Supply responses of cotton farmers in Ghana : case study of northwestern Ghana

This report describes a socioeconomic survey of farmers in the Wa area of northwestern Ghana and applies survey data to an economic model of factors affecting farmers" decisions to grow cotton. Against overviews of the history, importance, and problems of cotton production in Ghana, the authors present the findings of their survey of Ghana"s foremost cotton growing area. Credit is the major bottleneck. Although Ghana"s Cotton Development Board (CDB) provides free seeds, fertilizer, and insecticides, credit is needed for land preparation, weeding, harvesting, and hiring labor. Thus, up to 85% of farmers rely on personal savings to finance their farm operations; the remainder borrow from friends, family, and moneylenders. Risk aversion is the most salient characteristic of farmers" decisionmaking. Their desire for security leads them to diversify production and ration capital. The authors recommend that the CDB integrate credit and marketing services and expand extension efforts. Because of endemic farm labor shortages, rural development programs and wage policies should be reviewed. Of the variables employed in the authors" supply-response model, the prices of cotton and of guinea corn (a competing commodity) and a trend variable representing changes in extension services, technology, and other influences had the greatest effects on changes in cotton acreage and output. Thus, increasing cotton prices relative to those of competing commodities will elicit greater cotton production, other factors being unchanged. The CDB"s guaranteed minimum cotton price is a step in the right direction but would be more effective if periodically adjusted relative to the prices of competing commodities. The significance of the trend variable shows that government efforts to increase cotton production by providing free inputs have been very successful and should be intensified and expanded. Because commercial cotton production was only revived in Ghana in 1968, time series data used to test the model were limited and it is possible that not all significant variables were identified. A similar study should be made at a future date. Included in the report are a map, 11 tables, and an 11-item bibliography (1928-77).