Crop failure in Dalocha, Ethiopia: ActionAid's participatory emergency response

The national context

Since the early 1990s, having been a major food aid recipient for a decade, the Ethiopian government has sought to decentralise disaster management. This approach aims to link relief work with development objectives, restricting the distribution of free relief items to ‘unproductive’ members of the community. The new policy emphasises the use of the Employment Generation Scheme (food-for-work or cash-for-work programmes). National relief planning is being devolved to local rather than central government structures. The involvement of local communities in planning and implementation of relief interventions is encouraged more than before, but is still rare. This article demonstrates how NGOs can support the participation of local people in relief projects, and asks to what extent this approach is replicable in other situations and by other agencies.

ActionAid’s relief project in Dalocha

Dalocha is situated in the lowlands of the Rift Valley, 180 km south of Addis Ababa. Livelihoods depend mainly on rain-fed mixed farming. Most households experience an annual food gap of 3-4 months. The major food crops are maize and sorghum, with wheat and teff also grown mainly for cash. ActionAid-Ethiopia (AA-E) has worked since 1989 with communities in 34 rural Peasant Associations (PAs) of Dalocha woreda (government district administration, to which PAs report), with a focus on group formation as a basis for development activities.

In 1993, the harvest in Dalocha largely failed, due to excessive meher rains which then curtailed early. With continuing erratic rainfall and pest infestation, 1994 brought widespread food shortages. By June, 63 percent of households were reduced to one meal per day and community reciprocity was breaking down. The severe famine of 1985 had caused long-term damage to livelihoods, through the adoption of coping strategies such as migration and the sale of productive assets. The AA-E response in 1994 aimed to avert further asset erosion, protecting livelihoods as well as alleviating hunger, by providing affordable food as well as supporting farm production. Since it also aimed to reinforce, rather than undermine, ongoing development activities, it was crucial to encourage the participation of the beneficiary communities at each stage of the project.

Problem identification and community consultation

Community members were involved firstly in the information-gathering process. In order to assess the extent and severity of the problem, field staff divided into four teams, each covering one ‘block’ of the project area. As well as sample household surveys, a total of 168 focus group discussions took place during one week with a selection of local people. These individuals represented various groups, including: leaders from local iddirs (important social institutions which organise funerals), elders, religious leaders, PA executives, groups of women from poor households, and savings and credit groups. As a preliminary targeting exercise, each group listed poor households affected by acute food shortage. The lists were then cross-checked with existing data gathered through Participatory Rural Appraisal (PRA). These activities ensured the early involvement of different social groups in identifying needs and possible solutions. They also allowed comprehensive and balanced data to be collected very quickly. The continued involvement of community members in the design of the response ensured that the information-gathering was not purely extractive, but the beginning of a participatory process.

Designing the response

An all-day workshop was held with the group leaders from AA-E’s 102 community development groups, to verify the situation from the community’s perspective and consider response options. Discussion in groups was followed by a plenary session where each group presented their findings.

The farmers confirmed widespread hunger and related health problems; they described the increased cost of food grains and falling livestock prices, people trying to rent out farmland and eating weeds to eke out food supply. Lack of seed was also noted as a problem. They then suggested the following solutions:

• Soft loan for food, at zero interest, to be repaid over 2-3 years
• Wheat seed and fertiliser on credit to those who had lost their maize crop
• More/cheaper drugs available through health posts
• Blankets on credit
• Training/advice on savings and grain storage.

The participants suggested that assistance for group members should be channelled through the groups, and that for non-group members it should be organised through a respected local institution such as the iddir.

The value of this workshop lay not only in providing a clear outline of the problem and perceived needs from a community perspective, but also in allowing the community ‘ownership’ of the relief response. In discussions with the community today, especially those who were group members, they clearly believe the design and decisions were theirs.

“We spoke to the field workers and told them that unless we eat, it is useless to talk about development. There was cold weather and no food, so we asked for blankets and grain. We discussed it amongst ourselves and the leader went to a meeting.”
Afran group, Koro Gale PA

The community’s request for credit is interesting. AA-E shares the government’s concern that free relief can encourage ‘dependency syndrome’ and undermine development objectives. During discussions with community groups, AA-E field workers had already aired this dilemma with the community. At the beginning of the workshop it was again suggested that solutions should not be based on free handouts, and it seems likely that this informed the decision of the farmers to ask for credit. The communities still defend the credit approach as being appropriate, although in some cases there were repayment difficulties. (More recently, AA-E has chosen to use cash-for-work as an emergency response rather than credit, due to over-burdening with credit.)

Targeting – using local knowledge and community institutions

There were some differences in the percentages of people in need as suggested by the workshop groups. With variations in land holdings and family size across the area this was logical. For planning purposes, staff agreed to assume targeting of 65 percent of the overall population. This took into account farmers’ recommendations but also pre-existing PRA-generated data.

To target community members not already involved in AA-E groups, new committees, known as Disaster Prevention Committees (DPCs), were set up in each of the 34 PAs. Each comprised 7 members selected from respected local institutions: iddir, elders and religious leaders and one PA executive. Their primary responsibilities were selecting beneficiaries, organising grain distribution and collecting credit repayments. Formal terms of reference for the DPCs were drawn up, and workshops held to ensure clarity of roles and procedures.

The DPCs were new structures with a clear purpose, built on existing respected institutions. They created a balance, avoiding domination by one group or another, thus minimising bias. The community today acknowledge the importance of the role played by the DPCs.

“Without the DPCs it would have been impossible to select and distribute. It was well organised. These people know us, especially the iddir, and the PA has influence and knowledge. It helped to make the process fair.”
Elder (non-DPC), Inkat Lola PA

Targeting the poorest?

A question which AA-E had to address was whether or not to include the very poorest members of the community: those who are ‘unproductive’ and rely on others for their livelihood. In AA-E, it was decided that to support the very old or disabled would require a very different type of intervention: not one based on credit and aiming to support farm production. It was also feared that direct support to such individuals would undermine traditional community support mechanisms, and might transfer dependency to AA-E. However, another ActionAid programme, in north-east Ghana, designed an emergency response in 1994 consisting of grants for the poorest as well as credit for the creditworthy, based on suggestions by the community. As agencies seek to respond to emergencies in a more ‘developmental’ way, this issue may become a more common dilemma.

Involving all the community?

Although many different sections of the community were involved in needs assessment, targeting and project design, it was mainly group leaders and DPC members who took responsibility for implementation, management and monitoring. They organised distribution, and took difficult decisions on borderline cases during beneficiary selection and the final allocations of grain per household in each PA.

Whilst there is a need to gain the cooperation of local decision-makers in this way, it can be more difficult to engage some less empowered community members. For instance, in Dalocha, the respected members of community institutions are almost invariably men; thus few women were actively involved beyond the stage of information-gathering. At the time, only four groups out of the 102 were female-headed (although many group members were women). Whilst many have stated that their views and needs were adequately represented, it seems clear that to play a more influential role in an emergency project, they must already be doing so in the daily life of the community. This can be encouraged in development programmes.

A participatory approach: achievements and challenges

Community participation in the Dalocha project produced a swift, locally relevant response, with clear agency/beneficiary communications and effective, transparent targeting. It highlighted the utility of working through multiple community structures, and contributed to longer term objectives such as community empowerment. All these achievements, however, were supported greatly by the prior existence of AA-E’s programme in the area, established relationships with the communities, and a large, committed team of local field staff.

To what extent is this approach replicable by government agencies or local NGOs with fewer resources? Could it still be useful in more unstable communities, in conflict-related emergencies or where an agency has no prior local involvement? Consultation with local people in needs assessment exercises is already widely practised. In most situations, information-gathering can be made less extractive by allowing the community to define response options, encouraging community ‘ownership’ and ensuring a more appropriate response. However, the community’s perception of the agency may influence their suggestions. It should also be possible to listen to all groups, not only local leaders. The extent to which local knowledge can be used for targeting will vary, but would be more open to bias and inaccuracy in less cohesive communities, or where the agency/community relationship is new. Devolution of decision-making, implementation and monitoring can be perceived as positive development but also involves a loss of control. It can be challenging to develop joint monitoring and reporting systems which are workable but also geared deliver information required by management and donors. In ActionAid’s experience, the benefits outweigh the difficulties, but this might be more problematic in a local NGO with fewer staff.

Different facets of the participatory approach will be appropriate in different situations. More projects experimenting with this approach need to be studied in order to determine all the possibilities. We must also acknowledge that for community participation in relief to work at all, agencies and governments must truly want to recognise beneficiaries’ resources and capacities as well as their needs. This article demonstrates one such attempt and we encourage others to report on their own experiences.