Progress On the Economics and Adoption of Conservation Agriculture by Smallholders.

The Food and Agriculture Organization (FAO) of the United Nations and other organizations have been promoting conservation agriculture (CA) for a period of time now.  While the associated practices have long been employed by farmers they are increasingly being studied by social scientists seeking to understand their adoption and non-adoption. This paper reviews work on this topic carried out by the author and others over the last decade or so and then considers the prospects for new insights, based on the application of new statistical techniques in recent years. Key issues associated with the economics and adoption of conservation agriculture by smallholder farmers are highlighted initially. For example, what are the fundamental economic arguments for encouraging CA? How do the perceived economic advantages influence the viability of soil and water conservation (SWC) at the project and program level? Finally, what are the implications for adoption of CA? While an earlier synthesis of adoption research by the author revealed some underlying patterns of influence, the primary finding was that there are few universal variables that regularly explain the adoption of conservation agriculture. Recent progress in applying more sophisticated statistical methods to the study of adoption is reviewed using several diverse case studies. For example, duration analysis has been introduced to the study of on-farm technology analysis, allowing researchers to consider jointly the timing of adoption together with those factors influencing whether farmers adopt or not. Nested regression models allow the researcher to analyse the intensity of adoption (area affected) in addition to the choice to adopt at all. Finally, the use of discrete choice models promises further advances. As a form of trade off analysis, these approaches (known as ‘choice experiments’) present alternative scenarios to farmers comprising a set of attributes that have varying mixes of these attributes. Thus, it is possible to present complex technology and program options in a single bundle and to model the factors influencing the choices made by respondents. The potential insights that can arise from the use of these newer techniques are discussed.