ABSTRACT
Most research systems in developing countries were organised to serve commercial farmers operating in more favourable and homogeneous agroecological conditions than those in resource-poor farming contexts. Identifying mechanisms for making public institutions more accountable, transparent and subject to the demands of broad-based participation in the political process, is justified not only on the grounds that development loses much of its social meaning if it is not accompanied by democracy, enhanced equity and popular empowerment. The functioning of the agricultural technology system is subject to a series of contextual factors: political, economic and agroclimatic. NGOs can interact with government in both collaborative and conflictive modes. In much of their history in Latin America, this interaction has been conflictive both in the general political arena and in ATD in particular.
