targeting of sales (as noted in Arusha), was not at all clear. It is therefore 

RECOMMENDED that a

study be commissioned on the targeting impacts of SGR commercial operations during

7

food shortages 

(focusing on the questions of who benefited, at what cost, and to what extent the

N

private sector could have fulfilled the same function if there had been no restrictions on the

movement of grain).

1.2.c.  Focus area 1:  Dodoma Region

Dodoma, in the middle of Tanzania and poorly linked by road to surplus food production areas

(see Map 1), was one of the Regions worst affected by successive poor harvests in 1996/97, and

remains vulnerable in 1998/9.  The study team's visit here focused on village level targeting and

distribution, and on issues relating to the choice of NGO or government agencies as implementers

at this level.  Two Districts were visited, Dodoma Rural and Dodoma Urban (the rural area

surrounding the town).

Village Relief Committees were formed throughout the country, in accordance with a national

directive following the March 1997 assessment of food needs.  In the communities visited in

Dodoma,  these committees had been elected as instructed, including both women and men, but

their role and responsibilities were not clear. In particular, their authority in relation to local

government (led by Village Executive Officers) and local party leaders was not defined and was

inevitably weak in practice.  As one committee member in Hombolo (Dodoma Urban District)

described their role, it was limited to verifying the identity of people already on the distribution list.

It is likely that this varied from place to place, but if the Village Relief Committees are to play a

meaningful role in future targeted distributions, 

it is RECOMMENDED that the terms of

8

reference for Village Relief Committees be defined, and support provided in pilot vulnerable

areas (possibly through NGOs where appropriate) to develop their capacity and authority to

make decisions on relief distribution within the community

.  While there seems to be

widespread support (in the Districts, Regions and capital) for the 

idea

 of village committees,  the

reality is unlikely to be very effective without such support.

In the design of EMOP 5889 (the major WFP drought relief operation for 1997/98) a decision was

made to distribute through NGOs rather than local government, due to problems with the targeting

of earlier distributions (see Appendices 2 and 3 for similar comments in Uganda and Kenya).  In

Dodoma,  an important  sidelight was thrown on this issue by the experience of using two different

NGOs in different periods of distribution   one successful (NPA) and one less so in this case

(IIRO). Not all NGOs are equally able to manage food aid. Familiarity with the distribution area,

and established relationships with government and communities, are often rightly identified as key

advantages for NGOs involved in food aid distributions.  However, in the Dodoma case, NPA was

brought in to manage the relief operation and had no previous presence in the area. The key

factors for good targeting in this situation were primarily capacity related:

experience with food aid distributions (hence ability to anticipate and plan for difficulties);

staff capacity (in numbers and skills); and

  material resources (especially transport, to facilitate supervision and monitoring of

distributions).

Informants both in local government and villages (including members of village government and

committees) stressed the benefits of direct supervision of distributions by NPA staff together with

District officials, in order to cross check the selection of needy beneficiaries (whereas previous

distributions had been simply delivered to the village and left for local leaders to distribute

unsupervised).  A general point arising here is that targeting, whether done by government or other

agencies, requires both resources and skills.

The approach to targeting within the villages (stage 3) had already been established, and the local

population registered according to the government's three category guideline (as stated above),

A 11