Embed Size (px)
Citation preview
Health & Place 16 (2010) 1240–1247
Contents lists available at ScienceDirect
Health & Place
1353-82
doi:10.1
n Corr
E-m
adamb@
(K. Kosa1 Te2 Te
journal homepage: www.elsevier.com/locate/healthplace
Health impact assessment and evaluation of a Roma housing projectin Hungary
Agnes Molnar a,1, Roza Adany a,1, Balazs Adam a,1, Gabriel Gulis b,2, Karolina Kosa a,n,1
a Faculty of Public Health, Medical and Health Science Centre, University of Debrecen, 26 Kassai ut, 4028 Debrecen, Hungaryb Unit for Health Promotion Research, University of Southern Denmark, Niels Bohrs Vej 9, 6700 Esbjerg, Denmark
a r t i c l e i n f o
Article history:
Received 16 September 2009
Received in revised form
6 August 2010
Accepted 9 August 2010
Keywords:
Housing
Roma
Policy
Health impact assessment
Outcome evaluation
92/$ - see front matter & 2010 Elsevier Ltd. A
016/j.healthplace.2010.08.011
esponding author. Tel.: +36 52 460190x7743
ail addresses: [email protected] (A. Molnar),
dote.hu (B. Adam), [email protected] (G
).
l.: +36 52 460190.
l.: +45 6550 4212.
a b s t r a c t
An outstanding feature of marginalized Roma communities is their severely substandard living
conditions, which contribute to their worse health status compared to the majority. However, health
consequences of international and local-level housing initiatives in most cases fail to be assessed
prospectively or evaluated after implementation. This paper summarizes the result of a retrospective
health impact assessment of a Roma housing project in Hungary in comparison with the outcome
evaluation of the same project. Positive impacts on education, in- and outdoor conditions were noted,
but negative impacts on social networks, housing expenses and maintenance, neighbourhood
satisfaction and no sustained change in health status or employment were identified. Recommenda-
tions are made to improve efficiency and sustainability of housing development initiatives among
disadvantaged populations.
& 2010 Elsevier Ltd. All rights reserved.
1. Introduction
1.1. The Roma population in Europe
The number of European Roma is estimated anywherebetween 10 and 12 million, around 4 million of whom live inCentral and Eastern European member states of the EuropeanUnion (European Commission (EC), 2010; Ringold et al., 2005).The largest proportion of Roma lives in Bulgaria, Hungary,Romania and Slovakia, representing over 5% of the population ofthese countries (Council of Europe, 2002; Bernat, 2009). Scientificstudies of Roma are hindered by a number of issues such asmethodological problems the most serious being the definition ofwho is Roma and who is not; and ethical problems, namelyhistorically based distrust of Roma towards ‘official’ data collec-tion and means to overcome it.
Nevertheless, a number of studies described serious inequal-ities in health status between Roma and non-Roma to whichsocial exclusion, poverty, unemployment, low educational level aswell as ethnicity, discrimination and racism all contribute invarious degrees (Hajioff and McKee, 2000; Revenga et al., 2002;European Roma Rights Centre (ERRC), 2006; EC, 2007; Kosa and
ll rights reserved.
0.
[email protected] (R. Adany),
. Gulis), [email protected]
Adany, 2007). Housing and settlement issues are closely relatedwith all of these areas and are of particular significance for Roma.A large proportion of Roma have been identified as living incolonies or settlements, that is, segregated habitats characterizedby severely substandard conditions (EC, 2004).
The situation of Roma in the European Union was addressed bynumerous international and EU-level documents aimed at theprotection of minority rights, equal opportunities and efforts toimprove their situation. The governments of nine Central andSouth-Eastern European countries agreed to launch a large-scaleintegration programme titled ‘‘Decade of Roma Inclusion’’ in 2005(to which two more countries joined since), and expressedunprecedented political commitment to improve the welfare ofRoma in the areas of education, employment, housing and health(Decade of Roma Inclusion, 2005).
Half of the Decade of Roma Inclusion has already passed,justifying a close look at the achievements of this internationalinitiative. The question of whether the measures taken by thesecountries have indeed been efficient should be answered. Progressreviews of the Decade of Roma Inclusion investigated whetherthere are measures, programmes and policies in place, notwhether they work (Tashev and Hrabanova, 2007). A gap betweencommitments and implementation as well as lack of monitoringand evaluation of policies and programmes were noted by theInternational Steering Committee (Decade of Roma Inclusion,2008).
Health, in general, is not a cross-cutting issue within theDecade of Roma Inclusion but a separate field of action focusingon access to health care. Available documents regarding the
A. Molnar et al. / Health & Place 16 (2010) 1240–1247 1241
outcomes of the housing component of the Decade of RomaInclusion in Hungary described output and process rather thanimpacts on health and quality of life (Hungarian Ministry ofYouth, Family, Social Affairs and Equal Opportunities, 2006).
Health impact assessment (HIA) could be an appropriate toolto systematically explore probable health consequences ofinitiatives and implemented practices building on which policiescan be modified to optimize the health gains of populationsaccording to the Health in All Policies approach (Stahl et al., 2006).However, only one published health impact assessment (HIA) wasidentified that specifically targeted Roma people (Kosa et al.,2007). That study assessed the health impacts of a housing projectagainst eviction of a Hungarian Roma community.
An EU-funded project titled ‘‘Health Impact Assessment inNew Member States and Pre-Accession Countries’’ (HIA-NMAC)was launched in 2005 with the aim of collecting information andbuilding capacity related to HIA for new member states of the EU.One workgroup within the project was specifically chargedwith the aim of conducting pilot studies on the applicability ofhealth impact assessment in planning programmes for vulnerablepopulations. Four HIA case studies from 4 Central EasternEuropean countries (Bulgaria, Hungary, Lithuania and Slovakiawhich have also been involved in the Decade of Roma Inclusionwith the exception of Lithuania) were planned to be carried out inorder to integrate experiences with HIA related to Roma housingpolicies in these countries. 2 of these countries (Bulgaria andSlovakia) carried out prospective HIAs of national level housingprogrammes, whereas Lithuania – not having a national housingpolicy for Roma – completed a prospective HIA of a programme atmunicipality level. Hungary launched a comprehensive nationalprogramme in the framework of the Decade of Roma Inclusion toimprove the quality of life in segregated Roma habitats in 2005.The Hungarian case study included a retrospective HIA of thisprogramme at one location (Hencida) in 2006, and a subsequentevaluation of long-term health outcomes at the same locationfrom 2009.
The Hungarian case study of the HIA-NMAC project is thesubject of this paper aiming to investigate whether HIA wouldbe a useful tool to assess housing initiatives for vulnerablegroups such as Roma and to evaluate the effectiveness of ahousing project to relocate and integrate families within localcommunities.
1.2. The Roma population in Hungary
Roma constitute 1.9% of the population according to the lastcensus of 2001, whereas research estimates this proportionranging between 5.2 and 6.5% (Kertesi and Kezdi, 1998). Anenvironmental survey of segregated habitats carried out between2000 and 2005 in Hungary revealed that 16–25% of all Roma inHungary live in 758 substandard, poorly serviced, segregatedhabitats (colonies) around the country (Kosa et al., 2009). As totheir health status, 10% more of those over the age of 44 yearswho lived in colonies reported their health as bad or very badcompared to those in the lowest income quartile of the non-Romageneral population. Of those who used any health services, Romaliving in colonies experienced discrimination 8 times more oftenthan those in the general population (Kosa et al., 2007).
Hungary has various governmental decrees aiming at theintegration of Roma since 1997. A programme on Housing andSocial Integration was adopted in 2005 reflecting the country’scommitment to the Decade of Roma Inclusion (HungarianMinistry of Youth, Family, Social Affairs and Equal Opportunities,2006). The objectives of the programme were twofold: to improvehousing conditions and social integration through improved
access to education, employment, social services and health carefor those living in segregated areas. The first step of theprogramme was implemented through a governmental call fortenders inviting nine rural municipalities to apply which hadlarge colonies in bad conditions, Hencida being one of them.
1.3. The integration of the Roma population of Hencida
The village of Hencida is located in a north-eastern county ofHungary in the Berettyoujfalu district, a highly disadvantaged areaof Hungary in terms of socio-economic development. The villagehas 1325 inhabitants; the dependency ratio is more than 45%, wellabove the county average. The unemployment rate is 14%, nearlydouble the national average (among men 63%), of whom morethan 26% are skilled workers and 74% are unskilled. This villagewas chosen for our case study based on previous results of anenvironmental survey of segregated habitats of Roma in Hungarythat identified Hencida’s Roma colony as one of the mostdisadvantaged based on the severity of unfavourable environ-mental and housing conditions (Kosa et al., 2009). 2 segregatedcolonies existed within the village of 1318 inhabitants in 2001,giving home to 6.6–7.5% of the local population, all Roma.
The municipal government was invited to propose and won aproject with a budget of 224,000 euro to improve the conditions ofthe colonies and the quality of life of its inhabitants in 2005. Themunicipal government planned and carried out the projectsupported by a mentor and a project coordinator delegated bythe Ministry of Social Affairs for the duration of the project. Romawere included in the project through the representative of theRoma self-government during the planning of the project. 12families who had lived in life-threatening conditions wererelocated to other houses 10 of which were existing housespurchased and remodelled from the project budget, whereas 2 newhouses were built from the owners’ money and project support.Selection of the beneficiaries for relocation was based on twoprinciples: living in a house of life-threatening conditions, andfamilies be legally documented owners of their house. Propertyrights to the remodelled houses were based on barter contracts(8 houses) or rental contracts (4 houses). In addition, 46 familieshad their houses renovated, mainly from the outside. Water pumpswere installed in the gardens of 10 houses. Public places wererenovated, fifty trees were planted. Streets were restored andpavements were built in 1.2 km length. Drainage ditches wereconstructed and existing ones were cleaned. Information technol-ogy infrastructure was developed in the primary school. 84% ofthe Hencida project budget was spent on the housing projectapproximately half of which was used for remodelling 46 houses,the other half for relocation of 12 families and rehabilitation of thesettlement. 8.2% of the budget was allocated for management; 6.2%of the budget was spent on employment expenses mainly coveringthe employment of Roma public workers. 1.6% of the projectbudget was allocated to the local school for improving informationtechnology. Subsequent to the project, a school integrationprogramme was implemented and financed by the Roma EducationFund including several activities: assigning mentors to pupils, afterschool programmes, day-school and increased school supplies.
2. Methodology
2.1. Retrospective health impact assessment
Our health impact assessment is based on the broad modelof health of the Ottawa Charter according to which basicprerequisites such as shelter, education, social justice and equity,
A. Molnar et al. / Health & Place 16 (2010) 1240–12471242
among others, determine population health. The structural frame-work of the case study is based on the Gothenburg ConsensusPaper adopting a participative, multidisciplinary and intersectoralapproach (European Centre for Health Policy and WHO, 1999).The retrospective HIA was carried out after the end of the housingproject in Hencida in 2007. The team consisted of academicresearchers with expertise in HIA and relevant scientific areas asenvironmental health, epidemiology, health promotion, sociology,general medicine and law, as well as previous experience andknowledge on disadvantaged Roma communities.
2.1.1. Sources and methods of data collection
Structured and in-depth interviews were conducted in personand in some cases by telephone to assess health consequences ofthe local project. Information was collected from stakeholderssuch as the mayor and members of the municipal government;the president and members of the local Roma self-governmentrepresenting the local Roma community; the director of the localschool which Roma children of the village attend; the generalpractitioner and district child nurse of the district primary healthcare office that services all inhabitants of the village; professionalsof the village family and child help services; and the local projectcoordinator delegated by the Ministry responsible for implemen-tation. Questions were related to short-term impacts of theproject as well as management issues, that is, the process ofproject planning, capacity building, budgeting, involvement ofthe Roma community and communication between stakeholders.The interviews provided qualitative information on the actualsituation even in the case of the local professionals since ethnicityis not registered either in school, health or social care.
Additional data was gathered from different sources such asdocuments on the national programme and the local project inHencida; policy documents and research papers on Roma werealso reviewed. Evidence on the health impact of housing wasreviewed in the scientific as well as grey literature from previousHIAs on housing development.
2.2. Outcome evaluation
Outcome evaluation of the Hencida project was carried outfour years after the end of the project and 2 years after the abovedescribed retrospective HIA in order to assess long term healthimpacts of the project, and judge the accuracy of the predictionsof the HIA.
2.2.1. Sources and methods of data collection
Structured interviews with stakeholders were carried outrepeatedly; in addition, the mentor responsible for implementa-tion delegated by the Ministry and the regional public healthofficial of the National Public Health Service were also inter-viewed. Questions were focused on long term impacts andsustainability of the project four years after the implementation.Quantitative data were elicited from the GP and the district childnurse, professionals of the village family and child help services,the directors of the school and kindergarten, as well as the mayor.No baseline data from 2005 on the health status of thebeneficiaries could be elicited from the project documentationor the GP except data on disability pension.
An environmental survey was carried out by one of theresearchers to assess physical changes of the colony and itshouses. The survey tool was identical to the one used in a previousenvironmental health survey of segregated habitats that wascarried out at Hencida in 2001 (Kosa et al., 2009).
Most importantly, a questionnaire survey was conducted togather information from the beneficiary families that were
relocated from their previous dwellings. Adult members of nineof the 12 families were questioned. (One family dropped out ofthe project, two families were not found in their homes threetimes during the field research, due to fears from ‘‘officials’’ asreported by neighbours). The questionnaire consisted of 42questions on socio-economic, living and housing conditions andhealth status. Respondents had to assess their situation beforeand after the project. Five Roma beneficiaries living at 5 differ-ent streets whose houses were renovated outside were alsoquestioned.
Additional data were gathered from documents on the nationalprogramme and from the governmental evaluation of the projectin Hencida. National databases were used to gain basic demo-graphic and socio-economic data on the inhabitants of the village.
3. Results
3.1. Retrospective health impact assessment
Health impacts of the project are summarized according to thedeterminants of health.
3.1.1. Lifestyle
No connection between improved living conditions andchanges in risk behaviours such as smoking or alcohol consump-tion could be predicted. Decreased rates of infectious diseases andimproved nutrition could be predicted due to better cooking andstorage conditions. Adequate living conditions may indirectlyimprove the mental health of beneficiaries through decreasingstress. Roma representatives in Hencida reported decreased stressand improved self-esteem among participants.
3.1.2. Socio-economic determinants
3.1.2.1. Housing tenure and satisfaction. Decent housing provides afoundation to better socio-economic conditions through properconditions for personal hygiene and learning. Secure status interms of housing ownership or tenancy was perceived by thebeneficiaries, promoting improvement of mental health for bothparents and children.
3.1.2.2. Social networks. Social networks and emotional safetymight be strengthened reducing stress. Improved conditions forleisure-time and recreational activities may also have beneficialeffects on stress. Greater satisfaction with neighbourhood andeconomic activity may result in lower rates of ill health, disabilityand mortality. Nevertheless, relocation may bring negative socialand emotional consequences due to the breakup of supportiverelationships. Discrimination and racial harassment againstrelocated Roma from their new neighbours fundamentallydetermine their well-being in social housing developmentprojects. Some non-Roma citizens in Hencida complained thatRoma undeservedly benefited from the project. The selection ofbeneficiaries and property issues (who gets what and why othersnot) generated substantial debate and dissatisfaction amongparticipating Roma families as well, deteriorating family andsocial relationships. Even the president of the Roma self-govern-ment was heavily lobbying for including one of his family mem-bers as beneficiary which led to the exclusion of this family fromthe project. Positive impact on the occurrence and fear of crimenotably benefiting mental health, physical functioning and qualityof life could not be predicted in the Hungarian case.
3.1.2.3. Education. School attendance and performance improvedin the short term among Roma children, supporting the positive
A. Molnar et al. / Health & Place 16 (2010) 1240–1247 1243
relationship between housing conditions and better cognitivecapacity in childhood.
3.1.2.4. Employment, income, expenses. Increased rental or main-tenance and overhead costs may bring negative consequences ifincome levels do not change. In terms of income, adult malebeneficiaries were employed resulting in increased income duringthe housing project but their employment was terminated at theend of the project.
3.1.3. Outdoor conditions
Rehabilitation of urban centres or removal of communitiesfrom industrial areas may help reduce pollution and road trafficaccidents. Soil and water quality may improve due to decreasedindustrial pollution, removal of garbage deposits and weeds frompublic places as could be predicted in this case. The beneficialeffects of public sanitation and sewerage could not be predictedsince this was neither available nor planned anywhere in thevillage. Safer roads and beautified green spaces with trees providea better living environment and increased levels of safety andhealth.
3.1.4. Indoor conditions
Indoor air quality, temperature and dampness have tangiblehealth consequences in terms of cardiovascular, malignant andrespiratory diseases, particularly amongst children and elderlypeople. Decreased crowding and the elimination of rodents andparasites may contribute to a decreased rate of infectious andallergic diseases as well as improved mental health. Improvedindoor air quality, better cooking facilities, reduced dampness,allergens and crowding, instalment of running water and coveredcesspools as well as increased privacy could be predicted toimprove health status and decrease the number of home
Table 1Demographic and environmental characteristics of the population of Hencida.
2001
Demographic and environmental characteristicsNumber of inhabitants 1318
Number of Roma 88 (6.6%)a
51–100 (7.5%)b
Number of Roma colonies 1 colony in severe substandard
condition
Number of houses 31
Water mains Not available at the colony
Electricity Available at the colony
Gas mains Not available at the colony
Sewerage Not available in the village or at the
colony
Garbage deposit 4 designated, 13 illegal sites at the
colony
Carcass deposit 1 deposit near the colony
Waterlogged area Several streets in the colony
Access to paved road in 30 minutes Accessible
Access to public telephone Accessible
Housing conditionsHouses Built wall without insulation
Piped water Not available at the colony
Electricity Available in the houses
Heating Heated stove in one room
Drainage Outhouse
a Source: Census.b According to the Roma minority self government.c According to the local government.
accidents. However, this latter prediction could not be supported:two children of a family living in a renovated house were heavilyburnt one of whom died because of inefficient child supervisionwhile using the stove.
3.1.5. Access to and quality of health services
Positive impact on access to health services subsequent tohousing development such as more frequent contacts withgeneral practitioners and improved conditions for house callswere derived from HIAs as speculative but not predicted in ourcase, based on interviews after the end of the project.
3.2. Outcome evaluation
Basic demographic and environmental characteristics of thecommunity of Hencida are summarized in Table 1. Results of theenvironmental survey from 2001 made it possible to comparechanges in terms of environmental and housing conditions of theRoma colonies (Table 2).
As it can be seen from the outcome evaluation, the majorimpact of the housing project seems to be on education inasmuchas the number of persons completing primary school increased.However, it is uncertain to what extent this can be attributed tothe housing project itself or to secular trends. School attendanceenhanced during the project but shortly after the termination ofthe school integration programme irregular attendance of theprimary school became a key problem in a family of elevenmembers resulting in an administrative legal case. Two school-age children of this family were withheld in school to repeat theyear due to truancy.
In order to increase the employment of Roma people in theproject, a Roma building contractor without reference work insocial housing was charged with constructing houses and roads.However, beneficiaries’ employment was terminated at the end of
2010
1325
600c (45%)
Roma moved to inner streets of the village before the start of the project.
Roma houses are now scattered in the village and can be found in every street
(4 streets are mainly populated by Roma dwellers)
108
Available for all
Available for all
Available in 70% of the village houses
Not available in the village
No designated deposit in the village
No designated deposit in the village
One street in the village
Paved roads except 2 streets
Wired telephone system/high speed internet in the village
Built wall without insulation (adobe) 30%
Built wall with insulation 70%
Accessible in some gardens
Available in the houses
Heated stove in one room
Outhouse
Table 2Evaluation of the long-term outcomes of the 12 relocated families of Hencida.
Before the project (2005) After the project (2010)
Demographic characteristicsNumber of families 12 11 (one family were dropped out from the project)
Number of people 69 61 (4 births, 1 death, relocation)
Ethnicity Roma Roma
Age distribution 65% under 18 years (45 children) 68% under 18 years (42 children)
35% under 60 years (24 adults) 32% under 60 years (19 adults)
LifestyleNutrition � Inadequate conditions for cooking and storage, lack of
kitchen, unsafe cooking stoves
� Improper diet: low consumption of vegetables,
consumption of carbohydrate rich food, junk food,
undernutrition
� Kitchen, cooking possibilities, oven, pantry are available
� No change in terms of quantity and quality of nutrition
Risk behaviour � At least one smoker per household � At least one smoker per household
� Excessive alcohol consumption of the parents was reported
in case of one family
Socio-economic conditionsHousing tenure and
satisfaction
� 12 families are threatened with homelessness.
� Property owned by the families (precondition of
participation in the project)
� 10 houses were demolished and owners were relocated.
Debt was taken over on 2 newly built houses.
� Property owned by 7 beneficiary families, 4 families have
rental contracts, the property of 1 left house is under legal
consideration.
� 6 families: moderate satisfaction
� 3 families: it would had been better to stay at their
previous dwellings
Social network � Mutual support in the Roma community
� Traditions, cultural beliefs and attitudes
� Quarrels on property issues
� Moderate discrimination against Roma surrounded by the
non-Roma community
� Increasing inequalities among Roma—stratification,
improved attitude to non-Roma
� Quarrels mainly among relatives
� Increased fear of non-Roma
� Internal debates among Roma on the eligibility criteria and
the amount of support
� Dissatisfaction of non-Roma inhabitants
Education � 67% of adults completed less than 8 years of primary school
� 1 person completed vocational school
� School attendance in primary school is irregular in one
family
� 60% of adults completed less than 8 years of primary school
� 32% of adults completed 8 years
� 1 person completed secondary school
� 1 person completed vocational school
� School attendance in primary school is irregular in two
families
� 2 persons were enrolled to vocational school but dropped
out
Employment � 4 persons (2 women and 2 men) with permanent job
� 3 women on maternity leave
� 2 persons (1 woman, 1 man) receiving disability pension
� 3 persons (2 women and 1 man) permanently employed
� 2 women employed by the local government as temporary
public workers
� 3 women on maternity leave
� 1 person receiving disability pension
Income � 90%: 60–80 euro/month/person
� 10%: 140 euro/month/person
� 1 washing machine, 1 television and 1 bicycle in each
household
� 87%: 50–100 euro/month/person
� 13%: 100 euro/month/person
� Every family is eligible for regular social assistance (one or
two family members)
� Family allowance is provided after underage/school
children in 8 families, two mothers are eligible for maternal
leave payment and two for nursing assistance for caring a
disabled child.
� 1 washing machine and 1 bicycle can be found in every
household as at the start of the project. Two families
reported 3 televisions/household
Housing expenses Mean: 10–30 euro Mean: 50–70 euro
Criminalization No adults had or has been in jail 2 persons were in jail (robbery and violence)
Indoor environmentHouses � Lack of public utilities, insulation, drainage, running water;
90% of houses have electricity
� Dwellings are built with adobe walls, 12 in life threatening
conditions
� Crowdedness: 6 families with 4 or more members/room
� Electricity in all houses; running water has not been
installed in 4 of 12 houses
� Village houses built from adobe, 2 new houses built from
bricks; two rooms, kitchen, pantry, painted wall; 5 houses
with bathroom
� Decreased crowdedness: 2 families with 4 or more
members/room
A. Molnar et al. / Health & Place 16 (2010) 1240–12471244
Table 2 (continued )
Before the project (2005) After the project (2010)
Health status and health service usageFunctional limitations 50% limitation of 2 persons (1 woman, 1man) 50% limitation of 2 persons (1 woman, 1man)
Chronic diseases Cardiovascular: 3 (hypertension: 2, thrombosis: 1,
varicositas: 1)
Neuropsychiatric: 5 mentally retarded children, 2 with
epilepsy, 1 with brain tumour
Musculoskeletal: 2 (spinal hernia)
Cardiovascular: 5 (hypertension: 4, thrombosis: 1,
varicositas: 1)
Neuropsychiatric: 3 mentally retarded children, 2 with
epilepsy, 1 with brain tumour—the 2 other were taken
into social care
Musculoskeletal: 2 (spinal hernia)
Infections Scabies, louse and impetigo due poor personal hygiene in
children of 3 families
Scabies, louse and impetigo due poor personal hygiene in
children of 2 families
Injuries Scalding: 1 child Fracture: 2 adults, 2 children
Bicycle accident: 1 children
First visit for prenatal
care in the Roma
community of the
village
No data 30% before the 24th week of pregnancy, 50% between the
24th–30th weeks, 20% after the 30th week
Cervical screening During prenatal care or delivery During prenatal care or delivery
A. Molnar et al. / Health & Place 16 (2010) 1240–1247 1245
the project so there was no major long term impact on theiremployment status or income. Overhead expenses related to therelocation increased. Several negative impacts on social networkcould have been identified. Serious debates have been ragingwithin the Roma and non-Roma communities alike regarding themerit of the selected families and why others were left out. Of the12 families selected for relocation, one family sold their housewithout permission from the local government (a condition forinclusion), and heads of two other selected families weresentenced to prison because of robbery and violence.
Indoor conditions improved right after the project; thoughthese improvements were not sustained. The intervieweesclaimed to have carried out repair works on their houses, butthe researchers visiting the families saw deteriorated plaster-works and cracked walls in most of the houses; roofs with missingtiles, missing chimney in one and missing fences in anotherhouse. The presence of rodents was reported by one family.Though crowdedness seemed to have improved in the newlocations, most families keep heating only one room and still tendto gather in that room during wintertime. According to majoritystakeholders, Roma could not preserve the condition of theirhouses due to the lack of commitment and knowledge of how todo it. What can be known is that the building contractor offeringthe lowest budget had been chosen by the stakeholders forbuilding and construction works. This contractor later on wasreported to employ project workers in other profit-orientedventures and enter into secret deals with the families so as notto provide the housing quality he was supposed to produce.Quality control of the construction work was carried out but nowritten track of it was found. An official investigation against thiscontractor was reported to be currently ongoing. Based on theavailable information, it is impossible to decide to what extent theabove listed problems could be attributed to low quality buildingmaterials and work procedures and/or to improper maintenance.
Most beneficiary families reported no change in the frequencyof seeking medical care. Two families attributed their lessfrequent visits to the GP to their children growing up; whereastwo other families with newly diagnosed hypertensive membersreported an increased number of visits. Most families had theirchildren vaccinated with the mandatory vaccines for children(occasionally with delays) except for one family.
3.3. Comparison of the retrospective HIA with the outcome
evaluation regarding health outcomes
The predicted health outcomes of the HIA were mostlysupported by the evaluation of the project four years after itscompletion. Improved schooling, improved in- and outdoorconditions, increased costs related to housing, and no change inemployment or health behaviour (smoking or alcohol consump-tion) occurred among the beneficiaries of this project. Betterindoor conditions, as predicted, do not necessarily result in fewerhome accidents or better access to health and social care. Incontrast to our HIA, improved housing did not lead to betternutrition, lesser crowding or improved privacy.
As it was predicted in our HIA, no unequivocally positivechange in social relations was observed as a result of the project.Internal strives within the Roma community centred around theprinciples of distributing resources and choosing beneficiaries.Disagreements between the Roma and non-Roma communitiesoriginated from similar causes since the non-Roma also ques-tioned the principles of selecting beneficiaries, and remainedunconvinced of the justification of the positive discriminatoryintervention. Roma families would have preferred new locationsin mostly non-Roma neighbourhoods that were opposed by theirprospective neighbours. Most of the relocated families moved intoneighbourhoods that were primarily inhabited by Roma.
4. Conclusion
Health impact assessment offers an objective and multi-disciplinary tool for healthy public policy and health advocacy.HIAs utilize already existing epidemiological and social scienceknowledge as evidence base. Our case study comparing the HIAwith the outcome evaluation of a small-scale housing projectserves as evidence demonstrating the appropriateness andpredictive power of HIA for the assessment of health conse-quences of housing initiatives in Roma communities, but at thesame time underlines the importance of extending the evidencebase for future HIAs.
Since the publication of the Chadwick Report in 1842(Chadwick, 1842) a large body of evidence was collected on the
A. Molnar et al. / Health & Place 16 (2010) 1240–12471246
health benefits of decent housing (Shaw, 2004), and internationaldocuments define housing as a prerequisite of health (OttawaCharter, 1986). However, while the examined housing projecteliminated life-threatening conditions for the relocated benefici-aries and provided safe shelters for them, no unequivocal healthimprovements could be documented. In other words, it isuncertain whether any improvement in health status happened.Moderate improvements in the incidence of infectious diseases ormental health status might have occurred that could have beenproven by comparing properly documented before–after data butnot by retrospective data due to recall bias. If there was really noimprovement in health, that might have been due to variousreasons such as insufficient increase in the quality of housing, nochange in lifestyle, or increased stress due to social strifes becauseof the disputed selection principles of the project.
Based on our HIA and outcome evaluation, it can berecommended that housing projects planned for vulnerablecommunities be rigorously evaluated not only in terms of theirimpact on health but on its socio-economic determinants as well.Appropriate data should be collected before and after the project;a clear and detailed evaluation plan with indicators and baselinedata should be a condition for funding. Such evaluations arefundamental for making any judgement on the effectiveness ofthe project.
Improving health is not necessarily the primary or mostimportant argument for housing projects for Roma. Increasingeducational level – as it happened in this case – or employment orremoving indoor hazards might be sufficient justificationsespecially in light of these being major determinants of health.
Standards for housing improvements must be designed or atleast the minimally adequate housing conditions must be definedfor any such project along with carefully planned, implementedand documented quality control.
Planning any housing project should involve representatives,or optimally directly all members of the affected community;minority and majority alike. Clear principles of selecting bene-ficiaries should be agreed upon taking minority and majorityinterests into account. No time should be saved to come to anagreement that most stakeholders can accept. Principles ofselecting beneficiaries should be publicized and enforced; directparticipation of the beneficiaries should be expected in theconstruction of their houses. A steering group should follow theproject from beginning to end by employing sufficiently experi-enced and reliable personnel at the local level along withrepresentatives of minority and majority stakeholders.
Financing should match the aims and capacities. Cost-contain-ment should be balanced with quality considerations since cheapprojects might end up being very expensive or futile in the longrun. Overly ambitious aims as the renovation of 46 houses atHencida resulted in the scattering of funds, allowing only small-scale renovations and consequential debates among beneficiaries.
Whilst our case study of a housing project aimed at disadvan-taged Roma people demonstrated some positive impacts onschooling, no sustained effects on employment or health couldbe documented. The health of disadvantaged populations, amongthem Roma, cannot be sustainably improved by a housing projectalone though it is of primary importance; approaches addressingother major determinants of health, particularly employment,education and social support should also be developed (Commis-sion on Social Determinants of Health, 2008). Accountabilityshould be a fundamental principle of any project for vulnerablepeople, including those of the Decade of Roma Inclusion(McDonald and Negrin, 2010).
Based on the findings of this project, the authors suggest thatpriority should be given to small-scale, model or pilot housingprojects in rural and/or urban communities. These projects could
be expanded in the same community in phases, involving moreand more beneficiaries if previous phases can documentimprovements in health and/or quality of life. Ambitious, large-scale regional or national housing programmes should be basedon experience from such pilot projects.
Acknowledgements
This work was completed in the frame of the HIA-NMACproject (Health Impact Assessment in New Member States andPre-Accession Countries) funded by the European Commission,Health and Consumer Protection Directorate-General (Grantagreement no. 2004128), aiming to consolidate existing institu-tions and knowledge of HIA into one major network ofcollaborators by conduct of capacity building exercises, pilotstudies, methodological development and implementationexercise.
References
Bernat, A., 2009. Roma in Eastern Europe. Social Sciences Eastern Europe 2009(2). GESISLeibniz Institute for the Social Sciences, Service Agency Eastern Europe, Berlin./http://www.gesis.org/fileadmin/upload/dienstleistung/fachinformationen/series_ssee_01/Roma_in_Central_and_Eastern_Europe.pdfS (online 15 July 2010).
Chadwick, E., 1842. Report on the Sanitary Conditions of the Labouring Populationof Great Britain. HMSO, London.
Commission on Social Determinants of Health, 2008. Closing the Gap in aGeneration. Final Report. World Health Organization, Geneva. /http://whqlibdoc.who.int/publications/2008/9789241563703_eng.pdfS (online 5August 2010).
Council of Europe, 2002. Legal situation of the Roma in Europe. Doc. 9397 revised19 April 2002. /http://assembly.coe.int/Documents/WorkingDocs/Doc02/edoc9397.htmS (online 15 July 2010).
Decade of Roma Inclusion, 2005. Declaration of the Decade of Roma Inclusion./http://www.romadecade.org/index.php?content=77S (online 15 July 2010).
Decade of Roma Inclusion, 2008. Minutes and Summary of the XII. InternationalSteering Committee Meeting, 2008. /http://demo.itent.hu/roma/portal/downloads/Decade%20Documents/12th%20ISC%20Meeting%20Minutes%20February%2025-26%202008.pdfS (online 15 July 2010).
European Centre for Health Policy, 1999. Health Impact Assessment: mainconcepts and suggested approach. Gothenburg consensus paper. ECHP, WHORegional Office for Europe, Brussels. /http://www.hiaconnect.edu.au/files/Gothenburg_Consensus_Paper.pdfS (online 15 July 2010).
European Commission, 2004. The Situation of Roma in an Enlarged EuropeanUnion. European Commission Directorate General for Employment and SocialAffairs, Brussels /http://www.errc.org/cms/upload/media/00/E0/m000000E0.pdfS (online 15 July 2010).
European Commission High Level Advisory Group of Experts on the SocialIntegration of Ethnic Minorities and their Full Participation in the LabourMarket, 2007. Ethnic Minorities in the Labour Market An Urgent Call forBetter Social Inclusion. EC, Brussels /http://ec.europa.eu/employment_social/fundamental_rights/pdf/hlg/etmin_en.pdfS (online 15 July 2010).
European Commission, 2010. Communication from the Commission to the Council,the European Parliament, the European Economic and Social Committee andthe Committee of the Regions. The social and economic integration of theRoma in Europe COM/2010/0133 final. European Commission, Brussels./http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:52010DC0133:EN:HTMLS (online 15 July 2010).
European Roma Rights Centre, 2006. Ambulance Not on the Way: The Disgrace ofHealth Care for Roma in Europe. ERRC, Budapest /http://www.soros.org/initiatives/health/focus/roma/articles_publications/publications/ambulance_20061004/ambulance_20061004.pdfS (online 15 July 2010).
Hajioff, S., McKee, M., 2000. The health of Roma people: a review of the publishedliterature. Journal of Epidemiology and Community Health 54, 864–869.
Hungarian Ministry of Youth, Family, Social Affairs and Equal Opportunities, 2006.Housing and Social Integration Program of Roma Colonies (Roma telepekenel +ok lakhatasi es szocialis modellprogramja). Hungarian Ministry of Youth,Family, Social Affairs and Equal Opportunities, Budapest /http://www.romaweb.hu/doc/kormanyzat/telep/telprotelepek/telpro2005-06.pdfS (online15 July 2010).
Kertesi, G., Kezdi, G., 1998. A cigany nepesseg Magyarorszagon. (The Romapopulation in Hungary.) Budapest: socio-typo.
Kosa, K., Adany, R., 2007. Studying vulnerable populations: lessons from the Romaminority. Epidemiology 18, 290–299.
Kosa, K., Molnar, A., McKee, M., Adany, R., 2007. Rapid health impact appraisal ofeviction versus a housing project in a colony-dwelling Roma community.Journal of Epidemiology and Community Health 61, 960–965.
A. Molnar et al. / Health & Place 16 (2010) 1240–1247 1247
Kosa, K., Darago, L., Adany, R., 2009. Environmental survey of segregated habitatsof Roma in Hungary: a way to be empowering and reliable in minorityresearch. European Journal of Public Health. doi:10.1093/eurpub/ckp097 July17, 2009.
McDonald, C., Negrin, K., 2010. No Data—No Progress. Data Collection in CountriesParticipating in the Decade of Roma Inclusion 2005—2015. Open SocietyInstitute, New York /http://www.romadecade.org/files/downloads/General%20Resources/no-data-no-progress-20100628.pdfS (online 5 August 2010).
Ottawa Charter for Health Promotion, 1986. World Health Organization,Geneva. /http://whqlibdoc.who.int/hq/1995/WHO_HPR_HEP_95.1.pdfS (online5 August 2010).
Revenga, A., Ringold, D., Tracy, W.M., 2002. Poverty and Ethnicity: A Cross-CountryStudy of Roma Poverty in Central Europe. The World Bank Technical paper, No.531. The World Bank, Washington, DC. /http://www-wds.worldbank.org/external/default/WDSContentServer/WDSP/IB/2003/01/07/
000094946_02122404075867/Rendered/PDF/multi0page.pdfS (online 15 July2010).
Ringold, D., Orenstein, M.A., Wilkens, E., 2005. Roma in an Expanding Europe:Breaking the Poverty Cycle. The World Bank, Washington, DC /http://siteresources.worldbank.org/EXTROMA/Resources/roma_in_expanding_europe.pdfS (online 15 July 2010).
Shaw, M., 2004. Housing and public health. Annual Review of Public Health 25,397–418.
Stahl, T., Wismar, M., Ollila, E., Lahtinen, E., Leppo, K. (Eds.), 2006. Health in AllPolicies Prospects and Potentials. Ministry of Social Affairs and Health, Helsinki.
Tashev, T., Hrabanova, G., 2007. DecadeWatch: Results and Recommendations.Paper presented at the 10th meeting of the International Steering Committee(ISC) of the Decade of Roma Inclusion, Sofia. /http://demo.itent.hu/roma/portal/downloads/News/10th%20ISC/DecadeWatch.ISC.final.pptS (online 15July 2010).
