Behaviour under uncertainty, heterogeneous agents and herding
While current policy-focused macroeconomic models account for the fact the decisions of households, firms and financial intermediaries are critically influenced by uncertainty about idiosyncratic and aggregate shocks, the global financial crisis has made clear that homooeconomicus, representative-agent-type modelling assumptions cannot adequately capture shifts between behaviour in normal times versus boom and crisis periods. Furthermore, these models failed to provide warning of the magnitude and persistence of the adverse impact of the global financial crisis on the real economy. To make progress on these counts, it is crucial to incorporate the heterogeneity of households and firms, in regards to the shocks they are exposed to, how they respond to these shocks, and how their behaviour adapts to their relative effectiveness in responding to past shocks. To build models featuring micro-behavioural realism in decision making, an ideal starting place is to use laboratory experiments with human subjects, establishing and testing behavioural assumptions. A key advantage of such laboratory experiments is that the experimenter can in fully controlled form study the feedback mechanism between individual behaviour and aggregate outcomes. This includes, of course, giving particular scrutiny to issues such as the emergence of herding behaviour and boom-bust cycles, crisis dynamics, and how these are affected by regulatory schemes.