A plea for microsimulation Marc Hannappel Department of Sociology University of Koblenz-Landau 56070, Koblenz, Germany Email: marchannappel@uni-koblenz.de KEYWORDS Microsimulation; educational projections; ABSTRACT German demographic or educational projections are conven- tionally based on macrosimulations. Macrosimulation models use average values of the simulation parameters to compute the updating process. Additionally, the proportions of educational graduate rates stay constant in these models. In this paper I introduce a discrete event microsimulation model which is designed to project the graduate rates of the German population. With this simulation model it was demonstrated that the development of the graduation rates can become the subject of the simulation. I NTRODUCTION Since the late 1950s, a similar development can be observed in all western industrial countries. This development can be described as an increase of the educational demand and, as a consequence, an increase of the educational level of the society (Hradil 2006; Allmendinger et al. 2010; Geißler 2008). The proportion of students with a lower educational attainment decreased between 1955 and 2000, whereas the proportion of students with an intermediate and high educational attain- ment increased (Allmendinger et al., 2010). Afterwards, the development can be described as a polarization of educational attainment. The proportion of students with an intermediate educational attainment stagnates whereas the development of the rest continues as the patterns before (Hannappel, 2015). In the political and scientific discourse, this development is usually called “the educational expansion”. Germany occupies a middle rank in the PISA study (The Programme for International Student Assessment) (Deutsches PISA-Konsortium, 2001) which lead to a public and political controversy about the structure, the aims and the curriculum of the German education system. Hence, education policy becomes more and more important. Whereas the sociopolitical point of view focused on the development of the educational level of the population 1 , the educational policy focused on the educational participation of the next generation, e.g. the transition rates within the education system and the graduation rates. Therefore, a main point for policymakers is to provide 1 A widespread assumption is that an increase of the educational level leads to an increase of prosperity and social innovation capability (Anger et al. 2006; Becker 1994; Bildungsberichterstattung 2006; Picht 1964). an adequate educational infrastructure and a sufficient number of teachers (Klemm, 2012). However, educational reforms and teacher-training need time. Hence, education planning needs an anticipatory approach. To calculate the potential demand, policy makers need information about possible (or most likely) future developments of educational attainment. Educational projections of those developments are usually used as an information tool (Kultusministerkonferenz, 2002). Problem: The German ‘Kultusministerkonferenz’ (KMK) 2 regularly calculates educational projections since 1963 (Kul- tusministerkonferenz, 2002). The projection of the graduates of the education system are not calculated completely by the KMK. The calculation is based on a population projec- tion of the German Statistical Office (GSO). The GSO uses macrosimulation for the population projections (Statistisches Bundesamt, 2009). Macrosimulation models use mean values of certain parameters to project a system status (in the case of a population projection it is the age structure of a society) into the “future”. The parameters of the population projection are fertility, migration und mortality. The KMK uses these results to calculate the future graduates of the german education sys- tem (Kultusministerkonferenz, 2013). Additionally, the KMK uses current education transition rates and applies them to the calculated future population. At this point, however, there is a structural problem. The transition rates which are used for the calculation are percentage proportions of students who leave the school with a certain graduation. These proportions remain constant during the macrosimulation. Nevertheless, the previous development of the graduation rates never remain constant. From this point of view, the problem of this kind of projection is the structural separation of population projection and education projection. Objectives: In this paper I present a simulation method which solves the problem of the separation. Instead of a macrosim- ulation model, I use a microsimulation model to project possible developments of educational attainment. Whereas an expansion of microsimulation models and techniques can be observed during the last decade (Li and O’Donoghue, 2013), a similar development is not recognizable for Germany. Approach & Method: The projection of the educational attainment is based on an event-oriented dynamic microsimu- lation approach. Microsimulation models calculate single bio- 2 The KMK is a conference of the education ministers of the different German federal states. Proceedings 30th European Conference on Modelling and Simulation ©ECMS Thorsten Claus, Frank Herrmann, Michael Manitz, Oliver Rose (Editors) ISBN: 978-0-9932440-2-5 / ISBN: 978-0-9932440-3-2 (CD)