This study investigates the characteristics of technology adoption by small-scale farmers, notably the factors influencing the adoption of hybrid maize seed, inorganic fertilizer and machinery technologies. The study also on the basis of socio-economic and institutional factors, identifies the dimensions of small-scale farmers. Data for the study were obtained from a sample survey of 160 households in the Amangwane and Amazizi wards, located in the Okhahlamba magisterial district of KwaZulu-Natal during August 2000. The chief aim of this study is to generate empirical information that can be used to devise programs to encourage small-scale farmers to adopt agricultural technologies. The motivation of the research emanates from the fact that there is limited empirical information as to the actual adoption patterns of agricultural technologies by small-scale farmers. The nature and relative importance of factors associated with technology adoption is time and location specific. The study by using more recent and broader information builds on previous studies in order to complement technology adoption research on small-scale farmers. Understanding what factors influence the adoption of farm technologies and categories or dimensions of small-scale agriculture should provide information on policy options to stimulate technology adoption and improve growth in agricultural productivity. A categorical dependent variable was specified to identify farmers' adoption pattern of hybrid maize seed and fertilizer. Seventy-two farmers were adopters of both hybrid seed and fertilizer, 56 were adopters of either hybrid seed or fertilizer while 32 farmers were non-adopters. The results of binary logistic regression analysis indicate the adoption of hybrid maize seed and fertilizer is positively associated with, in order of importance, larger farms, older household heads, more value of livestock and better access to information sources. An index that indicates farmers' status of adoption of machinery technologies was constructed using a principal component analysis technique. The analysis showed that the adoption of machinery technologies can be represented by the single index which could be used as a dependent variable. A principal component regression analysis was subsequently used to determine factors contributing to the adoption of the machinery technology index. The results indicate that adoption was higher for (1) older and male headed households in general and residents of the Amangwane ward in particular; (2) operators of more arable land, owners of more livestock and earners of more non-farm income; and (3) households with large family labour, and households that made use of extension services and information sources. These results are consistent with hypothesised relationship between technology adoption and the predictors and are supported by previous empirical findings. Priority should be given to policies that alleviate the tenure insecurity problem on arable land and this in turn promotes a land rental market. This would involve an institutional change and legal infrastructural support services. Arable land holding is highly skewed within the communal setting and the state needs to address this equity issue on arable land through redistribution or reform policies. The state needs also to invest in public goods that alleviate the problems of private investors for example by encouraging credit providers or promoting rural financial markets to alleviate liquidity constraints and enhance adoption. Investment in farmer training and education should therefore, be seen as priority if higher adoption rates and an improvement in income are to be achieved. Inadequate and poor extension and information services imply an urgent need for the formation of community and farming associations and for the provision of extension services to groups of farmers. Investment in these areas may reduce the cost of technology transfer programmes. The results of a principal component analysis to identify the dimensions of small-scale farmers in communal areas of KwaZulu indicate that farmers fall into distinct categories. Component 1 is an emerging commercial and a more mechanised household while component 2 is a land-less farm household that is more educated and earns more non-farm income largely from contractor services. Component 3 is a non-farm female headed household that depends on income from land renting and non-farm jobs. Component 4 is a small intensive garden farmer, headed by a relatively educated female who has access to institutional services. Component 5 is relatively less educated, a female-headed and land-poor household that rents land and produces intensively. It is concluded that a single policy measure cannot do justice to the needs of all of the farmers since it would affect different households differently. An integrated and a comprehensive programme is needed that would promote agriculture; facilitate income transfer or safety nets to alleviate poverty and the relief of short-term stress; address the problems of tenure insecurity; overcome the gender inequalities in accessing resources; and restructure institutional supports by providing rural finance, and an extension and legal infrastructure.

Thesis (M.Sc.Agric.)-University of Natal, Pietermaritzburg, 2001.