Dynamic Microsimulation Modelling: A Survey and Critical Assessment IV

Abstract

The sub-field of dynamic microsimulation modelling, the inter-temporal simulation of micro units over time, was one of the original areas of microsimulation. A number of surveys have documented progress in the field at approximately 7 year intervals over time. Reviewing developments over the decade 2013-2023, we find that progress in the field is accelerating, as nearly half the papers published in the field have been published since the last survey article. It is timely therefore to provide a more recent survey and critical assessment of the direction of the field. This paper reviews the evolution of dominant methodological choices. However given the increased scale of the sub-field, we make use of bibliometric tools to assist in the analysis of the field. There has been a clear change in direction over the past 8 years. Previous research could be classified in two clusters, largely divided into closed, aligned, discrete time and open, non-aligned continuous time frameworks, both with family units of analysis. However, the dominant current direction in the field has been the introduction of more parsimonious model particularly in the health sphere. These papers typically use an individual unit of analysis, are non-aligned and have simpler behavioural and simulation structures, with a greater focus on the particular health policy application.