Uncounted: the hidden lives of Batwa womenBy Kathryn RamsayDespite bei - PDF document

brieÞng Uncounted: the hidden lives of Batwa womenBy Kathryn RamsayDespite being the original inhabitants of the equatorial forestsof Africa’s Great Lakes region, Batwaare, in official terms,practically invisible. Facing ongoing discrimination resultingin poverty, unemployment and poor access to education andhealth care, their situation is compounded by a lack ofacknowledgement of their struggles by their respectivegovernments. It is extremely difficult, frequently impossible,to find statistics and data about the Batwa communities inBurundi, Democratic Republic of Congo (DRC), Rwandaand Uganda; it is even harder to find gender-specific dataabout the situation of Batwa women and girls.Yet comprehensive and disaggregated data collection isvital to ensure that governments meet their obligations toprotect minorities and indigenous peoples underinternational law, and that development programmesrespond sufficiently and appropriately to the specific needs ofBatwa. The requirement is particularly great in relation toBatwa girls and women who, as previously documented byMinority Rights Group International (MRG),multiple forms of discrimination. MRG has worked with Batwa non-governmentalorganization (NGO) partners to establish areas of particularimportance, and highlight the need for more comprehensiveofficial data to help inform outreach programmes andpolicies. The unique investigations conducted in fourcountries, albeit on a small scale, identify and analyse someof the problems Batwa women and girls encounter, asconveyed by Batwa communities themselves. In particular,they focus on lack of access to education and worryingproportions of gender-based violence. BATWA WOMEN IN UGANDA. KATHRYN RAMSAY/MRG UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMENThe importance of data As the UN Forum on Minority Issues heard in 2008,standard data collection tends to ‘only count those whomwe care about’,often neglecting the specific experiences ofminorities and indigenous peoples, and particularly womenfrom these groups. The lack of information on the situationof Batwa women suggests scant concern about their plighton the part of governments, and little awareness orunderstanding of the problems and discrimination theyWithout knowledge or consideration of their situation,government policies are likely to exclude them altogether,or initiatives will fail to adequately address their issues. Thesame is true for the efforts of development agencies. Tocreate effective policies and programmes to improve thesocial and economic status of Batwa women,comprehensive information is required about the problemthe policy aims to change. Without it, the existing situationcould unintentionally be made worse. Accurate statisticaldata is therefore crucial in designing effective developmentpolicies, and is vital in measuring the impact of thosepolicies and programmes on the beneficiary community. Disaggregation of data In a development context, disaggregation of data by sex is awell-established principle, even if its implementation ispatchy. For example, the indicators measuring progresstowards the Millennium Development Goals (MDGs)should, as far as possible, be disaggregated by sex.means that it should be possible to identify the ‘gender gap’– where the goals are not being met for women and girls, ormen and boys – and steps that can be taken to address it. No similarly established principle exists for thedisaggregation of data by ethnicity. Disadvantageddoubly, of minority and indigenous women, are, therefore,hidden unless states choose to disaggregate data in this way.In the absence of disaggregated data, it is difficult toprove instances of discrimination, especially indirectdiscrimination (where apparently neutral provisions orpractices have a disproportionate disadvantage for membersof minorities or indigenous peoples). For example, aseemingly impartial requirement for all children to wearshoes to school could result in a disproportionate impacton a minority group that is poorer than othercommunities, leading to their children being unable toattend school. In such a case, statistics on the schoolattendance of different ethnic groups and data on reasonsfor non-attendance would demonstrate the discriminationand allow governments to improve the situation. To date, the only legal instrument incorporating thisprinciple of establishing indirect discrimination usingstatistics is in Europe.However, the four states underdiscussion in this briefing paper have a legal duty to ensureequality and non-discrimination under numerous Background on Batwa Batwa are indigenous peoples traditionally inhabiting theforests in the Great Lakes region of Africa. No officialfigures exist, but estimates place the population in Burundiat 30,000–40,000,Rwanda at 33,000and the easternDemocratic Republic of Congo (DRC) at 90,000.is the only one of the four countries with census data; the2002 census results showed Batwa numbered 6,705 or0.03 per cent of the population.Batwa are poor andmarginalized in all four countries. They have mostly losttheir traditional forest lands through a combination of a longprocess of deforestation to make way for farming andherding peoples, conflict leading to forced displacement, ormore recent expulsion from the forests in the name ofdevelopment or conservation. As a result, their poverty hasincreased, while the discrimination they face from dominantcommunities has become further entrenched. Batwa live onthe margins of society in poor-quality housing and havehigh levels of unemployment. Frequently, their onlyemployment is low-paid occasional work as labourers forneighbouring communities. Child mortality is high, andaccess to health care is low. As this briefing paperdiscusses, Batwa also have poor access to education andsuffer from high levels of violence. They are stereotyped bymajority communities, and resulting segregation means thatBatwa are often not permitted to eat or sit with othercommunities or use wells at the same time. Focus on Batwa women Batwa women and girls suffer multiple or intersectionalforms of disadvantage and discrimination. They arediscriminated against for being Batwa, and experiencegender discrimination as majority women do. They alsosuffer from particular discrimination on the basis of theirethnicity and gender combined, i.e. they have specificproblems because they are both Batwa and women. Thedisadvantage faced by Batwa women comes both fromoutside the Batwa community, from majority communities(for example, a prevailing myth that sex with a Batwawoman cures backache has been cited in the studies as afactor leading to rape of Batwa women by non-Batwamen), and from within the Batwa community, from Batwamen (for example men deciding whether or not theirdaughters are sent to school). UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMEN Burundi, Democratic Republic of Congo, Rwanda,Tanzania and Uganda 100 DEMOCRATIC RWANDA TANZANIA UGANDA Fort PortalBukobaButare NyaruguruRubavuLake Tanganyika Ingrained in Burundi is ‘an institutional unwillingness todifferentiate between Batwa and other poor non-Batwacommunities’,which similarly results in no officialstatistics on their situation. In the DRC, meanwhile, yearsof conflict have severely hampered any collection of data.Furthermore, logistical problems in all four countries renderdata collection challenging; many Batwa communities livein remote areas, making communication difficult.The transparent collection of official ethnic statisticsdepends on the prior recognition of the existence ofdifferent ethnic groups by the government. As in the caseof Rwanda, governments can fear that acknowledgingdifference or the existence of minorities may exacerbateor that data showing inequalities maylead to resentment between ethnic groups over who ismore disadvantaged.However, international law doesnot permit governments to deny the existence of minorityand indigenous groups to avoid implementing theirminority rights obligations.take into account that discrimination and marginalizationoccur with or without ethnic data, for example due tolong-held stereotypes; having no means of proving ordisproving cases of discrimination could in itself cause problems. Background to research When MRG began discussions with Batwa NGO partnerson designing a programme specifically to address theproblems faced by Batwa women, a major obstacle was thecomplete absence of official data to help demonstrate theextent of the problem to potential donors. It seemed thatthe lack of information on Batwa women rendered themvirtually invisible. The few figures that did exist camemainly from anecdotal evidence; however, such data didnot necessarily address Batwa women specifically. It was for this reason that MRG and its partnersdecided to focus their work on supporting Batwa NGOsand equipping them with improved skills to conductresearch on an issue of importance to Batwa women intheir respective countries. The Batwa NGOs working withMRG identified two priority research themes: education of Batwa girls violence against Batwa women. Methodology overview MRG worked with partners in Burundi, DRC, Rwandaand Uganda. The programme began with a regionalworkshop in September 2008 on data-collectionmethodology, whereby participants received training on theskills needed to design and carry out research. The sessionsincluded research ethics, sampling, various data collectionmethods and data analysis. The NGOs from each countryjointly designed studies tailored to the situation in theirrespective countries. This meant that they used differentmethodologies; however, it was not expected that theinternational instruments that they have ratified.duty to ensure non-discrimination should be interpreted toinclude a duty to collect and analyse data disaggregated byethnicity so that inequality can be highlighted andaddressed. Ethnic data: governmentsÕ role Across the four countries where Batwa live, lack of ethnicdata is a common problem, though the reasons differ.Given past ethnic conflict, the government in Rwandadoes not officially recognize the existence of any ethnicgroup, meaning that census data or other official surveysdo not exist to produce statistics showing the problemsexperienced by Batwa. That Rwanda disavows theexistence of any ethnic group also means that there is noofficial acknowledgement of the discrimination they face.Although the Rwandan government’s sensitivity towardsidentifying ethnic differences is understandable in light ofthe 1994 genocide, its approach is counter-productivebecause it fails to address the deep-rooted tensions thatcontinue to exist in Rwandan society, which may becomeexacerbated in the future.In Uganda, although the 2002 census included aquestion on ethnicity and provided figures on the numberof Batwa, their relatively small population means thatthey were categorized alongside 18 other ethnic groupswith fewer than 25,000 members in a group called‘others’.Therefore, where data on socio-economic status,for example, is disaggregated by ethnicity, it is still notpossible to see the exact situation of Batwa. UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMENBATWA GIRLS. UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMEN results could draw regionally applicable conclusions. Theteams surveyed villages, households, schools andindividuals (Batwa and non-Batwa; adults and children)between the end of 2008 and mid 2009. Each country’smethodology is explained in more detail below. Followingthe data collection and analysis, each project teamproduced a country-specific report. Their results aresummarized and discussed below; the full reports areavailable at http://www.minorityrights.org/6861/thematic-focus/gender.html. Education of Batwa girlsBackground MRG’s partners in Burundi and the DRC chose to focustheir research on education. They particularly wanted toexamine the reasons for the low level of enrolment ofBatwa girls in school and their high drop-out rates.Through a better understanding of the reasons for theproblems, they hope to improve their strategies fortackling the issue and convince the authorities anddevelopment agencies of the need to address the problemby initiating work where nothing is being done, orimproving the few initiatives that do exist. The methodology of each study was designedaccording to the information that partners felt was neededin each country, and while some similar conclusions canbe drawn, the results of the studies are country-specificand therefore not completely comparable. Resourceconstraints meant that both the Burundi and DRCresearch focused solely on the Batwa community, and didnot ask the same questions about other ethnic groups.This means that using the data to compare thedisadvantaged position of Batwa girls with that of othersections of society in the study areas is difficult. Anycomparisons between Batwa and other dominant ethnicgroups must be made using national education statistics. Burundi In Burundi, two NGOs – Unissons-nous pour lapromotion des Batwa (UNIPROBA, or Together for thePromotion of Batwa) and Union chrétienne pourl’éducation et le développement des déshérités (UCEDD,or Christian Union for the Education and Developmentof the Disadvantaged) – carried out the research betweenDecember 2008 and July 2009. They selected two provinces, Muyinga and Karuzi.They chose these areas because: they are geographicallyenclosed; larger numbers of Batwa live there; and theylack systematic development assistance. A preliminarysurvey formed the first stage of the process, namely toestablish the number of Batwa households in those areas,numbers of Batwa children, and numbers of Batwa boysand girls attending and not attending school. For themain survey, the research team selected five communeswithin the two chosen provinces that showed a lowernumber of children attending school (two in Muyingaand three in Karuzi). The second part of the study involved interviews withselected respondents in the five communes. Respondentswere Batwa girls, their parents, neighbours from otherethnic groups, teachers at the schools attended by Batwagirls, and administrative and religious authorities. Theaim of the interviews was to gather detailed informationshowing the reasons Batwa girls donot attend or completeprimary and secondary education. DRC The DRC research project was conducted between Januaryand June 2009 by a team representing five NGOs – theBatwa women’s organization Union pour l’émancipation dela femme autochtone (UEFA, or Union for theEmancipation of Indigenous Women), Action pour lapromotion des droits des minorités autochtones en Afriquecentrale (APDMAC, or Action for the Promotion of theRights of Indigenous Minorities in Central Africa),Association pour le regroupement et l’autopromotion despygmées (ARAP, or Association for the Regrouping andSelf-Promotion of Pygmies), Centre d’accompagnement desautochtones pygmées et minoritaires vulnérables (CAMV,or Support Centre for Indigenous Pygmies and VulnerableMinorities) and Collectif pour les peuples autochtones auKivu/RDC (CPAKI, or Collective for the IndigenousPeoples of Kivu).The research focused on five specific geographicalareas in South Kivu province (Bunyakiri, Kabare, Kalehe,Kalonge and Uvira), since they have higherconcentrations of Batwa than other parts of the province.The first stage involved mapping all the schools in thefive areas to find out the numbers of Batwa boys and girlsattending school. The research team visited 22 villagesand 18 schools in total. In the second part of the research, a questionnaire wasused to collect qualitative information on factors affectingBatwa girls attending school. Focus-groups of up to eightpeople were held, and between four and six people perschool or village were interviewed. Respondents wereteachers, Batwa and non-Batwa pupils, Batwa parents andkey individuals (local leaders). The names of keyindividuals to be interviewed were suggested by otherrespondents during the course of the study. Poor access to education In Burundi, the preliminary survey found 614 Batwahouseholds in the two provinces with 1,683 school-agechildren. Only 58 per cent of the school-age children wereattending school, and of these children, 65 per cent wereBatwa boys and only 35 per cent Batwa girls. UNICEFreports that the net primary school attendance ratio forboys overall in Burundi (2003–8) was 72 per cent andgirls 70 per cent. They also state that the net primary UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMEN in the population of the region. Fewer national statistics areavailable for the DRC too, which renders it more difficult toput the figures from the study into context. The availabledata shows that lack of education is a general problem in theDRC: World Bank figures state that only 51 per cent ofchildren in the DRC complete primary school.UNESCO data from 2007 show enrolment of girls inprimary school to be 39 per cent, which is almost equal tothe average percentage of Batwa girls enrolled at schoolacross the five study areas; however, without drop-out rateor attainment figures, it is difficult to draw definitiveconclusions on whether Batwa girls are really achievinglevels of education on a par with the national average. Causes of Batwa girlsÕ lack of access There were close correlations between the reasons forBatwa girls not accessing education in both Burundi andthe DRC. The Burundi study produced mainlyquantitative data showing the opinions of the variouscategories of respondents, whereas the DRC studyproduced more qualitative information in the form ofquotes from various respondents illustrating their views.Despite the different methodology of the two studies, it isclear that the main reasons why Batwa girls lack access toeducation are largely the same in both countries: poverty,the attitude of Batwa parents and early marriage. Poverty ‘They [Batwa parents] find it hard to pay the schoolfees because it’s parents who pay the teachers’ salaries inthe DRC.’ (Teacher, DRC) In Burundi, despite the fact that free primary school wasimplemented in 2005, resulting in an immediate 30 percent increase in enrolment according to the Minister ofEducation in 2006,the study found that an average ofabout 30 per cent of Batwa girls across the five communescited lack of materials (for example exercise books andclothing) as a reason for them not going to school. A lackof food was also cited by between 3 per cent and 9 per centof Batwa girls as a reason in three communes, but twocommunes reported high figures of 32 per cent and 20 percent respectively. More Batwa parents cited lack of food asa reason. In all communes except one, more than 30 percent of parents thought this factor constituted a seriousbarrier. In the final commune, the figure was 15 per cent.The DRC study found that poverty was mentioned as areason for lack of education by an overwhelming numberof interviewees of all categories across the geographical areaof the research. Teachers in the DRC noted that Batwaparents do not have the ability to pay school fees. Article43 of the 2006 Constitution abolished school fees;however, they continue to exist in practice. One teacherstated of Batwa parents, ‘They find it hard to pay theschool fees because it’s parents who pay the teachers’school attendance (2003–8) figure for girls as a percentageof boys is 97.These statistics show that school attendance by Batwachildren (and especially girls) is significantly lower thannational averages, and the gender disparity much greater. Inboth provinces, the drop-out rate of Batwa girls wasapproximately twice that of Batwa boys (on average 67 percent of Batwa girls had dropped out of school comparedwith 33 per cent of Batwa boys). Figures for the numbers of children who have neverattended school showed more variation between theprovinces. In Muyinga, of the 650 school-age Batwachildren, 248 (or 38 per cent) have never been to schooland, of those, 82 per cent are girls. In contrast, in Karuzi,of the 1,033 school-age children, a slightly smallerpercentage of children (34 per cent, or 348 children) hasnever been to school; however, more boys than girls havenever been to school (57 per cent of boys compared to 43It is interesting to note that the figures in Karuzi do notfollow the general expectation that fewer girls than boys areenrolled in school; however, it should also be highlightedthat more girls than boys dropped out of school in bothMuyinga and Karuzi. Unfortunately, it was beyond thescope of the study to investigate in more detail why theenrolment situation for Batwa girls in Karuzi was so muchbetter than for those in Muyinga, where Batwa girls werefar more likely than boys to have never attended school.In the DRC, fewer Batwa girls were in school thanBatwa boys. The difference was greatest in Kalonge, where84 per cent of Batwa children at school were boys andonly 16 per cent were girls. Most parity was found inBunyakiri, where 57 per cent of Batwa children at schoolwere boys and 43 per cent were girls. In the other areas,attendance of Batwa boys was between 59 and 61 per cent.Overall, 39 per cent of the Batwa children at school in thefive areas were girls. The study found that the 18 schoolssurveyed had 6,593 pupils, of which 235 were Batwa (144boys and 91 girls). Survey results: girls in education Batwa boys are almost twice as likely to attend schoolas Batwa girls in Burundi, while girls nationally are twiceas likely to go to school as Batwa girls.The drop-out rate of Batwa girls is twice that of Batwaboys in Burundi.In the DRC, 39 per cent of the Batwa children at schoolwere girls.This means that Batwa children make up 3.6 per cent of theschool population (boys 2.2 per cent and girls only 1.4 percent). The lack of accurate data on the overall ethniccomposition of the population of the DRC, and in particularthe ethnic make-up of South Kivu, means that it isimpossible to draw conclusions as to whether Batwa childrenare attending school in numbers proportionate to their total UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMEN salaries in the DRC.’ In addition, as in Burundi, othercosts are associated with school, such as suitable clothingand books. A lack of food also emerged as a problem in theDRC, with one parent saying, ‘The children alsocontribute in providing daily meals instead of going toschool because they say in order to be able to study wellyou have to eat.’ A key individual (local leader) stated ofBatwa families, ‘In a family with five children only twochildren at most are at school and they must be theyoungest, i.e. those who are not yet able to work and bringin something for the family.’ The responses also underlined how gender plays acrucial role in decisions about which children to educatewhen resources are scarce. Teachers in the DRC repeatedlysaid that Batwa girls were more likely than boys to be sentto work to earn money for the family. One teacher said, forexample, ‘Girls drop out of school more often than boysbecause girls are often sent to well-off families where theyserve as maids.’ This situation also makes girls morevulnerable to exploitation. In Burundi, 12 per cent ofBatwa parents in one commune mentioned that householdtasks take up girls’ time, preventing them from going toschool. Interviews with Burundian administrativeauthorities also identified that girls have to do householdchores that boys do not, and that where food is a problem,girls must work to help the family. Attitudes of parents The DRC study found that teachers and key individualsfrequently cited the lack of awareness of parents as a reasonfor Batwa girls not attending school. Teachers said of Batwaparents, ‘They don’t see the point of going to schoolbecause they don’t see immediate results,’ and, ‘There is notenough awareness among the pygmies as to why theyshould bring their children to school’. For one keyindividual, ‘Pygmies themselves discriminate in a waybetween their children because they do not accept that theyshould educate their girls … because they say that girlsdisappear and go and start other families elsewhere,’ acomment that reflects the patriarchal inheritance systemwhere only boys inherit from their father.In Burundi, both Batwa girls and Batwa parents wereasked how interested they were in school. In all communes,more than 92 per cent of Batwa girls had a ‘great interest’in school, with two communes at 98 per cent and one at99 per cent. Batwa parents in two of those communes alsoshowed strong interest in Batwa girls’ education (90 and 92per cent ‘great interest’). In the three remaining provinces,10 to 15 per cent of parents were indifferent to school forBatwa girls. Yet despite the high percentage of parents claiming agreat interest in Batwa girls’ education, significantnumbers agreed with the statement, ‘School is not goodfor Batwa girls.’ For example, in Buhiga commune, 78 percent of parents reported a ‘great interest’ in school forBatwa girls; however, 49 per cent also stated that school isnot good for them. Batwa girls reported the ‘ignorance ofparents who do not recognise the benefit of school’ as abarrier to their education in approximately the sameproportions as parents in the same communes felt that Early marriage ‘Women were created to be married, have children andto look for food and not to spend time in front of menstudying.’ (Mutwa parent, DRC)The attitude of Batwa parents in the DRC to education ofgirls is closely tied up with attitudes towards marriage andthe role of girls in society. One parent said, ‘Women werecreated to be married, have children and to look for foodand not to spend time in front of men studying.’ Anotherparent’s response was, ‘According to the Pygmies, a girl whogoes to school and reaches secondary level is considered afree woman, i.e. a whore, because she might not get ahusband within the community.’ The responses of otherparents and teachers indicate that the financial implicationsof marriage, i.e. a dowry given to the girl’s family,encourage early marriages and discourage education.According to one parent, ‘Pygmy parents don’t expect tolive long so their daughters must get married quickly forfear of being abandoned by the young men in the villagesbut also so as to use up the dowry before they die.’ Anotherparent said, ‘The more a girl studies, the less chance shehas of getting married,’ and a teacher’s view was, ‘Womenencourage their daughters to get married quickly so thatthey can get a few goats to eat as a dowry.’ BATWA WOMEN SINGING. KATHRYN RAMSAY/MRG In Burundi, the statistical data confirms that earlymarriage is a barrier to Batwa girls’ education. In twocommunes, more than 40 per cent of Batwa girls cited it asa reason preventing them from going to school and, in thesame two provinces, 35 per cent and 38 per cent of parentsmentioned it. In one commune, Mwakiro, only 6 per centof parents thought early marriage a problem in contrastwith 25 per cent of Batwa girls. In all communes, Batwagirls thought early marriage a greater problem than theparents did. School administrators in Burundi confirmedthat early marriage disadvantages Batwa girls and preventsthem from completing their education. Other factors The distance between home and school can be a factordiscouraging girls from going to school.However, theBurundi study concluded that it was not a major problemin their sample area. In three communes, only 1 per cent ofpupils had more than 10 km to travel every day, andalthough in the other two communes, 11 per cent and 90per cent of pupils had 11–20 km per day to travel, thisdistance was not thought to be a major barrier to Batwagirls attending school. In the DRC, the more precarioussecurity situation means that, although distance was notmentioned in every area, it did emerge as a bigger problemthan in Burundi. A DRC parent explained, ‘The distancebetween the village and the school scares children,especially girls who are afraid of being raped on the way byarmed gangs.’In addition, a few respondents in both countriesreferred to nomadism as a barrier. Traditionally, Batwa, ashunter/gatherers, moved within the forests; however, theloss of their lands has changed this practice. In Burundi,the study found that in almost all communes, nomadismno longer exists. Isolated cases of a nomadic lifestyleaffecting educational opportunities were seen in onecommune, where it was highlighted as a barrier toeducation by 2 per cent of Batwa girls.It is revealing to note discrepancies in responses ofadministrative authorities/key individuals compared toteachers and Batwa themselves, which highlightdiscriminatory attitudes towards Batwa. The DRC studyfound key individuals with very negative opinions of Batwachildren, for example stating, ‘They’re generally layabouts,’and, ‘Some say that Pygmy children are good for nothings.’In contrast, teachers stated, ‘With regard to the issue ofapplication and the level of these Pygmy children in class,the teachers of this school always said that these childrenare among the best in the school and are always among thetop in their class.’ However, the way teachers phrased theirresponses shows that they are aware of the stereotypes ofBatwa, for example one said, ‘Even those who left schooldid quite well compared to other children, consideringtheir reputation.’ It is, of course, impossible to knowwhether that teacher’s view of the ‘reputation’ of Batwachildren had any influence over the way that s/he treated UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMEN them in the classroom before they proved themselves to beas good as other children. It is clear, therefore, that the stereotypes held aboutBatwa children by society in general, including possibly bytheir teachers, are inaccurate. However, those discriminatoryattitudes may negatively affect their experience of school. Aparent cited ‘discrimination and marginalization at school’as a barrier to girls’ education.In Burundi, the design ofthe study did not focus on negative stereotypes; however, aswith the DRC, responses from teachers clearly show thatwhen they have the opportunity to study, Batwa girls do aswell in class, have the same IQ level as other children, andmix well with their peers. Conclusions The studies conducted in Burundi and the DRC haveresulted in new insights into the problems Batwa girls facein accessing education. The invisibility of Batwa womenand girls in the majority of surveys and national statisticsmeans that for the first time, detailed data on their schoolattendance and drop-out rates is available to highlight theirsituation. The investigations – both quantitative andqualitative – into the barriers to education faced by Batwagirls have also resulted in more nuanced understanding ofhow several factors interlink to prevent Batwa girls achievingtheir full potential. The qualitative information illustratesattitudes of the different actors and starkly outlines whatchallenges there are in addressing the situation.Both studies make recommendations for action toimprove the educational prospects of Batwa girls.Recommendations include ensuring that education forBatwa girls is free in reality, and providing assistance tofamilies so that they can pay for other costs associated withschool, such as books, uniforms and pens. Ensuring foodsecurity through provision of livestock, seeds or land isconsidered important in both Burundi and the DRC, sothat Batwa families will not need to withdraw girls fromschool to help provide for the family. Promoting income-generating activities for Batwa families is suggested for thesame reasons. The studies make recommendations on raising theawareness of Batwa parents on the importance of educationfor their daughters and encouraging them to prioritizeeducation. Regular monitoring to ensure all Batwa girls ofschool age are attending school and do not drop out wasThe recommendations do not refer directly topreventing early marriage; however, awareness-raising ofparents should also include discussion on the disadvantagesof early marriage,while efforts to tackle poverty withinBatwa families should reduce the financial demand for girlsto marry early to produce a dowry.Although the studies provide a valuable new insightinto the realities of life and lack of educationalopportunities for Batwa girls in Burundi and the DRC,they are small-scale, conducted with limited resources. A UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMEN larger-scale comparative study, covering a widergeographical area and other ethnic groups would beextremely useful to highlight the relative situation of Batwagirls compared to other girls in society. Both studiesrevealed discriminatory attitudes from teachers,administrative and other local authorities, but it wasbeyond the scope of the studies to investigate the impact ofthese negative attitudes on the education of Batwa girls.Social attitudes are known to impact on the educationalattainment of children and on their feelings towardsa subject that should be further examined. Background MRG’s partner organizations in Rwanda and Ugandaselected violence against Batwa women as the theme fortheir research. Anecdotal evidence suggested that violence,both from inside and outside the Batwa community was afrequent problem for Batwa women. The aim was todiscover more about the prevalence of violence and itscauses in order to suggest strategies for tackling the issue.The sensitivities surrounding the subject of violence againstwomen, in particular sexual violence, meant that themethodology of the research needed to be designedextremely carefully in order to encourage respondents tofeel comfortable in revealing information, to ensureconfidentiality, and avoid any negative consequences forwomen taking part in the research. Uganda The United Organization of Batwa Development inUganda (UOBDU) and the African International ChristianMinistry (AICM) conducted the research in Ugandabetween November 2008 and April 2009. The Batwapopulation in Uganda live in the south-west of thecountry; the study concentrated on three districts – Kabale,Kanungu and Kisoro – interviewing 120 people in total,with Kisoro providing a higher number of respondentsbecause of the larger number of Batwa living there. A list of names of all Batwa living in those areas wasdrawn up. Respondents were selected to take part in focus-group discussions, an individual interview or a keyinformant interview. Fifty-six Batwa were randomlyselected for focus-group discussions. Of these, 27 werewomen and girls and 29 men. Girls’ focus-groups were runseparately from the others, with a different set of questions,to ensure that the subject was addressed in a way suitableA large part of the study involved individual interviews.Forty-four Batwa women aged over 17 were randomlyselected for individual interviews about their views onviolence in their community and their own personalexperiences of violence. To provide some context, questionswere asked about their background and level of knowledgeof women’s rights. Of these 44 Batwa women, 89 per centwere married and 60 per cent had never been to school; 60per cent of them said they knew about some women’srights. When respondents were asked to name those rights,58 per cent knew about the right to property, 12 per centmentioned the right to education and the right to accessjustice from law courts; however, other responses included‘the right to do housework’, ‘the right to make mats’ and‘the right to take care of the husband’, clearly showing thatsome women were confusing rights with gender roles. Twenty ‘key informants’ (persons of influence) – 15 ofwhom were non-Batwa – were identified to give theirperspectives on violence against Batwa women; 16 were menand only 4 were women. It proved difficult to find womenkey informants. Key informants’ positions were governmentemployees (11), local councillors (2), NGO employees andcommunity resource persons (3), teacher (1) and Batwaleaders (3). The small numbers of women and Batwa amongthe key informants reflects the power dynamics in society;few Batwa, few women and fewer Batwa women have thelevel of education needed to take up positions in localgovernance or obtain government employment. Rwanda The Rwanda research was conducted between the end of2008 and mid 2009 by African Initiative for MankindProgress Organization (AIMPO), Communauté des potiersdu Rwanda (COPORWA, or Rwandan Potters’Community) and Association pour le développementglobal des Batwa du Rwanda (ADBR, or Association forthe Global Development of Rwanda’s Batwa). It focused onthree districts – Gasabo, Nyaruguru and Rubavu – selectedbecause they have higher numbers of Batwa resident there. The research team selected 95 Batwa women and girls,aged over 13, to interview, with slightly more coming fromNyaruguru and Gasabo. The majority of the women weremarried (57 per cent) with 23 per cent widowed and 20per cent single. As with Uganda, their level of education islow: 54 per cent had not been to school; 46 per cent hadreceived some primary education, although may not havecompleted primary school. They were asked about theirunderstanding of what constitutes violence against women. The most frequently mentioned form of violence wasrape, with physical violence and marginalization the nextmost numerous responses. In contrast to the Ugandanstudy, the Rwandan research focused on the respondents’beliefs and views on the prevalence and causes of violenceagainst Batwa women. The team decided not to ask thewomen about their own personal experiences of violencebecause of the sensitive nature of the subject. Types and prevalence of violence A shocking statistic emerged from the study in Uganda. Ofthe Batwa women responding to individual interviews, 100 per cent had experienced some form of violence. The studyfound that for the majority of respondents (64 per cent),the violence was ongoing or had taken place within thepast 12 months. In comparison, a survey from the UgandaBureau of Statistics on health in 2006 found that 70 perviolence at some time since the age of 15.The data is notdisaggregated by ethnicity; however, the national andregional averages indicate that Batwa women experiencemore violence than women from other communities. The UOBDU/AICM study asked the individualinterview respondents about the type of violence they hadexperienced. Fifty-seven per cent had been sexually abusedat some time, with 46 per cent having suffered maritalrape; 93 per cent had been slapped or hit; and 52 per centbeaten or kicked. Other forms of violence they mentionedincluded being threatened with a weapon (36 per cent),dowry-related violence (25 per cent), forced marriage (21per cent) and being detained against their will (18 percent). The results of the national survey found that 39 perIn south-westUganda, where Batwa live, the figure from the nationalsurvey of those experiencing sexual violence was slightlyhigher at 41 per cent. Again, the results suggest that Batwawomen suffer from more sexual violence than women from Violence against Batwa women in Uganda A staggering 100 per cent said they had experiencedviolence; for the majority, the violence was ongoing orhad taken place in the past 12 months.Fifty-seven per cent had been sexually abused at sometime, with 46 per cent having suffered marital rape.The participants in focus-group discussions and theindividual respondents all concluded that alcoholismand poverty were major causes, with husbands themain perpetrators. Rwanda The design of the Rwandan study means that there are nosimilar statistics on Batwa women’s personal experiences ofviolence. However, it produced figures on perceptions ofBatwa women regarding violence in their communities.When asked if they had heard of instances of violenceagainst Batwa women in their local area, 80 per cent saidthey had heard of cases. Only 7 per cent gave a definite‘no’ response, with 13 per cent preferring not to answerthe question. Fifty-seven per cent of respondents hadheard of instances of violence against Batwa womenduring the previous two weeks; 51 per cent had heard ofcases during the previous 12 months. A 2004 survey bythe Rwandan Ministry of Gender and Family Promotion(MIGEPROF) found that 32 per cent of respondents hadexperienced ‘community violence’ (violence perpetrated bysomeone not belonging to the victim’s family) in the UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMEN previous five years, and that 31 per cent had experiencedphysical domestic violence.The Rwandan study asked respondents about thetypes of violence they thought Batwa women and girlsexperienced in their local area. Fifty-four per centmentioned physical violence, 38 per cent psychologicalviolence, 35 per cent marginalization of natural resources,24 per cent rape, 23 per cent exploitation and 22 per centsexual violence. Four per cent did not believe violenceagainst Batwa women happened in their area. It isinteresting to note that ‘marginalization of naturalresources’ was the third highest mentioned form ofviolence, reflecting the fact that Batwa have lost their landand other resources (for example, access to areas to gatherclay for making pottery), and subsequently struggle tosurvive. A survey by the United Nations Development Fundfor Women (UNIFEM) in 2008 of four districts inRwanda found that 26 per cent of women had beenslapped, beaten or otherwise physically abused by theirhusbands, and 43 per cent had suffered marital rape.Again, direct comparisons of the data cannot be drawn,because the NGO study did not cover personalexperiences; however, if the 54 per cent of RwandanBatwa women citing physical violence are describingviolence within the family, it could be that Batwa womensuffer higher levels of domestic violence than othercommunities. The perception among Batwa women iscertainly that they suffer from more violence than otherwomen in Rwanda: 61 per cent believe the level ofviolence against Batwa women is greater than for otherwomen, and only 5 per cent believe it is less. Context of violence In Uganda, to the question of where the violence tookplace, responses included the home (67 per cent), bars (33per cent), fields/bushes (17 per cent) and schools (3 percent). Domestic violence was located in the home (86 percent), bars (33 per cent) and fields/bushes (6 per cent)according to respondents. In Rwanda, the domestic spherewas also the place mentioned most by respondents, with 56per cent saying they thought violence against Batwawomen occurred in the family, and 55 per cent saying itoccurred within a couple. Forty-five per cent mentionedthat violence happened in public places; 44 per centspecified schools; and 28 per cent reported that violencetook place in the community.When questioned about perpetrators of violence againstBatwa women in Uganda, answers from the individualinterviews corresponded with responses from focus-groupdiscussions. Respondents to individual interviews wereasked who had perpetrated the violence that they had citedin the interview. One hundred per cent of respondents said thathusbands perpetrated violence; 5–7 per cent alsomentioned each of the following categories – other family UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMEN members, the police, local defence personnel or delinquents– as perpetrators, showing that some Batwa women hadexperienced multiple occurrences of violence perpetrated bydifferent people. The Ugandan national survey found thatfor 63 per cent of married women who had experiencedphysical violence, the perpetrator was her husband/partner,and for 23 per cent it was a former husband/partner. Thesefigures suggest that Batwa women are more likely toexperience violence from their husbands/partners thanwomen in other communities. The national survey found that 17 per cent of marriedand unmarried women had experienced violence from ateacher, 0.5 per cent from an employer, and 0.4 per centfrom the police or soldier. These figures suggest that Batwawomen experience higher levels of violence from localauthorities (7 per cent from police or local defencepersonnel) than women from other communities. Theabsence of Batwa women reporting violence by teacherscould reflect their general lack of access to education. It isdifficult to compare the figures with complete certainty,though, because the UBODU/AICM study asked aboutperpetrators of violence in general (i.e. it included physical,sexual and psychological violence), whereas the nationalsurvey specified physical violence.The qualitative data obtained through the Ugandanfocus-group discussions revealed a more nuanced picture ofthe way Batwa women and men understand the situation.Participants in all focus-group discussions agreed thatperpetrators of violence were Batwa and non-Batwa, andthat Batwa husbands were the main perpetrators. Onegroup discussed how local council leaders indirectlyperpetrated rights violations through not respondingadequately to Batwa women’s reports of violence. Oneparticipant stated: ‘Some time ago women in our community used toreport their husbands to the chairman [of the LocalCouncil first level (LC1)]. When the chairman LC1met the husbands, they gave him some money and hekept on dodging the women until the time when thewomen gave up; the husbands again violated the rightsof their wives because they had not been punished.’ The issue of reporting of cases of violence will be discussedin more detail below.The Ugandan study asked the individual womenrespondents about the effects of the violence on theirhealth. Sixty-eight per cent reported some physical healthproblems, mainly related to sexual violence, includingunwanted pregnancy; 77 per cent reported psychologicalproblems. Those psychological problems included fear orworry (47 per cent), shame (47 per cent), humiliation (32per cent) and feeling life is useless (6 per cent). Forty per cent of respondents had sought medical helpafter experiencing violence. Of those who had not soughtmedical help, all felt they did not need medical attention.However, 43 per cent also mentioned that they did nothave any money for medical fees or transport to take themto medical facilities. Furthermore, 14 per cent said they didnot know where to go, and 14 per cent felt that medicalcare would not be beneficial. Causes of violence ‘When we go home from school, we find there is nosoap, no food and other things, so when boys or menpromise to give us these things some of us end up giving in.’ (Mutwa girl, Uganda)Both studies examined the views of respondents on whatcaused cases of violence against Batwa women. In Rwanda,the perception of respondents was that extreme poverty wasthe main cause. Sixty-seven per cent cited poverty, while 44per cent said violence occurred because of contempt forBatwa women. A number of other responses showedinteresting connections to the position of the Batwacommunity as a whole in Rwandan society. These included‘bad historical antecedents’ (26 per cent), ‘because we areless in number’ (25 per cent) and ‘because we are notrecognised as a people’ (25 per cent). The respondents alsocited a lack of education as a cause of violence (31 percent), and myths about Batwa women, in particular thathaving sexual relations with a Batwa woman will curebackache (15 per cent). In Uganda, the participants in focus-group discussionsand the individual respondents all concluded thatalcoholism and poverty were the major causes, with 82 percent of individual respondents citing alcoholism and 62 percent poverty. Respondents in the focus-group discussionsreported that men were more likely to beat or rape theirwives when drunk, and that drunkenness made women andgirls more vulnerable to abuse because they were less ableto resist advances from men. Some of the other reasons cited by respondents showthat there is a tendency for both women and men to blameBatwa women victims for the violence they experience. Forexample, individual interview respondents suggested ascauses of violence, ‘wife refusing to give food to thehusband’ and ‘women refusing to look after children’.Focus-group participants suggested that alcoholconsumption by women meant that they failed to do theirdomestic chores, which led to violence from theirIn contrast, alcoholism did not emerge as an issue inthe Rwandan study. The reason may be becauserespondents were not directly asked about it. It is unlikelyto be because it is not an issue at all, because theMIGEPROF study found that 43 per cent of women whohad experienced domestic violence attributed it to abuse ofalcohol by the perpetrator,while the UNIFEM studyfound that of the incidents of sexual violence experiencedby respondents, in 22 per cent of cases the perpetrator haddrunk alcohol. The Ugandan focus-group discussions around poverty asa cause of violence revealed that a lack of basic necessities,such as food, forced Batwa women and girls to acceptadvances from men. One Mutwa girl reported, ‘When wego home from school, we find there is no soap, no food andother things, so when boys or men promise to give us thesethings some of us end up giving in.’ The girls also said thatthose who were able to generate income through sellingitems, such as avocadoes, clothes or handcrafts, were lessvulnerable to violence because they could resist the advancesof men and boys exploiting their poverty. Justice and assistance for victims In Rwanda, 47 per cent of respondents said, when askedabout possible measures to stop violence, that reportingmethods should be improved. This figure suggests thatthere is a serious problem with Batwa women accessingeffective justice when they are victims of violence. Onerespondent mentioned that in a Batwa community thathad a woman as head, women and girls had sometimessought her help after experiencing domestic violence, uponwhich she intervened on their behalf with their partners. The study in Uganda asked the individual respondentswhether they had sought help after experiencing violence.Eighty-four per cent had approached the authorities,another agency or a member of their family. The majority ofrespondents (57 per cent) had approached the LocalCouncil. In the areas where Batwa live, Local Council firstlevel (LC1) officials are the only representatives of theauthorities nearby. No police are based in Batwa villages; itis the responsibility of the LC1 to deal with minor cases andrefer more serious cases to the police. Respondents alsomentioned approaching the church (25 per cent), theAbatakaor village council, NGOs and the police (each 18per cent). Sixteen per cent of respondents did not approachany agency or individual (including family members) forhelp. Of those, 71 per cent were too scared to ask for help,29 per cent were concerned about confidentiality and 14per cent did not know what help was available. In contrast,the national survey found that the majority of victims ofviolence approached their family (48 per cent), with only 18per cent going to a social service organization and 6 per centthe police. The responses of Batwa women suggest that they areaware that they should be able to live free from violence,and that they are trying to use available mechanisms toimprove the situation. However, other respondents infocus-group discussions commented that Batwa womenare shy and do not report violence because they have lowself-esteem and education. A further barrier to womenreporting violence included corruption by local authorityrepresentatives, who charged prohibitive ‘fees’ to take upcases even though it is illegal to do so. They thereforeeither do not attempt to report instances of violence incase they are asked to pay, or are frustrated in theirattempts to access justice. UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMEN On a more optimistic note, of the Ugandan respondentswho had sought help, 78 per cent had experienced somepositive change afterwards, such as their husband notbeating them again after he was fined or had receivedcounselling. One woman noted in a focus-group that, afterreporting her husband, ‘He was arrested and taken to thesub-county during broad daylight. Taking him to the sub-county while everybody was looking at him made him feelembarrassed and by the time he reached the sub-county hewas already a changed man.’ Conclusions This research is unique. No other studies into violenceagainst women in either Rwanda or Uganda have focusedon Batwa women and the specific violence they experienceas a result of belonging to an indigenous community andbeing women. The statistic that 100 per cent of Ugandan Batwawomen had experienced violence – significantly higher thannational averages – should inspire immediate strategies totackle gender-based violence to include Batwa women, andbe specifically tailored to address their situation. Many respondents from both studies suggestedawareness-raising on the issues of gender-based violencewould help reduce the levels of violence suffered by Batwawomen and girls. Fifty-two per cent of respondents inUganda indicated that mass sensitizations on gender-basedviolence were needed. In Rwanda, 38 per cent thought menshould be educated about violence against women, and 42per cent suggested women should be trained on their rights. Some suggestions focused on improving the justicesystem for victims. Almost half of Rwandan respondentsproposed improving the methods for reporting violence;Ugandan respondents suggested a variety of measures:punishing wrongdoers (9 per cent), culprits to pay heavyfines to their victims (5 per cent), imprisoning wrongdoers(2 per cent), and reducing court fees (2 per cent). Othersuggestions for reducing violence included: training teacherson violence against girls (Rwanda 25 per cent), providingmore safeguards for girls at school (Rwanda 33 per cent),stopping alcoholism (Uganda 14 per cent), and providingmore understanding for women who report violenceSeveral suggestions from respondents reflect the generaltendency for victims of gender-based violence either to beblamed for the violence they experience or to be blamed forfailing to prevent it. Respondents in Uganda indicated that‘girls should protect themselves’ (although they did not givedetails of exactly how they should do so) and that ‘womenshould stop moving at night’. A participant in a men’sfocus-group said: ‘Government should help us in these problems becausewe as parents have tried our level best and failed. Wehave young girls who disrespect us as their parents andalso move at night and we have failed to handle them, UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMEN maybe if police can arrest them, they can learn andbehave better.’ Violating women’s freedom of movement by preventingthem from going out at night does nothing to change theattitudes of those men who see any woman out of thehouse as a legitimate target for violence. Instead, it risksfurther stigmatizing women who go out after dark for anynumber of reasons, e.g. work, visiting friends, attending asick relative, as ‘deserving’ of any violence they experience.It is very important that strategies to tackle domesticviolence against Batwa women be implemented; however,they must be done in a sensitive manner without causingfurther prejudice against a community already stigmatizedby the majority. That violence against Batwa women inRwanda and Uganda also takes place in public spaces,including schools, shows that attention must be paid bothto violence from majority communities and from withinthe family. Since the studies reported here could only cover limitedgeographical areas, further research is needed for a fullerpicture to emerge. A comparative study of other ethnicgroups in Uganda would show whether there are similarlyhigh levels of violence in other communities in thatgeographical area, or whether the rates of violence againstBatwa women are significantly higher. It would also beuseful to have comparative data on whether women fromother ethnic groups experience similar problems to Batwawomen when reporting violence to local authorities, orwhether the marginalized position of Batwa in society leadsto increased difficulties for Batwa women in accessingFor a clear picture of the situation to emerge inRwanda, the sensitive issue of asking respondents abouttheir own experiences of violence would need to beaddressed. More qualitative data would also help clarify thereality for Rwandan Batwa women. Next steps These four new NGO studies on the experiences andproblems faced by Batwa women and girls are a movetowards ensuring their visibility through reliable data.However, they are only a small first step. Small-scale NGOstudies on specific issues are no substitute for official, large-scale surveys, conducted sensitively, respecting confidentialityand the principle of ethnic self-identification, and used totackle exclusion and discrimination. It is clear from the results of the research in all fourcountries that Batwa women and girls are marginalized BATWA WOMAN AND CHILD IN BURUNDI. AJANTA KAZA/MRG and suffer discrimination and worrying levels of violence;therefore, it is imperative that governments anddevelopment agencies recognize that, since their exclusionis disproportionate to that of other communities in thearea of education, and they suffer high levels of violence,they are also very likely to suffer increased marginalizationin other areas of life for which there is not yet any data.Robust official data must, therefore, be collected anddisaggregated to ensure a full understanding of, andsuitable responses to, their particular problems. RecommendationsData Collection Governments in the region should acknowledge thataccurate data disaggregated by ethnicity and gender isessential to develop and monitor effective programmesto improve the situation of minority and indigenouswomen and tackle discrimination against them. Aprerequisite for this work is for governments torecognize the diversity of the population within theirterritories and acknowledge the existence of minoritiesGovernments in the region should collect ethnic datasensitively and should work on a consultative basiswith minority and indigenous communities in orderto build their confidence in the process. Governmentsshould collect and use ethnic data in accordance withbasic principles. The reason for the data collection, theprocess, and what will happen to the data collectedshould all be transparent.International development agencies should collect anddisaggregate data by gender and ethnicity to ensurethat their programmes are effective in benefitingminority and indigenous women. Developmentagencies should systematically require such data in alltheir fields of work, thus encouraging improved datacollection across the region. UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMEN As members of one of the poorest and mostmarginalized communities in the Great Lakes Region,Batwa children, especially girls, should be identified andsystematically supported by governments to start schoolat the same age as other children, and to continue to thelevel they choose. This should include implementationof awareness-raising schemes to ensure that Batwaparents understand the importance of education fortheir daughters and discourage early marriage.Lack of school materials and hunger often causes Batwagirls to drop out of school. Governments andprogrammes to provide them with school materials andfood to prevent them leaving early, and ensure theirinclusion in existing programmes, for example theWorld Food Programme (WFP) scheme currentlyongoing in all four countries, which provides schoollunches for poor children. Violence Governments should ensure that their strategies onviolence against women include, and are culturallyrelevant to, Batwa women. Programmes should beimplemented in conjunction with Batwa communities,Batwa organizations and women’s rights organizationsto educate Batwa women about their rights, and Batwamen about women’s human rights. Public educationprogrammes should also address discrimination andviolence against Batwa women perpetrated by majorityThe operation of the justice system at the local levelshould be reviewed to ensure that Batwa women havefull and effective access to justice. Monitoring shouldtake place at the local level to ensure that Batwawomen are not denied access to the justice systembecause of lack of money or official corruption. Judges,police and government officials should receive trainingon women’s rights and violence against women,including domestic violence. UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMEN1MRG uses the term , which is the plural form of Twa. Pygmy is the term ordinarily used in the DRC, and thusappears in quotes from the DRC research, although in othercontexts it is considered derogatory. 2See, for example, Jackson, D., Twa Women, Twa Rights in theGreat Lakes Region of Africa3UNIPROBA Burundi estimate from 2000 in MRG, WorldDirectory of Minorities and Indigenous Peoples, URL:www.minorityrightsgroup.org/directory4CAURWA Rwanda survey in 2004 in MRG, World Directory ofMinorities and Indigenous Peoples, URL: www.minorityrightsgroup.org/directory5MRG, Erasing the Board, London, MRG, 2004, p. 9; 30,000 ineach of North Kivu, South Kivu and Ituri. 6Uganda Bureau of Statistics, , Kampala, Uganda Bureau of Statistics, 2006 (basedon 2002 information), p. 26, retrieved April 2010:http://www.ubos.org/onlinefiles/uploads/ubos/pdf%20doc7Jackson, 8‘Speaking truth to power’, statement to the UN Forum onMinority Issues, 1st Session on ‘Minorities and the right toeducation’, by Daniel Losen, Civil Rights Project, University ofCalifornia, Los Angeles, retrieved April 2010,http://www2.ohchr.org/english/bodies/hrcouncil/minority/9UN Statistics Division, ‘Millennium Development Goals Indicators’, January 2008, retrieved April 2010,http://unstats.un.org/unsd/mdg/Host.aspx?Content=Indicators/OfficialList.htm. 10The EU Race Directive provides that states may provide for‘indirect discrimination to be established by any meansincluding on the basis of statistical evidence’, Council Directive2000/43/EC of 29 June 2009, retrieved April 2010, http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:32000L11For example, the International Convention on the Eliminationof Racial Discrimination, International Covenant on Civil andPolitical Rights, International Covenant on Economic, Socialand Cultural Rights, and African Charter on Human and Peoples Rights.12Rwanda has a history of ethnic conflict between the Hutumajority and minority Tutsi populations (with Batwa oftencaught in the middle). The country experienced ethnicviolence, including massacres of Tutsis, during the 1960s and1970s. Tensions were not addressed and resurfaced in the1994 genocide. Failure to recognize the existence of ethnicgroups allows a situation to develop in which one group canenjoy a de facto situation of privilege, but no-one is allowed tochallenge it. For more, see MRG shadow report on Rwandasubmitted to the UN Human Rights Committee for its 95thSession, March 2009, retrieved April 2010,http://www2.ohchr.org/english/bodies/hrc/docs/ngos/13Uganda Bureau of Statistics, 14Warrilow, F., The Right to Learn: Batwa Education in the GreatLakes Region of Africa, London, MRG, 2008, p. 21.15Third Periodic Report submitted by Rwanda to the UN HumanRights Committee, UN Doc. CCPR/C/RWA/3, 95th Session,March 2009, para. 289: ‘Given the factors that led to theRwandan genocide of 1994, however, the Government refusesto recognize as a category communities seeking to identifythemselves as ethnic minorities or groups that claim to havebeen born with rights denied to the rest of the population.’16This position is not only held by the Rwandan government.According to the UN Committee on the Elimination of RacialDiscrimination, a government representative, ‘referring to theissue of disaggregated data on ethnicity, said that thegovernment’s policy was to avoid measures that fuelledsectarian fragmentation. Therefore, the government did notpublish statistics on ethnicity and religion because such datawould prompt people to make comparisons and drawdistinctions between various ethnic and religious groups.’ SeeUN doc., CERD/C/SR.1714, 2005.17UN Human Rights Committee, General Comment 23 on Article27: ‘The existence of an ethnic, religious or linguistic minorityin a given State party does not depend upon a decision bythat State party but requires to be established by objectivecriteria.’ UN Doc CCPR/C/21/Rev.1/Add.5, para. 5.2, retrievedApril 2010, http://www.unhchr.ch/tbs/doc.nsf/(Symbol)/18UNICEF, Country Statistics for Burundi, retrieved April 2010,http://www.unicef.org/infobycountry/burundi_statistics.html 19World Bank, Country Statistics for DRC, retrieved April 2010,http://ddpext.worldbank.org/ext/ddpreports/ViewSharedReport?&CF=&REPORT_ID=1305&REQUEST_TYPE=VIEWADVANCED&DIMENSIONS=62 20UNESCO, Education for All, Global Monitoring Reportthe DRC, retrieved April 2010, http://gmr.uis.unesco.org21Voice of America, Burundistruggles to implement free primary, 16 May 2006, retrieved April 2010, http://www1.voa22Constitution of the DRC, Article 43 includes ‘L’enseignementprimaire est obligatoire et gratuit dans les établissementspublics’ (Primary education is obligatory and free in publicschools), retrieved April 2010, http://www.presidentrdc.cd/23Ramsay, K., ‘The gender dimension of minority and indigenousState of the World’s Minorities andIndigenous Peoples 2009: Events of 20082009, p.24For more information on the impact of discrimination on theeducation of Batwa children, see Warrilow, 25These include negative health consequences of earlypregnancy for mothers and babies; psychological impacts oflack of choice in marriage and sexual relations; vulnerability toviolence especially when the husband is older; vulnerability toHIV/AIDS when he has multiple partners; denial of educationleading to fewer job opportunities in future; a lack ofknowledge of rights; and low self-esteem. See UNICEF, Marriage: Child Spouses, Innocenti Digest no. 7, March 2001,retrieved April 2010, http://www.unicef-irc.org/publications/pdf/26Ramsay, 27Uganda Bureau of Statistics, Survey 2006, p. 293, retrieved April 2010, http://www.measuredhs.com/pubs/pdf/FR194/FR194.pdf 30Rwanda Ministry of Gender and Family Promotion, ‘ViolenceAgainst Women’, Kampala, 2004, pp. 27, 48, retrieved April2010: http://repositories.lib.utexas.edu/bitstream/31UNIFEM, Baseline Survey on Sexual and Gender BasedViolence in Rwanda, 2008, p. 31, retrieved April 2010,http://www.unifem.org/attachments/products/baseline_32Rwanda Ministry of Gender and Family Promotion, 33UNIFEM, UNCOUNTED: THE HIDDEN LIVES OF BATWA WOMEN working to secure the rights ofminorities and indigenous peoples Uncounted: the hidden lives of Batwa women© Minority Rights Group International, May 2010 Thanks to Véronique Dion, Gender programme assistant, and Melissa Whitney-Long, MRG intern, whoassisted with research for this briefing paper. MRG gratefully acknowledges the support of Irish Aid and Cordaid. The contentsof this document are the sole responsibility of Minority Rights Group International and can under no circumstances beregarded as reflecting the position of Irish Aid or Cordaid. MRG would also like to thank the Canadian InternationalDevelopment Agency (CIDA) Conference Fund, which kindly supported one of the events that helped partners gain the skillsto carry out the research on which this report is based. Author: Kathryn Ramsay, Gender programme coordinator, MRG.Commissioning editor: Helen Kinsella. Report editor: Sophie Richmond. Production coordinator: Kristen Harrison.Minority Rights Group International(MRG) is a non-governmental organization (NGO) working to secure the rights of ethnic,religious and linguistic minorities worldwide, and to promote cooperation and understanding between communities. MRG hasconsultative status with the United Nations Economic and Social Council (ECOSOC), and observer status with the AfricanCommission on Human and Peoples’ Rights. MRG is registered as a charity, no. 282305, and a company limited by guaranteein the UK, no. 1544957.ISBN 978-190458494-0 This briefing is published as a contribution to public understanding. The text does not necessarilyrepresent in every detail the collective view of MRG or its partners. Copies of this study are available online atwww.minorityrights.org. Copies can also be obtained from MRG’s London office.Minority Rights Group International 54 Commercial Street, London E1 6LT, United KingdomTel+44 (0)20 7422 4200Fax+44 (0)20 7422 4201minority.rights@mrgmail.orgWebsitewww.minorityrights.org