project staff and beneficiaries raised a number of  points about  targeting which are helpful in

considering FFW as a food distribution mechanism:

Planning and preparedness

 are important:  IFSP E is now inviting communities to submit

project proposals at the same time as resources are requested from GTZ, so that  shelf 

projects  should be ready when the food arrives. In general, the technical and managerial

capacity needed to implement worthwhile projects quickly is one of several limitations on using

FFW in emergency situations (see section 2.2.b.).

  FFW in this particular area is regarded as 

 women's work 

, whereas work paid in cash is

 men's work  (this is not true everywhere, but it is fairly common).  On the one hand, this helps

to ensure that the food aid goes to women: on the other hand, there is a  risk of excessively

increasing their workload.  The IFSP E Annual Report for 1997 expresses concern about  the

social cost of FFW in terms of childcare and women's health, given the high rates of female

participation  [p.56]

Seasonal targeting

 is crucial.  IFSP E is careful to implement its projects when they do not

compete for labor with essential farming activities (i.e. primarily in the dry season, with only

part time work in the farming season).  This, too, is a general point important in the planning of

emergency FFW (again, see section 2.2.b.).

 Initial 

area targeting

   selecting the most vulnerable Sub locations   is regarded as the most

important level of targeting in IFSP E's work,  and was carried out at the beginning of the

project using agricultural and nutritional surveys. Self help groups from communities

elsewhere in the District can also apply for project support for specific activities.

Beneficiary / participant selection

 for the FFW activities is designed to be done partly

through self targeting and partly through community selection.  These strategies are discussed

in more detail in the following paragraphs.

Self targeting

 can be achieved through FFW in general by adjusting either or both of two factors:

the work requirement (the cost element for participants), and the payment rate (benefit element).

In Mwingi, IFSP E's experience is that the work requirement has a limited targeting effect: better 

off people may not want to do the work, but they sometimes register for participation and then

employ poorer people to work on the project for them (so the work is targeted on the poor, but the

food is shared with the better off).  This kind of arrangement may be out of tune with the intentions

of the project designers, but it seems to arise partly from traditional ways of organizing communal

work in this area.  According to local values, everyone should participate in community projects (by

contributing either labor, or money) if they expect to share in the results.  There was some concern

among project staff that trying to target FFW participation on the poor only could undermine this

ethos of communal work, as well as the local 

mwethya

 (mutual assistance) organizations which

are based on kinship and neighborhoods rather than economic status.  It was also noted that in

times of food shortage (which is exactly when FFW is most likely to be implemented) the self 

targeting effect is weaker, because everyone is looking for employment and food.

Setting self targeting 

payment rates

 for FFW is also more difficult in practice than theory suggests.

In principle, the value of the food payment should be at or below the minimum wage which the

participants could earn elsewhere, in order to encourage people with alternative income

opportunities to exclude themselves from FFW and take the better paid employment on the labor

market.   IFSP E's experience with this reflects two common difficulties.  The first is that wage

rates as well as employment opportunities in rural food insecure areas are highly seasonal, and

may also vary greatly from year to year depending on general conditions (in times of general food

shortage, both are likely to be low).   In Mwingi, the casual daily labor rate was estimated by IFSP 

E staff to vary between 60 /  and 100 /  per day.  A GTZ consultant found that actual wage rates in

the more remote areas in June/ July were between 30/  and 60/ , while the FFW ration was

estimated to be worth 60/  to 80/  during the same period [GTZ 1998, Evaluation of IFSP Eastern

A 43