist as part of the BMN program of
social assistance; or, by addressing the financial incentives that currently exist for municipalities to invest
in special education, and instead provide incentives to invest in integrated regular schooling; or, by
ensuring that the ongoing expansion of social workers will lead to an increase specifically in counseling
on early childhood development and nutrition. In other cases, the World Bank team recommends
expanding and further improving promising initiatives that are currently small scale such as the Roma
Health Mediator (RHM) program, or the home improvement and financial literacy training courses
provided by, for example, ETP Slovakia.
No policy measure can fully succeed if strong prejudices among parts of the majority population
against the Roma are not addressed simultaneously. To illustrate, in 2010, the World Bank carried out
222 qualitative interviews with government and civil society officials in Bulgaria, Romania, Czech
Republic, and Serbia. Respondents were asked to share their view of commonly held perceptions among
the general public. According to more than three-quarters of these officials in each of these countries, a
commonly held perception by the general public is that Roma are "lazy, lack will power, and prefer to live
off social assistance". Many people with whom the World Bank team met in Slovakia similarly mentioned
that these views are common, consistent with the 2008 European Social Survey finding that two-thirds of
respondents claim that the unemployed (regardless of ethnicity) are not seeking employment and 40%
report that social benefits make people lazy. On the other hand, results from the regional Roma survey
indicate that the vast majority of Roma wish at least a secondary education for their children, and value
200
lower paid, secure and full time employment over social assistance or over irregular employment with
more freedom. Yet, it is easy to see that a typical employer would not be keen on hiring a Roma, no matter
how successful might have been policy measures to improve Roma education and health outcomes.
Tackling Roma stereotypes is a vital component of inclusion strategies and need to be an integral
aspect of any measure as well as a goal in itself. Beliefs shape behaviors and therefore determine both
political outcomes and practical day-to-day decisions in applying policies and implementing programs.
Even in the absence of the measures advocated in this report, there are examples of mayors who have been
able to transform the lives of Roma in their municipality –at the same time improving the lives of all their
constituents - because they challenged stereotypes, forged partnerships with others also willing to
challenge stereotypes (e.g., social workers, school officials, psychologists), thus creating a virtuous circle.
Communication can have a significant impact on people‘s beliefs and behaviors, in particular through
targeted campaigns of communication to change behavior (CCB), which apply a marketing approach to
address social issues.140 The Slovak Government should join forces with civil society, media experts, and
international organizations in determining a roadmap for a comprehensive and creative communication
campaign aimed at fighting stereotypes and encouraging acceptance. For example, the Open Society
Foundation has a long history of supporting Roma inclusion programs in Slovakia and elsewhere,
including projects to fight stereotypes. The European Roma Grassroots Organization (ERGO) has
implemented a number of programs tackling stereotypes across Europe forging partnerships with
municipalities and creatively using (social) media and sports. And, the Roma Education Fund has been
working closely with teacher training colleges to create greater awareness among college students of the
challenges facing Roma children in school and foster interest in contributing personally to closing the gap
by seeking placement in schools with many Roma children. The Slovak government can take advantage of
these and other such campaigns.
Successful implementation will also require strengthening capacity at all levels, from policy
coordination to implementation of projects on the ground, and this implies building stronger
partnerships between regional and municipal bodies, and between public, civil society, and private
bodies. The Office of the Plenipotentiary for Roma Communities has started developing the necessary
capacity. It has prepared the National Roma Integration Strategy in a professional and consultative
manner. This capacity can be captured and built upon – regardless of who will politically oversee the
Office – and substantially expanded. In doing so, Slovakia can also build on international experiences
with integrated approaches to addressing the most vulnerable such as the Chile Solidario program. The
sizable allocation of technical assistance under the structural funds (4% of total ESF) can be deployed to
finance capacity building. For example, knowledge sharing and technical assistance to local actors such as
municipal governments requires strong regional level capacity; or, leveraging the knowledge and
experiences of promising NGO initiatives requires building stronger public-private partnerships whereby,
for example, NGOs with a proven track record are financially supported to work with municipal and
regional bodies to scale up promising initiatives.
140
CCB has been used widely and successfully to encourage environmentally responsible behaviors, as well as in
preventive health campaigns, for example, to promote basic hygiene and responsible sex. Its use to promote social
cohesion and mutual acceptance (e.g., in ethnically divided contexts) is growing and shows very promising results.
201
10.2 SUMMARY OF SECTORAL POLICY RECOMMENDATIONS
The policy recommendations for each sector can be summarized as follows:
Employment: the recommendations focus on specific measures to (a) improve job search incentives; (b)
improve efficiency of job search by allowing better matching of labor supply and demand; (c) improve
skills. These measures include:
(1) Reforming the system of activation allowances and focusing on providing in-work benefits;
(2) Strengthening job counseling and providing soft-skills job search support;
(3) Reaching out to non-State actors (firms, NGOs, churches etc.) and create a jobs platform; and,
(4) Providing second chance education, emphasizing on-the-job learning opportunities, including in
the social sectors where Roma are strongly underrepresented.
The report also calls for (d) investments in monitoring and rigorous evaluation of specific activation
measures, and systematically share best practices across municipalities.
Social Protection: the recommendations call for maintaining a strong safety net that continues to protect
the poor, but more so than is currently the case, include social protection benefits that explicitly promote
health and human capital investments, especially for children, and housing investments. Specifically:
(1) Incentivizing targeted investments in maternal health and early childhood development – from
conception to age 8;
(2) Strengthening the existing program providing support to poor young families conditional on
children participating in pre-school;
(3) Incentivizing secondary school completion, not early school leaving;
(4) Incentivizing job search incentives; and,
(5) Incentivizing investments into improved living conditions, especially by the most vulnerable.
Financial Inclusion: the recommendations call for measures building greater financial security:
(1) Promote financial literacy and debt management training;
(2) Improve access to financial services with a focus on savings facilitation, and linking savings
activities with human development outcomes; and,
(3) Take advantage of government social protection payment systems to promote financial inclusion.
Education: the recommendations focus on measures to (a) increase access to quality pre-school, moving
toward compulsory preschool from age 3 onwards, and improve home parenting; (b) promote integrated
regular primary schooling for all; and, (c) address early (secondary) school leaving. Specifically:
(1) Increase access to quality pre-school, moving toward preschool attendance from age 3 onwards,
and improve home parenting;
(2) Create the position of Roma school (or community) mediator;
(3) Provide school-based education services beyond regular classroom teaching – teaching assistants
and tutors;
(4) Provide educational support and opportunities outside regular school hours -after school and
summer programs;
(5) Provide added incentives that make a difference - cash transfers conditional on certain educational
achievements (e.g. participation in ECD and completion of secondary);
202
(6) Promote integrated regular primary schooling for all; move rapidly to close most special schools –
in part by changing the financial incentives currently in place for special education - and abolish
special classes in standard schools.
Housing: the recommendations focus on measures to (a) improve conditions – in situ - for the worst off
Roma living in slum areas, and (b) to help poor families move into better housing. The specific
recommendations include reform of the housing allowance and emphasize the desirability of incremental
approaches that will reach larger number of poor Roma households sooner with the expectation that
improvements will be gradual and take place over a period of time.
Specifically, to improve living conditions for poor households in situ (i.e., upgrading):
(1) Facilitate legal inclusion (provide technical assistance, legally register regardless of structure type,
encourage municipalities and NGOs);
(2) Improve access to basic services through infrastructural investments (access existing programs
such as State Housing Development Fund, support families manage utility debts, involve
beneficiaries in infrastructure upgrading); and,
(3) Improve support to poor residents to make home improvements (provide financial literacy and
home improvement training, delink housing allowance from legal residency status, ensure housing
allowance support is used for housing (improvements), increase availability of microfinance for
housing improvements);
Specifically, to help poor families move into better and integrated housing:
(1) Diversify government programs to enable poor households more choices than just low-cost social
housing to access better dwellings (incentives to use vacant housing in integrated areas (rent or
ownership), expand the availability of private formal rentals, provide strong social support when
families move into new neighborhoods; redesign the housing allowance so it can also be paid
directly to landlord);
(2) Involve families in the construction of new homes, which can be more cost effective than building
final finished units and build skills (provide families with well located land sites and building
materials, provide technical assistance to the families in construction).
In addition to these specific recommendations, there are also several cross-cutting recommendations
on housing, including synchronizing programs horizontally for a common goal (i.e. across different
housing programs but also programs in education, employment, health, etc., or commingling ERDF and
ESF funds) and target the poorest communities first using the Slovak poverty map being produced by the
National Statistics Office and the World Bank and the Slovak Roma Atlas being updated by UNDP.
Health: the recommendations focus on measures to (a) promote more effective use of existing health
services (including expansion of the Roma Health Mediators program), expand health knowledge and
awareness, or a combination of both; and, (b) measures that use entry points outside the immediate health
system (for example in education or housing) to promote greater health among Roma. As with the other
sectors, there can be important linkages with demand side incentives provided through the social benefit
system.
Specifically, measures promoting more effective use of existing health services, measures to expand
health knowledge and awareness, or a combination of both:
203
(1) Expanding the RHM program and improving its quality through the use of rigorous evaluation
methods.
(2) Better use of existing entry points within the health system – in particular ante natal care and birth
delivery - to promote greater health knowledge and awareness, as well as access to services,
among Roma. Birth delivery, but also ante natal care visits
(3) Exploring changing health provider payment systems and linking payments to results, i.e.
coverage of Roma population and regular check-ups could be considered.
(4) Promoting the use of mobile and wireless technologies can open up many opportunities ranging
from making appointments, providing reminders, as well as reminding patients about prescription
renewals or vaccinations
Specifically, using entry points outside the immediate health system to promote greater health
among Roma:
(1) Improving Health through the Education System - Preschool offers not only education benefits,
but can also offer many opportunities for health, promoting early childhood development broadly;
Schools offer excellent opportunities to educate on health (e,g. through the Roma peer program);
Promoting sports, especially among youth;
(2) Improving Health through the Housing and Infrastructure Sector - Improving infrastructure to
promote access to clean water, and reduce waste and indoor air pollution;
(3) Improving Health through the Social Protection System - Linking social assistance and family
benefits to results by making extra cash transfers for certain health outputs such as attending
health education sessions and getting adequate pre-natal check-ups.
Monitoring and Evaluation: the report recommends:
(1) Taking advantage of the Slovak poverty map being produced and the Slovak Atlas of Roma
Communities to improve targeting of inclusion programs;
(2) Bi-annually expand the EU-SILC survey implemented by the Statistical Office of the Slovak
Republic to include extra households from the poorest communities in Slovakia;
(3) Ensure that the programs being financed have results frameworks in place that clearly define
inputs, activities, outputs, and impacts;
(4) Undertake rigorous impact evaluations to learn what program works best; and,
(5) Consider building an inexpensive ‗municipal best practice‘ online knowledge portal.
EU Financing: the report recommends:
(1) Either develop one 2014-2020 Operational Programme for Roma Inclusion and finance it jointly
from the European Social Fund (ESF) and the European Regional Development Fund (ERDF), or
alternatively, Roma inclusion could become a separate Priority Axis within a larger Operational
Programme, again jointly financed by the ESF and the ERDF;
(2) Take advantage of the EU Structural Funds to strengthen the capacity of the Office of the
Plenipotentiary;
(3) To assist the local communities and municipalities develop high quality proposals;
(4) To simplify the use of ESF and ERDF, especially for locally led initiatives and small projects;
and,
(5) Use the poverty map for targeting Structural Funds financing.
204