Modelling for policy: It’s the sociality, stupid! – Gert Jan Hofstede

Policies for public issues should be both desirable and feasible. The role of agent-based models is to investigate whether a certain scenario is feasible. The models can ‘grow’ the scenario based on policy inputs and assumptions about system properties. They can be used interactively with stakeholders, allowing to obtain a sense of system behaviour and nonlinearities.

Human social behaviour is a major challenge for modelling in ABM. Should we model behaviours of individuals directly, such as movement through a city? Should we base ourselves on big data? Or will we create more relevant models if we model individual-level motives for behaviour, assuming relationships between these motives and behaviours, so that the model can compute the behaviours? What kind of motives, anyway? Or can we even generalise, use society-level constructs, and assume that all individuals share them?

In this lecture I will argue that finding ways to capture human sociality is a major assignment for ABM. We need to get this right in order to create relevant, reusable models.

About Gert Jan Hofstede

Gert Jan Hofstede (1956) is professor of Artificial Sociality at Wageningen University in the Netherlands. He works on agent-based models of sociality, often in the context of socio-technical or socio-ecological systems. Gert Jan co-authored some famous books on national culture with his father Geert, and published widely on simulation games. His Silico Centre Wageningen runs summer schools on agent-based modelling. Gert Jan’s background as a population- and behavioural biologist is visible in his foundational way of looking at human social behaviour, using generic theory.