In many southern African countries the role of national agricultural extension systems over the last two to three decades has been mainly to promote the hybrid maize and fertiliser packages of the Green Revolution. The assumption has always been that this would prove the saving of southern African smallholder agriculture and food security. Over the last eight or nine years another revolution has, however, swept the region: that of economic structural adjustment and reform programmes. Inefficient parastatal institutions have been a particular target for economic liberalisation programmes, and this has led to the demise of many former state-supported smallholder agricultural systems. Zambia is a case in point. A series of droughts in the early 1990s, and the disappearance of parastatals which had been supplying subsidised hybrid maize and fertiliser inputs to more remote parts of the country, have left large numbers of smallholder farmers facing a significant crisis. They are being forced to adapt their farming systems. The national extension system has found itself unable to meet this relatively sudden but huge demand. This had led to growing self-inquiry about how the institution might reform itself and become relevant in an era in which it is increasingly important for institutions to demonstrate effective performance if they are to maintain financial support.

This paper documents the start-up of an NGO project, the Livingstone Food Security Project, run by CARE in south east Zambia. The project, which pilots some features of an alternative extension model, is being carried out in collaboration with the Ministry of Agriculture, Food and Fisheries. In the first three seasons of the project, a series of participatory livelihood and needs assessment exercises have been conducted, village management committees (VMCs) have been established, and a seed loan scheme has been implemented through these institutions. These interventions have helped the satisfy the priority requirements of communities hit by successive droughts. Working with the VMCs and farmer extension facilitators, the project has achieved extensive coverage of farmers in its second and third seasons with small numbers of field staff. Collaboration with the government extension service is also increasing as the project provides training in participatory assessment approaches and community institution-building to field staff from districts outside the immediate project area. Technically, the project's broader farming systems and resource management challenges are just beginning in helping smallholder farmers develop more resilient production systems. However, with the role of the private sector in agricultural extension also increasing, the project is helping to demonstrate a partnership model of extension which may be more appropriate for the future, both in Zambia and elsewhere.