51

of the economy. However, this induction is not automatic. Induction

requires institutions, of which an extension system is one, to convey

specific technological needs of various producers to the researchers.1

If the extension system does not provide feedback about the needs of

small, subsistence farmers, it is unlikely that appropriate technologies

will be induced for them. In the absence of such pressures, the innovations

that are induced are more likely to be suited to the needs of large

aggressive farmers, who can influence the research system directly by

going to a university to talk with scientists or indirectly, through a

Ministry of Agriculture to which they have access.2 The farmers with

cash crops crucial for foreign exchange earnings (rubber in Malaysia,

cocoa in Ghana, coffee, sugar and coconuts in the other countries) seem to

have little problem conveying their needs to research institutes. But

the perspectives and needs of the poor, limited-resource, subsistence

farmer, unable to purchase machinery and chemicals, rarely inform the

work of research units. Thus, any extension system, and especially one

designed to serve the needs of rural poor, needs to emphasize feedback,

particularly from small farmers.
 The bias towards overemphasizing communication to the farmer,

instead of from the farmer, has several sources. First, it reflects

the structure of control within the extension system. If the

extension system is controlled and financed by farmers, the

agents usually are responsive to interest and needs of farmers,



 1Alain de Janvry, "Social Structure and Biased Technical Change in
Argentine Agriculture," in Hans Binswanger and Vernon Ruttan, Induced
Innovations (Baltimore: Johns Hopkins University Press, 1978), p. 311.

 2Rene Benalcazar R., "New Techniques, Agricultural Extension Services."