Intending the Unintended: the Act of Building Agent-based Models as a Regular Source of Knowledge Generation

Poverty is a complex issue that is rarely conducive to analysis in laboratory or field experiments. Effective interventions that aim to decrease or eliminate poverty require an understanding of the intricate web of associated social issues. The need for this increased comprehension necessitates the use of alternative robust means of analysis: one such being agent-based modelling. The strengths of agent-based modelling to disaggregate complex social behaviours and understand them are well known. However, while people have explored how the modelling process can prove to be fruitful, the usually unintended insight gained and the knowledge engendered during the model design process goes largely unnoticed. In this paper, we aspire to show precisely how the model building process is critical in leading to unintended knowledge generation for modellers by drawing from three US based examples where agent-based modelling was used to aid research into the effects of interventions that address poverty and human development through programs and issues facing low-income families. With these examples, we illustrate some of the means to harness new knowledge generated. In our discussion, we also highlight the advantageous nature of agent-based model design as an independent source of knowledge generation.