L’adoption des technologies agricoles : quelles leçons tirer des expérimentations de terrain ?

The Green Revolution, consisting of using High Yielding Variety (HYV) seeds together with high fertilizer doses, has been widely adopted under irrigated conditions, but generally not in rainfed areas that are prone to stresses like drought and flooding. This puzzling lag in technology adoption is holding back the role of agriculture for development in extensive regions of the world such as Sub-Saharan Africa and Eastern India, with high aggregate costs in terms of economic growth and human welfare. Field experiments have been particularly useful in addressing this adoption puzzle. Significant lessons have been learned on the roles of farmer behavior and of mediating factors such as credit, insurance, markets, and policies in constraining adoption. We use experimental field research in Eastern India to show that rainfed agriculture typically suffers from lack of effective supply of suitable HYVs, constraining adoption and resulting in low fertilizer use. Effective supply requires the existence of suitable HYVs (provided by research), the provision of information to farmers about these technologies (provided by extension services and social networks), and their local availability for adoption (provided by private agents in value chains). We consequently argue that solving the adoption puzzle for rainfed areas requires that governments and international donors increase investment in discovery-type agricultural research, support innovations in extension services, and promote the role of private agents in value chains. For donors this implies resolving a collective action problem in the provision of international public goods that remains pervasive. Codes JEL: O13, O30, Q16.