Changes in the Qualification Profile of Germany's Population: The Oncoming 30 Years

This project develops a simulation model which allows for predictions of the supply side of the labour market in 2050. Endogenous simulation is at the core of this project, external influences will not be implemented within the model, but can be added as constants or exogenous time series. Predictions are thus under the control of the user and allows for testing political interventions under controlled circumstances.

The fact that our simulation is to cover a rather long period makes it necessary to program new modules within a new microsimulation tool (but developing this tool — which is more or less complete now — is not part of the project). The implementation will also cover migration effects to enhance the predictive capacity of the model.

Dynamic microsimulation uses the units of an empirical data set to predict their attributes for the next time step. This prediction follows easily understandable rules (e.g. in the case of the age which is increased by 1 every simulated year). Dissolving households by divorce, founding new households by marriage, death and birth are simulated with the help of life and marriage tables. Other events, such as entering and leaving education and employment in different types of institutions is also simulated with empirically estimated transition probabilities.

In all cases we will not assume transition probabilities which are the same over the whole population, but these will differ according to the history of the simulated individuals and their embedding in households, regions, strata, professions etc., such that a major task to be fulfilled in the project lies in the collection of such transition probabilities. Unlike usual approaches to microsimulation, every simulation step will end up in an aggregation of data which will be used in the next time step when the simulated individuals make their education- and employment-related decisions.

With the German Socio-Economic Panel (SOEP) and the German Microcensus,  instrumenta are available which cannot only be used for initialising the simulation, but also for estimating most of the transition probabilities needed. And it can also be used for the validation of the model as the empirical panel can be compared to a simulated panel for more than twenty annual periods.

The model itself will be realised as program code within a simulation framework. Both framework and model will be available for policy evaulation and offer a comfortable frame for modeling and simulation which can also be used by other researchers.

The project will be supported by the simulation framework which has been developed by a group of students in their lab course and a diploma thesis

Project Members






  • [Troitzsch2004S] Details BibTeX entry
    Klaus G. Troitzsch: Simulationsverfahren. 2004. 33. 10. WISU - Das Wirtschaftsstudium. 10. pp 1256-1268, 1300-1302.


  • [Troitzsch2003SID] Details BibTeX entry
    Klaus G. Troitzsch: Simulation in den Sozialwissenschaften. in Barbara Orth, Thomas Schwietring, Johannes Weiß: Soziologische Forschung: Stand und Perspektiven. 2003. Leske+Budrich. Opladen. pp 353-363.