four hours a day on FFW, which should leave them time for working on their farms. The QAP FFW project proposal sees FFW as a substitute for ganyu, at least in part, and appears to propose that FFW be undertaken in the rainy season, as at other times of the year. However, the issue of FFW squeezing out work on fields is raised as an area for monitoring. There are significant differences in approach between MASAF and WFP in their programs. WFP's FFW projects are seen as providing alternative and more attractive sources of income, in particular to ganyu. People do not have to travel long distances to find work and can continue to work on their farms. MASAF is more a traditional PWP. where it is intended in the main phase of the project that participant arPe pai, a r2ch wg equivalent to the minimum rural wage and are expected to do a full day's wnrk. The ability to target is an area where PWPs are often said to have advantages. Certainly it is easy to target PWPs geographically and Malawi has an information system in FEWS which can provided up to date data on which to base this. It is also claimed that PWPs have a self- targeting element This is true in theory. Better-off people are unlikely to undertake hard physical labour for a relatively low wage However, the level of poverty is so high in Malawi and the availability of off-farm employment so relatively low, that even with careful EPA targeting, the extent of self-targeting is highly dependent on the wage offered. MASAF offers cash or a mixture of cash and food equivalent to the minimum rural wage. Even at this low wage, there is far more demand for participation than employment available Workers are employed on a first come first served basis. WFP is proposing a cash/food combination equivalent to 130 percent - 160 percent10 of the minimum rural wage. The demand for these is likely to be even greater. Given that it is politically unacceptable to implement a PWP with a wage below the minimum rural wage, it is likely that any PWP will attract many more participants than can be employed. participants are still likely to be poor, but they may not be the poorest. This means that some form of additional informal targeting or rationing must take place. An evaluation of the emergency WFP FFW projects has shown that different projects have adopted different approaches. Some have operated on a first come first served basis, in others physical productivity criteria have been adopted and in others there has been some attempt to include the most needy households (Dil, 1996). The problem of community level targeting should be recognized and tackled head-on (see Annex F). In their proposed QAP, WFP includes a second type of FFW project, which provides incentives to workers on community based self-help projects, where the community itself is the prime beneficiary of the project's output. It is proposed that the incentive offered (WFP do not call this a wage) would be around 20 percent below the minimum rural wage. This should be 10 It is difficult to state categorically the value of the WFP wage, as the value of the food component varies according to the time of year and whether the food is an alternative to ADMARC or free market purchases. G2