Review of methods for modelling systems evolution
MetadataShow full item record
Nicholson, C.F. ILRI, Nairobi (Kenya), Cornell University, Ithaca, New York (USA). 2007. Review of methods for modelling systems evolution. ILRI Targeting and Innovation Discussion Paper. no. 3. 123p. Nairobi (Kenya): ILRI.
Permanent link to this item: http://hdl.handle.net/10568/359
The objective of this review is to describe and evaluate conceptual, descriptive and mathematical modelling approaches for the evaluation of systems evolution, emphasizing the potential to include assessment of policy and technology impacts in systems with livestock. To achieve this objective, this document reviews basic concepts in modelling and prediction; provides an evaluative review of the characteristics of two models designed to predict future global food production and consumption over the next 15 to 25 years; describes selected conceptual frameworks useful for thinking about systems evolution; and compares eight modelling approaches - both descriptive and quantitative - that may be used to provide insights about systems evolution and its relationship to technology and policy. The literature reviewed herein focuses on methods that allow prediction of future behaviour of indicators over a time horizon relevant to help guide technology and policy development. This review also focuses on approaches applied at a scale between the farm and global levels over the time scale of 5 to 15 years. These also could be classified as modelling approaches applicable to the analysis of 'livelihood systems' or 'production systems', in which there is some reasonable degree of homogeneity in terms of the resources, agricultural and non-agricultural activities, and objectives of the agents in the system. Given that no review of modelling approaches can be exhaustive or comprehensive, examples are selected to illustrate the application of the approaches to agricultural livelihood systems modelling. To keep this task manageable, eight categories of methods have been selected and examples of their application provide an idea of the diversity of their application areas and variants. These eight categories are: Statistical analyses (other than time series); Time series analyses (distinct from analysis of time series data); Dynamic optimization; Dynamic computable general equilibrium models; Dynamic partial equilibrium models; Differential equations-based methods (e.g. system dynamics); and Agent-based models; Other simulation approaches (not associated with a particular approach).