Considerations for gender-responsive mathematical modeling of agriculture and natural resource management

Abstract

Models can play a key role towards improving understanding about critical environmental and development challenges, such as declines in soil fertility due to inadequate management practices, or drops in returns to labor as soil fertility declines. They serve as logical constructs that offer a simplified representation of complex systems to provide insights into a ‘perceived reality’ (McCarl 1984), or a specific part of their inner workings (Gershenfeld 1998). In particular, mathematical models focus on the effects of different components of the ‘target system,’ and describe or predict behaviors using mathematical concepts, theories, and language (ibid 1998). In the fields of agriculture and natural resource management (NRM), decision-making models, land use planning models, statistical relationship-based or process-based models may be used to analyze biophysical and/or socioeconomic phenomena.
The purpose of this brief is to identify considerations for integrating gender and social inclusion considerations in mathematical modeling focused on agriculture and natural resource management. The brief is not a guideline per se, nor is it exhaustive in terms of entry points for gender integration in mathematical modeling. Rather, it is intended to stimulate thinking on ways to engage with gender relations to develop models that can support analysis on innovations that promote equitable and sustainable agriculture and natural resource management.
There are many different ways to design and use models depending on the type of phenomenon to be analyzed and type of questions the model needs to answer. Econometric models can and often do include gender considerations or ‘gender’ as an explanatory variable. In contrast, models that describe only biophysical processes will not include gender per se in the model, but may nonetheless carry gender implications that require consideration (i.e. treating gender relations as an exogenous variable that shapes the biophysical world or that influences the impacts of biophysical change). Integrating gender considerations into models requires analyses that acknowledge the complex, shifting and context-specific nature of gender roles and relations (Doss et al. 2001, Gladwin et al. 2002). Drawing on the scarce literature on gender in modeling, we demonstrate below that gender relations require consideration across key phases of the modeling cycle, including when: 1) conceptualizing the model/framework; 2) collecting data to populate the model; and 3) interpreting model outputs.