A trait-based model of the potential demand for a genetically engineered food crop in a developing economy

We predict the potential demand of smallholder farmers for genetically transformed varieties of a food crop, the cooking banana of the East African highlands, Farmer demand for planting material is derived in an agricultural household model that accounts for variety traits and missing markets, The demand for candidate host varieties is predicted using a Zero-Inflated Poisson (ZIP) regression system. The fitted model is used to illustrate the sensitivity of farmer demand for improved planting material to (a) investments in research and development, represented by the effectiveness of gene insertion and expression, and (b) other public investments in education, extension, and market infrastructure that support diffusion, By comparing the characteristics of agricultural households we demonstrate that the choice of host variety can have social consequences, favoring one rural population compared with another, Clients for transgenic banana planting material are likely to be poorer, subsistence-oriented farmers in areas greatly affected by biotic constraints, A model of this type might be useful in assessing the investments needed to support the systematic dissemination of improved planting material. The approach can be generalized to other crop biotechnologies for smallholder farming systems, particularly in developing economies.