Parsimonious Multi-dimensional Impact Assessment

This article develops the conceptual and empirical foundations for a parsimonious, generic modeling approach to multi-dimensional (i.e., economic, environmental and social) impact assessments of agricultural technologies and environmental change. Joint distributions between technology adoption and outcome variables are characterized, and used to analyze the selection effects of adoption on a general class of impact indicators. The approach is implemented with a generic model that can be parameterized with low-order moments of outcome variables. A case study of adoption of a high-yielding maize variety in Kenya illustrates the model's use and confirms theoretical results.