Assessing adoption potential in a risky environment: The case of perennial pigeonpea

Perennial crops offer the opportunity to harvest from the same plant many times over several years while reducing labor and seed costs, reducing emissions and increasing biomass input into the soil. We use system dynamics modeling to combine data from field experiments, crop modeling and choice experiments to explore the potential for adoption and diffusion of a sustainable agriculture technology in a risky environment with high variability in annual rainfall: the perennial management of pigeonpea in maize-based systems of Malawi. Production estimates from a crop model for the annual intercrop system and data from field experiments on ratooning for the perennial system provided the information to create a stochastic production model. Data from choice experiments posed by a farmer survey conducted in three Malawi districts provide the information for parameters on farmers’ preferences for the attributes of the perennial system. The perennial pigeonpea technology appeared clearly superior in scenarios where average values for maize yield and pigeonpea biomass production were held constant. Adoption was fastest in scenarios where relatively dry growing seasons showcased the benefits of the perennial system, suggesting that perennial management may be appropriate in marginal locations. The potential for adoption was reduced greatly when stochasticity in yields and seasons combine with significant social pressure to conform. The mechanism for this is that low yields suppress adoption and increase disadoption due to the dynamics of trust in the technology. This finding is not unique to perennial pigeonpea, but suggests that a critical factor in explaining low adoption rates of any new agricultural technology is the stochasticity in a technology’s performance. Understanding how that stochasticity interacts with the social dynamics of learning skills and communicating trust is a critical feature for the successful deployment of sustainable agricultural technologies, and a novel finding of our study.