Abstract for: Population Health: Microfoundations for macrosimulation

In this paper we want to explore the question of whether it is possible to develop tractable models based around population groups with fidelity similar to models based around the individual. For clarity we will refer to the former as macro models, and the latter as micro models. We will start by presenting a model of disease incidence and mortality in both a micro and macro form. We will show ways in which the macro model can fail to capture the dynamics of the micro model, and techniques to mitigate such failures without introducing undue complexity in the macro model. We show that by selectively using more detailed formulation than is typical in macro models it is possible to get good concordance between micro and macro models using the same individual level parameter values in each. Second, using the relatively robust macro model formulations that support such concordance allows comparison with models based on different underlying micro assumptions.