Abstract

Joint seminar LISER – LSF, University of Luxembourg

A growing literature in macroeconomics and finance argues that the cross-sectional distribution of heterogeneous agents is decisive for aggregate economic outcomes, yet empirically substantiated information on the nature, origins, and temporal stability of behavioral heterogeneity is unavailable. Applying structural estimation to laboratory panel data from an intertemporal choice task, I obtain a parsimonious representation of behavioral heterogeneity and provide three insights: First, a small set of both rational and rule of thumb types explains 92% of all observed decisions. Second, there is evidence for type stability, i.e. many subjects’ types remain unchanged over several weeks. Third, time pressure affects the type distribution and halves the fraction of rational agents. Hence, this paper provides a microfoundation of the assumption of temporally stable heterogeneous agents, and it shows that generic features of the decision environment can fundamentally affect aggregate outcomes by changing the cross-sectional distribution of heterogeneous agents.