This panel will think critically about mathematical models as tools of governance: how the assumptions and data used in modeling reflect unexamined values, assumptions, discourses and political structures.
Governance decision-making processes, including setting priorities and allocating resources, are increasingly shaped by mathematical models that predict the impact of policies and investments. This panel sets out initial thinking for a critical anthropology of mathematical modeling, exploring modeling as a technology and generator of evidence or knowledge, and examining how specific assumptions and data may be shaped by social and political contexts. While the power and authority invested in modeling continues to grow, there has to date been relatively little exploration of models as abstractions of more complex realities, embedded in algorithms, concealed in code. Potential themes to be explored include some of the following, or others: How are the choices made in designing models unconsciously reflective of what modelers believe is important, visible and measurable? How is time encoded in modeling, and how does the assumed predictive power of modeling shape and influence the future and produce other consequential social outcomes? How do broader economic and political forces, gender or other inequalities, institutional missions, donor agendas, and pressures to demonstrate scalability impact the development of models? How might models also be part of production chains that consume knowledge, shaping housing, credit, recidivism, development aid and epidemiology? Can these processes be rendered more transparent, accountable? What might anthropologists contribute to these discussions regarding the connections among technology, knowledge, and society?