Abstract

Spatial dynamic microsimulations are a powerful tool for analyzing complex dynamics in populations. They consider regional and individual dependencies by modeling socioeconomic systems at the micro-level like persons or households. The basic idea goes back to Orcutt (1957) who criticized that it is impossible to aggregate individuals' heterogeneity and the complexity of mutual influences on a higher level.

For regionalized simulations, survey data often do not suffice as a base population due to small regional sample sizes and lack of geographical information. Furthermore, not all desired variables may be jointly available in a single survey. This necessitates the creation of a synthetic micro data set or the expansion of an existing population using statistical methods.
Additionally, statistical methods are required to model transitions between states and occurrences of events, like births, the formation of partnerships, or regional mobility, using model-based predictions and transition probabilities.

Challenges and opportunities of microsimulations are presented exemplary for the regionalized dynamic MikroSim model developed in the DFG research unit FOR2559. The MikroSim base population thereby contains a geographically allocated synthetic representation of the approximately 82 million German inhabitants and 40 million households generated based on anonymized distributions of the German register and micro-level survey data. This allows the small-scale analysis of social developments under various scenarios. The utility of microsimulations for decision-making is illustrated in different policy-relevant areas such as demographic change, within-town planning of primary schools, and the impact of climate change.