Community Managed Targeting – Tanzania

Post-script

by Jeremy Shoham, ENN

DFID commissioned an evaluation of the community-managed targeting programme in Singida and Dodoma regions of Central Tanzania in May 1999. As a member of the evaluation team, I can confirm the very positive impression of the programme portrayed in the above article. All stakeholders that I interviewed, e.g. beneficiaries, village relief committee members and staff at different levels of government, were satisfied with the programme. They believed that the system did target the neediest households and that it was equitable, transparent and accountable. The system seemed to work even when only small amounts of food were available, for example, in Iramba district where limited food availability meant that only the poorest 20% of households could be targeted. The system compared very favourably with the one implemented the year before which most people were dissatisfied with. The previous system operated through village government structures and was not at all transparent. The result was that certain groups were favoured above others and that more or less everyone was given a ration so that the per capita receipts of food aid were very small and not that useful to those with genuine large food deficits.

There also appear to have been some unintended benefits of this community-managed programme. First, women became more prominent in public life at village level and were seen by the men to have done an excellent job. Many male members of the community stated that they would vote for some of these women in subsequent village government elections. Up until now women have always been in a minority on village government committees. Second, regional governments of both Singida and Dodoma were hoping to retain these village relief committees for food security work during non-emergency periods, i.e. following up with vulnerable households to determine how the community could support these families.

While not wishing to minimise the success of this programme, there were of course some difficulties:

1. as stated in the article, the limited availability of food aid determined that village relief committees (VRCs) were directed to register only 60% of households. In many villages the perception was that up to 80% of households fitted the selection criteria. The exclusion of some families therefore placed considerable strain on VRCs and in a few isolated cases this led to insecurity at distribution sites;
2. population figures for each village were mainly taken from district level data. These were often wildly inaccurate (usually under-estimates) so that amounts of food allocated to villages were too small. VRCs then either reduced the per capita ration or under-registered, e.g. only registered some members of a household;
3. in a minority of cases, particularly in Dodoma region, under-registration was carried out in a discriminatory manner. For example, households of the same size would have different numbers registered by the committee, i.e. those favoured by the committee would tend to have more registered. In Singida region under-registration was systematic and therefore fairer, e.g. if there was a shortage of food, infants in households would not be registered;
4. Members of the VRC would spend between 7-14 days in their activities. This undoubtedly detracted from their ability to carry out their own work. Many VRC members wanted some form of community payment, e.g. child care, cooked meals. Another form of payment suggested was that the committee should be allowed to keep the food sacks which have a considerable monetary value.

The success of this community managed targeting contrasts with other recent experiences of community targeting, e.g. amongst pastoralists in Turkana and recently in South Sudan, (see Field Exchange 6).

There may be a number of factors specific to the Tanzanian context which predisposed to the success of this type of community managed targeting system. These are as follows:

• a well-established structure of village government and tradition of village level public meetings and decision making
• the participatory approach is consistent with established national policy and initiatives already taken in the region and strong high level government support at central and regional levels deflecting some local political resistance
• peace and stability in an area of small cohesive village units (3-5,000) and an absence of any serious inter-communal division
• no excessive stress - the intervention was preventive, i.e. to preserve livelihoods, and not a response to an acute crisis
• availability of a pool of government extension staff and some NGO personnel experienced in working at village level often with participatory approaches