'It is difficult to escape what is linked to survival': sexual exploitation and food distribution in Burundi

CARE International has been a key partner of the World Food Programme (WFP) since the outbreak of Burundi’s civil war in 1993, distributing emergency food aid to refugees, returnees, internally displaced persons, female-headed households, orphans and other vulnerable people in 16 of Burundi’s 17 provinces. In 2005, CARE distributed over 31,000 tons of food to over 800,000 beneficiaries.

As the security situation in the country has improved, the programme has moved from generalised emergency feeding to semi-regular ‘targeted distributions’. WFP and the government allocate food resources based on agricultural production and food security data (collected on a quarterly basis with the Food and Agriculture Organisation (FAO)). Implementing partners and local government officials are supposed to identify households that meet pre-established vulnerability criteria, and are thus included in the beneficiary lists. According to WFP’s criteria, vulnerable groups include the handicapped, people suffering from incurable diseases, people with caring responsibilities for children, orphan heads of households and people without access to land.

There are several challenges associated with this approach to food distribution. First, given the highly vulnerable state of the Burundian population after 13 years of civil war, and the fact that the vulnerability criteria are fairly imprecise, the number of households that meet the criteria almost always exceeds the resources available. Second, leakage associated with the distribution of humanitarian aid is extremely common: widespread corruption, a lack of transparent, participatory local governance structures and a culture of impunity at all levels mean that village heads, local politicians, merchants and others routinely use their power to have themselves or members of their families inscribed on the lists in place of ‘real’ vulnerable individuals. While community-based targeting techniques could help address much of this problem, WFP has been slow to embrace such new approaches. The fact that the recipients of the targeted distributions are not fixed, and distributions rarely occur in the same village twice, also complicates efforts to create or empower beneficiary committees or other local bodies that could participate in targeting.

Between May and December 2004, CARE carried out a study of the impact of food aid on community power relations and social networks. This study, which echoed anecdotal evidence gathered by teams in the field, confirmed that there were many irregularities associated with the creation of beneficiary lists, and that food distribution was often an opportunity for powerful cliques within the village to affirm or consolidate their power over more vulnerable members of the community. At the same time, CARE was implementing gender and diversity training for all staff and developing a code of conduct which, while being sensitive to the local context, would address issues of sexual harassment and exploitation, both within the mission and vis-à-vis project beneficiaries. This led CARE to question whether sexual harassment and exploitation were among the ways in which power over food aid was being exercised. CARE conducted a study between October 2004 and June 2005 in order to document whether sexual relations were being used as a means to access food aid, to identify the reasons and mechanisms behind such abuse if it was taking place and to develop strategies to reduce the risk to beneficiaries.

Methodology

Sexual harassment and exploitation are obviously very sensitive topics, particularly within the cultural context of Burundi. Indeed, the 2004 study of power relations, which used traditional qualitative research methods, did not produce any direct evidence of such practices. For the research into sexual exploitation, CARE decided to use an experimental approach to answer the question of whether sex was used as a means to obtain food aid. Partnering with a local theatre group called Tubiyage (‘Let’s talk about it’), which has extensive experience in facilitating community discussions on ethnic conflict, sexual violence, HIV/AIDS and other sensitive subjects, the research team used interactive theatre techniques to introduce the subject in focus groups and public forums, and to elicit testimonials from victims.

Special use was made of the ‘invisible theatre’ technique: while beneficiaries were gathered and waiting at a food distribution point, two or three actors mingled with the crowd and posed as members of the general population. A ‘situation’ with a sexual violence theme was created involving the actors and members of the crowd who, not realising there were actors present, intervened to resolve the situation. Afterwards, the actors revealed themselves and facilitated a discussion regarding the community’s response to the situation. The invisible theatre sessions were complemented by 18 focus groups with female beneficiaries, as well as semi-structured interviews with administrators, village sages (bashingantahe), local association leaders, priests and merchants.

Findings

The interactive theatre sessions produced strong reactions from the crowd: murmurs of agreement and giggles. The actors overheard comments such as ‘that happens, that happens; you know how much they dishonour us!’. In the focus group discussions and semi-structured interviews, both victims and perpetrators confirmed that sexual harassment and exploitation were present in the food aid process. Exploitation took place in secret and was never discussed openly, and certainly not during the public validation of beneficiary lists, where irregularities are supposed to be identified. Widows and other single women, either without husbands or without grown-up sons, were found to be particularly vulnerable, as they had no adult males in the household to protect their reputation, and no money to bribe the village heads to include them on the lists.

Fear that they would be excluded from the lists was the main factor which led women to submit to requests for sexual favours:

The chief came to my house and asked if I could share a beer with him. Afterwards, he told me that he wanted me to be a beneficiary but that I had to be available for him. I accepted because I didn’t have any other means and I had already missed the food aid twice.

The chief came to my house with the list of beneficiaries. He showed me all the names that had been erased and tells me that my name will be erased if I don’t have sexual relations with him.

If the chief comes and you refuse, he will definitely erase you from the list.

This fear is well-founded. As one woman noted:

The administrator asked me to have sexual relations with him, but I refused. From that day on I could not benefit from the WFP food aid or from any other services of the municipality.

Perpetrators are generally those who establish the beneficiary lists: this public function gives them a power that they abuse. Women interviewed suggested that, along with demanding sex as a last resort when a women cannot pay a bribe, chiefs use the development of beneficiary lists as an occasion to target the most beautiful women and girls for sexual harassment. Local officials justified requests for money, and sometimes for sex, by saying that chiefs needed to be compensated for the time spent preparing the lists. As one local village head put it: ‘if the woman does not say no, the man will take advantage of the situation’.

The participants in the theatre presentations and focus groups also unanimously confirmed the presence of bribes and other forms of corruption. Current practice is for those who establish the lists to put themselves first, followed by their family members and those who are able to buy beer for the individual in charge of the list. Focus group participants employed a Kirundi proverb to explain that the richest benefit the most from food aid: ‘the breeders of the herd [i.e., the richest] receive that which is meant for the poor’.

Recommendations

Participants in the study suggested procedures to reduce the incidence of sexual harassment and exploitation of food aid beneficiaries. These included:

• always having an employee of WFP or CARE present during the creation of lists to ensure transparency;
• electing mixed committees of beneficiaries, including women, to monitor list creation and food aid distribution;
• involving local associations such as women’s groups, church groups and village development committees in the creation of lists;
• asking secondary school students who can read and write to create the lists;
• ending the involvement of the local administration in the creation of lists; and
• ensuring that list validation is done publicly in every village, with the active participation of women and young people.

These suggestions confirm what CARE knows from other settings: supporting transparent local structures and encouraging the meaningful participation of women will reduce sexual exploitation and other forms of corruption and abuse. The role of the local administration remains a complicated question; while the study identified local officials as the primary drivers of abuse, experience shows that side-stepping them completely can cause significant conflict between participatory structures (such as beneficiary committees) and local authorities, and leads them to actively undermine project activities. In the context of Burundi’s post-conflict transition, where local officials have been elected for the first time, the government and donors such as WFP are insisting on more, not less, collaboration with the local authorities.

The study has proved to be a powerful tool for advocacy with WFP. Since sharing its findings CARE has been allowed to devote more human resources to monitoring the development and public validation of lists, and the agency has been experimenting with new approaches, including separate validations with men and women and involving local partners, such as the Burundian Red Cross and the Catholic Church Diocese Committees, who are helping CARE agents to monitor targeting and list development at the village level.

CARE hopes that involving diverse local actors and promoting multiple mechanisms by which beneficiaries can express themselves can help to reduce sexual exploitation and corruption, and enable food aid to reach the most vulnerable people without becoming a source of power and a tool of abuse.

Nona Zicherman is the Transition Programs Sector Coordinator, CARE Burundi. Her email address is: Nzicherman@care.org.bi. This article is drawn from the study Using Innovative Approaches To Better Understand Sexual Harassment and Exploitation within the Food Distribution Program, CARE International in Burundi, June 2005. Thanks to CARE UK for providing financial support for the study.