Targeted interventions which benefit specific groups identified as belonging to the poorest sections of the population are especially relevant in countries with narrowly concentrated pockets of poverty (as opposed to low-income countries with widespread poverty). Experience with anti-poverty projects has shown, however, that benefits intended for the poorest are often captured by the middle-class. Careful targeting is difficult and costly. For food subsidies, especially in the case of urban areas where the potential of diversion is greater, targeting can be achieved by choosing carefully the commodities to be subsidized or by locating distribution outlets in poor neighbourhoods.
Apart from difficulties in targeting, donors have experienced a number of difficulties in undertaking projects which directly address poverty problems at the grass-root level. This has much to do with the administration-intensive nature of projects whose disbursement might be quite small, problems of sustainability, the importance of involving local communities in enhancing the effectiveness of poor people at the local level in representing their own interests, and the possible reticence of recipient governments, as well as the need to take into account local realities. Building a significant technical assistance component into capital projects has been found to help cope with problems of sustainability. Many donors have opted to work through NGOs and other intermediaries; but in that case, further problems of impact and replicability have been found.
Reasons why many aid projects have proved to be ineffective include: (a) lack of commitment to help the poor (resulting in diversion of benefits to the non-poor); (b) macro-economic and sectoral policies (e.g. pricing policies in the agricultural sector and land-use policies in the urban areas) at variance with the objectives of the project; (c) frequent shifts in donors' concerns and policies which have often overshadowed the need for stable long-term strategies, and lack of coordination among donors; (d) neglect of the institutional and managerial aspects of poverty-oriented projects; and (e) inadequate participation of the beneficiaries (local community organizations, organized groups of poor workers, borrowers, women, etc.) and little attention given to socio-cultural and political factors in recipient countries.
In the absence of reliable quantitative data, qualitative information can also be used, at least to identify and localize the poor and target programmes for them. For instance, populations in certain regions, certain ethnic groups, displaced and refugee populations, households headed by women, and such functional groups as smallholder farmers, the landless in rural areas, nomadic pastoralists and small-scale fishermen, are known to be specially vulnerable to poverty.
Although there is no general agreement, areas where projects are considered by some to have been relatively successful are: agricultural research (particularly on crops grown predominantly by the poor) and extension; the introduction of high-yield varieties of food grains in Asia and especially India (the "Green Revolution"); irrigation that benefits many small farmers, especially in Asia; basic infrastructure in rural areas, especially farm-to-market roads in many low-income Sub-Saharan African countries; primary education, basic health care and nutrition; rural public schemes, and relief from natural and man-made disasters.