Open Source Tools for 'Heterogeneous Agent' Modeling

`Representative Agent' models in economics assume that differences across people (say, between poor and rich) do not matter for broad economic outcomes. Such models performed poorly in the Great Recession; but progress in `heterogeneous agent' modeling has been held back because the tools for working with such models are primitive. The Econ-ARK/HARK project provides a Python-based set of tools that make it much easier to get started, and to make rapid progress, in handling the computational challenges of taking account of the differences between agents (in preferences, in incomes, in beliefs) that matter for overall economic outcomes.

 Speaker: Christopher Carroll, Johns Hopkins University
 Speaker: Matthew White, University of Delaware