Food-for-Work (FFW) was conceived as both a short-run assistance program for meeting basic food needs of low income households, and as a long-run developmental tool for building infrastructure and for providing income to ease capital constraints on farm production. However, it was feared that FFW might divert labor from own-farm production and reduce the level of locally produced food crops. The purpose of this dissertation was to empirically examine these hypotheses in the Ewalel and Marigat locations of Baringo District, Rift Valley Province, Kenya.

A househoId-firm model that integrated both production and consumption concerns of FFW was developed. The model was block recursive. First, production decisions were made by maximizing net returns (net income) subject to production constraints. This output (income) was then substituted into the budget constraint, and household utility was maximized subject to this budget constraint and to a total time constraint.

The data used in the study was drawn from a representative sample of 300 households were randomly selected in Marigat-Ewalel locations. Of these, 100 were found to be participants in the FFW Project supported by the UN/FAO World Food Program. Food items provided to the program in the study area are maize, beans, and vegetable oil.

A two-year linear programming model was developed for the production segment of the model. ln this model, three crops under two technologies and two types of livestock were used. The household consumption component of the model was specified econometrically using systems of demand equations, the Almost ldeal Demand System. Seven commodities including FFW items, five foods, non-food and leisure, were used in the system. The analysis was conducted for both participant and non-participant households to compare levels of production activities, employment, income, and consumption patterns with and without the FFW program.

The production component of the analysis revealed that the following results were associated with FFW in the study area: (a) augments own-farm output by contributing to the minimum nutrient 1 requirement, (b) eases the capital-constraint by the second year of participation, (c) increases the marketable surplus from both own-crop and livestock production, (d) increases hired labor in farm production, (e) causes a shift from maize to millet production, and (f) increases savings. As a result, the net income for the representative farm households with FFW is 52% higher than those without FFW; and participation in the FFW program declines by 11% from year 1 to year 2. Thus, disincentive effects on own-farm employment and output were not found in this study. In fact, according to the model used, the FFW Program could be expanded by either increasing the monthly participation hours or the number of participants without resulting in any production disincentive.

The results of the entire household-firm model, which reveals the changes in consumption resulting from participation in FFW and changes in income, were derived in elasticity form. Most of the benefits to the representative participant households, as compared to non-participants, take the form of increased consumption of food items. Thus, the primary effects of FFW are to insure participants increased consumption and saving without creating disincentives to either own-farming or to local agricultural production.