Food systems are complex.  Farmers are heterogeneous – their access to information, their cultural connection to other farmers, their relative position of power and independence, can differ enormously, even within limited geographical areas.  They sell into different markets, under different incentives, and at different levels of exposure to production and price risks. Making purposeful changes (in policies and market incentives, for example) in such a complex system is difficult – predicting the reaction of the system, positive or negative, is impossible with conventional means. Unintended consequences abound, and mistakes in design can take years to manifest – in some cases, they can be detected only when it’s too late to fix them.  

The TELLUS project seeks to address this complexity via Complex Human-Environment System Simulation (CHESS). Working with an international group of food system experts at the Food System Economics Commission, and the Food and Land Use Coalition, the project seeks to better understand and develop policy proposals for a) increasing the adoption of regenerative practices among farmers, b) improve resilience of the agricultural system, and c) explore the role of market concentration in facilitating / inhibiting these effects.

The TELLUS model

TELLUS is an Agent-Based Model, constructed in NetLogo by Toby Pilditch. It simulates a population of individual farms, each with their own (heterogeneous) land, endowments, knowledge, and psychology. These farms are placed within both an economic (market) and an ecological context, selecting farming practices that (within the evolving constraints of physical and informational constraints) maximise their particular set of preferences.

The model allows us to explore how these behaviours change over time as farmers adapt to their changing contexts (and how their contexts evolve as a consequence of their actions). TELLUS allows us to look not only at the changes in practice adoption rates among farmers, but also the ecological impact of these decisions, the effect on livelihoods (e.g., farmer exits from the system), and production levels. This allows us to explore the interplay between multiple UN Sustainable Development Goals (e.g., zero-hunger, responsible consumption and production, poverty), and develop policy proposals accordingly. Lastly, through the inclusion of shocks and demand patterns as inputs, various forms of resilience (e.g., recovery rates, robustness) and market concentration impacts can be investigated and incorporated into the valuation of policy proposals. 

​